ABSTRACTS FOR THE 2018 NAEMSP SCIENTIFIC A · CategoryofSubmission:Cardiac Background: Advanced...

IPEC_A_1377791 TFJATS-ipec.cls September 21, 2017 21:47 Trim Info: 8.25in × 11in

ABSTRACTS FOR THE 2018NAEMSP SCIENTIFIC ASSEMBLY

Oral PresentationAbstracts (1–30)

1. Timing of Advanced Airway Placementafter Out-Of-Hospital Cardiac Arrest:Earlier Is Better

Justin Benoit, Jason McMullan, Henry Wang,Changchun Xie, Peixin Xu, Kimberly Hart,Christopher Lindsell, University of CincinnatiCategory of Submission: Cardiac

Background: Advanced airways (e.g.,endotracheal tubes, supraglottic airways)are frequently placed by Emergency MedicalServices (EMS) in patients with out-of-hospitalcardiac arrest (OHCA). However, the opti-mal timing of advanced airway placementduring the sequence of resuscitation eventsis unknown. We hypothesized that earlieradvanced airway placement would be asso-ciated with increased probability of return ofspontaneous circulation (ROSC). Methods:This secondary analysis of ROC PRIMED studydata included adult, non-traumatic, OHCApatients with advanced airway placement byEMS prior to ROSC. Patients were excludedif EMS witnessed the arrest or arrest timewas unknown. The primary exposure vari-able was time from EMS arrival to advancedairway placement. The outcome variable wasROSC. A Cox proportional hazards modelwas constructed to estimate the probability ofROSC as a function of the time to advancedairway placement using non-linear penalizedsplines. The Cox model was stratified by initialcardiac rhythm, accounted for resuscitationduration, and adjusted for Utstein variablesincluding age, sex, bystander interventions,and EMS response time. Patients were rightcensored at time of hospital arrival or EMStermination of resuscitation. Results: A totalof 7,547 OHCA patients were evaluated. Meanage was 67 years (standard deviation 15),69% were male, 38% had an initial shockablerhythm, and 49% received bystander CPR.Median EMS response time was 6 minutes(interquartile range 4–7). Time from EMSarrival to advanced airway placement was0–5 minutes (12%), 5–10 (36%), 10–15 (29%),15–20 (14%), 20–25 (5%), 25–30 (2%), and >30(2%). Median time from EMS arrival to ROSCwas 19 minutes (interquartile range 14–25).Time to advanced airway placement wassignificantly associated with ROSC based onthe Cox model. For initial shockable rhythms,

PREHOSPITAL EMERGENCY CARE 2017;0:1–51

doi: 10.1080/10903127.2017.1377791

the probability of ROSC was 59%, 55%, 51%,45%, 39%, and 33% with airway placement at5, 10, 15, 20, 25, and 30 minutes, respectively.For non-shockable rhythms, the probabilityof ROSC was 43%, 40%, 35%, 30%, 25%, and20% at the same airway intervals. Conclusions:EMS advanced airway placement for OHCAhas a time-dependent association with ROSC.Early advanced airway placement is associatedwith increased ROSC, regardless of initialcardiac rhythm.

2. EMS Agencies with High Rates of FieldTermination of Cardiac Arrest Care AlsoHave High Rates of Survival

John Summers, Christopher Berry, AnneKnorr, Mark Olaf, Douglas Kupas, GeisingerHealth System Category of Submission:Cardiac

Background: The relationship between fieldtermination of resuscitation (FTOR) and sur-vival from cardiac arrest is unknown. Wehypothesized that EMS agencies with more fre-quent FTOR would be more likely to opti-mize resuscitative efforts on scene and wouldalso have better patient outcomes. Methods:The Cardiac Arrest Registry to Enhance Sur-vival (CARES) identified out-of-hospital car-diac arrests (OOHCAs) occurring from 2013 to2016. A priori, EMS agencies were included ifthey submitted at least 80 cases during thisperiod. Subsequently, agencies were dividedinto quartiles based upon FTOR frequency. Thetop and bottom quartiles were identified as high(HFTAs) and low field termination agencies(LFTAs). Generalized estimating equation mod-els were used to compare HFTAs and LFTAs.Results: Seventy agencies were classified asHFTAs (treating 31,486 OOHCA patients) and70 agencies were classified as LFTAs (treating27,314 OOHCA patients). FTOR was performedon 51.6% HFTA patients and on 7.1% of LFTApatients. The mean patient age was 62.1 yearsand 61.2% were male. HFTAs were more likelyto have patients with a shockable rhythm (OR =1.16, 95%CI 1.1–1.3, p = .003) and who receivedbystander CPR (OR = 1.52, 95%CI 1.3–1.7,p < .001) than LFTAs. HFTAs had higher pro-portions of ROSC (35.4% vs. 26.4%, OR = 1.38,95%CI 1.2–1.6), survival to discharge (12.5% vs.8.5% OR = 1.46, 95%CI 1.3–1.7), and favor-able neurologic outcome in survivors (86.7% vs.77.9%, OR = 1.84, 95%CI 1.4–2.4) than LFTAs,all p < .001; These results remained signifi-cant after controlling for patient characteris-tics like age, shockable rhythm, and bystanderCPR. When compared to LFTAs, HFTAs spentgreater time at the scene before patient trans-port (25 min vs. 16 min, 95%CI 6.3–9.0, p <.001)and were more likely to administer drugs topatients (92.0% vs. 86.7%, 95%CI 1.0–2.1, p =.04). Conclusions: EMS agencies with the high-est rates of FTOR also have higher rates ofROSC, survival, and good neurologic outcome.HFTAs spend more time on scene before patient

transport, suggesting they may not have aculture of “scoop and swoop” for OOHCA care.Additional studies are needed to identify anysubgroup of OOHCA that may benefit by trans-port for care at a hospital.

3. Prehospital Delivery of DeathNotifications Associated with HigherRates of Occupational Burnout AmongEMS Professionals

Remle Crowe, Rebecca Cash, Madison Rivard,Abraham Campos, Brian Clemency, RobertSwor, Eric Ernest, Ashish Panchal, The NationalRegistry of Emergency Medical TechniciansCategory of Submission: Operations, Qual-ity, Safety, Systems, Disaster

Background: EMS professionals often under-take the difficult task of notifying familieswhen a death occurs in the prehospital set-ting. However, many do not receive relatedtraining, which may exacerbate the associatedstress. The emotional strain that accompaniesdeath notifications has been linked to burnoutin other healthcare settings, yet this has notbeen examined in EMS. Our objective was toassess the relationship between death notifica-tion, training and work-related burnout amongEMS professionals. We hypothesized that aftercontrolling for training, delivering death notifi-cations would be associated with higher oddsof burnout. Methods: We analyzed data froma cross-sectional electronic survey adminis-tered in April 2017. A sample size calcula-tion approximated that 1,300 responses wereneeded to make estimates with 95% confidence.Assuming an 11% response rate from previouswork, we randomly selected 19,330 nationally-certified EMS professionals. Inclusion criteriaconsisted of EMTs or higher, practicing in non-military settings. We assessed burnout usingthe validated Copenhagen Burnout Inventoryand providers self-reported training and thenumber of adult death notifications deliveredin the past 12 months. We conducted mul-tivariable logistic regression modelling usingconfounders selected a priori from previousresearch: certification level, experience, agencytype, and call volume. We used the Hosmer-Lemeshow goodness-of-fit test to assess modelcalibration. Results: We received 2,333/19,330responses (response rate:12.1%) and 1,514 (65%)met inclusion criteria. Over half (53%, n =780) delivered at least one death notificationin the past 12 months, while one-third (32%, n= 468) exhibited burnout. A step-wise increasein burnout prevalence was noted as numberof death notifications increased. The prevalenceof burnout was 23%, 36%, and 51% for thosewho delivered 0, 1–5, and 6 or more death noti-fications, respectively. After adjustment, deliv-ering one or more death notifications wasassociated with 47% greater odds of burnout(OR:1.47, 95%CI:1.12–1.94). Meanwhile, train-ing was associated with reduced odds ofburnout (OR:0.60, 95%CI 0.47–0.77). Conclu-

1

https://crossmark.crossref.org/dialog/?doi=10.1080/10903127.2017.1377791&domain=pdf&date_stamp=2017-09-21

https://doi.org/10.1080/10903127.2017.1377791


2 PREHOSPITAL EMERGENCY CARE XXXXX 2017 VOLUME 0 / NUMBER 0

sions: After adjustment for provider charac-teristics including experience, delivering deathnotifications was associated with higher odds ofburnout, while training was protective. Impor-tant limitations include response bias, recallbias and the cross-sectional nature of this evalu-ation. Prospective work is needed to explore theunderlying causes of this relationship.

4. Incivility among Nationally CertifiedEms Professionals Is Linked toWorkforce-Reducing Factors

Rebecca Cash, Remle Crowe, Kim White-Mills, Madison Rivard, Ashish Panchal, TheNational Registry of Emergency Medical Techni-cians Category of Submission: Operations,Quality, Safety, Systems, Disaster

Background: Incivility is defined as nega-tive interpersonal acts that violate norms forsocial interaction, ranging from breaches ofetiquette to outright harassment. In otherhealthcare settings, incivility has been linkedto negative individual and organizationaleffects, although scant literature exists con-cerning incivility in the unique, high-stressEMS setting. Our objective was to describe theassociation between incivility and stress, careersatisfaction, turnover intentions, and work-place absence among EMS professionals. Wehypothesized that exposure to incivility wouldbe linked to poorer personal and occupationalwell-being. Methods: Based on a sample sizecalculation, 38,000 nationally-certified EMSprofessionals received an electronic ques-tionnaire. Incivility was measured using anEMS-adapted Workplace Incivility Scale (WIS).Stress was measured using the depression anx-iety and stress scale (DASS). Satisfaction wasmeasured using a 4-point Likert scale and highself-reported workplace absence was classifiedas missing 10 or more days of work in the previ-ous 12 months. Non-military, practicing EMTsor higher were included in the analysis. Mul-tivariable logistic regression was conductedusing a priori selected covariates based ondirected acyclic graphs to obtain adjusted oddsratios and 95% confidence intervals (OR, 95%CI) to examine the association between expe-riencing incivility and the outcomes of interest.Results: A total of 3,741 EMS professionalsresponded (response rate = 10.3%), with 2,815(75%) meeting inclusion criteria. Most weremale (70%) and white, non-Hispanic (87%)with 54% certified at the EMT level. Incivilitywas experienced at least once per week by47% of respondents. Exposure to incivility wasassociated with greater odds of dissatisfactionwith EMS (4.70, 3.48–6.35), dissatisfactionwith a main EMS job (6.68, 4.99–8.93), dissat-isfaction with immediate supervisors (11.04,8.21–14.85), increased stress (5.31, 4.04–6.98),intent to leave one’s job or the EMS professionin the next 12 months (3.99, 3.17–5.02 and 3.55,2.48–5.09, respectively), and workplace absence(1.38, 1.06–1.81). Conclusions: About half ofnationally-certified EMS professionals wereexposed to regular incivility. Exposure to inci-vility was associated with workforce reducingfactors such as career dissatisfaction, stress,turnover intentions, and workplace absence.Further research is needed to understandhow organizational climate and interpersonalbehaviors in the workplace affect individualemployees and EMS workforce stability.

5. Statewide Trends in Out-of-HospitalCardiac Arrest Related to Drug Overdose

Samuel Beger, Gabriella Smith, VatsalChikani, Daniel Spaite, Samuel Keim, TerryMullins, Taylor George, Bentley Bobrow, Uni-versity of Arizona College of Medicine – PhoenixCategory of Submission: Student, Resident,Fellow

Background: Along with out-of-hospital car-diac arrest (OHCA), opioid abuse and over-dose (OD) have become major public healthproblems in the US. While opioid-relateddeaths have increased in the US, recent tem-poral and regional trends in the proportion ofOHCAs related to overdose (OD-OHCA) arelargely unknown and may impact treatmentstrategies and outcomes. Objective: To assesstrends in incidence, process of care, and out-comes of OD-OHCAs compared to presumedcardiac etiology arrests (C-OHCA). Methods:Statewide observational study utilizing anUtstein-style database, along with detailedreview of EMS first care reports linked with hos-pital records and vital statistics data between2010 and 2015. The proportion and 95% Con-fidence Intervals were calculated to comparethe rate of arrests between OD-OHCAs vs. C-OHCAs. Multivariate logistic regression wascarried out to compare survival between thetwo groups. Results: There were a total of21,658 confirmed OHCAs during the studyperiod. After excluding non-C-OHCAs/non-OD-OHCAs, 18,988 cases remained. Overall,18,001 (94.8%) of arrests were C-OHCA and 987(5.2%) were OD-OHCA. There was a signifi-cant increase in the proportion of OD-OHCAsbetween 2010, 4.6% (95% CI = 3.8–5.4) and2015, 6.4% (95% CI = 5.7–7.3). Mean age forOD-OHCA was 38.8 yrs compared to 64.2 yrsfor C-OHCA (p < 0.0001) and location of OD-OHCAarrests was more likely residential 66.6%vs. 54.0% (p < 0.0001). Shockable rhythm waspresent in 7.0% of OD-OHCAs vs. 22.6% of C-OHCAs (p < 0.0001). Bystander CPR was per-formed in 49.4% of OD-OHCAs vs. 48.3% ofC-OHCAs (p < 0.5231). Overall survival to dis-charge in the OD-OHCA group was 18.6% vs.11.9% in the C-OHCA group (p < 0.0001). Afterrisk adjustment, there was an aOR of 2.0 (1.6–2.5) for survival to hospital discharge in the OD-OHCA group compared to the C-OHCA group.Conclusions: This statewide study found a sig-nificant upward trend in the proportion of OD-OHCAs as well as differences in populationdemographics and epidemiology. Given thevarying etiology, location, and age, it is surpris-ing that the bystander CPR rates were nearlyidentical. It is likely that regional variations inOD-OHCAs exist and emergency medical sys-tems should track data to optimize their preven-tion and resuscitation efforts.

6. Death by Suicide: The EMS ProfessionCompared to the General Public

Bentley Bobrow, Micah Panczyk, RobynBlust, Paula Brazil, Taylor George, VatsalChikani, Chengcheng Hu, Daniel Spaite,Arizona Department of Health Services Categoryof Submission: Operations, Quality, Safety,Systems, Disaster

Background: EMS professionals face high lev-els of chronic physical/emotional stress andPost Traumatic Stress Disorder related to pre-hospital care. Suicide has been linked to otherfirst responder professions, such as law enforce-ment, presumably related to multiple chronicstressors. While high-profile anecdotal EMTsuicide cases and national survey data on sui-cidal ideation/attempts have received muchattention, there is a paucity of data on EMTsuicide completions. We sought to determinethe statewide proportionate mortality ratios ofsuicide completions among EMTs compared tothe general public (GP) in Arizona. Methods:Observational study of adults (�18 yrs; 1/2009–12/2015. The Arizona Vital Statistics Infor-mation Management System-Electronic DeathRegistry was queried with manual reviewof decedent occupation free-text fields. Thesedata were compared to the non-EMT cohortaggregate of all other occupations combined.

Suicide was defined based on ICD-10 E-Codes.The proportionate mortality ratios (PMRs) forsuicide were compared between the groups,after adjusting for age, sex, race, and ethnic-ity. Results: There were a total of 349,793 GPdeaths (all causes) of which 7,775 (2.2%) wereby suicide. EMT death total was 1,205 EMT–63 (5.2%) by suicide. Demographics of suicide:Mean age: GP–48.7 yrs; EMT - 43.4 yrs (p =0.023); Male: GP–77.3%; EMT–88.8% (p = 0.029);White non-Hispanic: GP–80.0%; EMT–73.0%(p = 0.166). The crude odds ratio (OR) for EMTsuicide was 2.43 (95%CI = 1.88–3.13) comparedto the GP. The adjusted OR (aOR) for EMT sui-cide was 1.39 (95%CI = 1.06–1.82) compared tothe GP. The top three mechanisms of suicideamong EMTs and the GP in Arizona, respec-tively, were firearm (67% vs. 57%), suffocation(24% vs. 21%), and poisoning (9.5% vs. 17%).Conclusions: In this statewide analysis, EMTsin Arizona had a significantly higher propor-tional mortality ratio of deaths due to suicidecompared to the general population, even aftercontrolling for age, sex, race, and ethnicity. Thisis the first study that we are aware of to com-pare EMT suicide completions with the generalpublic. Hopefully this information will increaseawareness and spur studies to elucidate under-lying causes and evaluate the effectiveness ofinterventions.

7. Assessment of the Rapid ArterialOcclusion Evaluation (Race) Scale inReal-World Practice for Prediction ofLarge Vessel Occlusion and Reducing Timeto Thrombectomy

Peter Antevy, Brijesh Mehta, Ashutosh Jad-hav, Joy Sessa, Randy Katz, Hoang Duong,Andrey Lima, Gina Dimartini, LakotaWoodall, Ryan McTaggart, Ronil Chandra,Thabele Leslie-Mazwi, Joshua Hirsch, AlbertYoo, Tudor Jovin, Raul Nogueira, MemorialHealthcare System Category of Submission:Medical

Background: Prehospital identification ofpotential large vessel occlusion (LVO) strokepatients may lead to faster triage and treat-ment. We examined whether the Rapid ArterialOcclusion Evaluation (RACE) scale can bereliably implemented in a real-world settingwith multiple EMS agencies and lead to rapidtreatment. Methods: A prospective study wasperformed at a high volume comprehensivestroke center. In the first phase, eight EMSagencies were educated on use of the RACEscale using an online training video. All EMSstroke alerts were recorded. When EMS RACEscore was 5 or higher, the neurocath lab teamwas alerted prior to EMS arrival as part of aparallel workflow. Upon emergency depart-ment arrival, the following characteristics weretracked: NIHSS score, RACE score, CT findings,presence of LVO and workflow time metrics.Results: During the study period (January 2016to June 2017), RACE score was provided for 797of 1498 EMS stroke alerts (53%). Higher pre-hospital RACE scores correlated with NIHSSscores. LVO was found in 13% of patients withan available RACE score. A RACE score of 5or higher was able to identify 64% of all LVOpatients (sensitivity: 64%; specificity: 72%;PPV: 30%; NPV: 93%; accuracy: 71%; Youden’sindex). However, of the 260 patients with RACEscore 5 or higher, only 68 patients (26%) werefound to have LVO while 29 patients (11%) hadICH; among 499 patients with RACE score lessthan 5, LVO was present in 38 patients (8%).When an EMS stroke alert with high RACEscore triggered early alert of the neurocath labteam, median door to groin puncture time forthrombectomy was 68 minutes compared to91 minutes for cases with sequential work-flow. Conclusions: The RACE scale can be


NAEMSP 2017 ANNUAL MEETING ABSTRACTS 3

successfully implemented across EMS agenciesand results in faster door to groin puncturetimes. While a RACE score of 5 or higher isassociated with greater likelihood of LVO, thereare a significant number of false positives. Fur-ther refinement of prehospital stroke severityscales is warranted to improve the accuracy ofthis approach.

8. Effecting Neurologically-IntactSurvival for Children withOut-of-Hospital Cardiac Arrest

Paul Pepe, Paul Banerjee, Amninder Singh,Latha Ganti, University of Texas Southwest-ern Medical Center Category of Submission:Pediatric

Background: EMS crews commonly limit on-scene care for pediatric out-of-hospital cardiacarrest (POHCA) patients, typically attempt-ing to provide treatment while transporting.Hypothesis: Neurologically-intact survivalfor children can be improved by deferringtransport and prioritizing on-site care usingstrategies that expedite on-scene drug deliveryand intubation with tightly-controlled ventila-tion. Methods: Data for all consecutive POHCAcases between January 1, 2012 and April 30,2017 were collected prospectively (compre-hensive Utstein-style registry). In 2014, newtraining prioritized on-scene resuscitationstrategies (Phase I) that expedited drug deliv-ery and intubation with controlled ventilation(e.g., rates ∼6/min). In 2016, techniques todose/prepare drugs while responding wereintroduced (Phase II). Neuro-intact survivalin 2012–13 (Phase 0, pre-changes) were thencompared to Phase I and II outcomes. Through-out the study, protocols followed the 2010American Heart Association guidelines. Noother relevant modifications were madesystem-wise. The modified training includedpsychological and skills-enhancing tools toprovide greater confidence in providing on-scene care. Results: EMS crews managed 143consecutive POHCA cases over the 5.33-yearstudy period throughout which the majorityof children continued to present in asystole,including those resuscitated. In resuscitatedpatients, the interval from vehicle arrival on-scene to the first epinephrine administration fellfrom 16.5 minutes (2012–2013) to 7.3 minutes(Phase I) and 5.0 minutes (Phase II). Childrenreceived intubation and intraosseous insertionin much greater frequency on-scene in Phase Iand II with no other significant differences interms of age, sex, etiology, response intervals,or sequence of drug infusions. Rates of survivalto hospital discharge with intact neurologicalstatus did improve immediately: 23.2% (13/56)in Phase I and 34.7% (17/49) in Phase II versus0 of 38 for the pre-change calendar years of2012–2013 (p < 0.0001; 2-tailed Fisher’s exacttest). By 2017, the mean time to epinephrineadministration had fallen to 2 minutes forresuscitated patients and 3.33 minutes for allpatients. Conclusions: Although a historically-controlled study, the sudden appearance ofneuro-intact survivors following the renewedfocus upon on-scene care was profound, imme-diate and sustained. Beyond skills-enhancingstrategies, physiologically-driven techniquesand supportive encouragement from leader-ship, pre-arrival psychological and clinicaltools were also likely contributors to theobserved outcomes.

9. Motivations for Exiting the EMSProfession Differ between EMTS andParamedics

Madison Rivard, Remle Crowe, Rebecca Cash,Jeremy Miller, Ashish Panchal, The NationalRegistry of Emergency Medical Technicians

Category of Submission: Operations, Quality,Safety, Systems, Disaster

Background: Understanding motivations forexiting the workforce is important to improverecruitment and retention of EMS profession-als. Factors influencing the choice to leaveEMS have not been explored by provider level.Our objective was to describe and comparethe most important factors in the decisionto leave EMS among EMTs and paramedics.As education requirements and practice set-tings vary between EMTs and paramedics, wehypothesized that reasons for leaving EMSdiffered by certification level. Methods: Thiswas a cross-sectional analysis of an electronicquestionnaire deployed in June 2017 to allnationally-certified EMTs and paramedics whodid not renew National EMS Certification dur-ing the 2016–2017 recertification period end-ing on March 31, 2017. Since National EMSCertification is not required to renew a licensein all states, participants were asked if theywere currently practicing in EMS. Inclusioncriteria consisted of those who reported notworking in EMS. Z-tests of proportion witha Bonferroni adjustment for multiple compar-isons were used to evaluate differences inreasons for leaving EMS between EMTs andparamedics. Results: We received 4,793/51,344responses (response rate = 10%) and 2,703 metinclusion criteria. Most were EMTs (85%, n =2,291) and 15% were paramedics (n = 412).For EMTs, the most commonly selected rea-son for leaving EMS was the pursuit of furthereducation (22%), while paramedics most com-monly cited a desire for better pay and ben-efits (20%). There was more than a two-foldincrease in the proportion of paramedics thatselected illness/injury/disability compared toEMTs (13% vs. 6%, p < 0.001). Three timesas many paramedics selected stress/burnoutcompared to EMTs (9% vs. 3%, p < 0.001).Only 5% of EMTs listed retirement as the mostimportant factor for leaving EMS comparedto 14% of paramedics (p < 0.001). Exclud-ing those who left for retirement, 68% ofEMTs stated they intended to return to EMS,compared to 32% of paramedics (p < 0.001).Conclusions: Important factors related to leav-ing EMS differed by provider level. Of concern,a larger proportion of paramedics reportedillness/injury/disability or stress/burnout astheir primary reason for leaving the profes-sion compared to EMTs. Additionally, fewerparamedics reported an intention to return toEMS. Limitations include potential responsebias and confounding.

10. Do Age Appropriate Vital Sign CutPoints Improve the Predictive Ability ofthe Physiologic Criteria of the FieldTriage Decision Scheme for IdentifyingChildren Who Need the Resources of aTrauma Center

E. Brooke Lerner, Jeremy Cushman, MohamedBadawy, Amy Drendel, Courtney Jones,Manish Shah, David Gourlay, Medical Collegeof Wisconsin Category of Submission: Trauma

Background: Prior research found the FieldTriage Decision Scheme’s (FTDS) physiologicstep is a moderate predictor of pediatric traumacenter (TC) need. Predictive ability could behindered by the current use of adult valueswhen defining abnormal vital signs. Our objec-tive was to determine the accuracy of the FTDSphysiologic step when traditional cut pointsare compared to age-specific cut points foridentifying children needing TC resources.Methods: A prospective study of all injuredchildren �15 years, regardless of severity, trans-ported by EMS to pediatric TC was conducted

in three mid-sized cities. EMS providers wereinterviewed to obtain patient demographicsand presence or absence of each FTDS criteria.Children were considered to need a TC if theymet a published consensus definition. Outcomedata was obtained through structured hospitalrecord review. The over- and under-triage ratesand positive likelihood ratios (+LR) were cal-culated using traditional and age-specific cutpoints for the physiologic step, as well as forsystolic blood pressure (SBP), and respiratoryrate (RR). Results: EMS and outcome data wereavailable for 9,484 children. 2% of all patientsneeded the resources of a TC. 11% of patientsmet the physiologic step when traditional cutpoints were used and 23% when age-specificcut points were used. Using the traditionalphysiologic criteria, 46% of children needing aTC would have been under-triaged and 10%over-triaged (+LR 5.44, 95%CI 4.75–6.24). Usingthe age-specific physiologic criteria, 40% wouldhave been under-triaged and 22% would havebeen over-triaged (+LR 2.69, 95%CI 2.40–3.01).The traditional RR cut point had a +LR of 3.12(95%CI 2.39–4.07). The age-specific RR cut pointhad a +LR of 1.86 (95%CI 1.56–2.22). The tra-ditional SBP had a +LR of 5.28 (95%CI 3.35–8.34). The age-specific SBP had a +LR of 6.10(95%CI 3.54–10.00). EMS did not obtain RR in16% and SBP in 28% of cases. Conclusions: Theaccuracy of the physiologic step of the FTDS isnot improved by using age-specific criteria. Therate of under-triage is decreased while the rateof over-triage is increased.

11. Comparative Effectiveness ofAntiarrhythmics for Out-of-HospitalCardiac Arrest: A Systematic Review andNetwork Meta-Analysis

Shelley McLeod, Romina Brignardello-Petersen, Andrew Worster, John You, AllaIansavichene, Gordon Guyatt, SheldonCheskes, Schwartz/Reisman Emergency MedicineInstitute, University of Toronto Category ofSubmission: Cardiac

Background: The objective of this systematicreview, direct pairwise meta-analysis andnetwork meta-analysis (NMA) was to assessthe use of antiarrhythmic drugs for patientsexperiencing out-of-hospital cardiac arrest(OHCA). Methods: Electronic searches ofMedline, EMBASE, and Cochrane CentralRegister of Controlled Trials were conductedand reference lists were hand-searched.Randomized controlled trials (RCTs) inves-tigating the use of antiarrhythmic agentsadministered during resuscitation for adult(� 18 years) patients suffering non-traumaticOHCA were included. Two reviewers inde-pendently screened abstracts, assessed riskof bias of the included studies, and extracteddata for the following outcomes: return ofspontaneous circulation (ROSC), survival tohospital admission, survival to hospital dis-charge and survival to hospital discharge withgood neurologic status. Direct and indirectevidence were combined in a NMA using afrequentist approach with fixed-effects modelsand reported as relative risks (RR) with 95%confidence intervals (CIs). For each pairwisecomparison, the certainty of direct, indirect,and network evidence was assessed using theGRADE approach. Results: 8 RCTs involving4,464 patients were combined to comparethe effectiveness of five antiarrhythmic agents(amiodarone, bretylium, lidocaine, magnesium,and sotalol) and placebo administered duringresuscitation following OHCA. Lidocainewas associated with a statistically significantincrease in ROSC compared to placebo (1.15;95% CI: 1.03–1.28) and was also superior tobretylium (1.61; 95% CI: 1.00–2.60) for ROSC.



When compared to placebo, both amiodarone(1.18; 95% CI: 1.08–1.30) and lidocaine (1.18;95% CI: 1.07–1.30) were associated with astatistically significant increase in survival tohospital admission (certainty of the evidencewas high). However, no antiarrhythmic wasstatistically more effective than placebo forsurvival to hospital discharge or neurologicallyintact survival, and no antiarrhythmic wasconvincingly superior to any other for anyoutcome (certainty of the evidence was lowor very low). Conclusions: Amiodarone andlidocaine were the only agents associated withimproved survival to hospital admission inthe NMA. For the outcomes most importantto patients, survival to hospital discharge andneurologically intact survival, no antiarrhyth-mic was convincingly superior to any other orto placebo.

12. Emergency Medical Services ProviderPerspectives on Pediatric Calls: AQualitative Study

Jessica Jeruzal, Lori Boland, Monica Frazer,Jonathan Kamrud, Russell Myers, CharlesLick, Andrew Stevens, Allina Health Emer-gency Medical Services Category of Submission:Cardiac

Background: Previous survey results in ourambulance service indicate that 9-1-1 responseto incidents involving children are particularlydistressing for emergency medical services(EMS) clinicians. This qualitative study wasconducted to increase understanding about thedifficulties of responding to pediatric calls andobtain information about how organizationscan better support EMS providers in managingpotentially difficult calls. Methods: Paramedicsand emergency medical technicians from a sin-gle U.S. ambulance service were invited to par-ticipate in focus groups about responding to 9-1-1 calls involving pediatric patients. A total of17 providers from both rural and metro serviceregions participated in six focus groups held incommunity meeting spaces. A semistructuredfocus group guide was used to explore: (1)elements that make pediatric calls difficult, (2)pre-arrival preparation practices, (3) experi-ences with coping after difficult pediatric calls,and (4) perspectives about follow-up resourcesand support. Focus groups were audio recordedand transcripts were analyzed using standardcoding, memoing and content analysis meth-ods in qualitative analysis software (NVivo).Results: Responses about elements that makepediatric calls difficult were organized intothe following themes: (1) the social value ofchildren, (2) clinical difficulty of pediatric calls,(3) type or nature of the call, (4) interactionswith parents, and (5) location/scene chal-lenges. With regard to pre-arrival preparation,participants often cited mentally reviewingprotocols, equipment location and dosages,and discussing assignment of on-scene taskswith their partner. The use of retrospective callreviews among peers was highlighted acrossthe topic areas as a high-value, commonly usedmethod for coping with difficult pediatric callsthat also serves as a learning or preparationtool. Suggestions for additional supportiveresources included: increasing opportunitiesfor external feedback (e.g., from hospital-basedstaff); additional, more frequent pediatric clin-ical training; institutionalization of structuredrecovery time after difficult calls; and improvedstorage and labeling of pediatric equipment.Conclusions: This study provides qualitativedata about the difficulties of responding topediatric calls and resources needed to supportclinicians. Findings from this study will beused to guide EMS leadership in designingand implementing institutional initiatives toenhance wellbeing among EMS clinicians.

13. Effectiveness of PrehospitalHypertonic Saline for HypotensivePatients: A Systematic Review andMeta-Analysis

Ian Blanchard, Armghan Ahmad, Karen Tang,Paul Ronksley, Diane Lorenzetti, GeraldLazarenko, Eddy Lang, Christopher Doig, HStelfox, Alberta Health Services/University of Cal-gary Category of Submission: Professional

Background: The optimal prehospital fluidfor the treatment of hypotension is unknown.Hypertonic fluids, meaning that the composi-tion of solutes is higher to that of the humanbody, may increase circulatory volume andmute the pro-inflammatory response of thebody to injury and illness. The purpose of thisstudy was to determine whether in patientspresenting with hypotension in the prehospi-tal setting (population), the administration ofhypertonic saline (intervention), compared toan isotonic fluid (control), improves survival tohospital discharge (outcome). Methods: In thisPROSPERO registered review, searches wereconducted in Medline, Embase, CINAHL, andCENTRAL from the date of database inceptionto November, 2016, and included all languages.Two reviewers independently selected random-ized control trials of hypotensive human par-ticipants administered hypertonic saline in theprehospital setting. The comparison was iso-tonic fluid, which included normal saline, andnear isotonic fluids such as Ringer’s Lactate.Assessment of study quality was done usingthe Cochrane Collaborations’ risk of bias tooland a fixed effect meta-analysis was conductedto determine the pooled relative risk of sur-vival to hospital discharge. Secondary out-comes were reported for fluid requirements,multi-organ failure, adverse events, length ofhospital stay, long term survival and disabil-ity. Results: Of the 1,160 non-duplicate cita-tions screened, 38 articles underwent full-textreview, and five trials were included in thesystematic review. All studies administered afixed 250 mL dose of 7.5% hypertonic saline,except one that administered 300 mL. Two stud-ies used normal saline, two Ringer’s Lactate,and one Ringer’s Acetate as control. Routinecare co-interventions included isotonic fluidsand colloids. Five studies were included in themeta-analysis (n = 1,162 injured patients) withminimal statistical heterogeneity (I2 = 0%). Thepooled relative risk of survival to hospital dis-charge with hypertonic saline was 1.02 timesthat of patients who received isotonic fluids(95% CI: 0.95,1.10). There were no consistentstatistically significant differences in secondaryoutcomes. Conclusions: There was no signifi-cant difference in important clinical outcomesfor hypotensive injured patients administeredhypertonic saline compared to isotonic fluid inthe prehospital setting. Hypertonic saline can-not be recommended for use in prehospital clin-ical practice for the management of hypotensiveinjured patients based on the available data.

14. Are EMS Provider CharacteristicsAssociated with Appropriate Responsesduring Violent Encounters?

Donald Garner, Mallory DeLuca, RemleCrowe, Rebecca Cash, Madison Rivard,Jefferson Williams, Ashish Panchal, JoseCabanas, Wake County EMS Category ofSubmission: Professional

Background: Violence against EmergencyMedical Services (EMS) providers is increasing.Little is known regarding providers’ responseduring threatening encounters. Recognitionand management of threatening situations iskey to provider and patient safety. Our objec-tive was to evaluate the association betweenprovider characteristics and response to

escalating threats of violence during EMS calls.We hypothesized that providers with greaterEMS experience and training would be morelikely to escape threatening situations. Meth-ods: EMS providers of a large county-based sys-tem participated in specially-developed patientcare simulations. Each scenario escalatedthreats of violence so that providers shouldescape the scene for safety. Trained evaluatorsrecorded performance per provider on 51standardized data elements including time, de-escalation attempts, and escape. Our primaryoutcome was whether the provider escapedbefore the scenario ended. Our secondaryoutcome was whether a provider made an ade-quate de-escalation attempt. Descriptive statis-tics and univariable odds ratios (OR, 95%CI,p-value) were calculated. Results: We evalu-ated 272 EMS providers as individual membersof two-person crews, with <3% missing data.Overall, 55% (n = 145/263) made an adequatede-escalation attempt and 55% (n = 147/268)escaped the unsafe scene. Of those who did notescape, nearly half (44%, n = 53/120) also didnot make an adequate de-escalation attempt.EMS experience (p = 0.31) and military back-ground (p = 0.39) were not associated withodds of de-escalation. A two-fold increase inodds of adequately attempting de-escalationwas observed for providers with Crisis Inter-vention Training (CIT) (2.13, 1.15–3.93, p =0.02). As experience increased, a stepwisedecrease in the proportion of providers thatescaped was noted (p-trend = 0.01). Providerswith 20-plus years of EMS experience had 64%lower odds of escaping (0.36, 0.17–0.76, p <0.01; referent:<5 years experience). Providerswith military experience (0.38, 0.18–0.84, p =0.02) or CIT (0.37, 0.20–0.67, p < 0.01) also hadreduced odds of escaping. Conclusions: Nearlyhalf of EMS providers failed to escape a sim-ulated scene with threat of physical violence.Experienced providers and those with militaryor CIT training had lower odds of escaping.Limitations include that these results wereobtained in a training environment. Futureresearch should focus on developing trainingto improve recognition of failed de-escalationand the need to escape an unsafe scene.

15. Performance Characteristics of theModified Rapid Arterial OcclusionEvaluation Scale (MRACE) To PredictLarge Vessel Occlusion

Hinnah Siddiqui, Denisse Sequeira, Mar-cus Robinson, Christian Martin-Gill, FrancisGuyette, Department of Emergency Medicine, Uni-versity of Pittsburgh School of Medicine Categoryof Submission: Student, Resident, Fellow

Background: Stroke is a leading cause of dis-ability in the United States. The most debilitat-ing strokes are caused by large vessel occlusion(LVO), and patient outcomes are improvedthrough delivery of time-sensitive endovascu-lar therapies at comprehensive stroke centers(CSC). The Rapid Arterial Occlusion Evaluation(RACE) scale can identify patients with LVOand facilitate triage to CSCs, with publishedsensitivity of 68% and specificity of 85% at scoreof �5. We aimed to demonstrate the implemen-tation feasibility and performance of prehospi-tal mRACE scale, which does not assume thelaterality of aphasia and agnosia symptoms, toidentify LVO. Methods: The mRACE scale wasimplemented in 12 EMS agencies, scoring bothaphasia and agnosia regardless of lateralityof symptoms to improve the ease of trainingand capture of atypical symptoms. Trainingconsisted of a didactic presentation with bothvideo and hands-on demonstrations of patientscenarios. A step-by-step scoring guidedparamedics through the exam. mRACE datawere collected prospectively and documented



upon completion of the prehospital electronichealth record. A project coordinator obtainedin-hospital data elements for those individu-als transported to UPMC facilities. Analysisincluded descriptive statistics and performancecharacteristics (sensitivity, specificity, positivepredictive value (PPV), and negative predictivevalue (NPV). Results: From December 2015 toJuly 2017, a prehospital mRACE scale was com-pleted for 780 patients with suspected stroke.Complete in-hospital data were available for517 (66%). Of these, 186 had a mRACE scaleof �5. There were 188 (36%, CI 32–40%) caseswith final diagnosis of ischemic stroke of which65 (12.6% CI 10–16%) had LVO. This yielded75.3% (CI 72–79%) sensitivity, 68.6% (CI 65–73%) specificity, 56.3% (CI 45–67%) PPV, and83.8% (CI 75–90%) NPV with a ROC AUC of0.76 in the identification of LVO. Conclusions:Implementing the prehospital mRACE scaleto identify patients with LVO is feasible andperforms similarly to the RACE scale withoutneed to discriminate laterality of symptoms.Further research is necessary to determine ifimplementation of the mRACE scale leads toincreased interventions for patients with LVOand subsequent decreased morbidity.

16. Effects of Failed DefibrillationAttempts on Waveform Characteristics ofthe Ventricular FibrillationElectrocardiogram

Jacob Thomas, David Salcido, JamesMenegazzi, Department of Emergency Medicine,University of Pittsburgh School of MedicineCategory of Submission: Student, Resident,Fellow

Background: The morphology of the electro-cardiogram (ECG) of the ventricular fibrillation(VF) waveform during cardiac arrest can bequantified using signal analysis (QECG). Stud-ies have shown that QECG measures may bepredictive of defibrillation success. We soughtto quantify the effect of failed rescue shocks onthe QECG values for patients with VF in out-of-hospital cardiac arrest (OHCA). We consid-ered a failed shock to be one in which the ECGrhythm was VF prior to and after the shock. Wehypothesized that failed rescue shocks wouldlead to worsened QECG measures. Methods:Electronic defibrillator data were taken fromnon-traumatic, EMS-treated OHCA cases fromthe Resuscitation Outcomes Consortium (ROC)Continuous Chest Compression trial. For eachshock, QECG values amplitude spectrum area(AMSA), median slope (MS), centroid fre-quency (CF), and detrended fluctuation analy-sis (DFA) were calculated for the closest artifact-free 3 second gap in chest compressions priorto and after the shock. We used custom-builtMATLAB programs to perform QECG calcu-lations. QECG values were compared using apaired t-test for the pre- and post-shock val-ues. Correlation coefficients were also calcu-lated between the time from shock to post-shock window and the change in QECG values.Results: Out of 5,195 total shocks, 1,399 shockswere analyzable. 520 were the first shock. Forall shocks, AMSA increased from 4.83 to 5.60(p-value < 0.01). MS increased from 2.36 to2.44 (p-value = 0.01). CF increased from 7.05to 7.16 (p-value < 0.01). DFA did not showany change: 1.28 to 1.27. For only first shocks,similar results were observed. No correlationappeared between time to post QECG mea-surement and the change in QECG values.Conclusions: For all the QECG measures exceptfor DFA, a slight improvement in value wasobserved. While statistically significant, thesechanges may not be physiologically or clinicallymeaningful. Possible explanations include: (1)These may be a result of the CPR delivered inbetween the shock and the post-shock QECG;

(2) The pre-shock values started very low whichmay make decreases difficult to detect from afloor effect; and (3) Modern bi-phasic wave-forms may be less harmful than those previ-ously studied.

17. Epidemiology of Mortality in PatientsTransported by Emergency MedicalServices (EMS)

Ian Blanchard, Dan Lane, Tyler Williamson,Brent Hagel, Gerald Lazarenko, Ian Phelps,Darren Sandbeck, Damon Scales, EddyLang, Christopher Doig, Alberta HealthServices/University of Calgary Category ofSubmission: Professional

Background: Outside of key conditions suchas cardiac arrest and trauma, little is knownabout the epidemiology of mortality of alltransported EMS patients. The purpose of thisstudy was to describe characteristics of EMSpatients who after transport, die in a health carefacility. Methods: EMS transport events overone year (April, 2015–2016) from a BLS/ALSsystem serving an urban/rural population ofapproximately 2 million were linked with in-hospital datasets to determine overall, emer-gency department (ED), and in-patient mortal-ity. Medical Priority Dispatch System (MPDS)determinant, age in years (> = 18 years -adult, <=17 years - pediatric), gender, day ofweek, season, time (categorized in six hourperiods), and the highest mortality MPDScards, paramedic clinical impressions, and EDdiagnoses (International Classification of Dis-ease v.10 - Canadian) are presented. Anal-yses included two-sided t-test or chi-squarewith alpha < 0.05. Results: A total of 239,534EMS events resulted in 159,507 patient trans-ports; 141,114 were included for analysis afterduplicate removal (89.1% linkage). Of 141,114patients, 4,269 died (3.0%; 95%CI 2.9%, 3.1%).There were 724/4,269 deaths in the ED (17.0%)and 3,545/4,269 died as in-patients (83.0%). Theproportion of overall mortality by MPDS deter-minant was Echo (24.6%), Delta (3.9%), Charlie(3.4%), Bravo (1.1%), Alpha (2.1%), and Omega(1.1%). For adults the mean age of survivorswas less than non-survivors (59.2 vs. 75.8; p< 0.001), but pediatric survivors were olderthan non-survivors (8.8 vs. 2.8; p < 0.001).Males had increased mortality (3.3%) comparedto females (2.8%)(p < 0.001). Mortality didnot change by day of week (p = 0.573), butdid by season with increased ED mortalityin the winter (p = 0.004). The highest over-all mortality occurred with patients presentingbetween 0600–1200 hours (3.9%), and the lowestbetween 0000–0600 hours (2.3%)(p < 0.001). TheMPDS cards with the highest overall mortalitywere 9-cardiac/respiratory arrest (34.4%), 33-interfacility transfers (7.1%), 6-breathing prob-lems (5.8%), and 28-stroke/transient ischemicattack (4.3%). The highest overall mortalityfor paramedic clinical impressions were car-diac arrest (76.4%), respiratory arrest (18.0%),hypovolemia/shock (11.4%), and stroke/CVA(10.9%). The ED diagnoses with the highestoverall mortality were related to neoplasms(19.8%), circulatory system (12.4%), respira-tory system (7.4%), and infections (6.0%). Con-clusions: Significant in-hospital mortality dif-ferences were found between event, patient,and clinical characteristics. These data provideimportant foundational and hypothesis gener-ating knowledge regarding mortality in trans-ported EMS patients that can be used to guideresearch and training.

18. Epidemiology of Infections And Sepsisin a Large, Canadian Emergency MedicalServices (EMS) System

Daniel Lane, Ian Blanchard, GeraldLazarenko, Christopher Oleynick, Laurie

Morrison, Hannah Wunsch, Sheldon Cheskes,Steve Lin, Refik Saskin, Damon Scales, Insti-tute of Health Policy, Management and Evaluation,University of Toronto Category of Submission:Student, Resident, Fellow

Background: Sepsis is a life-threatening syn-drome caused by a dysregulated immuneresponse to infection. Early recognition andintervention are critical to improve patientoutcomes. In modern healthcare systemsparamedics often encounter patients with sep-sis before other clinicians, offering an importantopportunity for earlier sepsis care. The purposeof this study was to estimate the incidence andexamine characteristics of patients with infec-tions, and sepsis transported by paramedics.Methods: A one-year cohort of all adults(>= 18 years) transported by a BLS/ALS EMSsystem servicing a rural/urban populationof approximately 2 million was linked to in-hospital administrative databases(emergencydepartment[ED] and inpatient). Infection, andsepsis cases were classified based on ED infec-tious disease diagnosis code, and an existingsepsis algorithm based on ED diagnosis codesand EMS clinical information. Clinical charac-teristics including age (years), Glasgow ComaScore (GCS)<15, tachypnea (>22/minute), andfever (> = 37.8 Celsius), and operational factorssuch as prehospital time (minutes), transportdistance from municipality to hospital, andhigh-priority Medical Priority Dispatch Sys-tem (MPDS) determinant (Echo/Delta) wereevaluated in adults (> = 18 years) and com-pared to patients not meeting sepsis criteria.Two sided t-test or difference of proportionwere used with statistical significance <0.05.Results: 131,174 unique adult encounters weresuccessfully linked to in-hospital databases(89% linkage rate). The one-year incidenceof infections, and sepsis were 11% and 2.1%,respectively. A minority of all patients withinfections presented with fever (18%), abnor-mal GCS (22%) or tachypnea (32%). Comparedto other patients, adults with sepsis were morelikely to have an abnormal GCS (60% vs. 16%,p < 0.001), tachypnea (48% vs. 20%, p < 0.001),or fever (25%vs.4%, p < 0.001). They were gen-erally older(mean 75 vs. 60 years, p < 0.001),and more likely to have a high priority MPDSdeterminant (38% vs.31%, p < 0.001). Sepsispatients had longer prehospital intervals (mean44 vs.39 minutes, p < 0.001) despite shortertransport distances(15/9.3 vs.16/9.9 km/miles,p = 0.004). The in-hospital mortality ratefor patients with infection was 6.8% (95%CI,6.4–7.2), and 19% for sepsis (95%CI, 18–21).Conclusions: Infections and sepsis are commonamong paramedic-transported patients, andparamedics spend a considerable time withthese patients prior to arriving in the ED. Thesepatients frequently have altered vital signs,suggesting earlier recognition may be feasible.The in-hospital mortality of these patients issignificant, supporting the need for furtherresearch into opportunities for prehospitalidentification and intervention.

19. Combined PrehospitalHypoxia-Hypotension “Depth-DurationDose” and Mortality in Major TraumaticBrain Injury

Daniel Spaite, Chengcheng Hu, BentleyBobrow, Vatsal Chikani, Bruce Barnhart,Joshua Gaither, P. David Adelson, KurtDenninghoff, Amber Rice, Chad Viscusi,Duane Sherrill, Samuel Keim, University ofArizona Category of Submission: Trauma

Background: Our previous work has shownthat the depth-duration doses of prehospi-tal hypoxia (SpO2 < 90%) and hypotension[SBP < 90 mmHg], separately, are strongly asso-ciated with mortality in Traumatic Brain Injury



(TBI). However, hypoxia and hypotension areobviously not mutually exclusive. Hence, thenext logical step in evaluating the influenceof the “dose” of these physiological anoma-lies in TBI is to identify the combined risk.Methods: We evaluated major TBI cases(moderate/severe) enrolled in the EPIC Study(NIH-1R01NS071049) before TBI guidelineimplementation (N = 16,711; 1/07–9/14).Definitions: hypoxia dose-SpO2 depth <90% integrated over time (min); hypotensiondose-SBP < 90 mmHg integrated over time.Both dose variables were then transformedto achieve approximate normality. Logisticregression was used to determine the associa-tion between odds of death and nonparametricfunctions of the (transformed) hypoxia andhypotension doses. The combined fitted effectsof both hypoxia and hypotension then yieldedthe dose score. Results: After exclusions [age<10 (6.8%), transfers (28.4%), SBP >200 (2.1%),missing SBP/SpO2/time (12.8%), only onerecorded SBP or SpO2 (8.3%),] 6682 casesremained (median age = 40; male = 70%).Mortality increases consistently across thequartiles (Q) of unadjusted dose score (Nohypoxia or hypotension-5.6%; Q1–16.5%; Q2–20.8%; Q3–35.8%; Q4–43.2%). In the adjustedmodel, the mortality increase is remarkablymonotonic (indeed, nearly linear) with increas-ing dose score. Across the entire range ofdose, an increase of one standard deviation ofadjusted dose score is associated with a 63%increased odds of death (aOR = 1.63) amongpatients with either hypoxia or hypotension orboth. This result is strongly supported by thehighly significant, monotonically-increasingrelationship between the separate hypoxiaand hypotension doses and their adjusteddeath rates. Conclusions: Both hypoxia andhypotension depth/duration appear to have aprofound and additive influence on TBI mortal-ity. The influence of hypoxia and hypotensionon outcome (both separately and combined)appears to be far more complex than the currentliterature reflects (only being assessed dichoto-mously as present or not). Future TBI studiesshould account for both the depth and durationof prehospital hypoxia and hypotension.

20. Association Between Induction andSedation Agents and Post IntubationHypotension in Trauma Patients

Frederick Brown, Francis Guyette, ChristianMartin-Gill, Jonathan Elmer, Department ofEmergency Medicine, University of PittsburghSchool of Medicine Category of Submission:Trauma

Background: Medications used for rapidsequence intubation and post-induction seda-tion may cause hypotension, resulting insecondary injury and worse outcomes aftertrauma. We identified patient and treatmentcharacteristics associated with post-intubationhypotension. Methods: We retrospectivelyreviewed charts from consecutive patientsundergoing transport between January 2001and June 2016 by STAT MedEvac, a multistatecritical care transport service with >10,000missions per year. We identified adult traumapatients intubated by a flight crew member. Theprimary outcome was early post-intubationhypotension, defined as a systolic blood pres-sure (SBP) <90 mmHg within 15 minutesof intubation. We used logistic regression toidentify predictors of post-intubation hypoten-sion. Results: During the study period, 4701adult trauma patients were intubated andtransported. Mean age was 44 years, 26% werefemale, mean pre-induction heart rate was99 bpm (SD 25), SBP was 137 mmHG (SD 32),SpO2 was 96% (IQR 95, 100), and respiratoryrate was 18 (SD 7). A total of 14% of patients

experienced post-intubation hypotension.Patient factors independently associated withhypotension were age (adjusted odds ratio(aOR) 1.03, 95%CI 1.02–1.04), female gender(aOR 1.40, 95%CI 1.06–1.85), lower SBP (aOR0.96, 95%CI 0.96–0.97), higher heart rate (aOR1.01, 95%CI 1.00–1.01) and lower SpO2 (OR0.97, 95%CI 0.95–0.98). Paralysis with rocuro-nium (aOR 1.83, 95%CI 1.26–2.67) compared tosuccinylcholine was associated with increasedodds of post-intubation hypotension while pre-intubation treatment with lidocaine (aOR 0.70,95%CI 0.52–0.94), post-intubation treatmentwith fentanyl (aOR 0.23, 95%CI 0.18–0.30) andpost intubation administration of normal saline(aOR 0.30, 95%CI 0.10–0.88) were associatedwith a decreased occurrence of post-intubationhypotension. Conclusions: In trauma patientsundergoing critical care transport, multiplepatient factors and modifiable treatmentsincluding administration of rocuronium forintubation were independently associatedwith hypotension. Additional investigation isneeded to confirm this effect and identify otherpatient and treatment factors associated withpost-intubation hypotension. In the interim,current protocols and clinical practice shouldbe reviewed.

21. Prehospital Lactate: A SeverityIndicator in Early Sepsis Management

Kurt Isenberger, Aaron Burnett, JeffreyAnderson, Adam Mayer, Sandi Wewerka,Joseph Pasquarella, Ralph Frascone, RegionsHospital Category of Submission: Medical

Background: Serum lactate levels can rapidlydictate clinical awareness of shock and promptintervention in sepsis. Collecting serum lac-tate levels is restricted to the hospital setting.The objective of this study was to investi-gate whether a prehospital lactate value (PL)improves time to intervention upon ED arrival.We hypothesized that EMS communication ofa PL value to the ED physician would result inimproved time to antibiotics and an in-hospitallactate (IL) order. Methods: This prospective,observational study included patients with aprehospital impression of infection based onSIRS criteria. A POCT PL was collected byparamedics prior to ED arrival, and reportedduring presentation. ED metrics were collectedand compared to a previously collected con-trol group (CG) of patients presenting with-out a PL but who met early sepsis criteria. Coxregression models were used to estimate hazardratios (HRs) with 95% confidence intervals (CIs)for time to physician order for an antibiotic orin-hospital lactate test. Results: Study patients(age 60–96) included 170 with PL measures and269 controls. The PL group was older on aver-age (mean age, 69 vs. 54; p < 0.001), and wasmore likely to have expired in the hospital (10%vs 5%; p = 0.027). An antibiotic was orderedfor 104 prehospital lactate patients (61%) and216 controls (80%). In a Cox model adjustedfor age and gender, the CG had a nearly two-fold faster rate of time to antibiotic order (HR= 1.93; 95% CI: 1.50, 2.48), relative to prehos-pital lactate patients. Among the PL patients,time to antibiotic order did not differ by PL(>2 vs. �2 mmol/L, p = 0.545). Time to ILorder did not significantly differ between CGand patients with PL >2 mmol/L (p = 0.811),but time to IL order was twice as slow forpatients with PL �2 mmol/L (vs. CG, HR =0.47; 95% CI: 0.33, 0.68). Conclusions: In thissmall study, the availability of PL value didnot improve time to antibiotics. Patients witha normal PL had a prolonged time to antibi-otics compared to the CG, and a higher deathrate. A larger study is required to validate theseresults.

22. Prevalence of Mortality Due toRebound Toxicity after “Treat andRelease” Practices in Prehospital OpiateOverdose Care: A Systematic Review andMeta-Analysis

Jennifer Greene, Brent Deveau, Justine Dol,Micheal Butler, Dalhousie University Categoryof Submission: Medical

Background: Death from fentanyl overdosewas declared a public health crisis in Canadain 2015. Traditionally, patients who have over-dosed on opiates that are managed by emer-gency medical services (EMS) are treated withthe opiate antagonist naloxone, provided venti-latory support and subsequently transported tohospital. However, certain EMS agencies haveallowed paramedics who have reversed an opi-ate overdose to refuse transport, if the patienthas the capacity to do so. The safety of thispractice has not been examined by a systematicreview. Therefore, our intent is to examine theavailable literature to determine the prevalenceof mortality and serious adverse events within48 hours of EMS treat and release due to sus-pected rebound opiate toxicity after naloxoneadministration. Methods: A systematic searchwas preformed on May 11, 2017 in PubMed,Cochrane Central, Embase, and CIHAL usingsearch strategies developed with the aide of ahealth sciences librarian. No search limits wereapplied. Included studies were hand searched.Two authors conducted the screening, selec-tion and data extraction process. Discrepan-cies were resolved via discussion. A modifiedQUIPs tool was used to evaluate risk of bias.Analysis for prevalence of outcomes were pre-formed. Results: A total of 1,401 records werescreened after duplicate removal. Eighteen fulltext studies were reviewed with eight selectedfor inclusion. Included studies had a low riskof bias. The prevalence of mortality within48 hours was so infrequent that it could notbe quantitatively meta-analyzed. There were4/4912 (0.00081%) total reported deaths of sus-pected rebound etiology from included patientsacross all studies. Only one study reported onadverse events of patients released on scene.This study found no incidence of adverse eventsfrom their sample of 71 released patients. Con-clusions: Mortality or serious adverse events inthe included studies due to suspected reboundtoxicity in patients released on scene post EMStreatment with naloxone was rare. Despite lim-ited studies, the prevalence rate was so low thatwe concluded that this practice may be safe interms of mortality and may be considered analternative of traditional transport. Additionalprospective studies need to be preformed tostrengthen knowledge around adverse events.

23. Prehospital qSOFA Score as Predictorof Sepsis and Mortality Eileen Shu, CrystalIves Tallman, Megann Young, William Frye,Leyla Farshidpour, Danielle Campagne,UCSF-Fresno; Department of Emergency MedicineCategory of Submission: Medical

Background: The quick Sequential [Sepsis-related] Organ Failure Assessment (qSOFA)score was proposed in 2016 as a rapid wayto identify adult patients with suspected infec-tions who are likely to have poor outcomes. A2017 study showed that qSOFA was correlatedwith hospital admission, ICU admission, hospi-tal length of stay, and inpatient mortality. How-ever, to our knowledge, the ability of the qSOFAscore to predict patient outcomes has not beenevaluated in the prehospital setting. We hypoth-esize that prehospital qSOFA scores are corre-lated with up-triage (change to a higher acu-ity triage zone in the emergency department),presence of sepsis, ICU admission, and in-hospital mortality. Methods: We conducted a



retrospective observational study using prehos-pital ambulance vital signs to calculate qSOFAscores for all adult medical patients that pre-sented in September 2016 to a large academicemergency department in Fresno, CA. Informa-tion from the electronic health record (EHR)was used to determine up-triage, presence ofsepsis, hospital admission, ICU admission, andin-hospital mortality. Results: A total of 1,903adult medical patients were transported byambulance to the emergency department dur-ing the study period. Of these, 151 patients(7.93%) were prehospital qSOFA positive. Apositive prehospital qSOFA score was corre-lated with emergency department diagnosis ofinfection (29.1% vs. 15.2%; p < 0.001), hos-pital admission (55.0% vs. 33.4%; p < 0.001),ICU admission (9.93% vs. 2.22%; p < 0.001),admission diagnosis of sepsis (19.2% vs. 3.08%;p < 0.001), and in-hospital mortality (6.62%vs. 0.74%; p < 0.001). A positive prehospi-tal qSOFA score was not associated with up-triage (7.95% vs. 5.82%; p = 0.291); however, itwas correlated with final triage to a high acu-ity zone in the emergency department (35.8%vs. 8.96% p < 0.001). Conclusions: PrehospitalqSOFA is correlated with the diagnosis of infec-tion and sepsis. Furthermore, it is correlatedwith poorer patient outcomes including needfor hospital admission, ICU admission, and in-hospital mortality. However, a positive prehos-pital qSOFA score in isolation does not appearto be more useful than the current triage pro-cess in the emergency department to identifypatients who should be triaged to a high acuityzone in the absence of other patient factors.

24. Prehospital Provider Attitudes andBeliefs Regarding Pediatric SeizureManagement: A Multicenter, QualitativeStudy

John Carey, Jonathan Studnek, Lorin Browne,Malcolm Leirmoe, Daniel Ostermayer, TylerMiller, Diaa Alqusairi, Thomas Grawey,Stephanie Schroter, E. Brooke Lerner, ManishShah, Baylor College of Medicine, Pediatrics,Section of Emergency Medicine Category ofSubmission: Pediatric

Background: Seizures have the potential tocause significant morbidity and mortality,and are a common reason EMS are requestedfor a child. A pediatric prehospital seizureevidence-based guideline (EBG) was publishedand has been implemented as protocol inmultiple EMS systems. Knowledge trans-lation and protocol adherence in medicinecan be incomplete. In EMS, systems-basedfactors and providers’ attitudes and beliefsmay contribute to incomplete knowledgetranslation. The purpose of this study wasto identify EMS provider-reported attitudes,beliefs, barriers, and enablers to adhering toEBG-derived seizure protocols in multipleEMS systems. Methods: This was a qualitativestudy utilizing 30-minute semi-structuredinterviews of paramedics who recently trans-ported actively seizing 0–17 year olds in twodifferent urban EMS systems. Interviewersexplored the providers’ decision-making dur-ing their recent case and regarding seizuresin general. Two investigators used NVivosoftware, the grounded theory approach,and constant comparison to independentlyanalyze transcribed interview recordings untilthematic saturation was reached. Results:Several overarching themes emerged from the32 paramedics that were interviewed. Enablersincluded dosing/protocol references, training,provider knowledge about preferred routes,predefined provider roles, options to use dif-ferent routes, online medical control, multiplecrews on scene, and physical accessibility ofmedication on scene. Systems barriers included

equipment availability, controlled substancemanagement, infrequent training, few pediatriccalls, unclear definition of a treatable seizure,and incongruent protocol and reference tooldosing. Personal barriers included fear of res-piratory depression, confusion about dosing,and misconceptions about preferred routes,febrile seizure management, and accuratemethods of weight estimation. Paramedicsshared other opinions about management:preference for intranasal vs. intramuscularmedication, how transport distance affectsmanagement, use of online medical control,and the need to manage bystanders. Providerssuggested system improvements to addressequipment, medication, protocol, and traininglimitations. Conclusions: Paramedics identi-fied many standardized strategies EMS systemsused that enabled pediatric seizure protocoladherence, as well as numerous systems-basedand personal barriers to adherence. Providersidentified solutions to address the barriers.Conducting research on EMS protocol changes,policy modifications, and training that addressthe barriers identified in this study mayenhance understanding of how to optimizepediatric prehospital seizure outcomes.

25. Analysis of Dosing Errors Made byParamedics During Simulated PediatricPatient Scenarios after Implementation ofState-Wide Pediatric Drug DosingReference

John Hoyle, Glenn Ekblad, Tracy Hover, BillFales, Richard Lammers, Dena Smith, WesternMichigan University, Homer Stryker, MD School ofMedicine Category of Submission: Pediatric

Background: Medication errors occur at ahigh rate for prehospital pediatric patients.Epinephrine dose errors have been 60%. Toreduce errors, Michigan implemented a pedi-atric dosing reference (PDR), with doses listedin milliliters, the requirement that doses bedrawn into a smaller syringe from a pre-loadedsyringe using a stop cock and dilution of drugsto standard concentrations. The purpose ofthis study was to evaluate the prevalenceof medication errors by paramedics treatingpediatric patients after the implementation of astate-wide PDR. Methods: 8 EMS agencies com-pleted 2 validated, pediatric scenarios: infantseizing and infant cardiac arrest. Agencies wereprivate, public, not for profit, for profit, urban,rural, fire-based, and third service. Simulationstook place in a simulation center or mobilesimulation unit. EMS crews used their regularequipment with sham drugs and were requiredto carry out all the steps to administer a drugdose. Two evaluators scored crew performancevia direct observation and video review. A doseerror was defined as > = 20% difference com-pared to the weight-appropriate dose. Descrip-tive statistics were utilized. Results: 80 simula-tions have been completed and initial analysishas been conducted using descriptive statistics.The majority of crews were EMTP/EMTP. Incardiac arrest scenarios, 8/20 (40%; 95% CI18.5%, 61.5%) epinephrine doses were incor-rect. In 0/20 doses, there was no cross check ofthe drug volume prior to administration. Therewere 6, ten-fold overdoses and one, ten-foldunderdose. In seizure scenarios, 5/11(45%;95% CI 16%, 74.9%) benzodiazepine doseswere incorrect (2 underdoses, 3 overdoses); 2/9(22%; 95% CI 0%, 49.4%) drug dilutions wereincorrect resulting in large dosing errors. In1/10 cases (10%; 95% CI 0%, 28.6%) the crewwas unable to dilute D50 to D25. Unrecognizedair bubbles were frequently entrained in theadministration syringe resulting in underdoses.In 11/20 (55%) of cases there was an error usingthe length-based tape for weight determina-tion. Conclusions: Epinephrine dose errors

have decreased since implementation of PDR,but frequent ten-fold errors still occur. Crosschecks of drug doses do not occur. Errors occurwith dilution and length-based tape use. Errorreduction strategies are needed for pediatricprehospital drug administration.

26. Training in Prehospital DeathNotifications Linked to Improved ProviderComfort and Preparation

Abraham Campos, Rebecca Cash, RemleCrowe, Madison Rivard, Brian Clemency,Robert Swor, Ashish Panchal, Eric Ernest,Department of Emergency Medicine, Univer-sity of Nebraska Medical Center Category ofSubmission: Student, Resident, Fellow

Background: Death notifications in the prehos-pital setting are difficult situations that requiretraining. However, this training is not uni-formly included in initial EMS education, andthe proportion of providers prepared for thistask is unknown. Our objective was to describethe prevalence of death notification training byprovider level and its association with prepa-ration and comfort in performing this task.We hypothesized that fewer EMTs receivedtraining and that training was associated withgreater preparation and comfort. Methods: Anelectronic questionnaire was sent to a randomsample of 20,000 nationally-certified EMS pro-fessionals in April 2017. Participants reporteddeath notification training received during ini-tial or continuing education and adult deathnotifications performed in the past 12 months.Level of comfort and preparation in deliver-ing adult death notifications was rated usinga 4-point scale. Inclusion criteria were prac-ticing, non-military EMTs or higher. Certifica-tion level was grouped into advanced life sup-port (ALS:paramedic/intermediate/AEMT) orbasic life support (BLS:EMT). Odds ratios (OR,95%CI, p-value) were calculated to estimate theassociation between training and provider com-fort and preparation. Results: There were 2,333responses (12% response rate), and 1,514(65%)met inclusion criteria. Most respondents hadperformed at least one adult death notificationin the past year (ALS: 87%, BLS: 78%, p < 0.001).Equal proportions of ALS and BLS (51% ver-sus 52%, p = 0.58) respondents received deathnotification training during an initial course,however fewer BLS respondents received addi-tional training (BLS: 30% versus ALS: 44%, p <0.001). A larger proportion of BLS respondentsdid not receive any death notification training(BLS: 40%, ALS: 32%, p = 0.005). Over one-third(34%) of those without training had performedan adult death notification in the past year.After controlling for certification level, trainingwas associated with increased odds of reportinggreater comfort (2.20, 1.77–2.75, p < 0.001) andpreparation (6.05, 4.73–7.74, p < 0.001) in per-forming death notifications. Conclusions: Mostrespondents delivered a death notification inthe past year; however, one-third of these EMSproviders had not received training. Trainingwas associated with greater comfort and prepa-ration in delivering death notifications. Limita-tions include recall bias attributed to self-report.Future work should focus on barriers to receiv-ing death notification training.

27. Reduction in Cervical SpineImmobilization Is Not Associated withMissed Injuries

Jennifer Gibson Chambers, Michael O’Brien,Brian Clemency, University at Buffalo Categoryof Submission: Student, Resident, Fellow

Background: Previous studies have demon-strated EMS providers can correctly deter-mine which patients have a cervical spineinjuries and patients arriving at the emergency



department via EMS without a cervical collarrarely have serious cervical spine injuries. Ina recent study, we demonstrated that imple-mentation of a spinal motion restriction (SMR)protocol was associated with decreased cervi-cal collar use. We sought to determine if thisdecrease was associated with an increase inserious cervical injuries among patients trans-ported without cervical collars. Methods: Thiswas a secondary analysis of a retrospectivechart review of patients transported by a sin-gle large, commercial EMS agency with a dis-patch for motor vehicle collision to one of threehospitals. EMS and hospital data were reviewedfor all calls during a 6-month period before(January–June 2015) and a 6-month period after(January–June 2016) the protocol change. Fisherexact test was used for statistical comparisonsbetween time periods. Cervical spine injuriesidentified on CT were considered serious ifthe patient required operative intervention, dis-charge in an immobilization collar or cervicalspine injuries present in patients who died asa result of traumatic injuries. Results: Therewere 1,614 patient records identified, 819 underthe immobilization protocol and 796 under theSMR protocol. Cervical collar use decreasedfrom 66.8% to 59.3% (p = 0.002). There wasno significant difference between time periodsin proportion of male patients, average age,or subtype of motor vehicle accident. No sig-nificant change was observed in the rate ofCT cervical spine imaging (51.0% before and52.5% after, p = 0.55). Serious cervical spineinjuries were identified in 2.2% before and 2.4%of imaged patients after SMR (p = 0.99). Allpatients with serious cervical spine injury wereplaced in cervical collars by EMS providers, asensitivity of 100%. The specificity was 14.0%before and 18.7% after SMR (p = 0.10). Conclu-sions: Despite decreased use of cervical collarsunder the SMR protocol, there were no motorvehicle accident patients with serious cervicalfractures transported without a cervical collarin either period. These findings may not gener-alize to other mechanisms of injury.

28. Psychometric Properties of a Survey onPatient Safety Culture (SOPS)-Based Toolfor EMS

Remle Crowe, Rebecca Cash, MadisonRivard, William Gilmore, Alex Christgen,Tina Hilmas, Lee Varner, Amy Vogelsmeier,Ashish Panchal, The National Registry ofEmergency Medical Technicians Category ofSubmission: Operations, Quality, Safety,Systems, Disaster

Background: Measuring and improving orga-nizational safety culture has been linked topositive safety outcomes in EMS, yet few eval-uation tools exist for this unique setting. TheAgency for Healthcare Research and Quality’s(AHRQ) Surveys on Patient Safety Culture(SOPS) are widely used to assess safety culturein various healthcare settings and results areincluded in a national comparative databaseto allow for benchmarking; however, thereis no SOPS instrument specific for EMS. Ourobjective was to evaluate the psychometricproperties of an EMS-adapted tool based onexisting SOPS domains. We hypothesized thatthe reliability and validity of the EMS toolwould be similar to existing SOPS instruments.Methods: We developed and cognitively testeda 37-item instrument adapting 10 domains fromthe SOPS instruments and one new domaincapturing the unique EMS aspect of communi-cation while enroute to a call. We administeredan electronic survey to all 332,584 nationally-certified EMS professionals. Analysis inclusioncriteria consisted of EMTs or higher practicingin non-military settings. We evaluated domainstructure using confirmatory factor analysis

(CFA) using a polychoric correlation matrix forordinal data. We used prior SOPS thresholdsto assess fit (0.90), factor-loadings (0.4), andfactor variances (0.5). We assessed domainreliability and validity using Cronbach’s alpha(cutoff:0.6) and Pearson’s correlation coeffi-cients (r; cutoff:0.3). Results: We randomly split23,765 responses into equally-sized calibrationand validation datasets. The CFA supportedthe 11-domain model with a comparative fitindex = 0.94, exceeding the 0.90 threshold.Item factor-loadings all exceeded 0.4 (range:0.51–0.98). Three domains exhibited factorvariances below the 0.5 threshold: staffing,communication about incidents, and handoffs.Cronbach’s alpha was above 0.6 for all domains(range:0.65–0.88). Predictive validity was sup-ported as all domain composite scores werecorrelated with the outcome variables of overallsafety rating (r = 0.44–0.72) and frequency ofevent reporting (r = 0.31–0.48). Results fromthe validation dataset confirmed the presentedcalibration results. Conclusions: Overall, theEMS-adapted tool demonstrated adequate psy-chometric properties, and the reliability andvalidity of the tool were consistent with exist-ing SOPS instruments. Important limitationsinclude potential response bias and the inabil-ity to aggregate data at the agency level. Futurework should focus on agency-level data testing.

29. Explaining Disparities in Field Triageof Older Adults: Factors that InfluenceEMS Destination Decisions AND Reasonsfor Over- and Under-Triage

Courtney Jones, Jeremy Cushman, JuliusCheng, Martina Anto-Ocrah, Nancy Wood,Heather Lenhardt, Molly McCann, SuzanneGillespie, Ann Dozier, Jeffrey Bazar-ian, Manish Shah, University of Rochester,School of Medicine and Dentistry Category ofSubmission: Trauma

Background: The Field Triage Decision Scheme(FTDS) is designed to identify severely injuredpatients and guide EMS providers’ selection ofa destination hospital, but a minimal amountis known regarding the real-world applicationof these criteria. We aimed to identify the fac-tors that influence EMS destination decisions,the extent to which EMS decisions align withthe FTDS, and explore EMS provider-identifiedreasons for over- and under-triage of olderadults. Methods: We conducted a prospectivemulti-center study, encompassing all four hos-pitals within a county, one of which was a ver-ified Level I trauma center which serves a ninecounty region of over one million people. Weenrolled all older adults aged 55 or older whosustained an injury of any severity and weretransported by EMS. Research staff adminis-tered a standardized interview-based surveywith the patient’s EMS provider. FTDS criteriawas used as the gold standard to assess pat-terns of destination decisions and adherence toprotocol. We used descriptive statistics to char-acterize the study sample and used chi-squaretests to assess factors that influenced destina-tion decisions and agreement between EMSdecisions and the FTDS. Proportions were usedto quantify reasons for under- and over-triage.Results: Data from 4,295 patients were ana-lyzed. The median age was 75 years and 59%were female. Using the FDTS as a gold standardfor destination decisions, 1,584 patients (43.8%)were over-triaged and 285 (42.0%) were under-triaged. There were only 2 patients (2%) whomet the mechanism of injury criteria who wereunder-triaged, compared to 154 (41.1%) and 141(47.2%) who met the physiologic and anatomiccriteria who were under-triaged, respectively.Of those who were over-triaged to the traumacenter, the most frequently cited reason byEMS was patient request (60.5%). Of those who

were under-triaged to the non-trauma centersthe most frequently cited reasons by EMS waspatient request (61.4%) and proximity of thenon-trauma center (13.7%). Conclusions: EMSprovider destination decisions are influencedby mechanism of injury, but a substantial pro-portion of patients who meet the physiologicand anatomic criteria of FTDS were under-triaged. Both under- and over-triage appear tobe heavily influenced by patient preference.

30. Does Mechanism of Injury PredictTrauma Center Need for Children? E.Brooke Lerner, Mohamed Badawy, JeremyCushman, Amy Drendel, Courtney Jones,Manish Shah, David Gourlay, Medical Collegeof Wisconsin Category of Submission:Pediatric

Background: To determine if the Mechanismof Injury step of the Field Triage DecisionScheme (FTDS) is accurate for identifying chil-dren who need the resources of a trauma cen-ter (TC). Methods: EMS providers transportingany injured child �15 years, regardless of sever-ity, to a pediatric TC in three midsized commu-nities over 3 years were interviewed. Collecteddata included EMS observed physiologic con-dition, suspected anatomic injuries, and mech-anism. Patients were considered to need aTC if they met a consensus-based definition.Data were analyzed with descriptive statisticsincluding positive likelihood ratios (+LR) and95% confidence intervals (95%CI). Results: Atotal of 9,484 provider interviews were con-ducted and linked to hospital data to obtainpatient outcome. Of those, 215 (2.3%) neededa TC. A total of 1,485 enrolled patients wereexcluded from further analysis because theymet the physiologic or anatomic steps of theFTDS. Of the remaining 7,999 cases, 61 neededa TC. The mechanisms sustained by the remain-ing cases were 35.5% fall (15 needed TC), 28.5%motor vehicle crash (MVC) (26 needed TC),7.1% struck by a vehicle (8 needed TC), 0.2%motorcycle crash (MCC) (none needed TC),and 28.8% had a mechanism not on the FTDS(12 needed TC). Among those who fell greaterthan 10 feet, 2 needed a TC (+LR 2.67; 95%CI:0.73–9.79). Among those in a MVC, 42 werereported to have been ejected and none neededa TC. While 63 had reported intrusion >12inches and 1 needed a TC (+LR 1.40; 95%CI:0.20–9.69). There were 34 reported as havinga death in the same vehicle, and 2 neededa TC (+LR 5.41; 95%CI: 1.37–21.00). Conclu-sions: Over a quarter of the children who needthe resources of a TC are not identified in thephysiologic or anatomic steps of FTDS. Themechanism of injury step of the FTDS doesnot include over a quarter of the mechanismsexperienced by children transported by EMSfor injury. Use of the mechanism step does notappear to greatly enhance identification of chil-dren who need a TC. More work is needed toimprove the identification of children who aneed the resources of a TC.

Poster PresentationAbstracts (31–206)31. Appropriate Needle Length forEmergent Pediatric Needle ThoracostomyUtilizing Computed Tomography

Maria Mandt, Kathleen Adelgais, Kari Hayes,Fred Severyn, Children’s Hospital ColoradoCategory of Submission: Pediatric

Background: Needle thoracostomy is a life-saving procedure. Advanced Trauma LifeSupport guidelines recommend insertion ofa 5 cm, 14-gauge needle for pneumothorax



decompression. High-risk complications canarise if utilizing an inappropriate needle size.No study exists evaluating appropriate needlelength in pediatric patients. Utilizing com-puted tomography (CT), we determined theneedle length required to access the pleuralcavity in children matched to BroselowTM

Pediatric Emergency Tape color. Methods:Three investigators reviewed chest CTs ofchildren < 13 years of age obtained between2010–2015. Patient exclusions included thosewith a chest wall mass, muscle disease, pectusdeformity, anasarca, prior open thoracotomy,inadequate imaging, or missing height docu-mentation. We established four groups basedupon BroselowTM color as determined byrecorded height. Investigators, trained by apediatric board-certified radiologist, obtainedstandardized CT measurements of chest wallthickness at four points: right/left secondintercostal space at the midclavicular line(ICS-MCL) and right/left fourth intercostalspace in the anterior axillary line (ICS-AAL).Our outcome was the median chest wall thick-ness and interquartile ranges (IQR) for eachBroselow grouping and anatomic site. Results:To date, 225 chest CTs have been reviewed.Median patient age was 5 years and 52.4% weremale. Children measuring Broselow Gray/Pink(<68 cm), had a median chest wall thicknessat the right ICS-MCL of 1.5 cm (IQR 1.3 cm,1.9 cm), left ICS-MCL 1.6 cm (IQR 1.5 cm, 2 cm),right ICS-AAL 1.7 cm (IQR 1.5 cm, 1.9 cm), leftICS-AAL 1.6 cm (IQR 1.4 cm, 2.2 cm). Childrenmeasuring Broselow Red/Purple (68.1–90 cm):right ICS-MCL 1.8 cm (IQR 1.5 cm, 1.9 cm), leftICS-MCL 2 cm (IQR 1.7 cm, 2.1 cm), right ICS-AAL 1.8 cm (IQR 1.6 cm, 2.2 cm), left ICS-AAL1.6 cm (IQR 1.3 cm, 2 cm). Children measuringBroselow Yellow/White (90.1–115 cm): rightICS-MCL 2.1 cm (IQR 1.5 cm, 2.3 cm), leftICS-MCL 1.9 cm (IQR 1.6 cm, 2.3 cm), right ICS-AAL 1.8 cm (IQR 1.7 cm, 2.1 cm), left ICS-AAL1.7 cm (IQR 1.5 cm, 2.1 cm). Children measur-ing Broselow Blue/Orange/Green (>115.1 cm):right ICS-MCL 2.4 cm (IQR 2.1 cm, 2.9 cm),left ICS-MCL 2.4 cm (IQR 2.1 cm, 2.9 cm),right ICS-AAL 2.1 cm (IQR 1.7 cm, 2.9 cm),left ICS-AAL 2.1 cm (IQR 1.6 cm, 2.9 cm).Conclusions: Median chest wall thicknessvaries little by height or location in children <13 years of age. The standard 5-cm needle istwice the chest wall thickness of most children.

32. Descriptive Analysis of DefibrillationVector Change for PrehospitalRefractory Ventricular Fibrillation

Matthew Davis, Andrew Schappert, JayLoosley, Kristine VanAarsen, Shelley McLeod,Sheldon Cheskes, Department of Medicine, Divi-sion of Emergency Medicine, Western UniversityCategory of Submission: Cardiac

Background: Patients in ventricular fibrilla-tion (VF) who do not respond to standardAdvanced Cardiac Life Support treatments aredeemed to be in refractory VF (rVF). The idealprehospital treatment for patients with rVFremains unknown. Double sequential externaldefibrillation (DSED) has been proposed as aviable option for patients in rVF. Although themechanism by which DSED terminates rVFremains unknown, one theory is that the changein defibrillation vector that occurs may con-tribute. Our objective was to describe clinicaloutcomes for patients presenting in rVF dur-ing out-of-hospital cardiac arrest (OOHCA) forthose who underwent vector change defibril-lation, compared to those who received stan-dard treatment. Methods: This was a retro-spective chart review of adult (�18 years)patients presenting in rVF during OOHCA over15 months beginning in March 2016. Patientswho underwent vector change defibrillation

had a change in pad position (anterior-anteriorto anterior-posterior) after 3 or more consecu-tive shocks. Termination of rVF was defined asthe absence of VF after a vector change or stan-dard defibrillation during the next rhythm anal-ysis. Results: There were 372 OOHCA, with 25(6.7%) patients meeting our definition of rVF. Ofthese, 16 (64.0%) patients (median age 62 years,81.3% male) had vector change after a median(IQR) of 3 (3.0–4.0) paramedic defibrillationattempts. Median (IQR) time to vector changedefibrillation was 8.8 (7.1–11.1) minutes. Eight(50%) patients had termination of rVF after thefirst vector change shock, 6 (37.5%) had prehos-pital return of spontaneous circulation (ROSC)and 5 (31.3%) patients survived to hospital dis-charge. Of the 9 rVF patients who did not havevector change, median age was 63 years and88.9% were male. The median (IQR) numberof defibrillations within this group was 5 (4.5–7.0). All patients remained in VF after the fourthdefibrillation. Prehospital ROSC was achievedin 3 (33.3%) patients. Three patients (33.3%)survived to hospital discharge. Conclusions:This is preliminary evidence that vector changedefibrillation in patients with rVF may resultin VF termination. A randomized controlledtrial is warranted to test whether or not vectorchange has a role in the termination of rVF.

33. Benchmarking EMS Compass StrokePerformance Measures Using a LargeNational Dataset

Jeffrey Jarvis, Dustin Barton, Lauren Sager,Nick Nudell, Williamson County EMSCategory of Submission: Operations,Quality, Safety, Systems, Disaster

Background: Prehospital stroke alerts havebeen promoted as a means of facilitating rapidED treatment of acute strokes. These alerts aredependent upon the performance of validatedstroke screening tools and assessment of bloodglucose to eliminate a common stroke mimic.EMS Compass has identified several perfor-mance measures on this topic. No work hasbeen done to calculate a national performancebenchmark for these measures. These bench-marks would be useful in system improvementefforts. We sought to describe national perfor-mance on these measures for the first time.Methods: Using anonymous data from 9-4-1consenting agencies in a large commercial EMSelectronic health record (ESO Solutions), weidentified records of patients felt to have acutestrokes who were transported from the sceneof a 9-1-1 call. From these records, we calcu-lated the proportion of all patients who had astroke screen and blood glucose documented.For each of these measures, we also calculatedthe 95% confidence interval. Results: Over a 61/2-year periods, we identified 168,854 patientswith 9-1-1 calls who had an impression of acutestroke. Of these, 88,751 patients or 52.6% (52.3–52.8%) had a stroke scale documented. Addi-tionally, 140,294 patients, or 83.1% (82.9–83.3%)had a blood glucose documented. Conclusions:In this study, we calculate the first nationalbenchmarks of two important clinical perfor-mance measures on stroke care described byEMS Compass. Importantly, there was poor per-formance of stroke screens with only 52.6%of all 9-1-1 calls for stroke having them doc-umented. At 83.1%, agencies performed bet-ter with blood glucose documentation. Theseresults provide initial benchmarks and providea starting point for improvement of both themeasures, documentation systems, and clinicalperformance.

34. Effect of Instructor’s Real-TimeFeedback During LaypersonCardiopulmonary Resuscitation Trainingon Quality of Cpr Performances: AProspective Cluster Randomized Trial

So Yeon Kong, Sang Do Shin, Kyoung JunSong, Tae Han Kim, Gwan Jin Park, Depart-ment of Emergency Medicine, Seoul NationalUniversity Hospital Category of Submission:Cardiac

Background: It was reported most bystanderCPR does not meet high quality CPR criteria,strongly implying an urgent need for newstrategies to assist in the delivery of qualitybystander CPR. The aim of this randomizedtrial was to assess the effectiveness of instruc-tor’s real-time, objective feedback during CPRtraining compared to a conventional feedbackin terms of trainee’s CPR quality. Methods:We performed a cluster randomized trial ofcommunity CPR training at Nowon DistrictHealth Center in Seoul. CPR training classeswere randomized into either intervention(instructor’s objective real-time feedbackbased on Laerdal QCPR Classroom) or control(conventional feedback) group. Laerdal QCPRClassroom software is a real-time feedbackdevice, which monitors quality of real-timeCPR performances of multiple trainees simulta-neously. During each training session, traineesperformed a total of five CPR. The primaryoutcome was the total score, which is anoverall measure of chest compression quality.Generalized linear mixed models were usedto analyze the outcome data from baseline tofifth CPR session, accounting for both cluster-and individual-level covariates. Results: Atotal of 77 training sessions (1,894 trainees)were randomized into 37 intervention (996trainees) and 40 control (898 trainees) groups.At baseline, both groups had equal overall CPRquality scores (78 in both groups). During thecourse of the training, QCPR feedback signif-icantly increased trainees’ overall quality ofCPR performance compared with conventionalfeedback (p < 0.01). In terms of changes frombaseline to last session, trainees in the interven-tion group demonstrated significant improve-ments on overall quality of CPR compared withthose in the control group (QCPR feedback� = 11.64 (95% CI 9.75–13.53) ; Conventionalfeedback � = 6.96 (5.16–8.76); p < 0.001). A sta-tistically significant difference between the twogroups was observed for change in compres-sion depth from baseline to fifth CPR sessionwith a mean change of 4.51 mm in the interven-tion group and 2.72 mm in the control group (p< 0.001). Conclusions: Considering the rate ofchest compression, we did not observe a statis-tically significant different between two groups(p = 0.06). In this prospective randomizetrial, instructor’s objective real-time feedbackresulted in improved overall CPR quality.

35. Confirming the Safety and Feasibilityof a Bundled Resuscitation TechniqueInvolving a Head-Up/Torso-UpMechanical Chest Compression Techniquefor Cardiopulmonary Resuscitation

Paul Pepe, Kenneth Scheppke, Peter Antevy,Daniel Millstone, Charles Coyle, Craig Pru-sansky, Sebastian Garay, Johanna Moore, Uni-versity of Texas Southwestern Medical CenterCategory of Submission: Operations, Qual-ity, Safety, Systems, Disaster

Background: Strategies to lower intracranialpressure (ICP) and improve cerebral/systemicperfusion during CPR have become a recentfocus for resuscitation researchers. One experi-mental method to lower ICP has been to elevatethe head/chest during CPR combined with theuse of devices to enhance venous return to thethorax. The purpose of this study was to evalu-ate both the safety and clinical feasibility of sucha bundled technique that includes mechani-cal CPR devices used at an angle. Methods:The EMS system catchment (pop. 1.4 million)is geographically expansive with broad ethnic



diversity, extremes of age and socioeconomicsand low frequency of bystander CPR. Throughan established Utstein-style registry, all out-of-hospital cardiac arrest (OOHCA) cases (allrhythms) were followed over 3.5 years (Jan-uary 1, 2014 through June 30, 2017; n =2,285).EMS crews were previously using the Lucas©device and impedance threshold device (ITD),but, after April 1, 2015, they also: (1) applied O2while deferring positive-pressure ventilationand ITD application several minutes; (2) raisedthe backboard ∼20° (head/torso-up) followingITD application; and (3) solidified a pit-crewapproach for device application. With neuro-intact discharge data not available until 2015,“short-term” survival (sustained resuscitationby EMS to hospital admission) was used forconsistent comparisons. Quarterly reports wererun to identify any periodic variations or incre-mental effects during protocol transition (inQuarter 2, 2015). Results: There were no com-plications/difficulties in using the head/torso-up position (n = 1,319). Of 806 consecutiveOOHCA cases attended between January 1,2014 and March 31, 2015, quarterly (all-rhythm)survival rates remained constant (mean 17.87%,range 15–20%) but rose steadily during thetransition period with an ensuing sustaineddoubling (35.18%; range 30–40%) for the nexttwo years (July 1, 2015 through June 30, 2017).Outcomes improved across subgroups whileresponse intervals, indications for initiatingCPR and bystander CPR rates were unchanged.EMS resuscitation rates in 2016 and 2017 werefound to be proportional to neuro-intact dis-charge. Conclusions: The head-up/torso-upCPR bundle was not only feasible, but was asso-ciated with an immediate, steady rise in EMSresuscitation rates during the transition phasewith a subsequent sustained doubling of sur-vival chances, making a compelling case thatthis bundled technique may improve OOHCAoutcomes in future clinical trials.

36. Intraosseous Access Use in Chemical,Biological, Radiation, and NuclearPersonal Protective Equipment TimCollins, Clinical & Medical Affairs, TeleflexMedical Category of Submission:Operations, Quality, Safety, Systems,Disaster

Background: To determine comparisons ofsuccess rates and ease-of-use ratings in achiev-ing intraosseous access in both wearing andnon-wearing of Chemical, Biological, Radia-tion and Nuclear (CBRN) personal protectiveequipment (PPE) in a cadaver model. Methods:Using a cross over design, eight experiencedparamedics inserted an intraosseous (IO)device (Arrow EZ-IO©) into a cadaver spec-imen wearing their standard prehospitalclothing. The sample then crossed over andapplied CBRN PPE and repeated IO insertions.IO insertion times were recorded and assessedfor clinical accuracy both before and after crossover with wearing CBRN PPE. Data collectioninvolved the sample completing a confidentialquestionnaire assessing self-perceived ease-of-use scores for IO access measured in Likertscales (0–10). Qualitative data was capturedfollowing structured focus group interviews.Results: The results found no statistical differ-ence between ease-of-use scores for IO accessbetween wearing or non-wearing CBRN PPE.No difference in determining land marking forIO insertion (M 9 vs. 8.75 p = 0.726), humeralsite insertion (M 9.13 vs. 8.75 p = 0.593),administration of IO saline flush (M 9.25 vs.8.75 p = 0.405), holding and manipulatingdriver (9.13 vs. 8.75 p = 0.593), and trocarremoval (9.25 vs. 8.75 p = 0.405). The meanease-of-use scores were found to be lower inCBRN group but not significant, focus groupdiscussions stated that PPE had some restric-

tions but effective EZ-IO insertion could still beachieved. Insertion times (25 secs SD 3.46 vs.34.38 secs SD 4.17 p = 0.0002) were statisticallylonger with wearing CBRN PPE. However,focus group discussion stated that it wouldtake significantly longer to achieve intravenous(IV) access and that IO was an effective andfaster option compared to IV during a CBRNincident. Conclusions: Intraosseous access canbe effectively and promptly achieved whilewearing CBRN PPE. IO access took an addi-tional 9.4 seconds while wearing CBRN PPEwhich can provide fast and efficient vascularaccess during a CBRN incident.

37. Reprioritization of 9-1-1 EmergencyMedical Calls Using Historical ClinicalData

Veer Vithalani, Sabrina Vlk, Steven Davis,Neal Richmond, Office of the Medical Director;MedStar Mobile Healthcare Category of Submis-sion: Operations, Quality, Safety, Systems,Disaster

Background: Emergency Medical Services(EMS) systems often utilize a structuredapproach to 9-1-1 call-taking and emergencymedical dispatch (EMD). One such system,Medical Priority Dispatch System (MPDS),categorizes 9-1-1 calls into EMD codes based onproblem and severity, with response prioritiesand resources determined at the local levelthrough a predetermined response matrix.In this study, we propose a methodology forutilizing historical clinical data to increase theaccuracy of 9-1-1 call prioritization of patientswith time-sensitive critical illness. The primaryobjective is to increase the number of patientswith time-sensitive critical illness who receivethe highest-priority response (“Priority 1”).The secondary objective is to decrease thenumber of Priority 1 responses to patientswho do not have time-sensitive critical illness.Methods: All 9-1-1 calls-for-service in a largeEMS system, between December 1, 2015 andNovember 30, 2016, were included. Electronicpatient care reports (ePCRs) were analyzedfor time-sensitive critical illness, including anypatients in cardiac or respiratory arrest or whorequired airway management or electrical ther-apy (pacing, cardioversion, or defibrillation).The percentage of calls with time-sensitivecritical illness was calculated for each of the382 EMD codes in the MPDS. In our proposedresponse matrix, any codes which had at least1% of patients with time-sensitive critical illnesswere assigned a theoretical Priority 1 response.Results: Out of a total of 119,287 actual calls-for-service, 30,123 (25.2%) were assigned a Priority1 response through the current response matrix;1,205 (4%) of these patients had time-sensitivecritical illness. Utilizing our proposed method-ology, these same calls-for-service would haveresulted in 25,441 (21.3%) Priority 1 responses,including 1,333 (5.2%) patients with time-sensitive critical illness. The net result wouldhave been an overall 15.5% decrease in Priority1 responses, and a 10.6% increase in Priority 1responses to patients with time-sensitive criti-cal illness. Conclusions: Historical clinical datamay be used to increase the accuracy of callprioritization of patients with time-sensitivecritical illness, while simultaneously increas-ing operational efficiency and 9-1-1 resourceutilization.

38. Kink in the Stroke Chain of Survival: IsEMS Appropriately Prenotifying the ED ofSuspected Strokes?

Jeffrey Nusbaum, Nachiketa Gupta, AlecGlucksman, Michael Redlener, Kevin Munjal,Mount Sinai Hospital Category of Submission:Student, Resident, Fellow

Background: The purpose of this study wasto determine rates of prenotification in a largeurban setting among patients suspected by EMSof having had an acute stroke and to deter-mine factors associated with appropriate preno-tification. Methods: This was a retrospectivecohort study of all patients with a dischargediagnosis of CVA, TIA or intracranial hemor-rhage who arrived by EMS between January 1and December 31, 2015 at three urban hospitals.Patients transferred from another acute carefacility were excluded. “Get with the Guide-lines” data was matched to data from the pre-hospital care reports. Appropriate prehospitalnotification was defined by any reference in theEMS narrative or hospital record to advancednotification of the patient’s arrival. Logisticregressions were used to determine factors thatmay have been important for EMS prenotifica-tion and whether prenotification was associatedwith higher rates of tPA administration. Analy-sis was done using the R-statistical computingsoftware. Results: During the study period, 379patients presented via EMS; 126 arrived within3.5 hours of their last known normal (LKN).EMS suspected a CVA in 107 (85%). Prenotifi-cation was given in 52 of 107 instances (49%).Shorter EMS LKN times were associated withincreased rates of prenotification (p < 0.01).Prenotification was more likely in patients withhigher NIHSS (p = 0.01). For the elements ofthe Cincinnati Prehospital Stroke Scale (CPSS),prenotification was 24% higher in patients withslurred speech (p = 0.01), 24% higher with armdrift (p = 0.01), and 20% higher with facialdroop (p = 0.04). In a multivariate logisticregression including the three components ofthe CPSS, slurred speech was the most influ-ential factor for prenotification (p = 0.09), fol-lowed by arm drift (p = 0.14), and facial droop(p = 0.56). With appropriate prenotification,there was a 17% increase in likelihood of receiv-ing tPA (p = 0.06). Conclusions: Prehospitalproviders are not consistently providing preno-tification. In our cohort, EMS prenotified the EDin patients with more severe and recent onsetsymptoms. Similar to other studies showingimproved time interval metrics with prenotifi-cation, our study suggests that prenotificationwas associated with higher rates of tPA admin-istration. There may be a benefit to dedicatingresources toward EMS education on the role ofprenotification in the stroke chain of survival.

39. Interaction Effects of Communitiesand Advanced Airway Management onSurvival after Out-of-Hospital CardiacArrest; Multi-Level Analysis

Dongsun Choi, So Yeon Kong, Tae HanKim, Jeong Ho Park, Kyoung Jun Song,Young Sun Ro, Ki Ok Ahn, Sang Do Shin,Seoul National University Hospital, Department ofEmergency Medicine Category of Submission:Cardiac

Background: Chest compression and adequateventilation are essential for oxygen delivery inout-of-hospital cardiac arrest (OHCA) patients.The association between prehospital advancedairway management (AAM) and survival out-comes was inconsistent. We hypothesized thatdifferences in the application of prehospitalAAM between regions due to medical resourcewould have an effect on the effectiveness ofthe AAM. The aim of this study was to inves-tigate whether the effect of prehospital AAMon outcomes between regional EMS systems offour Asian cities. Methods: We used a PAROS(Pan-asia resuscitation outcome study) registry.We identified patients with OHCA of pre-sumed cardiac etiology who were resuscitatedby emergency medical services in four Asiancities between 2012 and 2014. OHCA patientswere witnessed by EMS personnel and age



under 18 years were excluded. The main expo-sure variables were AAM. The primary end-point was survival discharge and neurologicrecovery. We compared outcomes between theAAM and non-AAM groups using multivari-able logistic regression with an interaction termbetween AAM and four Asian cities (Osaka,Seoul, Singapore, and Taipei), after adjusting forpotential confounders. Results: Among 27,375patients, 16,510 patients were included in thefinal analyses. Survival discharge and neuro-logic recovery was better in the non-AAMgroup (8.7% vs. 5.1%, 4.9% vs. 2.0%) than inthe AAM group (adjusted odds ratio [aOR] 0.58[95% confidence interval (CI)0.59–0.68]). In theinteraction model for the survival discharge, theaORs for AAM of Osaka and Singapore was 0.43(95% CI 0.35–0.52) and 0.31 (0.17–0.58), respec-tively. In the interaction model of Seoul andTaipei, the association between AAM and sur-vival to discharge were statistically insignifi-cant (aOR 0.99 ; 95% CI 0.75–1.30) and aOR1.04; 95% CI 0.69,1.55, respectively) . The sim-ilar results showed for neurologic recovery.Conclusions: Regional EMS system modifiedthe effect of AAM on outcomes for patients withOHCA.

40. Rapid Cycle Deliberate Practice andCoaching of Specific InterventionsImproves Cardiopulmonary ResuscitationQuality Measures in Teams of EMSProviders

Christopher Berry, Pamela Humphrey,Anthony Halupa, Stephen Taylor, JarrettShugars, Douglas Kupas, Geisinger HealthSystem Category of Submission: Cardiac

Background: High-quality cardiopulmonaryresuscitation (CPR) skills are paramount forgood survival from cardiac arrest, but previousstudies have suggested that CPR quality is oftenpoor. The purpose of this study was to eval-uate changes in the quality of EMS providerCPR skills using rapid cycle deliberate prac-tice (RCDP) of specific teaching interventionswith real-time feedback. Methods: A record-ing mannequin, with feedback blinded fromparticipants, was used to evaluate CPR qual-ity metrics on 3- or 4-person teams of EMSproviders. All participants were certified EMSproviders who also had previous CPR educa-tion and certification. CPR quality metrics wereassessed, before and after educational inter-ventions, using a 5-minute resuscitation casesimulating adult cardiac arrest with ventricu-lar fibrillation. The intervention included usingcoaching and RCDP to teach techniques of palmlift, two-person two-thumbs-up bag-mask ven-tilation, upstroke ventilation during continuouscompressions, and chest compressions duringdefibrillator charging. CPR metrics included:compression fraction, compression rate, per-centage of compressions with full depth andfull recoil, percentage of compressions between100–120 per minute, ventilation rate, percentageof ventilation of adequate volume, and lengthof longest pause. Outcomes were comparedthrough Paired Samples t-tests using bias-corrected bootstrapping, resampling 1000 timeswith 95% confidence intervals. Results: Thesample consisted of 67 providers divided into18 teams. There were significant improvementsfor the following metrics of CPR quality whencomparing the pre- and post- interventionmeasures: percentage of compressions between100–120 per minute (39.5% vs. 78.5%; p = .001,ȵ2 = 0.60), compression fraction (78.8 vs. 92.3;p = .006, ȵ2 = 0.52), percentage with full recoil(52.7% vs. 85.6%; p = .001, ȵ2 = 0.60), percent-age with adequate ventilation volume (38.5%vs. 57.4%; p = .002, ȵ2 = 0.55), and longest pausein compressions (16.6 sec vs. 6.2 sec; p = .004,ȵ2 = 0.48). Conclusions: The use of RCDP and

coaching with real-time mannequin feedbackled to significant improvements in measures ofCPR quality in teams of EMS providers.

41. Influence of Patient Race onAdministration of Analgesia by StudentParamedics

Bill Lord, Sahaj Khalsa, University of theSunshine Coast Category of Submission:Professional

Background: Disparities in healthcare are asso-ciated with factors that include social status,age and race or ethnicity, with evidence show-ing African American individuals receive fewerprocedures and poorer-quality medical carethan white individuals. Disparities in the man-agement of pain have been shown to be associ-ated with race. However, there is limited dataregarding the influence of race on analgesiaprovided by paramedics. As such, this studyaims to investigate associations between patientrace and student paramedic management ofpain, using a null hypothesis of no difference.Methods: This retrospective cohort study useda contiguous dataset of all student paramedicrecords entered in the FISDAP Skill Trackerdatabase between January 1, 2014 to Decem-ber 31, 2015. Cases were extracted if aged 16to 100 years, the patient was alert and theprimary or secondary impression was trauma(abdominal, chest, extremity, neck-back, multi)or burns. Head injury was excluded as thisis a contraindication to analgesia in some set-tings. The primary outcome of interest wasthe interaction between patient race and stu-dent paramedic administration of any analge-sia for cases meeting inclusion criteria. Sec-ondary outcomes of interest were associationsbetween age and gender and analgesia admin-istration. The adjusted logged odds of patientsreceiving any analgesic was tested with bino-mial logistic regression using a stepped mod-elling approach. Results: A total of 59,962 caseswere available for analysis; median age was50 years (IQR 39 years), 50.2% were female(n = 30,077). The most common cause of traumawas fall, representing 50% (n = 26,053) of cases.14.1% of patients received any analgesia (n= 8,425). Caucasian patients have significantlyhigher logged odds of receiving analgesia thannon-Caucasian patients (p < 0.001). When anal-gesic administration is adjusted for age cate-gory and gender, African Americans have thelowest logged odds of receiving any analge-sia when compared to Caucasian patients (OR0.65, p < 0.001). Conclusions: The results indi-cate inequality in the provision of analgesia bystudent paramedics based on patient race. Thissuggests a need for education that addressescognitive and affective biases that can affectclinical judgements, and EMS audit of cases toidentify disparities in care based on race.

42. Paramedics Providing Palliative Careat Home: Patient and Family Satisfaction

Alix Carter, Judah Goldstein, Marianne Arab,Michelle Harrison, Barbara Stewart, MireilleLecours, Carolyn Villard, James Sullivan,Dalhousie University Category of Submis-sion: Operations, Quality, Safety, Systems,Disaster

Background: Paramedic crisis and symptommanagement for patients receiving palliativecare with the goal to treat in place repre-sents a novel approach to care. A new clini-cal practice guideline, additional medications,and a training program Learning EssentialsApproach to Palliative Care (LEAP) Mini forParamedics were implemented in two provin-cial EMS systems. Our objective was to deter-mine the impact of this new model of careon patient/family satisfaction and to describe

their experience with paramedic palliative sup-port at home. Methods: The perspective offamilies/patients registered in a provincial pal-liative care registry in Nova Scotia and PEIwas gathered in a mixed methods approachfrom June 1, 2016 to August 31, 2016. Uponenrollment in the program, a survey wasmailed. Six months after an episode of care(allowing for grieving time) a semi-structuredtelephone interview using a validated guidewas conducted. Overall satisfaction was mea-sured using a 5-point Likert scale. Respon-dent characteristics are reported descriptively.Open ended-questions were analyzed by the-matic content analysis. Results: A total of225 registration surveys were distributed, 67(30%) were returned. Of those, 49 (73%) werecompleted by the family. For the interviews,families were contact sequentially (8 declined,22 disconnected telephones, 32 unansweredcalls with two call attempts). Eighteen fam-ilies completed the interview. Three themesemerged from the pre-encounter survey: regis-tering meant fulfilling loved one’s care wishes,providing peace of mind, and feeling preparedfor emergencies. Post-encounter, 14/18 fam-ilies rated the care received as “excellent,”and all indicated that symptoms were helped.Seven families indicated that without the pro-gram, they would have had to be in hos-pital. Five themes emerged: 24/7 availability,professionalism of paramedics, compassion ofparamedics, relief of symptoms, and a plea forprogram continuation. Thematic saturation wasreached with minimal divergence of comments.Conclusions: The model of paramedics provid-ing palliative support in the home resulted inhigh patient/family satisfaction; registering inthe program, prior to any call for assistance,provides peace of mind and a feeling of beingprepared. Families particularly note the valueof 24/7 availability, success in relief of symp-toms, and the degree of compassion and profes-sionalism of paramedics.

43. Amplitude Spectrum Area Changesduring Cardiopulmonary Resuscitationafter Different Durations of UntreatedCardiac Arrest in a Porcine Model ofVentricular Fibrillation with aConcurrent Acute Myocardial Infarction

Giuseppe Ristagno, Francesca Fumagalli,Weilun Quan, Giovanni Babini, RobertoLatini, Yongqin Li, IRCCS–Istituto di RicercheFarmacologiche Mario Negri, Milan, ItalyCategory of Submission: Cardiac

Background: Amplitude spectrum area(AMSA) is a predictor of successful defib-rillation (DF). In this study, we investigatedthe effect of high quality cardiopulmonaryresuscitation (CPR) on AMSA in relationshipwith the duration of untreated ventricularfibrillation (VF) in a preclinical porcine modelwith a concurrent acute myocardial infarction.Methods: An established model of myocardialinfarction followed by VF and CPR was used.Forty-four pigs were subjected to different VFdurations: 8–10 minutes (short), n = 14; 12 min-utes (intermediate), n = 21; and 13–15 minutes(long), n = 9. Continuous mechanical CPR(Lucas, PhysioControl) with ventilation withoxygen and epinephrine administration (1 mgat 2 minutes of CPR) was performed for 5 min-utes prior to a 150 J DF attempt. AMSA andchanges in AMSA during CPR (dAMSA), inrelationship with the duration of untreatedVF, coronary perfusion pressure (CPP), andepinephrine administration were evaluated.Results: Overall AMSA decreased from 13.7± 0.8 mVHz to 6.5 ± 1.7 mVHz during the15 minutes VF (dAMSA −7.2 ± 2.5 mVHz, p <0.01), while it increased to 17 ± 1.2 mVHz after5 minutes of CPR (dAMSA 10.5 ± 3.5 mVHz,



p < 0.01) independently of the duration ofuntreated VF (dAMSA: 10.7 ± 1.7, 11.3 ± 1.5,and 14.7 ± 1.5 mVHz, for short, intermediate,and long VF duration, respectively, p = NS).AMSA during CPR was significantly correlatedwith CPP (r = 0.46, p < 0.01). AMSA increasedsignificantly during the first 2 minutes of CPR,as compared to the subsequent 2 minutesafter epinephrine (dAMSA 6.2 ± 0.8 vs. 3.6 ±0.6 mVHz, p < 0.01). Interestingly, dAMSAimproved similarly in the 3 VF duration groupsduring the first 2 minutes of CPR, but afterepinephrine the magnitude of dAMSA con-tinued to increase only when the durationof untreated VF was � 13 minutes, while itdecreased for shorter durations. Conclusions:High quality CPR allowed for AMSA increasesindependently of the duration of untreated VF.However, epinephrine administration furtherimproved dAMSA only in the instance oflonger durations of VF, while it seemed to havea detrimental effect for a shorter duration.

44. Prehospital Evidence-Based GudielineImplementation Methodology: ASystematic Literature Review

Jennifer Fishe, Remle Crowe, RebeccaCash, Nikiah Nudell, Christian Martin-Gill,Christopher Richards, University of FloridaCOM Jacksonville, Department of EmergencyMedicine Category of Submission: Opera-tions, Quality, Safety, Systems, Disaster

Background: As prehospital research advances,evidence-based guidelines (EBGs) are increas-ingly implemented into EMS practice. How-ever, incomplete EBG implementation mayhinder improvement in prehospital patient out-comes. To inform future EBG efforts, this studyreviews and summarizes existing evidencepertaining to prehospital EBG implementationmethodologies. Methods: This study is a sys-tematic literature review followed by the Grad-ing of Recommendations, Assessment, Devel-opment, and Evaluation (GRADE) method-ology. PubMed®, EMBASE®, Scopus®, andGoogle Advanced SearchTM were searchedwithout language or publication date filtersfor articles addressing prehospital EBG imple-mentation. Conference proceedings, textbooks,non-English articles, and articles that did notaddress prehospital EBG implementation wereexcluded. GRADE was applied to remainingarticles independently by three of five mem-bers of the Prehospital Guidelines ConsortiumResearch Committee. Variations in ratings wereresolved by consensus. Study characteristicsand salient findings are reported. Results: Thesystematic literature review produced 1,375articles, with 41 meeting inclusion criteria. Mostarticles described EBG implementation (N =24, 59%), or implementation barriers (N = 13,32%). Common study designs were statementdocuments (N = 12, 29%), retrospective cohortstudies (N = 12, 29%), and cross-sectionalstudies (N = 9, 22%). Using GRADE, evidencequality was rated low (N = 18, 44%), or verylow (N = 23, 56%). Salient findings included: (1)EBG adherence and patient outcomes dependupon successful implementation, (2) publishedstudies generally lack detailed implementationmethods, (3) implementation takes longer thanplanned (mostly for EMS education), (4) EMSsystems’ heterogeneity affects implementation,and (5) multiple barriers limit successful EBGimplementation (e.g., financial constraints,equipment purchasing, coordination withhospitals and regulatory agencies). The studyfound no direct evidence for best prehospitalEBG implementation practices, includingcomparisons of implementation methods, orof methods in different contexts (e.g., urbanversus rural, ALS versus BLS). Conclusions:While numerous implementation barriers are

well described, there is a paucity of evidencefor optimal prehospital EBG implementationmethods. For scientific advances to reachprehospital patients, future prospective studiesshould compare implementation methodolo-gies in different prehospital contexts. EBGprojects should publish reproducible imple-mentation methods, with “lessons learned”compiled in an easily accessible repository.Funding priorities should include implemen-tation research to ensure the efforts of EBGdevelopment translate into practice.

45. Are There Disparities in Dispatch CPRInstruction Receipt and CPRPerformance?

Amanda Amen, Patrick Karabon, BrianMcNally, Cherie Bartram, Kevin Irwin, Kim-berly Vellano, Robert Swor, Oakland UniversityWilliam Beaumont School of Medicine Categoryof Submission: Student, Resident, Fellow

Background: Dispatch-assisted cardiopul-monary resuscitation (DA-CPR) has beenshown to improve rates of bystander CPR(BCPR), which enhances survival in Out ofHospital Cardiac Arrest (OHCA). Our objec-tives are to evaluate whether there are racialand socioeconomic disparities in the receiptof DA-CPR instructions and subsequent CPRperformance. Methods: We performed a retro-spective review of the Cardiac Arrest Registryto Enhance Survival (CARES) dispatch registryfrom January 2014 to December 2016. Datawas collected from a convenience sample ofdispatch agency supervisor audits of 9-1-1OHCA audio recordings in one state. Elementsrelated to dispatcher CPR instruction, andbarriers to bystander CPR performance wererecorded. Demographics including patient race(white, black or other) and Utstein data werecaptured from the parent CARES database.These data were merged with census tractdata regarding socioeconomic status (SES)of each incident location. The effects of raceand SES were analyzed to determine theirassociation with two outcome variables: callerreceipt of DA-CPR instructions and subsequentperformance of CPR. Multivariate logisticregression analysis was performed. Results:We identified 1,872 cases from 23 dispatchagencies that had dispatch, Utstein, and censustract data. The population was predominantlywhite (70.0%), male (66.0%), with an averageage of 63.5 +/−18.7. DA-CPR instructions weremore commonly associated with an incidentthat occurred in a private residence (ORadj3.8, 95% CI (2.5–5.8)) or in highest incomequartile census tracts [ORadj: 1.65; 95% CI(1.01–2.72)]. Older patient age [ORadj: 0.99;95% CI (0.98–0.99)] and black race [ORadj:0.61; 95% CI (0.39–0.98)] were negatively asso-ciated with receipt of DA-CPR instructions.Subsequent performance of CPR after DA-CPRinstruction was more common in witnessedarrests [OR 2.0, (95% CI 1.3–3.0)] and negativelyassociated with black race [ORadj: 0.31; 95% CI(0.16–0.58)] but not significantly different bysocioeconomic or demographic characteristics.Conclusions: Although this preliminary studyis limited by incomplete demographic anddispatch data, we identified racial disparitiesin provision of DA-CPR instructions andsubsequent CPR performance. These findingsvaried minimally by SES or other demographiccharacteristics.

46. Utilization of Emergency MedicalResources at Mass Gathering Events at anUrban University with a Collegiate-BasedEmergency Medical Services Agency

Emma Ordway, Neil Sarna, Lindsey DeGe-orge, Jose Nable, Georgetown University

Category of Submission: Student, Resident,Fellow

Background: Mass gathering events (MGEs)challenge medical directors and emergencymedical services (EMS) agencies with pro-viding appropriate and sufficient medicalresources. This study aimed to examine EMSresource utilization during MGEs at a medium-sized urban university with a collegiate-basedbasic life support (BLS) agency, and howsuch utilization may be associated with spe-cific attributes of these events. Methods: Allemergency medical dispatches for the studiedon-campus EMS agency during MGEs wereincluded for analysis in this retrospective study,covering MGEs from January 1, 2012 throughSeptember 1, 2016. This collegiate-based agencyis the sole provider of medical standby detailsat its university. Environmental factors such astemperature, location (indoor vs outdoor), esti-mated event size, and event type were analyzedfor each MGE based on data from standby dutylogs and the National Weather Service. Linearregression, logistic regression and bivariate cor-relations were used to determine correlationalrelationships between environmental factorsand patients-per-event presentation rates (PPR)to EMS during these events. Results: No callsfor service occurred for any events with lessthan 500 attendees, while at least 1 call for ser-vice occurred at 6.1% of events with 500–1000attendees and at 24.5% of events with over1000 in attendance. Neither heat nor humiditywas found to be significant predictors of PPR,with p-values of 0.72 and 0.65, respectively.However, in the subset of events that attractedmore than 1,000 people and were outdoor non-sporting events, the linear regression of PPRand temperature had a Pearson’s CorrelationCoefficient of 0.983 and a p-value of 0.017.Outdoor non-sporting events, as compared toindoor non-sporting events, had an increasedlikelihood of calls for service (OR 4.4, p =0.18). Outdoor sporting events, as compared toindoor sporting events, were also more likelyto have requests for EMS (OR 6.1, p = 0.005).Conclusions: This study highlights thatenvironmental features such as estimatedcrowd size, location, event type, and outdoortemperature can possibly be used to predictEMS resource utilization at MGEs. Universityadministrators, event organizers, and EMSagencies can potentially prepare medical plansfor such mass gatherings by pre-assessing theseevent attributes.

47. Simple Feedback Form Improves Qualityof Out-of-Hospital CPR

Ben Weston, Jamie Jasti, Melissa Mena,Jackson Unteriner, Kelly Tilotson, Ziyan Yin,Mario Colella, Tom Aufderheide, MedicalCollege of Wisconsin Category of Submis-sion: Operations, Quality, Safety, Systems,Disaster

Background: Despite medical advances andhealth awareness campaigns, the incidence ofprehospital cardiac arrest remains high whilesurvival rates remain low. Excellent prehospitalcare is tantamount to survival and high qualityCPR is a vital contributor to positive outcomes.A quality improvement program was recentlyimplemented to provide simple, goal basedfeedback to prehospital providers after eachcardiac arrest resuscitation. Expanding uponan earlier preliminary study, we aim to assesswhether the provision to prehospital providersof a simple CPR feedback form led to improvedquality metrics in out of hospital cardiac arrestresuscitations. Methods: This before and afterretrospective review evaluated data from aquality improvement program in a midsizedurban community with BLS and ALS providers.Two 9-month periods, one before and one after



the implementation of the form were evaluated.Metrics measured included the means and ratesof goal achievement for compression depth,rate, and fraction as well as preshock pausetime. Results: A total of 439 before encountersand 621 after encounters were evaluated includ-ing those of BLS and ALS providers. Overall,significant differences were found in the meancompression depth (5.0 cm vs. 5.5 cm; p <0.001), compression fraction (79.2% vs. 86.4%;p < 0.001), compression rate (109.6/min vs.114.8/min; p < 0.001) and preshock pause time(18.8 sec vs. 11.8 sec; p < 0.001). Additionally,improvements were noted in goal achievementfor compression depth (48.5% vs. 66.6%; p <0.001), compression fraction (68.1% vs. 91.0%;p < 0.001), and preshock pause time (24.1% vs.59.5%; p < 0.001). No significant difference wasfound in goal achievement of compression rate.Conclusions: We found that the introductionof a simple CPR feedback form to prehospitalproviders was associated with improvement inprehospital CPR quality.

48. Cumulative Success of PrehospitalAdvanced Airway Management in aNational Cohort

Jeffrey Jarvis, Dustin Barton, Henry Wang,Williamson County EMS Category ofSubmission: Medical

Background: Repeated attempts at AdvancedAirway Management (AAM) are associatedwith increased risk of adverse events. Thereare few current descriptions of the number ofattempts needed for success. We sought to char-acterize cumulative AAM success rates in anational cohort of Emergency Medical Services(EMS) agencies. Methods: We used 9 years ofdata from ESO Solutions, a national EMS elec-tronic health record system. We included allencounters with attempted AAM. We examinedthe following subsets: (1) cardiac arrest intu-bation (CA-ETI), (2) medical non-arrest intu-bation (NA-ETI), (3) rapid-sequence intuba-tion (RSI), (4) sedation-assisted ETI (SAI), and(5) some type of supraglottic airway (SGA).Using binomial proportions with exact confi-dence intervals, we determined the cumulativesuccess rates for each attempt. We also identi-fied rates of first-pass success (FPS) and over-all success (OS), and the number of attemptsneeded to reach OS. Results: A total of 61,793patients from 552 EMS agencies underwentAAM efforts, including 38,063 CA-ETI, 19,138NA-ETI, 7,229 RSI, 3,095 SAI, and 9,993 SGA.The number of AAM attempts per patient var-ied (median 1, range 1–10). CA-ETI perfor-mance was: FPS 71.4% (95% CI: 70.8–71.9%),4 attempts to reach the OS threshold of 91.7%(91.4–92.1). NA-ETI performance was: FPS66.3% (95% CI: 65.4–67.2%), 3 attempts to reachthe OS threshold of 80.4% (79.6–81.1%). RSIperformance was: FPS 75.9% (95% CI: 74.9–76.9%), 5 attempts to reach the OS thresh-old of 96.3% (95.8–96.7). SAI performance was:FPS 66.9% (95% CI: 65.2–68.6%), 4 attemptsto reach OS threshold of 86.9% (85.6–88.1%).SGA performance was: FPS 88.8% (95% CI:88.1–89.4%), 5 attempts to reach OS thresh-old of 93.2% (92.6–93.6%). Conclusions: In thisnational series, first pass prehospital AAM suc-cess rates have improved from prior studiesbut are still low. Multiple attempts are com-mon and often unsuccessful. These results mayguide protocols limiting AAM attempts.

49. Benchmarking EMS CompassPerformance Measures Using a LargeNational Dataset: Pediatric Care

Jeffrey Jarvis, Dustin Barton, Lauren Sager,NIck Nudell, Williamson County EMS Cate-gory of Submission: Operations, Quality,Safety, Systems, Disaster

Background: Children make up ∼10% of allEMS transports, often require weight-baseddosing, and are commonly affected by res-piratory issues. A subset of the EMS Com-pass performance measures addresses pediatriccare, including documentation of weights, vitalsigns, and treatments for dyspnea. No bench-marks of these measures have been done on anational scale. We aim to describe these mea-sures using a large national cohort. Methods:Using a 6 ½ year sample of 9-4-1 EMS agenciesusing the ESO electronic health record (EHR),we calculated compliance rates among trans-ported 9-1-1 patients under 15 for the followingmeasures: (1) documented weight, (2) at leastone SpO2 and RR documented for those withany respiratory illness, (3) at least one dose of abeta-agonist given to those with asthma, and (4)at least one dose of beta-agonist given to thosewith asthma and an SpO2 <90%. For measuresrequiring administration of a medication, onlyALS providers were included. For each mea-sure, a rate and 95% Confidence Interval werecalculated. Results: There were 524,856 patientsanalyzed. Of these, 287,719 [54.8% (54.7–55.0%)]had a documented weight. There were 43,067children with a respiratory impression, 37,689of these [87.5%, (87.2–87.8%)] had at least oneSpO2 and Respiratory Rate documented. 6,202children had an impression of asthma and 4,336of these [69.9% (68.8–71.1%)] received a beta-agonist. Of those children with an impressionof asthma, 755 were hypoxic and 635 [84.1%(81.5–86.7%)] of them received a beta-agonist.Conclusions: These are the first benchmarkdata drawn from a large, national datasetagainst the EMS Compass measures. Theseresults provide a starting point for qualityimprovement efforts and suggest areas forimprovement in pediatric care. Only 55% ofchildren had documented weights which areneeded for correct medication dosing andonly 83% of hypoxic asthmatics received abeta-agonists. This highlights opportunities forimprovement.

50. AEDS on Wheels: A Pilot Programme toEquip Taxis with AEDS

Alexander White, Desmond Mao, VernonKang, Marcus Ong, Singapore General HospitalCategory of Submission: Cardiac

Background: We aimed to determine the fea-sibility of improving AED utilization rates andtime-to-first-shock times by equipping taxiswith AEDs. Methods: This is a prospectiveobservational feasibility study conducted inSingapore, a densely populated SoutheastAsian nation with 5.54 million people on a landarea of 719 square kilometres. There are 3,300licensed SMRT taxis in Singapore. 155 drivers ofthe SMRT taxi company were recruited, trainedand certified in CPR+AED skills. They werethen assigned to 100 taxis equipped with AEDsand displaying AED decals on taxis’ windowsand interior. A phone app alerted drivers tocardiac arrests within 1.5 km. Drivers receivingthe alert would choose to accept or declineto respond. Upon arrival, the drivers eitherprovided AED to lay bystanders on scene orapplied it themselves. If paramedics arrived atthe scene first, taxis would be notified to standdown. Post-incident, drivers were requiredto document incident and submit AED forcheck-up and maintenance. Results: FromNovember 2015 to July 2017, more than 2,400activations were sent to a total of 71 drivers.A total of 24 taxi drivers accepted 192 alerts tomobilize. Of these mobilizations, 22 taxi driversarrived at scene of 105 potential out-of-hospitalcardiac arrest cases prior to ambulance arrival.The mean time of activation-to-acceptance ofthe case was 1.17 minutes (95%CI 0.90–1.43).The average time of activation-to-arrival at

scene was 6.01 minutes (95%CI 5.24–6.78).Where the taxi arrived at scene, the averagedistance from case location to the initial loca-tion of the taxi was 763 meters (95%CI 654–871). Taxis that were closer to the incidenthad a higher likelihood of arriving before theambulance (763 meters vs. 955 meters, P-value= 0.041). A total of 10 drivers were “SuperResponders” as they had arrived at the scenethree times or more. Conclusions: A voluntary“AEDs on Wheels” program can be an excitingfeature of a public AED program (PAD). Thismode of mobilizing AEDs has a high likelihoodof utilization, increases the reach of AEDs, andimproves time-to-first shock, all of which areimportant components of successful PAD.

51. Medical Command Training forEmergency Medicine Residents: AnOverview of Medical Command Education,Oversight, and Evaluation

Abagayle Renko, Nicholas Julius, ChaddNesbit, Penn State Milton S. Hershey Medical Cen-ter Category of Submission: Student, Resi-dent, Fellow

Background: Training Emergency Medicine(EM) residents provide medical oversight asa requirement for EM residency accreditationthrough the ACGME; yet, no standard curricu-lum from which to train residents to developthis essential skill exists and literature describ-ing the current state of resident medical com-mand training is limited. We sought to assessthe state of medical command training in EMresidency programs. Methods: A thirty ques-tion survey was created and distributed elec-tronically through email via the Research Elec-tronic Data Capture (REDCap) program. Thesurvey contained questions regarding demo-graphics, general facility and program descrip-tors, medical command training procedures,personnel providing command, resident over-sight, and feedback. Descriptive statistics werecollected and analyzed using chi-squared testsfor categorical variables. Results: A total of109 surveys were completed (54.5% responserate), and 96 of those programs (88.1%) reportedthat their residents do receive formal medicalcommand training. A majority of those pro-grams begin medical command training dur-ing their residents’ first (42 programs, 43.8%)or second (40 programs, 41.7%) year of resi-dency. Most programs do not have required for-mal classroom-based (56 programs, 57.7%) oronline-based (75 programs, 77.3%) training. EMphysicians are the primary individuals provid-ing training (91 programs, 93.8%). Most pro-grams allow their residents to begin givingmedical command in their second year of res-idency (52 programs, 54.7%). A majority of pro-grams do not have a system in place to trackhow many medical command calls their resi-dents take (63 programs, 66.3%), nor do theyassign dedicated medical command shifts totheir residents (85 programs, 89.5%). Most pro-grams allow their residents to issue medicalcommand orders without the presence of anattending physician (62 programs, 65.3%). Amajority of programs indicated that their res-idents are provided feedback on their perfor-mance for their command call management(83 programs, 85.6%) and most programs indi-cated that medical command calls by residentsare not routinely audited (51 programs, 53.4%).Conclusions: Most EM residencies train theirresidents in providing medical command, yetthere is wide variation in how this is accom-plished. Further research and analysis arerequired to make recommendations for a moreuniform system of resident command training.

52. Near Misses in a Two-Tiered SuburbanEMS System: A Descriptive Study onDown-Triaged Patients Who Are Taken



Emergently to the Operating Room orAdmitted to Intensive Care Units

Joslyn Joseph, Joshua Bucher, David Feldman,Albert Ritter, Frederick Fiesseler, MorristownMedical Center Category of Submission: Oper-ations, Quality, Safety, Systems, Disaster

Background: A two-tiered EMS system has theadvantage of incorporating volunteer, public,and private BLS ambulances into the system todecrease response times and spread resourcesfurther. An ALS unit who responds to a scenemay down-triage or “release” to BLS if noALS interventions are warranted outside ofBLS scope of practice to allow their unit tostay in service. To date, no studies have eval-uated the characteristics of high-risk patients“released” to BLS and then taken to the Oper-ating Room (OR) or admitted to the IntensiveCare Unit (ICU). In order to make safer triagedecisions, we sought to describe this “near-miss” mistriaged population of patients whowere ultimately deemed to be critically ill byEmergency Departments and had the poten-tial to decompensate quickly. Methods: Set-ting: A suburban two-tiered EMS system inwhich ALS units evaluate approximately 14,000patients per year. Patients: All patients from2007–2015 "released" to BLS, transported toan Emergency Department, and subsequentlyadmitted to an ICU, Cardiac CatheterizationLab, or OR. Protocol: Demographics, historyof present illness, vital signs, GCS, disposition,final diagnosis, and interventions done priorto EMS arrival and by EMS personnel wereextracted via chart review and 95% ConfidenceIntervals (CIs) calculated when appropriate.Results: Out of 17,639 patients from 2007–2015who were evaluated by ALS and triaged to BLS,372 patients (2%) were mistriaged to BLS. Theaverage age of patients was 66.4 years CI (61.0–71.7) and 52% were female. The most com-mon mistriaged final diagnosis category wasNeurological, 24% CI (23.3–24.7), followed byGI/Abdominal Emergencies 15%(14.3–15.7%).Sepsis was mistriaged 10.2% CI (9.5–10.9) ofthe time, and 9 patients, 2.4% CI (2.3–2.6) weretaken emergently to the Cardiac Catheteriza-tion Lab. Conclusions: This is the first step toinvestigate this phenomenon unique to two-tiered EMS systems. From our study, we canconclude that more education is needed to rec-ognize prehospital Neurological and Abdomi-nal/GI Emergencies to avoid near misses in thefuture. More research is also needed to deter-mine which patients, if any, had poor outcomesas a result of being mistriaged to make triageprotocols safer for our patients.

53. Astounding Rates of Suicidality in EMSProviders: A Hidden Epidemic

Al Lulla, Jyotirmoy Das, Ghady Rahhal,Rebecca Dougherty, Bridgette Svancarek,Washington University in St. Louis Category ofSubmission: Student, Resident, Fellow

Background: EMS providers experience severeworkplace stress, which increases their risk ofsuicidality. Past suicidal thoughts and attemptshave been established as placing individualsat high risk for future suicidal behavior. Wesought to assess the severity of the problemof suicidality in EMS providers and to iden-tify potential factors that place individuals athigher risk. Methods: We administered a 19item online survey to a convenience sampleof 16 EMS agencies and 1,688 EMS providers.In order to assess for suicidality, the SuicideBehaviors Questionnaire Revised (SBQ-R) wasutilized. SBQ-R assesses 4 dimensions of suici-dality using a Likert scale. These 4 dimensionsare (1) lifetime suicidal ideation and/or suicideattempt, (2) frequency of suicidal ideation overpast 12 months, (3) threat of suicide attempt,

and (4) likelihood of future suicidal behavior.In prior studies, a SBQ-R score of 7 or greaterhas been validated as an effective predictor ofsuicidal behavior. The SBQ-R score has pre-viously demonstrated ability to identify indi-viduals at risk for suicide with 93% specificityand 95% sensitivity. We used Pearson’s chi-square to determine the relationship betweensuicidality and gender, age, shift-length, hoursworked per week, years in EMS, race, practicesetting, service type, family history of suicide,and knowing an EMS provider who commit-ted suicide. Results: We received 289 completedsurveys analyzed less than 2 weeks after surveydistribution. 30.8% (89) [95% confidence inter-val (CI): 25.5–36.1%] of individuals had SBQ-R scores greater than or equal to 7, reflectingsuicidality. The strongest predictors of suicidal-ity were family history of depression or suicide[OR = 3.0 (1.8–5.1)], and working in a hospital-based service [OR = 2.0 (1.0–4.0)]. Gender,age, race, practice setting, shift length, hoursworked per week, years in EMS, and know-ing an EMS provider who committed suicidewere not found to be statistically significant pre-dictors of suicidality. Conclusions: High ratesof suicidality exist within the EMS community;however, further research on risk factors andpotential solutions needs to be conducted.

54. Identification of Sepsis in thePrehospital Setting: An ObservationalStudy of Paramedic Sepsis ScreeningStrategies

Daniel Lane, Ian Blanchard, GeraldLazarenko, Laurie Morrison, Steve Lin,Hannah Wunsch, Sheldon Cheskes, RefikSaskin, Damon Scales, Institute of Health Policy,Management and Evaluation, University of TorontoCategory of Submission: Student, Resident,Fellow

Background: Sepsis is a life-threatening syn-drome where earlier recognition and promptintervention is critical to improving patientoutcomes. In modern healthcare systems,paramedics encounter many sepsis patientsfirst, offering an opportunity for earlier detec-tion. The purpose of this study was to providethe incidence of paramedic reported suspicionof infection, and to compare the accuracy ofpublished paramedic screening strategies forsepsis within a cohort of Emergency MedicalServices (EMS) patients. Methods: A previouslypublished systematic review that identifiedstrategies for paramedic identification of sep-sis was updated and used as the source forparamedic screening strategies. A one-yearcohort of EMS data linked to in-hospitaladministrative databases (n = 131,745;89%linkage rate) was used for the cohort of EMSpatients. Sepsis was identified by EmergencyDepartment (ED) International Classification ofDiseases v.10 Canadian (ICD-10CA) diagnosiscodes, and EMS clinical information. Theincidence of paramedic documented suspicionof infection in patients diagnosed with sep-sis in the ED, and the sensitivity, specificity,positive, and negative likelihood ratios (LR)for each of the screening strategies, using therecommended score threshold as originallypublished where applicable are reported.Results: Paramedics documented suspicionof infection in 350 of 2,713 [13%(95% Confi-dence Interval) 2–14%] sepsis patients. Twelveparamedic sepsis screening strategies wereidentified in the literature. The PRESS, HEWS(score of > = 2), and Robson scores had thehighest sensitivities [0.98(0.98–0.99), 0.87(0.86–0.88), 0.74(0.72–0.76) respectively], and lowestnegative LR [0.08(0.04–0.08), 0.27(0.24–0.30),and 0.39(0.37–0.42), respectively] for ruling outsepsis. The PSP score (high risk) and SepsisAlert strategies had high specificity [0.98(0.98–

0.98] and 0.99(0.99–1.0)], and positive LR[19(17–22) and 13.6(11.6–16.0)] for ruling insepsis, but lower sensitivity [0.34(0.33–0.36)and 0.07(0.06–0.08)]. Comparing the qSOFAscore recommended in the Sepsis-3 definitionto the previously recommended SIRS score,qSOFA was better for ruling in sepsis [pos-itive LR 9.1(8.5–9.7) vs. 2.7(2.6–2.8)], whileSIRS was better for ruling out sepsis [neg-ative LR 0.67(0.65–0.70) vs. 0.74(0.72–0.75)].Conclusions: Paramedics had low rates ofdocumented suspicion of infection in sepsispatients. Paramedic screening strategies mayhelp to identify sepsis, but the choice of strategywill depend on whether the goal is to correctlyrule out versus rule in these diagnoses.

55. Preliminary Impact of AddingFollow-Up Home Visits On Call VolumesGenerated by EMS “Super-Users” Enrolledin a New Mobile Integrated HealthProtocol

Roger Stone, Jamie Baltrotsky, Alan Butsch,Ashley Robinson, Barry Reid, MontgomeryCounty MD Fire Rescue Services Categoryof Submission: Operations, Quality, Safety,Systems, Disaster

Background: Rising EMS call volumes tax EMSresources in many jurisdictions. A significantcontributor to volumes includes the frequent9-1-1 callers, some of whom may return homefrom hospitals with limited resources. Aftera new partnership in 2015 between EMS andour County’s HHS agency helped facilitateservices for 9-1-1 “Super-users”, our previousstudy found a preliminary association withreduced call volumes. Our agency has nowpartnered with discharging hospitals to starta new home visit program under Maryland’snew Mobile Integrated Health (MIH) Protocolbeginning March 2017. We wished to establishif this additional intervention was associatedwith a reduction of EMS call volumes fromenrollees in the protocol. Hypothesis: Initiationof follow-up home visits by our paramedicsand hospital outreach nurses has an impact onEMS utilization by a selected group of enrolled9-1-1 super-users. Methods: After our EMS-HHS partnership identified 9-1-1 super-users,we recruited a voluntary cohort to enroll in theMIH program. We retrospectively measuredusing CAD and EMS records cumulative callvolumes for the group of new enrollees, 90, 60and 30 days before and after the home visitsprogram started. Results: A cohort of Patients(N = 10) was enrolled in the MIH protocol andscheduled for home visits beginning March 1,2017. Cumulatively, those patients generated63, 53, and 30 calls during the periods 90, 60,and 30 days, respectively, prior to the home vis-its. Thereafter, those calls decreased to 7, 8, and18 calls for the periods of 30, 60, and 90 days,respectively, after visits began. The changeyields 9-1-1 call reductions of 77%, 85%, and71% during the post intervention three months.Conclusions: We believe super users in ourlarge system benefit from a coordinated pro-gram of EMS partnerships with public healthagencies and hospitals. A new partnershipwith Hospital Outreach and the initiation offollow-up home visits had preliminary impactson call volumes generated by the enrollees overa 30–90 day period. More studies are needed toprospectively prove value, sustainability andbest practices of these programs, and whichinterventions during home visits make themost difference.

56. Multi-Disciplinary Community HealthCare Interventions Reduce EMSUtilization by Elders

Joseph Petrosino, Jeffrey Boyd, JoanneMcGovern, James Dziura, Gina Stover,



Fangyong Li, Geliang Gan, Ryan Carter,Sandy Bogucki, Department of EmergencyMedicine, Yale University School of MedicineCategory of Submission: Professional

Background: Previous studies in a small, subur-ban town showed that more than half of elderswho fall and require lift assists will activate the9-1-1 system again within 30 days. Community-based interventions involving paramedics, vis-iting nurses, and primary care providers sub-stantially reduced the frequency of repeat EMSand lift assist calls. This study was designed toevaluate these findings across larger and morediverse populations of elders at risk for falls.Methods: For this non-randomized, prospec-tive study, informed consent to follow subse-quent health care utilization was obtained from2,265 participants residing in AMR’s regionalresponse areas. Participants chose to have nointervention, or the interventions that includedsequential home visits by a research paramedicevaluating disability and home safety, a visitingnurse assessing for home health care require-ments and eligibility, plus an offer of free trans-portation for a primary care provider visit. Par-ticipants were enrolled during (1) an EMS callfor lift assist, (2) an ED visit, or by (3) self-referral. Subsequent EMS calls were capturedby matching identifiers in our study databasewith those in AMR’s call records. Outcomescompared the proportion of participants thathad at least one EMS call during the 30 or90 days following enrollment (noninterventiongroup), or completion of the intervention, usinga generalized estimating equation approach,in SAS. Results: As of May 31, 2017, 980non-intervention group participants completed>30 days of follow up, and 652 completed>90 days, vs. 1,285 intervention group partic-ipants completing > 30 days, and 980 com-pleting >90 days. At 30 days, the interventiongroup showed a 51% reduction, and at 90 days,a 38% reduction in the proportion of partici-pants with at least one EMS response (both p< 0.001), compared with nonintervention. EDenrollees benefitted most, with a 45% reductionin EMS utilization at 30 days (p < 0.001), anda 25% reduction at 90 days (p = 0.009). Con-clusions: This study demonstrates short-termeffectiveness of our multidisciplinary commu-nity health care interventions at reducing EMSutilization by the elderly. ED enrollees, by ourmetrics the group that was most disabled atbaseline, appeared to benefit the most from theinterventions.

57. Effect of Sodium BicarbonateAdministration During Out-of-HospitalCardiac Arrests on End-Tidal CO2Readings in Considering Termination ofResuscitation

Brandon Morshedi, Alysha Joseph, RayFowler, University of Texas – SouthwesternCategory of Submission: Cardiac

Background: The administration of sodiumbicarbonate (NaHCO3) during out-of-hospitalcardiac arrests (OHCA) has been shown to pro-duce increased end-tidal capnography (EtCO2)during resuscitations. Traditionally, EMS sys-tems may authorize termination of resusci-tation (TOR) efforts after prolonged periodswhere the EtCO2 remains <10 mmHg. How-ever, if NaHCO3 has been administered, theEtCO2 may be elevated, possibly resulting inprolonged resuscitation efforts. The purposeof this study was to determine the effects ofthe administration of NaHCO3 during OHCAon EtCO2. We hypothesized that there wouldbe no observable difference in EtCO2 readingsbetween OHCA TOR patients who receivedNaHCO3 and those who did not. Methods:A retrospective analysis was performed on allOHCA TOR patients in a large, urban EMS

system between January 2013 and December2016. The off-line and on-line medical con-trol databases were queried to identify allpatients for whom the Provider Impressionwas “Cardiac Arrest.” The records were indi-vidually examined to determine the EtCO2readings and whether these patients receivedNaHCO3. Results: A total of 182 OHCA caseswere selected which had a documented EtCO2,with 93 receiving NaHCO3 and 89 not receiv-ing NaHCO3. The results were analyzed usinga Welch’s t-test. A significant difference wasfound in EtCO2 readings between the twogroups, with a mean EtCO2 of 26 mmHg inthe NaHCO3 group and a mean of 19.7 mmHgin the non-NaHCO3 group, with a p-value of0.026. A subgroup analysis showed that whencomparing the 50 highest EtCO2 readings, thesignificance was even greater, with a meanof 39.9 mmHg in the NaHCO3 group and27.7 mmHg in the non-NaHCO3 group, witha p-value of 0.0018. When the EtCO2 readingswere below 20 mmHg, there was no significantdifference. Conclusions: There are no widelyaccepted guidelines for the technique of TOR inOHCA patients. The value of employing EtCO2readings in TOR decisions is unclear. This studyindicates that administering NaHCO3 duringOHCA will significantly elevate the EtCO2, andNaHCO3 administration complicates the utilityof EtCO2 when levels are above 20 mmHg. Fur-ther study of the use of EtCO2 in TOR decisionsis necessary.

58. Feasibility of RecordingOut-of-Hospital Cardiac ArrestTreatment Via Use of a MobileSmartphone Application

Samuel Sondheim, Joseph Devlin, WilliamSeward IV, Aaron Bernard, Richard Feinn,David Cone, Frank H. Netter MD School ofMedicine, Quinnipiac University Category ofSubmission: Student, Resident, Fellow

Background: Given the demanding nature ofout-of-hospital cardiac arrest (OHCA) resus-citations, recordings of the times of interven-tions in EMS patient care reports (PCRs) areoften inaccurate. The American Heart Associ-ation developed Full Code Pro (FCP), a smart-phone application designed to assist providersin recording the timing of interventions per-formed. Through OHCAsimulations, this studyassessed the group size necessary to use theFCP recording functions accurately and safelywithout compromising patient care. Programevaluation was based on participant feedbacksurveys, data accuracy, delays between record-ing and performing interventions, and delays incare attributed to using the application, strat-ified by group size. Methods: Simulations ofa standard OHCA scenario using the Gau-mard TraumaHal mannequin and a dedicatediPhone 5 pre-loaded with FCP version 3.4 wererun with group sizes of 2–6 participants, withgroup sizes determined by participant avail-ability. Participants included Connecticut certi-fied paramedics and paramedic students whohad completed the respective coursework. Aseven-item feedback survey using a Likert scaleestablished participant feedback on the applica-tion. Videos of the simulations were analyzed toassess for delays. One-way ANOVA with trendanalysis was used to test if outcomes differedby group size and if differences tended in onedirection in parallel with group size. Results:There were 37 simulations including 142 partic-ipants. The feedback survey questions achieveda Cronbach’s alpha of 0.91 signifying high reli-ability, and demonstrated a linear trend sup-porting greater satisfaction with FCP as groupsize increases (p < 0.001). Similarly, increasinggroup size displayed linear trends with greaternumbers of interventions recorded (p = 0.009)

and fewer missed and false recordings (p =0.002). Delays revealed significant linear trends(p = 0.018 for delays in recording and p <0.001 for delays in care), as increasing groupsize corresponded with lesser delays. Greatestimprovement was noted to be between groupsof 3 and 4 participants. Conclusions: OHCAsimulations using FCP demonstrate increasedprovider comfort, increased recording accu-racy, and decreased delays as the group sizeincreased. While the application may improverecordings for PCRs and future research, thedata suggest a sufficient number of providers(>3) should be present to achieve reliable datawithout compromising patient care.

59. Influence of NeighborhoodSocioeconomic Status on Disparities inEmergency Medical Services Use andQuality of Prehospital Care for IschemicStroke

Timmy Li, Manish Shah, Adam Kelly, JeremyCushman, David Rich, Edwin van Wijngaar-den, Gina Lovasi, Courtney Jones, North-well Health Category of Submission: Student,Resident, Fellow

Background: A minimal amount is knownregarding the effect of neighborhood socioe-conomic status (nSES) on emergency medicalservices (EMS) use and quality of prehospi-tal stroke care. We assessed the associationbetween nSES and EMS use, decision delaytime, and quality of prehospital care amongstroke patients. Methods: A retrospectivecohort study was performed using the GetWith The Guidelines-Stroke registry at twohospitals to identify patients with a hospitaldiagnosis of ischemic stroke between 2012 and2016. Registry data were merged with datafrom EMS medical records and the UnitedStates Census Bureau. Patient addresses weregeocoded and a one-kilometer buffer was cre-ated around each patient’s address to representtheir neighborhood. Census data from eachbuffer were used to create a composite nSESscore, which was categorized into quartiles.Multivariable log-binomial regression modelsassessed the associations between nSES and 1)EMS use, and 2) decision delay time to calling9-1-1. Among EMS patients, we also assessedassociations between nSES and (1) dispatchedEMS level of care, (2) EMS response time, (3)EMS on-scene time, (4) Cincinnati PrehospitalStroke Scale (CPSS) assessment, and (5) hospi-tal prenotification by EMS. Results: Of 1,472patients, 48% were aged 50–74 years, 50% werefemale, 73% were white, and 59% used EMS.Compared with patients in the highest nSESquartile, patients in the lowest nSES quartilewere 20% less likely to use EMS (risk ratio (RR):0.80; 95% confidence interval (CI): 0.67, 0.95).EMS providers performed the CPSS on 65%of patients. Patients of lower nSES were lesslikely to have a CPSS performed: risk ratios,compared with the highest nSES quartile, were1.72 (95% CI: 1.14, 2.60), 2.91 (95% CI: 2.00,4.21), and 3.39 (95% CI: 2.30, 4.99) for nSESquartiles 2, 3, and 4 (lowest nSES), respectively.nSES was not significantly associated withother outcomes. Conclusions: Among a sam-ple of ischemic stroke patients, 41% did notuse EMS and those of lower nSES used EMSmore frequently. EMS providers performed theCPSS assessment less frequently on patients oflower nSES. Understanding reasons for theseobservations is vital to improving the qualityof prehospital stroke care.

60. Paramedic Recognition of ParoxysmalSupraventricular Tachycardia

Spencer Sample, Colleen Shortt, Erich Hanel,Michelle Welsford, Michael G. DeGroote Schoolof Medicine, McMaster University, Hamilton,



Ontario Category of Submission: Student,Resident, Fellow

Background: Paroxysmal supraventriculartachycardia (PSVT) is a common group ofarrhythmias that Advanced Care Paramedics(ACPs) can often manage with vagal maneu-vers, adenosine, and/or cardioversion, pro-vided that they correctly identify the rhythm.The purpose of this study is to determinethe accuracy of ACP identification of PSVT.Methods: Following ethics approval, all callswith patients �18 years with a 12-lead ECGavailable, who were assessed by ACPs withina region of western Ontario between July 2015and December 2015 and had a documentedheart rate>150 bpm, were included. Paramediccall reports were retrospectively reviewed forstudy data, including documentation of ACPidentified PSVT. The reference standard wasconsensus between a fellow and prehospitalphysician who adjudicated each ECG for thepresence of PSVT in a blinded, independentfashion. In the event of a disagreement, athird, blinded prehospital physician was usedfor consensus. Results: Of the 442 patientsincluded, 197 (45%) were male and the medianage [Interquartile range(IQR)] was 70.0 years(58.0–82.8). ACPs identified 74 (16.7%) patientsas having PSVT. Of these, 48.5% had a his-tory of previous arrhythmia, compared to31.9% of patients with no ACP identifiedPSVT (p = 0.026). They were also significantlyyounger [median(IQR) = 63.0 (47.0–72.0)]compared to those without ACP identifiedPSVT [median(IQR) = 72.0 (61.0–85.0)] (P <0.0001). Sensitivity of ACP identified PSVTwas 97.3% (95%CI:85.8–99.9%) and specificitywas 90.6% (95%CI:87.3–93.3%). The positivepredictive value (PV) of ACP identified PSVTwas 48.6% (95%CI:41.1–56.3%), the negativePV was 99.7% (95%CI:98.1–99.9%), the positivelikelihood ratio (LR) was 10.4 and the negativeLR was 0.03. Moderate inter-rater agreementwas seen between initial ECG interpretations[387 (87.6%)] (kappa = 0.42, 95%CI:0.29–0.54)by the fellow and prehospital physician, whileagreement was higher (good) between thetwo prehospital physicians [49/55 (89.1%)(kappa = 0.70, 95%CI:0.48–0.92)]. Conclusions:These results indicate that ACPs are adeptat identifying PSVT, but are prone to falsepositives. Given the relatively good sensitivityand specificity seen in this investigation, futurestudies should investigate ACP recognition ofspecific rare arrhythmias (antidromic accel-erated atrial fibrillation) that may requiredifferent management including avoidance ofadenosine.

61. Police Department Tactical Medicine(TACMED) Program Impact on TraumaPatient Mortality: Review of a LargeUrban EMS and TACMED System

Elliot Ross, David Wampler, Avery Kester,Xandria Gutierrez, Crystal Perez, LaurenReeves, Alejandra Mora, Joseph Maddry,Craig Manifold, San Antonio Uniformed ServicesHealth Education Consortium Category ofSubmission: Student, Resident, Fellow

Background: Tactical Emergency MedicalServices (TEMS) is a growing subspecialty ofprehospital care. Tactical providers are ideallysuited to provide care at the point of injury inareas traditional EMS cannot enter. A minimalamount is currently known regarding theclinical impact of these programs. This studyexamines patient outcomes of those treated by aPolice based TEMS system vs. traditional EMS.Methods: Study inclusion criteria consisted oftrauma patients where police were dispatchedand EMS was staged and were then transferredto a Level I trauma hospital. All patients thatdied at the scene or enroute were excluded. The

computer automated dispatch (CAD) systemwas used to identify all cases from 2011–2015.The TEMS and EMS records for cases meetinginclusion criteria were extracted. Demograph-ics, injury description, prehospital index (PHI)scores, disposition, and interventions werecollected. Hospital disposition and outcomedata were linked using the regional traumaregistry. Using gender, injury year/type, age,and ISS a case-match controlled comparisonbetween EMS and TEMS records (2:1) wasconducted. Chi-square (or Fisher’s Exact) testfor categorical and t-test (or Wilcoxon) forcontinuous variables. Results: Of the 122,707CAD events, only 2243 met inclusion criteria.Seventy TEMS records and 140 EMS casematched controls were included. Majority weremale (90%) civilians (99%) with a median ageof 31. Sixty percent of patients were injuredsecondary to a shooting, 30% stabbing, and10% assault. Moderate to severe bleeding wasencountered in 75% of patients, and 46% sus-tained major trauma (PHI � 4). TEMS providershad a shorter response time compared to EMSproviders; 6 vs. 13 minutes, p < 0.0001. Cohortshad similar PHI scores and intervention per-formance rates. Final hospital disposition andhospital resource utilization were comparable.Both had similar number of ventilator, ICU,and hospital days. There was no differencein mortality rates. Conclusions: In this study,TEMS providers exhibited shorter responsetimes and performed medical interventions atsimilar rates to traditional EMS. Although nodifferences in patient outcomes were noted,all patients who died prior to hospital arrivalwere excluded. Future studies are needed todetermine how response time impacts the rateof preventable death.

62. Optimizing Deployment of MechanicalCPR Does Not Improve OHCA OutcomesWhen Compared with Manual CPR

Brandon Oyler, Louis Gonzales, Jeff Hayes,Mark Escott, Jose Cabanas, Paul Hinchey,Lawrence Brown, Dell Medical School at theUniversity of Texas Category of Submission:Cardiac

Background: Deploying mechanical CPR inout-of-hospital cardiac arrest (OHCA) is logisti-cally challenging. Inefficient deployment mightexplain reports of unfavorable OHCAoutcomesassociated with mechanical CPR. We hypoth-esized that in an EMS system with optimizeddeployment, sustained ROSC and survival tohospital discharge will not differ for OHCApatients managed with and without mechani-cal CPR. Methods: In 2015, we initiated a qual-ity improvement process to choreograph andoptimize deployment of mechanical CPR. Allprimary first response agency (attending ±75%of OHCAs) field personnel attended in-persontraining and practical exercises emphasizinghigh quality traditional CPR, timely defibril-lation, airway management / ventilatory sup-port and first-round medication administrationbefore initiating mechanical CPR. We then ana-lyzed all adult, non-traumatic OHCA attendedby the first response agency during 2016. Dur-ing the study period, mechanical CPR deviceswere deployed on some—but not all—firstresponse units; use of mechanical CPR wasbased primarily on availability and/or whetherpatients achieved ROSC after initial resusci-tation attempts. We therefore used propensityscore matching to select cases with and with-out mechanical CPR that had similar patientdemographics and arrest characteristics. Weexcluded OHCAs with sustained ROSC follow-ing only CPR or defibrillation without medica-tion administration, terminations of resuscita-tion without meaningful resuscitation attempts(including DNRs), and EMS-witnessed arrests.

All prehospital data were obtained from theEMS electronic health record; hospital out-comes were obtained from receiving hospitals.Results: Of 444 eligible OHCAs, 227 receivedmechanical and 217 received traditional CPR.Crude ROSC (29.1% vs. 39.2%) and survivalto discharge (5.7% vs. 13.8%) were lower withmechanical CPR, but mechanical CPR caseswere also less likely to be witnessed arrestsand less likely to present with a shockablerhythm. In the propensity score analysis of 187patients with mechanical CPR well-matched to187 patients with traditional CPR, both ROSC(29.2% vs. 39.5%; difference: −10.3%; CI: −0.7%to −19.9%) and survival to discharge (7.0% vs.14.1%; difference: −7.1%; CI: −0.9% to −13.1%)remained significantly lower for patients receiv-ing mechanical CPR. Conclusions: In an EMSsystem with optimized deployment, mechani-cal CPR was associated with decreased ROSCand decreased survival to discharge.

63. Gender Disparities in the PrehospitalSetting among Known St-SegmentElevation Myocardial Infarction Patients

Krystal Baciak, Stephen Sanko, Marc Eck-stein, University Of Southern California-Los Ange-les County And Los Angeles Fire DepartmentCategory of Submission: Student, Resident,Fellow

Background: Identification of a ST elevationmyocardial infarction (STEMI) in the prehospi-tal setting has been shown to decrease door-to-balloon time and mortality. Up to 20% of STEMIpatients do not present with typical symptomsand gender disparities exist in the prehospi-tal setting in the assessment of patients ulti-mately found to have ACS. Our hypothesis iswomen are more likely to have delayed STEMIcare than men. Methods: This is a retrospectivecohort study of 9-1-1 patients who were trans-ported by a single large urban EMS providerto STEMI-Receiving Centers (SRC) from Jan-uary 2011 to December 2015 and were diag-nosed with a STEMI, had emergent PCI, andwere found to have a culprit coronary arteryobstruction. Our primary outcome was EKG-to-balloon time (E2B). Our exclusion criteriawere: interfacility transfer, age under 18, inabil-ity to calculate E2B, and missing gender data.Our secondary outcomes were: time intervalsfrom 9-1-1-call through device time. Results:Of the 2,778 patients eligible for analysis, 2,148patients were included in final analysis afterapplication of the exclusion criteria. Womenhad longer on-scene times, longer times from9-1-1-call to arrival at the SRC, time from firstmedical contact (FMC) to balloon, and timefrom 9-1-1 call to EKG (P < 0.001). Time fromfirst medical contact to cath lab arrival waslonger in women, but did not reach statisti-cal significance (P < 0.002) using a very con-servative Bonferroni-corrected p-value. Therewere no statistically significant differences inwhether or not a prehospital EKG was per-formed or transmitted, whether a prehospi-tal EKG indicating STEMI was noted, whetheror not aspirin was given, transport time, timefrom EKG to cath lab arrival, EKG-to-balloonor door-to-balloon (p > .001). Conclusions: Ourstudy demonstrates women are more likely tohave delayed times from 9-1-1-call to hospitalarrival, FMC to balloon, and time from 9-1-1 callto EKG, but do not have a delayed E2B or door-to-balloon time. Limitations include short trans-port times, a single urban EMS service, and theretrospective nature of the study.

64. Statewide Retrospective Analysis onthe Characteristics of EMS Refusals ofCare

Novneet Sahu, Patrick Matthews, RossMegargel, Rutgers - New Jersey Medical School



Category of Submission: Student, Resident,Fellow

Background: Improving EMS systems of carerequires a better understanding of out-of-hospital refusals of care. There is a paucity ofdata on EMS refusals of care. Studies over thepast three decades have shown widely vary-ing results on the characteristics, demograph-ics, and rates of EMS refusals of care. The pur-pose of this study is to analyze, at the state level,the characteristics, demographics, and rates ofEMS refusals of care to provide a platformfor identifying targets to help improve EMSsystems of care. Methods: Delaware statewideEMS data for all refusals and transports werequeried for the calendar year of 2016. Age, gen-der, dispatch reason, time of year, and locationwere aggregated and retrospectively analyzedthrough descriptive statistics and multi-variatelogistic regression. Results: Of the 155,303 EMSincidents, 12,244 (7.9%) resulted in refusals ofcare. Patients 65 years and older had a smallerpercentage of refusals than adults 18–64 yearsold and children <18 years old (6.4% vs. 8.8%vs. 10%, p < 0.001). Men had a greater refusalrate than women (8.5% vs. 7.3%, p < 0.001).Diabetes-related problems (36.2%) and motorvehicle accidents (28.5%) resulted in the high-est rates of refusal of care (p < 0.001). The high-est percentage of overall refusals occurred dur-ing mid-summer (8.8%, p < 0.001). Locationsof care which include places of recreation andbodies of water had the highest refusal rates(45.6% p < 0.001). Conclusions: In this popu-lation, geriatric patients had lower refusal pro-portions; whereas, prior studies suggested thatgeriatric refusal numbers are greater than otherage groups. The greater refusal rate amongmen is consistent with previous literature. Priorstudies have shown the highest rates of refusalsfor motor vehicle accidents and other trauma,however, diabetes-related problems comprisedthe highest percentage of refusals in this popu-lation. Mid-summer time of year and places ofrecreation also comprised high percentages ofrefusals and further investigation is necessaryto identify root causes of these patterns.

65. Feasibility of Point-of-CareUltrasound (Pocus) in Out-of-HospitalCardiac Arrest (OHCA) by NoviceUltrasonographers

James Fitzgibbon, Emily Lovallo, MarekRadomski, Jeremiah Escajeda, ChristianMartin-Gill, Department of Emergency Medicine,University of Pittsburgh School of MedicineCategory of Submission: Cardiac

Background: Point-of-care ultrasound(POCUS) may be a useful tool to predictsurvival and guide interventions in out-of-hospital cardiac arrest (OHCA), yet a paucity ofdata exists on its prehospital use by users withlimited ultrasound experience. We aimed todetermine the feasibility of using POCUS dur-ing OHCA by resident and fellow physiciansstaffing a 24/7 prehospital response vehicleand identify barriers to its use. Methods: Wedeployed a portable ultrasound device (iViz, bySonosite) for use by prehospital physicians forOHCA in an urban EMS system. All physiciansreceived POCUS education as part of graduatetraining, and were provided an instructionalvideo on use of the iViz device. POCUS usewas limited to identifying cardiac motionduring pulse checks, without interruptingresuscitation, and the results could be usedto guide management at the physicians’ dis-cretion. Data were recorded prospectively bysaving video and still images on the device andthrough a custom electronic form within thepatient care report (emsCharts). The primary

measure was the frequency of use of POCUSduring OHCA. Secondarily, we characterizedimage quality by expert (ultrasound fellowshiptrained) faculty review (using kappa statisticfor agreement), and identified barriers to theuse of prehospital POCUS. Results: FromNovember, 2016 to March, 2017, 348 physicianfield responses were reviewed, including127 cases of OHCA, and 56 (44%) cases withPOCUS performed. Still or video images wererecorded in 48 (86%) cases and video in 34(61%) cases. From video images, agreement inidentifying cardiac motion between prehospitalphysician and expert reviewer occurred in 91%of cases (K=0.82). Reasons cited for not usingPOCUS included return of circulation soonor before arrival, prioritizing interventions,provider preference, not having the ultrasounddevice, mechanical failure, and cessation ofresuscitation per advanced directives. Conclu-sions: Use of POCUS by novice prehospitalphysician ultrasonophraphers to detect wallmotion in OHCA is feasible and correlateswith expert interpretation. Several avoidablebarriers to the use of prehospital POCUS maybe addressed through additional educationalinterventions and increased familiarity with thedevice.

66. Air Versus Ground Transfer toComprehensive Stroke Center in Patientswith Large Vessel Occlusion Stroke

Ali Shams, Chris Kanaan, RebbecaGrysiewicz, Chris Kazmierczak, LauraSteucher, Robert Swor, Beaumont HealthCategory of Submission: Student, Resident,Fellow

Background: Optimal treatment ischemicstroke caused by a large vessel occlusion (LVO)involves timely transfer from a primary strokecenter to a comprehensive stroke center (CSC)that can offer mechanical endovascular therapy.Transfers are either done via air or ground,however data have not shown a clear benefitof one method of transfer over the other. Theobjective of this study was to compare air vsground transfer times from decision to transferto definitive care in patients with LVO strokestransferred to a single CSC. Methods: Thisis a cohort study of patients transferred to asingle suburban CSC (January 2015–December2016) from seven primary stroke centers withina 15-mile radius with the diagnosis of LVOstroke. Key time intervals including transporttime, time from decision to transfer (accessto sending hospital EMR allowed characteri-zation of this time point), and time intervalsfrom arrival to first ED to interventional skinpuncture (access) and reperfusion at the CSCwere recorded. Non-parametric statistics wereused for comparisons. Median and range arereported. Results: There were 30 inter-hospitaltransfers from within a 15-mile radius. Of these16 were by air and 14 were ground transfers.Air transport times were significantly shorter(16.5 vs. 30.0 minutes, p = 0.013). There wasno difference between transfer decision-CSCED arrival between air and ground (65.5 vs.67.5 minutes, p = 0.967, respectively). Inhospital processes for air and ground transferpatients at the CSC were shorter but not signif-icantly so, CSC door-table (2.5 minutes vs. 16.5,p = 0.44) and CSC door to access (28.0 vs. 40.5,p =. 44). Time interval for air and ground trans-fer were not different for arrival to 1st ED toaccess (155.5 vs.172, p = 0.118 or arrival - reper-fusion (208.5 vs. 211, p = 0.495). Conclusions:In our small pilot study, despite shorter trans-port times, there was no significant differencebetween air and ground transfer from decisionto transfer to CSC arrival, or time from firsthospital to access or reperfusion. Assessmentof unmeasured intervals are needed to assess

the optimal method for inter hospital transferof critical patients.

67. Patient Preferences toward EmergencyMedical System Provider Attire

Jesse Olsen, Jeffrey Lubin, Khaled Iskan-darani, Penn State College of Medicine Categoryof Submission: Operations, Quality, Safety,Systems, Disaster

Background: In a health-care landscape drivenby patient satisfaction and quality assurance,preferences towards provider attire has becomea topic of interest. Uniforms afford essentialvisual clues for personnel identification; recentresearch demonstrates attire impacts patientpreferences for both nurses and physicians inemergency settings. In emergency medical sys-tems, teams rely on trust for effective and suc-cessful responses. In the context of EMS person-nel, no studies have addressed patient percep-tion of attire. This prospective study addresseshow EMS attire influences patient perceptionof care through five different variables: like-ability, trust, confidence, willingness to confide,and intelligence. Methods: Over six weeks inthe Emergency Room at Penn State Hershey,165 surveys were completed evaluating a teamof two EMS providers. Participants surveyedviewed one of three two-minute videos of anEMS team responding to a patient with chestpain. In each video EMS personnel wore a dis-tinct outfit: a blue tee shirt, a white button-upshirt or turnout gear. Participants subsequentlycompleted a six question survey addressingproviders on a 5-point Likert scale. Attireswere compared using a two tailed Kruskal-Wallis test, a non-parametric equivalent of anANOVA. Results: Of 165 surveys completed,87.5% of responders rated EMS attire as impor-tant. No differences in responses were foundrelated to patient age, gender or ethnicity. Anal-ysis of the likert data, showed no significantdifferences with respect to perceived providertrust, smartness, likeability or confidence. How-ever, participants answered significantly loweron the Likert scale for willingness to discussconfidential information with the providers inthe turnout gear compared to the other twoattires at an alpha of .0057. Conclusions: Basedon our results, EMS provider attire does notimpact patient perceived quality of care. Lowerresponses were found for turnout attire, possi-bly from a lack of association of EMS providerswith fire gear. Studies drawing a larger sample,and those that analyze more outfits or aspectsof provider appearance would lend support tothis conclusion. Our study was small, limitedby length of the videos, and number of outfitstested but our results conclude attire as a minorfactor in EMS responses.

68. Multivariable Analysis of FactorsAssociated with EMS Non-transports

Rickquel Tripp, Jonathan Elmer, FrancisGuyette, Christian Martin-Gill, Departmentof Emergency Medicine, University of PittsburghSchool of Medicine Category of Submission:Operations, Quality, Safety, Systems,Disaster

Background: Emergency response withouttransport confers a risk of negative patient out-comes, increased liability, and non-payment.Yet, few rigorous studies have identified riskfactors for non-transports. We aimed to identifydemographic and clinical characteristics pre-dictive of non-transports using a large databaseof out-of-hospital EMS responses. Methods: Weretrospectively reviewed consecutive patientcare records from 21 urban, suburban, and ruralEMS agencies in Western Pennsylvania fromApril 2013 to December 2016. We identifiedage, gender, race, ethnicity, level of transport,



last vital signs (BP, RR, HR, SPO2, and GCS),loss of consciousness (LOC), abnormal mentalstatus (AMS), medical category, and timeof day. We excluded cases of cardiac arrest,interfacility/scheduled transports, EMS assist,no patient encountered, and patients aged<18 years or unknown age. For non-transports,we described the incidence of protocol-definedabnormal vital signs (HR <50, >100; SBP <100,>200; DBP <50, >100; RR <12, >24; SpO2<95; GCS <15), LOC, and AMS. We usedunadjusted and adjusted logistic regressionto identify independent predictors of non-transport. Results: We identified 385,908 casesmeeting study criteria, with 35,266 (9.1%) non-transports. Patient characteristics were: medianage 59 years (IQR 41–77), 55.6% female, 16.8%Black, 0.7% Hispanic, and 96.3% advanced lifesupport (ALS). Incidence of abnormal vitalswere HR (N = 4435, 12.6%), SBP (N = 539,1.5%), DBP (N = 1324, 3.8%), RR (N = 159,0.5%), SpO2 (N = 1543, 4.4%), and GCS (N =834, 2.4%). There were 785 (2.2%) with LOCand 2031 (5.8%) with AMS. In adjusted mul-tivariable analysis, we identified associations(OR, 95%CI) with non-transports and malegender (1.08, 1.00–1.16); ALS (1.60, 1.04–2.47);and morning [6:00–11:59] hours (0.79, 0.70–0.88) and evening [18:00–23:59] hours (1.07,1.03–1.11), compared to overnight [0:00–5:59]hours. Medical categories most associated withnon-transports were trauma (2.37, 1.79–3.14),dizziness/syncope (1.80, CI 1.47–2.20), andallergic reaction (OR 1.54, CI 1.33–1.79). Race,ethnicity, LOC, and AMS were not associ-ated with the incidence of non-transports.Conclusions: Patients not transported byEMS often have abnormal heart rate andare associated with complaints of trauma,dizziness/syncope, or allergic reaction. Thisinformation can guide patient refusal protocolsand future research on outcomes of these at-riskpatients.

69. Among Stemi Patients, Inferior STElevation Is Not Associated with a HigherFrequency of Hypotension after FieldNitroglycerin

Nichole Bosson, Jayson Morgan, BenjaminIsakson, Amy Kaji, Atilla Uner, KatherineHurley, Timothy Henry, Marianne Gausche-Hill, James Niemann, LA County EMS AgencyCategory of Submission: Cardiac

Background: Patients with inferior STEMIinvolving the right ventricle are believed to beat higher risk for hypotension after nitroglyc-erin (NTG). The objective of this study wasto determine if inferior STEMI is associatedwith increased risk of hypotension upon EDarrival in patients treated with NTG by EMS.Methods: Consecutive adult patients with sus-pected STEMI transported by EMS to one ofthree participating PCI-capable hospitals wereprospectively identified and maintained in alog during an 18-month period. Investigatorsreviewed records for initial field and ED vitalsigns, field NTG treatment, and hospital out-comes. Inter-rater reliability was assessed ona random 10% sample of records using thekappa statistic. Patients with a hospital diagno-sis of STEMI treated with NTG were included.Patients with hypotension on EMS arrival wereexcluded. Inferior STEMI was defined as ST-elevations in the inferior leads on the prehos-pital ECG. The frequency of ED hypotension,defined as a triage SBP less than 100 mmHg, inpatients with inferior STEMI was compared topatients with other STEMI. Patients were fur-ther stratified by lesion location. The frequencyof hypotension was compared with Fisher’sexact test and change in SBP with Hodges-Lehmann’s median difference. Results: Of 239patients with STEMI, 46 were excluded for ini-

tial hypotension and 38 did not receive NTG;thus, 155 comprised the study cohort. Medianage was 61 years with 71% male. Hypotensionoccurred in 3 patients (4%) with inferior STEMIand 10 patients (14%) with other STEMI, RR0.3 (95%CI 0.1, 1.0) p = 0.02. Inter-rater relia-bility was excellent, kappa 0.93 (95% CI 0.80,1.0). Mean decrease in SBP was −15 ± 23 mmHgand −10 ± 22 mmHg in patients with inferiorand other STEMI, respectively, median differ-ence in the decrease in SBP −4.5 mmHg (95%CI −12.0, 3.0). Compared to patients treatedwith PCI in any other location, hypotensionafter NTG among patients with proximal or midRCA lesions was similar, RR 1.0 (95%CI 0.9,1.1) p = 0.6. Conclusions: In this cohort, whencompared with other STEMI patients, thosewith inferior STEMI had a lower frequencyof hypotension after field NTG; RCA lesionlocation was not associated with an increasedrisk.

70. Characteristics of Emergency MedicalTechnician Graduates Unsuccessful onthe National Certification CognitiveExamination

Rebecca Cash, Remle Crowe, MadisonRivard, Ashley Larrimore, William Krebs,Jeremy Miller, Ashish Panchal, National Reg-istry of Emergency Medical Technicians Categoryof Submission: Operations, Quality, SafetySystems, Disaster, Disaster

Background: Research on EMT student perfor-mance has focused on pass rates and character-istics related to success. Conversely, a minimalamount is known regarding EMT graduateswho were unsuccessful at passing the examina-tion. The objective of this study was to describedemographics and test-related performance ofgraduates unsuccessful on the computer adap-tive National EMT Certification examination.We hypothesized that the majority of candi-dates who are unsuccessful on the examina-tion are close to the passing standard (max-imum length testers) and would be likely toretest. Methods: National EMT Certificationcognitive examination results for graduates ofnon-military EMT education programs from theclass of 2013 were analyzed as a cross-sectionalevaluation. The computer adaptive test termi-nates when the 95% confidence interval sur-rounding the estimate of the candidate’s abilityis entirely above or below the passing standard.Test length ranged from 70 to 120 questions.Unsuccessful testers were defined as candidateswho had a grade of fail or incomplete (didnot finish the examination) on their first exam-ination attempt. Chi-square tests were used tocompare demographics of candidates and toassess for differences in retesting between min-imum and maximum length testers. Results: Atotal of 59,560 EMT graduates from the classof 2013 attempted the National EMT Certifica-tion cognitive examination and 33% (n = 19,899)were unsuccessful the first attempt. The pro-portion of males and females who were unsuc-cessful did not differ (males: 34%, n = 12,642;females: 33%, n = 6,187, p = 0.05). More thanone-third of unsuccessful candidates receivedthe maximum number of questions (36%, n =7,128) while 40% (n = 7,985) received the min-imum number of questions. Of those unsuc-cessful on the first attempt, 66% (n = 13,111)attempted a second examination. More maxi-mum length testers attempted a second exam-ination compared to minimum length testers(72%, n = 5,156 vs. 60%, n = 4,763, p < 0.001).Conclusions: Two-thirds of first-time candi-dates unsuccessful on the National EMT Certi-fication cognitive examination attempted a sec-ond examination. A greater proportion of thoseclose to the passing standard (maximum lengthtesters) retested. Future work is needed to bet-

ter understand the reasons behind candidateretesting including personal and educationalexperiences.

71. Interfacility Transport of thePregnant Patient: A 5-Year RetrospectiveReview of a Single Academic Center BasedCritical Care Transport Program

Philip Nawrocki, Asa Margolis, Shawn Brast,Matt Levy, Johns Hopkins Lifeline Category ofSubmission: Student, Resident, Fellow

Background: Interfacility transport of pregnantpatients involves unique challenges and con-siderations. Data from the National EmergencyMedical Services Information System (NEM-SIS) dataset indicate that 0.6% of all EMStransports and 0.6% of interfacility transportsinvolve pregnant patients. Limited informa-tion exists surrounding the safety and adverseevents of this patient population in the out-of-hospital setting. This study aimed to exam-ine clinically significant adverse events thatoccur during the interfacility transport of preg-nant patients. Methods: A retrospective reviewof quality assurance data was performed. Thestudy population consisted of pregnant patientstransported to the labor and delivery units oftwo hospitals within an academic quaternary-care hospital system between January 2012 andDecember 2016. Primary outcomes (adverseevents) were defined as: hypotension, respira-tory distress, exacerbation of hypertensive dis-ease of pregnancy (preeclampsia, eclampsia),need for vasoactive medications, dysrhythmia,intubation or unintended extubation, changein mental status, need for restraints, cardiacarrest or death, and delivery during transport.Use of online medical direction and reason forconsultation were secondary outcomes of inter-est. Results: Our critical care transport systemperformed 30,181 total interfacility transportswithin the five year study period. 709 patients(2.4%) met inclusion criteria. Clinically signif-icant adverse events occurred during 32/709patient transports (4.5%). The most frequentevents were: exacerbation of hypertensive dis-ease requiring intervention (25), hypotension(4), and altered mental status (2). There werezero instances of cardiac arrest, death, or deliv-ery. Conclusions: Interfacility transport of preg-nant patients is a common occurrence thatinvolves unique challenges and risks. Withinthe experience of this critical care transport pro-gram, significant adverse events were identi-fied in 4.5% of transported patients over a 5-year period. This data will help guide the train-ing of prehospital providers and the formationof protocols to mitigate and respond to theseevents. Notable limitations include the use ofdata from a single system, absence of scenetransports, and use of paramedic/nurse crewconfiguration.

72. Advanced Provider Response Unit(APRU), an Answer to Low-Acuity 9-1-1Calls?

Saman Kashani, Stephen Sanko, MarcEckstein, USC Keck School of Medicine, Dept ofEmergency Medicine, Los Angeles Fire DepartmentCategory of Submission: Student, Resident,Fellow

Background: The Los Angeles Fire Depart-ment (LAFD) has experienced an unsustainableincrease in 9-1-1 calls. Over the past 2 years,call volume rose by 14%; vastly higher than thehistorical rate of increase of 1–2%. To addressthe increasing call volume, while still pro-viding care for the citizens of Los Angeles,the LAFD launched the Advanced ProviderResponse Unit (APRU), a specialized ambu-lance staffed by a licensed advanced practice



provider (APP) and a firefighter/paramedicwith the mission of treating and releasingpatients on scene and providing linkage tofurther care. This is a description of the first19 months of service. Methods: This was a ret-rospective review of LAFD electronic healthrecords from January 2016 to August 2017 in theLos Angeles area. The APRU was active 4 daysa week for approximately 82 weeks. Enrolledpatients were either low-acuity 9-1-1 callers,identified through monitoring 9-1-1 radio traf-fic or housed (i.e., non-homeless) 9-1-1 fre-quent users identified from prior LAFD healthrecords. Summary descriptive statistics werecollected. Results: The APRU was linked to1,079 incidents over approximately 328 days ofservice (mean 3.3 incidents/day). Of these inci-dents, there were 127 cancellations, 88 foundno patient, 13 refused care, and another 12were ineligible for APRU care. The remain-ing 839 were treated (77.8%). Of those treated,379 (45.2%) were treated and care was trans-ferred to another transporting unit, 360 (42.9%)were treated and released on scene, and 100(11.9%) were treated and transported. Of the 100transported by the APRU, 58 were transportedto a non-emergency room with 55 transporteddirectly to mental health clinics and 3 to a sober-ing center. Of the 360 treated and released onscene, the APRU spent an average of 23 min-utes on scene (minimum 1 minute, maxi-mum 1 hour 15 minutes, median 20 minutes).Conclusions: The LAFD APRU has shownpromise in decreasing costly EMS transportsand ED care. Furthermore, by leveraging thediagnostics skills of the APP, patients canbe treated and released on scene or medi-cally cleared for alternate destinations. Furtherresearch is needed to study this novel type ofEMS care.

73. Predictive Value of Each ComponentField Triage Guidelines on HospitalOutcome in EMS-Treated TBI

Sola Kim, Sang Do Shin, Kyoung Jun Song,Young Sun Ro, Jeong Ho Park, Seoul NationalUniversity Hospital Category of Submission:Student, Resident, Fellow

Background: Unbiased estimates for fieldtriage guideline performance are importantin optimizing trauma systems and improvingoutcomes among seriously injured patients.The accuracy of each triage component has notbeen evaluated in traumatic brain injury (TBI)patients. Based on evaluation as a diagnostictest, we considered the standard to be mortal-ity and disability, which is the final hospitaloutcome. The aim of this study is to predict theperformance of each component of field triageguidelines on hospital outcomes in TBI patients.Methods: This was a cross-sectional observa-tional study using a nationwide, prospectiveregistry of severe trauma patients treated byemergency medical services (EMS) providersin 10 provinces in Korea. The study populationwas adult TBI patients between January 2013and December 2013. The main exposure waseach component of field triage set by theAmerican College of Surgeons Committee onTrauma and Centers for Disease Control andPrevention as determined by EMS provider.The primary outcome was hospital mortalityand secondary outcome was disability at dis-charge. Disability is defined as new disabilityor worsened Glosgow Outcome Scale (GOS)including death than pre-event GOS. Sensitiv-ity, specificity and area under the curve (AUC)were calculated. Results: Total 5,133 patientsmet the field triage guidelines. 21.5% died, and51.4% of patients had disability. The sensitivityand specificity for mortality of the physiologic,anatomic and mechanical criteria were 91.4%and 47.3%, 20.0% and 93.15%, 57.8% and

89.3%, respectively. Among each componentof criteria, altered mentality showed highestsensitivity and AUC for mortality, which was89.2% (95% CI 87.4 to 91.0) and 0.699 (95%CI0.687 to 0.711). Amputation and chest wallinstability in anatomic criteria showed highestspecificity for mortality, 99.8% (95%CI 99.6%to 99.9%). Altered mentality showed highestsensitivity and AUC for disability, which was75.9% (95% CI 74.3% to 77.5%) and 0.671 (95%CI0.658 to 0.684), respectively. Conclusions: Thephysiologic criteria of field triage guidelinesshowed high sensitive for mortality. Anatomicand mechanical criteria showed low sensitivityand high specificity. The trend was similar fordisability. Altered mentality of physiologiccriteria showed highest sensitivity and AUCamong each component of field triage scheme.

74. Effect of Chest Compression ParameterVariation on Waveform Characteristics ofthe Ventricular FibrillationElectrocardiogram

David Salcido, Matthew Sundermann, Alli-son Koller, Rena Sufrin, John Kucewicz,Pierre Mourad, Graham Nichol, JamesMenegazzi, Adeyinka Adedipe, Departmentof Emergency Medicine, University of PittsburghSchool of Medicine Category of Submission:Cardiac

Background: The ventricular fibrillation (VF)electrocardiogram (ECG) waveform is knownto deteriorate over time if untreated, recoverwith CPR, and to predict defibrillation suc-cess. VF ECG measures could inform CPRquality feedback algorithms based on patientphysiologic response. Objectives: Investigatethe effects of chest compression rate, depthand duty cycle (DC) on VF ECG waveformcharacteristics in a swine cardiac arrest model.Methods: Twelve mixed-breed domestic swinewere sedated (ketamine & xylazine), anes-thetized (fentanyl) and paralyzed (vecuro-nium), followed by endotracheal intubationand mechanical ventilation. Animals wereinstrumented with a battery of physiologicalsensors, including multi-lead ECG (BioAmp,ADInstruments, Inc), recorded continuouslywith a high-fidelity data acquisition unit (Pow-erLab, ADInstruments, Inc) at 1000 Hz. Ventric-ular fibrillation was induced with a 3-second100 mA transthoracic shock. After 7 minutes,animals were randomized to receive continu-ous CPR with a custom robotic device using1 of 6 pre-programmed, 2-phase CPR schemesthat varied 1 parameter in 5 x 1-minute intervalsper phase while holding the other 2 parametersfixed. Frequency (AMSA) and slope-based (MS)quantitative ECG characteristics of artifact-filtered ECG were calculated from 3-secondsegments at the end of each 1-minute inter-val and compared between rate, depth and DCschemes, as well as experimental phases. Cor-relations between CPR parameter settings andECG characteristics were calculated. Results:Compression rate showed a low-to-moderatecorrelation (0.454) with change in MS in PhaseI, however neither DC nor depth showed a cor-relation with either AMSA or MS. In ANOVAmodels, MS differed between CPR groups at theend of Phase I (p = 0.046) but not AMSA, sug-gesting limited response of quantitative ECGmeasures after extended time intervals. Con-clusions: In this study only chest compressionrate in early phase CPR appeared to be relatedto quantitative characteristics of the VF ECG.

75. Variation in the Characteristics ofPatients with Acute Stroke Arriving byEMS Versus Those Arriving by PrivateVehicle

Robert O’Connor, Karen Braden, JosephCarrera, Nicole Chiota-McCollum, Elizabeth

Hundt, George Lindbeck, Karen Johnston,University of Virginia School of Medicine Cate-gory of Submission: Medical

Background: We conducted this study to iden-tify differences between patients arriving byEMS versus those arriving by private vehiclewith acute ischemic and hemorrhagic stroke.Determination of these differences may allowfor refinement of public education on thetimely treatment of acute stroke. Methods: Thisstudy was conducted at an academic medi-cal center that is an accredited comprehen-sive stroke center. Consecutive patients withacute stroke were enrolled between January2015 and May 2017, and were categorized bymode of arrival (EMS vs. private vehicle). Thetype of stroke (hemorrhagic vs. ischemic) wasidentified and the NIHSS measured in the EDfor all stroke patients, with the ICH Scoreand Hunt & Hess Score determined for ICHand SAH respectively. Age, gender, PMH ofstroke, and “time last known well” were iden-tified. Statistical analysis was performed usingthe Yates corrected Chi-Square, Mann-Whitney,and Kruskal-Wallis tests. Results: A total of 935patients were enrolled with 716 (77%) arrivingby EMS and 219 (23%) arriving by private vehi-cle. Of these, 636 (68%) had ischemic strokes,190 (21%) had ICH, 92 (10%) had SAH, and17 (2%) were not classified. Ac greater propor-tion of ICH (93%) and SAH (93%) patients thanischemic stroke patients (69%) arrived by EMS(p < 0.001). Patients arriving by EMS had sig-nificantly higher NIHSS (9.2 vs. 2.7, p < 0.001),ICH scores (1.7 vs. 0.3; p < 0.001), and Hunt& Hess scores (2.8 vs. 2.0) than those arrive bycar. The “last known normal” time was signifi-cantly lower for the EMS arrival group (mean= 547 minutes; median = 211 minutes) thanthe private vehicle group (mean = 1,407 min-utes; median = 715 minutes; p < 0.001). Demo-graphic data and prior history of stroke weresimilar based on mode of arrival. Conclusions:Stroke patients arriving by EMS have signifi-cantly higher NIHSS, ICH score, and Hunt &Hess score and significantly shorter time from“last known well” than those arriving by car.Because a significant proportion of ischemicstroke patients arrive by car, targeted publiceducation efforts should focus on identificationof stroke patients with longer symptom dura-tion and those with lower NIHSS.

76. Carotid Blood Flow Is Dependent onRate And Duty Cycle during CPR Cardiac.

Joshua Lampe, Karen Moodie, Jeffrey Gould,Christopher Kaufman, Norman Paradis, Fein-stein Institute for Medical Research Category ofSubmission: Cardiac

Background: We have previously presenteddata that blood flow generated by piston-typemechanical chest compressions (CC) is sensi-tive to changes in the inter-compression pausetime, which changes both chest compressionrate and duty cycle. We sought to clarify thedependence of CC generated blood flow onchanges in CC rate and duty cycle during pis-ton type CPR. Hypothesis: We hypothesizedthat the observed dependence of CC generatedblood flow on changes in intra-compressionpause time is due to the change in CC dutycycle. Methods: CPR was performed on fivedomestic swine (∼30 kg) using standard phys-iological monitoring. Blood flow was mea-sured by Doppler in the right common carotidartery. Ventricular fibrillation (VF) was electri-cally induced. CC were started after 5 minutesof untreated VF. CC were delivered at a rate of125 or 50 compressions per minute (cpm) with aduty cycle of 45% or 27% for each rate, and at adepth of 2” for a total of 6 minutes after 2 min-utes of “break-in” CPR (increased depth from1 inch to 2 inches). CC rate or duty cycle were



changed every 1.5 minutes. Results: At a rate of125 CPM, CC delivered at a duty cycle of 45%generated roughly twice the carotid blood flowin L/min of CC delivered at a duty cycle of 27%(0.157 ± 0.086 L/min vs. 0.075 ± 0.04 L/min,respectively). However at a rate of 50 CPM,blood flow was not dependent on duty cycle(45%: 0.045 ± 0.015, 27%: 0.037 ± 0.015). Thisrelationship appeared to be conserved whenblood flow was compared at the level of L percompression. Conclusions: The results of theseexperiments suggest that carotid blood flow isdependent on both rate and duty cycle. Thesedata suggest that the dependence of CC gen-erated blood flow on intra-compression pausetime cannot be assigned to either the change inrate or duty cycle, but is a combination of botheffects. These data highlight possible mechanis-tic differences between piston and vest CPR.

77. Rearrest Incidence and Post-ROSCRhythms after Prehospital Return ofSpontaneous Circulation inOut-of-Hospital Cardiac Arrest

Amber Rice, Joshua Gaither, Daniel Spaite,Vatsal Chikani, Sean Wentworth, TylerVadeboncoeur, Taylor George, Terry Mullins,Bentley Bobrow, University of ArizonaCategory of Submission: Cardiac

Background: Limited out-of-hospital cardiacarrest (OHCA) studies have found that rearrestafter return of spontaneous circulation (ROSC)is both common and independently associ-ated with lower survival. To better understandprehospital rearrest after ROSC, we sought todescribe rearrest cardiac rhythms for adultswith OHCA of presumed cardiac etiology in anexpanded and more recent sample of OHCAs.Methods: Cases were identified from Septem-ber 2008 to December 2015 from three EMS sys-tems in Arizona. Minute-by-minute post-ROSCand rearrest rhythms were grouped into Utsteincategories by two emergency medicine trainedphysicians after analysis of continuous defibril-lator ECG data (E Series, ZOLL Medical). Rear-rest was defined as 1 minute of lethal arrhyth-mia or crew restarting CPR for any lengthof time, indicating loss of pulses. Descrip-tive statistics were used to describe the dis-tribution of post-ROSC and rearrest rhythms.Results: Of 1,603 adult OHCA patients, therewere 409 cases of ROSC (25.5%) and 350 wereincluded in this analysis. Cases were excludedif age <18 (2), non-cardiac etiology (34), ROSCwas not achieved prior to ED arrival (1,194)or adequate electrocardiograph (ECG) rhythmrecordings were not available (23). There was atotal of 4,009 minutes of ROSC (not includingrearrest) with 7 distinct post-ROSC rhythms.Sinus rhythms predominated after achievingROSC, with sinus tachycardia representing thegreatest percentage (52.15%) of all rhythms.A smaller percentage of minutes were seenof sinus rhythm (21.14%), sinus bradycardia(5.00%), V-tach (4.41%), idioventricular (8.91%),atrial fibrillation/flutter (1.88%), and junctionalrhythms (1.31%). Almost half of ROSC patientsin this sample (45%) sustained at least oneepisode of rearrest and 22 patients (7%) sus-tained multiple rearrests. The most commonrearrest rhythms in this sample were pulselesselectrical activity (62.3%) and VT/VF (32.6%).Conclusions: This study demonstrates thatrearrest is common after ROSC in cases of pre-hospital OHCA. In this analysis, a wide varietyof both post-ROSC and rearrest rhythms wereobserved. This information helps prepare EMSrescuers for rearrest and provides the poten-tial for targeted interventions to prevent OHCArearrest.

78. Effect of Early Detection by Dispatcheron Survival Outcomes afterOut-of-Hospital Cardiac Arrest

Seo Young Ko, Sang Do Shin, Kyoung JunSong, Ki Jeong Hong, Young Sun Ro, So YeonKong, Tae Han Kim, Seoul National UniversityHospital, Seoul, Republic of Korea Category ofSubmission: Cardiac

Background: Dispatcher-assisted cardiopul-monary resuscitation(DA-CPR) is an importantintervention to improve outcomes of out-of-hospital cardiac arrest. We studied theassociation between the time to detect cardiacarrest by dispatcher and outcomes in out-of-hospital cardiac arrest (OHCA). Methods: Weconducted a cross-sectional study. All adultOHCAs of presumed cardiac etiology andbystander witnessed between 2013 and 2015were analyzed. The main exposure of interestwas time from EMS call to detection of cardiacarrest by dispatcher. Patients with unknowntime to detection by dispatcher or extremelylonger detection time (>20 minutes), andunknown outcomes were excluded. Time todetection of cardiac arrest by dispatcher wasclassified into the early (0–90 seconds), middle(91–180 seconds), and late (181–1,200 seconds)groups. The primary outcome was survivalto discharge and secondary outcome wasgood neurological recovery. Multivariablelogistic regression analysis was performed,adjusting for patient, arrest, environmental,and dispatcher factors. Results: Of 83,083OHCAs, 6,539 (7.9%) patients were instructedDA-CPR between 2013 and 2015. A Total of6,383 (7.7%) patients were enrolled, excludingcases who did not receive bystander CPR. Therates of DA-CPR performed were 28.7%, 43.0%,and 28.3% in early, middle, and late detec-tion groups, respectively. Overall, survival todischarge occurred in 635 (9.9%) OHCAs andgood neurological outcome was observed in441 (6.9%) patients. After adjusting for potentialconfounders, longer time to recognize cardiacarrest was associated with decreased odds ofsurvival to discharge for both middle (AOR0.74, 95%CI 0.59–0.91) and late groups (AOR0.75, 95%CI 0.59–0.94) compared with earlygroup. There was no significant associationbetween recognition time and good neurolog-ical outcome [Middle vs Early AOR(95% CI):0.81 (0.63–1.04), Late vs Early AOR (95%): 0.79(0.60–.03), Late vs Middle AOR (95% Cl): 0.98(0.76–1.26)]. Metropolitan status was significanteffect modifier (p < 0.001). In non-metropolitanareas, compared to the early group, AORs (95%Cl) for survival to discharge were 0.65 (0.49–0.85) in the middle group, 0.68 (0.51–0.90) inthe late group. In metropolitan areas, there wasno significant association between recognitiontime and survival to discharge [Middle vs EarlyAOR (95%CI): 0.91(0.64–1.30), Late vs EarlyAOR(95%Cl): 0.88(0.60–1.29)]. Conclusions:The shorter duration from the EMS call to recog-nition of cardiac arrest by dispatcher was asso-ciated with favorable outcomes after OHCA.

79. Impact of Real Time Chest CompressionFeedback Increases with Application ofthe 2015 Guidelines

Kenan Kunstal, Tifany Hoyne, Sara Wat-tenbarger, Stacie McCauley, Laurel Linder,Daniel Davis, ZOLL Medical Category ofSubmission: Cardiac

Background: Cardiac arrest survival is depen-dent upon chest compression quality. Targetparameters for compression depth and ratebecame more specific from the 2010 Guide-lines [�2 inches, 80–120/min] to the 2015Guidelines [2.0–2.5 inches, 100–120/min].Real-time audiovisual feedback (RTAVF) mayimprove compression guideline adherence,but the impact of RTAVF with application ofmore specific targets is unknown. Hypothesis:Dependence on RTAVF to achieve compressionguideline adherence will increase with appli-

cation of the more specific 2015 Guidelines.Methods: Data were collected as part of abenchmarking program conducted at multipleU.S. hospitals. Compression rate and depthwere recorded using standard compressionmannequins and RTAVF defibrillators (RSeries, ZOLL Medical). The program includedsubjects enrolled before (n = 756) and after (n= 995) introduction of the 2015 Guidelines,with target compression parameters modifiedaccordingly. At baseline subjects performed2 min of continuous compressions with RTAVFfeedback disabled. After a brief RTAVF orien-tation, subjects repeated 2 min of continuouscompressions with feedback enabled. The2010 Guidelines cohort and 2015 Guidelinescohort were compared with regard to the per-centage of compressions meeting appropriaterate/depth targets with and without use ofRTAVF. Results: An increase in compressionguideline adherence was observed with useof RTAVF for both the 2010 Guidelines cohort[60.3% to 96.0%, OR 15.9 (10.8–23.6), p <0.01] and the 2015 Guidelines cohort [16.7%to 95.0%, OR 94.4 (67.9–131.2), p < 0.01].The proportion of subjects requiring RTAVFto achieve adherence increased from the 2010Guidelines cohort to the 2015 Guidelines cohort[36.1% vs. 79.3%, OR 6.8 (5.5–8.4, p < 0.01), p< 0.01]. There were no statistically significantdifferences between the 2010 Guidelines cohortand the 2015 Guidelines cohort with regardto the proportion of subjects that could not becorrected [3.6% vs. 4.0%, OR 1.1 (0.7–1.9), p =0.63] or became nonadherent [0.4% vs. 1.0%,OR 2.6 (0.7–9.3), p = 0.16] with RTAVF. Conclu-sions: The use of RTAVF increases adherence tochest compression guidelines, particularly withapplication of the narrower 2015 Guidelinestargets for compression depth and rate.

80. Direct Transport to ComprehensiveStroke Center May Not ExpediteReperfusion of Large Vessel OcclusionStroke

Ali Shams, Chris Kanaan, RebbecaGrysiewicz, Chris Kazmierczak, LauraSteucher, Robert Swor, Beaumont Health Cat-egory of Submission: Student, Resident,Fellow

Background: A body of knowledge has evolvedthat has demonstrated improved survival andfunctional outcome from LVO strokes withtimely mechanical endovascular therapy. Todecrease time to care, EMS policy makers havebegun to develop methods to identify andtriage EMS LVO stroke patients directly to com-prehensive stroke centers (CSC). Our objectivewas to assess whether time to definitive carefor LVO stroke patients is decreased in patientswho present directly to a CSC compared topatients who are transferred from a primarystroke center. Methods: We performed a cohortstudy of patients admitted to a single suburbanCSC (July 2015 –December 2016) with a diagno-sis of LVO stroke. Patients presented directly tothe CSC, or were transferred by air or groundfrom a primary stroke center. Time intervalsfrom arrival at either first hospital or CSCto interventional skin puncture (access) andreperfusion at the CSC were recorded. Trans-fer distance was calculated using Google Maps.Because we sought to assess impact of triagewithin a regional EMS system, we includedpatients transferred within a 15-mile radius.Non parametric statistics were used for compar-isons. Median and range are reported. Results:We had a total of 62 cases admitted to ourCSC, with 54 transported within 15 miles. Ofthese, 25 patients were direct transports (15via EMS and 10 via private car) and 29 weretransferred from 7 hospitals. As expected, trans-ferred patients had shorter times from CSCarrival to access and reperfusion [median, 30.5



(6–216) vs. 156 (30–248), p < 0.001, and 69 (25–288) vs. 209 (99–315), p < 0.001], respectively.When comparing first hospital arrival to out-comes, direct transport patients had a smalldecrease in time to access [156 (30–248) vs. 171(115–384), p = 0.03] and no difference to reper-fusion [208.5 (25–288) vs. 209 (142–412), p = 1.0].Conclusions: Using this small pilot data set,direct transport of LVO patients to a CSC hada minimally shorter time to access and no dif-ference in time to reperfusion compared tothose transferred. EMS systems need to criti-cally assess the benefit of direct transport vsearly transfer as a component of strategies tooptimize care for LVO stroke patients.

81. Benchmarking EMS Compass CardiacPerformance Measures Using a LargeNational Dataset

Jeffrey Jarvis, Dustin Barton, Lauren Sager,Nick Nuddel, Williamson County EMS Cat-egory of Submission: Operations, Quality,Safety Systems, Disaster, Disaster

Background: Early defibrillation of shockablecardiac arrests, aspirin and 12 lead ECG, and acombination of nitroglycerin and non-invasivepressure ventilation (NIPPV) in acute decom-pensate heart failure has been shown to pro-vide meaningful clinical benefit. There has notyet been work done to provide benchmarkson these measures based on large nationaldatasets. We aim to describe national perfor-mance on these measures. Methods: Usinga 6 1/2-year convenience sample of recordsfrom 9-4-1 consenting EMS agencies usingESO Solutions electronic health record (EHR),we calculated compliance with the follow-ing performance measures: the average timefrom dispatch to first defibrillation in shock-able rhythms, the proportion of these providedwithin 5 minutes, the proportion of patientsover 35 with non-traumatic chest pain whoreceived both aspirin and a 12 lead ECG, andthe proportion of patients with acute decom-pensated heart failure (ADHF) as defined bySBP > 200 and either a RR > 30 or an SpO2< 90 who received both NTG and NIPPV. Fortimes, we provide the average, median andinterquartile rank. For proportions, we also cal-culated the 95% confidence interval. Results: Of11,144 cardiac arrests with an initial shockablerhythm, 1,630 or 14.6% (14.0–15.3%) were defib-rillated within 5 minutes. The average time tofirst shock was 13.65 min, IQR 9.0(6.4, 13.2).There were 533,127 patients over 35 with non-traumatic chest pain. Of these, 199,123 or 37.4%(37.2–37.5%) received both aspirin and a 12 leadECG. There were 2,612 patients with ADHF and2,100 or 80.4% (78.9–81.9%) of these receivedboth NTG and NIPPV. Conclusions: There wasa low rate of rapid defibrillation pointing outthe difficulties with achievement of this metricwithout non-EMS (public) support. There wasalso poor compliance with a chest pain bun-dle of aspirin and 12 lead ECG use. On theother hand, there was much better use of NTGand NIPPV in ADHF. These data provide base-line performance benchmarks for use in systemimprovement.

82. Characteristics Associated withSuccess on the National AEMTCertification Examination

Madison Rivard, Rebecca Cash, Remle Crowe,Jeremy Miller, Ashish Panchal, The NationalRegistry of Emergency Medical Technicians Cat-egory of Submission: Operations, Quality,Safety Systems, Disaster, Disaster

Background: Advanced emergency medicaltechnician (AEMT) certification, the providerlevel between emergency medical technician(EMT) and paramedic, was first issued on a

national level in 2011. While characteristics ofexamination success at other provider levelshave been explored, little is known regardingthe AEMT level. Our objective was to examinethe association between AEMT graduate char-acteristics and success on the National AEMTCognitive Examination. We hypothesizedthat prior EMT experience, program entranceexams, course-ending final exams, and examfee payor would be associated with success.Methods: We performed a cross-sectionalevaluation of all first-attempt National AEMTCertification cognitive examination resultsfrom October 2016 to April 2017. Upon com-pletion of the examination, a brief, voluntaryquestionnaire was administered assessinggraduates’ characteristics and experiences.Descriptive statistics were calculated, and theassociation between characteristics reportedby graduates and success on the exam wasassessed using univariable logistic regressionmodels (OR, 95%CI). Results: In the studyperiod, 3,835 AEMT graduates attempted thecognitive examination and 2,372 completedthe post-test questionnaire (response rate= 62%). Among those who completed thequestionnaire, 56% (n = 1323) were successfulon the first attempt. Compared to those withno EMT experience prior to enrollment in anAEMT program, those with one to five years ofexperience had greater odds of passing (1.37,1.10–1.71), while more than five years of EMTexperience was not significantly associatedwith examination success (1.09, 0.84–1.42).Attending an AEMT program that required anentrance exam was not associated with oddsof success (0.85, 0.69–1.05). However, respon-dents who were required by their programto complete a final course-ending cognitiveexamination exhibited higher odds of successcompared to those who did not (2.18, 1.78–2.65). Compared to those who paid for theirown exam, there was no difference in oddsof passing for those whose employers (1.21,0.99–1.49) or programs (1.16, 0.85–1.58) paidsome/all of the exam fees. Conclusions: PriorEMT experience and program course-endingcognitive examinations were significantlyassociated with increased odds of successon the National AEMT Examination. Futurework should examine the impact of programentry requirements and program curriculumcomposition on graduate performance.

83. Change in Quantitative VentricularFibrillation Over Bouts of ChestCompressions in CPR

Matthew Sundermann Sundermann, DavidSalcido, James Menegazzi, Department of Emer-gency Medicine, University of Pittsburgh School ofMedicine Category of Submission: Student,Resident, Fellow

Background: Chest compressions (CC) givenduring cardiac arrest generate blood flow to thebrain and other vital organs, but the effect of CCis dependent on their performance characteris-tics. Quantitative ECG (QECG) features of theventricular fibrillation (VF) waveform correlatewith myocardial perfusion levels during car-diac arrest and therefore may be a good qualitymetric. We hypothesized that there would be anassociation between change in QECG measuresand CC characteristics. Methods: CC processand associated continuous prehospital ECGdata were retrospectively extracted from defib-rillator downloads obtained from the continu-ous chest compression (CCC) trial of the Resus-citation Outcomes Consortium (ROC). Caseswere included if they had at least one defibril-lator file with a bout of CC bounded by ana-lyzable ECG signal segments, and amountedto 25,210 bout-gap intervals spanning 1,099unique cases. For each bout, the QECG mea-

sures AMSA, MS, LAC, and DFA were calcu-lated for the starting and ending ECG segmentsaround the bout, and CPR performance met-rics were calculated for the intervening bout ofCC. CC process metrics included rate, depth,duty cycle, fraction, bout duration, dosed rate,dosed depth, and dosed duty cycle. We thenanalyzed the relationship between CC metricsand QECG by regressing the change in QECGmeasures from the start each bout to the endof each bout against the CC process parametersfor that bout in multivariable models includ-ing bout duration and patient characteristics.Results: CC rate was associated with change inQECG value and was significant for change inMS (t = 2.13, coefficient 8.92, p = .0330). Allother associations between chest compressionparameters and dQECG were not significant.Conclusions: These results suggest a limitedrelationship between CC process metrics andQECG measures during resuscitation of out-of-hospital cardiac arrest.

84. The Utilization of a Province Wide EMSSystem by Children and Youth withMental Health Complaints

Aaron DeRosa, Michael Zhang, Judah Gold-stein, Carl Jarvis, Md Shamsuzzaman, Uni-versity of Prince Edward Island, Atlantic RegionalTraining Centre Category of Submission: Stu-dent, Resident, Fellow

Background: Children and youth EmergencyDepartment (ED) and hospital based mentalhealth(MH) service use is increasing in Canadaand the United States. This may extend to theEMS setting. Our objective was to describetrends and characteristics of EMS utilizationby children and youth with MH complaints.Methods: We conducted a retrospective pop-ulation based quantitative descriptive study,using secondary data from the provincial EMSdatabase. Patients 5 to 18 who utilized EMSfor MH related complaints between 2010 and2015, inclusive, were used in the analysis. Wedescribed prevalence, demographics, and oper-ational characteristics. MH calls were based onchief complaint or clinical impression relatingto MH and resemble the Canadian ED shortlist of Diagnosis under Mental and BehaviouralDisorders. Continuous and discrete variablesreported as n, mean, SD; Categorical as n, %.Results: Our electronic query retrieved 16,169EMS responses for children and youth; of which2108(16%) were related to MH. The mean agewas 16.26(SD1.699) and most MH calls werefemale (n = 1238, 59%). There was a 27%increase in total MH calls over the 6 yearstudy period compared to a 9% increase inall EMS calls in the same age group. Femaleshad the largest increase (47%) in MH relatedcomplaints over the study period. The major-ity of patients were single users (n = 1436,68%), whereas, 180 repeat users accounted for503(24%) responses, ranging from 2–13 inci-dents over the study period. Most patients weretransported (n = 1920; 91%). The two mostcommon conditions addressed by paramedicswere overdose/poisoning (n = 1747; 83%), anddepressed/suicide (n = 250; 12%). Anxiety (n =257; 35%) was the most prevalent charted co-morbidity, followed by Attention-Deficit Dis-order/ Hyperactivity Disorder (n = 207,28%).When categorizing patients over a calendaryear 1635 patients were low users (1 call peryear), 108 patients were medium users (2–4calls per year), 8 patients were high users (5–14 calls per year). Conclusions: We observedan increasing trend in MH related EMS ser-vice use by children and youth. The majorityof patients are transported by paramedics tothe ED. This trend should be considered whendeveloping EMS policies, programs, and train-ing for paramedics.



85. Push Dose Epinephrine Use in CriticalCare Transport

Francis Guyette, Gabriela Galli, NealMcQuaid, Jonathan Elmer, Christian Martin-Gill, Department of Emergency Medicine, Univer-sity of Pittsburgh School of Medicine Category ofSubmission: Student, Resident, Fellow

Background: The use of push dose epinephrine(PDE) is becoming increasingly common inthe treatment of profound hypotension in theprehospital setting. However, no quantitativeresearch has been done to analyze the patientpopulations receiving this treatment. We aim todescribe the population of patients treated withPDE as compared to hypotensive patients nottreated with PDE. Methods: We performed aretrospective cohort study to describe the useof PDE in a critical care transport system. Weevaluate the use of PDE for management ofprehospital hypotension from January 2015 toApril 2017. We reviewed prehospital and in-hospital medical records for patients treatedand transported by a multi-state air medicalservice that incorporated PDE into its proto-cols (epinephrine 100 mcg IV/IO every 2 min-utes for SBP < 70 mmHg or peri-arrest state).Patients were selected if they were hypotensiveand met inclusion criteria for PDE use in thecurrent protocol. We compared pretreatmentcharacteristics and vital signs for patients fol-lowing an index event (SBP < 70). We utilizednon parametric (rank sum tests) and chi-squareto identify differences between the cohorts.Results: 1862 eligible (SBP < 70) cases wereidentified, PDE was administered to 23%. Casesvs. controls differed by age, PDE median age65 (IQR 55–76) No PDE 61 (IQR 50–72), butnot gender or race. Patients receiving PDE weremost commonly treated for post arrest, cardio-genic shock, trauma and sepsis. Patients receiv-ing PDE were also more likely to be intubated(PDE 32.4%, No PDE 14.3% Pr0.00) and vaso-pressor dependent (PDE 32.8%, No PDE 15.5%Pr0.00) prior to flight crew arrival. Patients alsodiffered with respect to Lactate level (PDE 8.2(IQR 4.5,9) vs. No PDE 3.7 (IQR2.3,7) and pre-treatment crystalloid (PDE 1000 IQR 500,2600)vs. No PDE (PDE 1000 IQR 200,2000). Otherpretreatment variables (HR, SpO2, RR) did notdiffer. Conclusions: Prehospital administrationof PDE in our system is administered in onlya fraction of patients meeting protocol crite-ria. PDE administration is associated with intu-bation, vasopressor use, increased lactate, andcrystalloid compared to patients not receivingPDE possibly indicating a selection or indica-tion bias.

86. Adherence to Recommendations forPrehospital Cardiac Arrest Care Acrossan EMS System of Care: How Well Are WeImplementing Guidelines?

Jonathan Kamrud, Lori Boland, AndrewStevens, Jessica Jeruzal, Charles Lick, AllinaHealth Emergency Medical Services Category ofSubmission: Cardiac

Background: To evaluate adherence to Amer-ican Heart Association (AHA) recommenda-tions for optimal care for out-of-hospital car-diac arrest (OHCA) across the spectrum ofprehospital care by analyzing care renderedby bystanders, dispatchers, first responders(FR), and emergency medical services (EMS)providers within a system of care. Methods:A total of 294 OHCA events treated by a sin-gle ambulance service in Minnesota in 2014–2015 occurred before ambulance arrival in adultpatients who sustained non-traumatic arrest,and had complete data available for bystander,dispatch, first responder, and EMS care ele-ments. An adherence index (AI; range = 0 to6) was calculated based on successful delivery

of six care elements aligned with AHA recom-mendations: dispatcher provided instructionsfor CPR when possible, bystander or FR initi-ated chest compressions (pre-ambulance CPR),bystander or FR placed an AED (pre-ambulanceAED), compression fraction during EMS CPR> 0.8, compression rate during EMS CPR of100–120/minute, and number of pauses >10 secin duration during EMS CPR was < 3. Onlythe first 10 minutes of compressions were con-sidered for EMS CPR criteria. Data sourcesincluded audio recordings of dispatch calls,the Cardiac Arrest Registry to Enhance Sur-vival (CARES) registry data, and transthoracicimpedance data tracings. Results: Adherenceto individual guidelines was generally high:dispatcher instructions for CPR = 100%, pre-ambulance CPR = 93%, pre-ambulance AED =72%, compression fraction = 84%, compressionrate = 91%, and number of pauses >10 sec < 3= 81%. Care was delivered in accordance withall six criteria (AI = 6) in 52% of events (n = 153)and the AI was � 5 in 78% of events (n = 228).The number of events with AI � 5 increasedfrom 70% among 2014 cases to 83% among 2015cases (p = 0.009). Conclusions: Adherence tothe guidelines for optimal prehospital OHCAmanagement that were studied is very high inthis system of care and appears to be increas-ing. Identified opportunities for improvementinclude increasing pre-ambulance AED use andreducing pauses during EMS CPR.

87. Recognition of Out-of-HospitalCardiac Arrest during Emergency Calls byCommunity Level Public Awareness ofCardiopulmonary Resuscitation: AMulti-Level Analysis

Sun Young Lee, Young Sun Ro, Sang Do Shin,Kyoung Jun Song, Ki Jeong Hong, SoyeonKong, Tae Han Kim, Seoul National UniversityHospital Category of Submission: Cardiac

Background: In dispatcher-assisted car-diopulmonary resuscitation (CPR) program,dispatchers’ recognition of out-of-hospitalcardiac arrest (OHCA) is the first step to initiatebystander CPR. This study aimed to investigatewhether the community CPR awareness isassociated with recognition of arrest, provi-sion of CPR instruction, and bystander CPR.Methods: A nationwide population-basedobservational study was conducted with adultOHCA patients with cardiac etiology between2013 and 2015. Exposure was communitylevel awareness of CPR using the nationalKorean Community Health Survey databasecategorized into quartile (the lowest, lower,higher, and the highest) groups. Endpointswere recognition of arrest, provision of CPRinstruction, and bystander CPR. Multi-levellogistic regression analysis was performedfor study outcomes. Adjusted odds ratios(AORs) per 10% increment in community CPRawareness were calculated adjusting for poten-tial confounders. Results: A total of 43,875OHCAs were included in the final analysis.Of those cases, 20,182 cases (46.0%) were rec-ognized during the emergency calls and CPRinstructions were given in 17,804 (40.6%) cases.Compared with the lowest CPR awarenesscommunities, AORs (95% CIs) for arrest recog-nition were 1.06 (0.96 to 1.17) in lower, 1.12(1.02 to 1.23) in higher, and 1.19 (0.99 to 1.22)in the highest CPR awareness communities.For CPR instruction, AORs (95% CIs) were1.13 (1.00 to 1.27) in lower, 1.25 (1.08 to 1.44)in higher, and 1.25 (1.09 to 1.44) in the highestCPR awareness communities. Bystander CPRwas done in 21,973 cases (50.1%) and thecommunities with the highest CPR awarenessshowed higher bystander CPR rate (51.4%)than other communities (higher 50.9%, lower48.3%, and the lowest 46.4%) (p < 0.01). AORs

(95% CIs) per 10% increment in CPR awarenesswere 1.05 (1.00 to 1.11) for arrest recognition,1.11 (1.06 to 1.17) for CPR instruction, and 1.07(1.03 to 1.11) for bystander CPR. Conclusions:Although the dispatcher’s recognition rateof OHCA was not different according to thecommunity CPR awareness level, dispatchersprovided more CPR instruction in communitieswith higher CPR awareness level. Finally, morebystander CPR was provided to the patients inhigher CPR awareness communities.

88. Effect of Text Message Alert System forTrained Citizens on BystanderCardiopulmonary Resuscitation andSurvival to Discharge in a MetropolitanCity: A Before-After Population-BasedStudy

Sun Young Lee, Sang Do Shin, Kyoung JunSong, Ki Jeong Hong, Young Sun Ro, SoyeonKong, Tae Han Kim, Seoul National UniversityHospital Category of Submission: Cardiac

Background: Bystander cardiopulmonaryresuscitation (CPR) is a key factor to improvesurvival outcomes in out-of-hospital cardiacarrest (OHCA) patients. A text message (TM)alert system for trained citizens was imple-mented to increase bystander CPR in thecommunity. This study aimed to determinethe effects of a TM alert system on bystanderCPR rate and survival outcomes after OHCA.Methods: A before-after population basedstudy was conducted with resuscitationattempted OHCAs between 2014 and 2015in the study districts of Seoul, South Korea.Seoul implemented a TM-alert system as acommunity intervention in May, 2015. Theintervention group was defined as OHCA casesthat occurred from May to December in 2015,and the historical control group was definedfrom the same period (May to December) in2014. Endpoints were bystander CPR rate andsurvival to discharge. Multivariable logisticregression analysis was used to evaluate theeffect of TM alert intervention compared withhistorical control group. Results: A total of1,124 OHCAs were analyzed, with 560 OHCAcases in the intervention group and 564 OHCAcases in the historical control group. BystanderCPR was performed in 141 patients (25.1%) in2014 and 119 patients (21.3%) in 2015 (p-value= 0.14). Survival to discharge was observedin 31 patients (5.5%) in 2014 and 56 patients(10.0%) in 2015 (p-value = 0.57). The adjustedodds ratios (95% CI) of bystander CPR andsurvival to discharge for intervention groupcompared to control group were 0.80 (0.60to 1.06) and 0.94 (0.57 to 1.54), respectively.Conclusions: The text message alert system forCPR trained citizens was not associated witha significant increase in bystander CPR andsurvival to discharge rates.

89. Comparison of Manual vs. MechanicalChest Compression Quality duringPrehospital Cardiac Resuscitation

Joshua Gaither, Amber Rice, Chengcheng Hu,Robyn McDannold, Margaret Mullins, DanielSpaite, Tyler Vadeboncoeur, Taylor George,Terry Mullins, Bentley Bobrow, University ofArizona Category of Submission: Cardiac

Background: Cardiopulmonary resuscitation(CPR) quality is strongly linked to outcomes fol-lowing out-of-hospital cardiac arrest (OHCA).Manual CPR quality varies and has risk toproviders. We hypothesized that use of amechanical CPR device might provide higherquality CPR than manual CPR during tech-nically challenging periods of OHCA resusci-tation, including the packaging, loading, andtransporting of patients. Methods: Cases ofOHCA at a single site from 10/2008–10/2016



were identified. Two CPR quality metrics,chest compression fraction (CCfr) and CC rate(CCra), were measured using accelerometer-based technology (E & X-Series), and com-pared between 3 groups: packaging (terminal5 minutes on scene), loading (terminal 3 min-utes on scene), and transport. Mechanical CPRwas performed using AutoPulse® (ZOLL Med-ical), while most cases of manual CPR wereperformed with real-time audiovisual feedback(Real CPR Help®). Manual CPR [metronomerate of 100 beats per minutes (bmp)] andmechanical CPR (set CCra of 80 bpm) werecompared by the median proportion of time inwhich CCra was within +/−5 bmp of the targetrange (pCCra) and the mean CCfr is reportedusing the Wilcoxon rank-sum test. Results: 357cases were reviewed and 239 excluded: no ageor age <18 years (6), medical or unknownlocation (31), non-cardiac etiology (87), dataunavailable (115), leaving 118 included. No sig-nificant difference in CCfr was noted betweenthe two groups during transport (p = 0.47). Incases with mechanical CPR, CCfr was higherduring packaging 85.0 vs. 74.5 (p = 0.0043) andloading 86.0 vs. 72.2 (p = 0.001) than in caseswith manual CPR. With mechanical CPR, CCrawas more frequently within the target rangeduring all study periods 0.4 versus 0.8 (p =0.001), 0.3 vs. 1 (p = 0.0021), and 0.5 vs. 0.8 (p= 0.0002). Conclusions: In adults with OHCA,use of a mechanical CPR device was associatedwith higher CCfr during patient packaging andloading and a higher proportion of time withinthe target CCra rate during all time periods.Use of mechanical CPR may improve CPR qual-ity without exposing providers to the risks ofperforming manual CPR during the packaging,loading, and transport of OHCA patients.

90. Is Prehospital Epinephrine UsedAppropriately in Pediatric Anaphylaxis?

Joslyn Joseph, Brian Walsh, David Feldman,Morristown Medical Center Category of Submis-sion: Pediatric

Background: Anaphylaxis is an acute, life-threatening condition that requires immediaterecognition and treatment. The goal of therapyshould be early recognition and treatmentwith epinephrine to prevent progressionto life-threatening respiratory compromiseor cardiovascular collapse. More prehospitalproviders, parents, and school nurses, are beinginstructed in using epinephrine. We soughtto determine how often epinephrine is usedin children and, more importantly, how oftenit is administered correctly for anaphylaxis.Methods: Setting: A suburban two-tiered EMSsystem in which ALS units evaluate approx-imately 600 patients under age 13 per year.Patients: Children less than 13 years old overa 5-year period for whom ALS was dispatchedfor “Allergy/Anaphylaxis”. Protocol: Demo-graphics, history of present illness, vital signs,and interventions performed prior to EMSarrival and by EMS personnel were extractedusing chart review. The percentage of patientswith 95% confidence intervals (“CI”) who weregiven epinephrine prior to EMS arrival, byEMS, and overall were calculated. Anaphy-laxis was defined as acute cutaneous and/ormucosal involvement after antigen exposureplus one of the following: respiratory compro-mise, cardiovascular compromise, or persistentGI symptoms. Appropriate treatment wasdefined as epinephrine being administeredwhen the patient’s clinical syndrome met thedefinition of anaphylaxis, or being withheldwhen the clinical syndrome did not meet thedefinition. The percentage of patients whowere treated appropriately was then calculatedwith CI. Results: Out of 2,750 ALS calls forpatients under 13 years old, 287 (10.4%) were

for “Allergy/Anaphylaxis.” The average ageof patients was 6.5 years and 63% were male.59% (CI: 54–65) of these patients receivedepinephrine - 49% (CI: 44–55) prior to EMSarrival, and 10% (CI: 6–13) by ALS personnel.The percent of patients who received appro-priate treatment was 62% (CI: 56–66%). Of theinappropriate treatments, epinephrine wasgiven inappropriately 30% (CI: 24–35%) of thetime, and was withheld inappropriately 9%(CI: 5–12%) of the time. Conclusions: Despiteincreasing incidence and public awarenessof life-threatening allergic reactions, bothlaypeople and prehospital providers struggleto diagnose and treat anaphylaxis in pediatricpatients. More education is needed to recognizethis disease process and treat it appropriately.

91. Pediatric Out-of-Hospital CardiacArrest Outcomes before and afterImplementation of a StandardizedResuscitation Tool

Scott Alter, Lisa Clayton, Richard Paley,Richard Shih, Florida Atlantic UniversityCategory of Submission: Pediatric

Background: Pediatric out-of-hospital car-diac arrest (POHCA) occurs infrequently, yetrequires the same urgency as for adults. There-fore, it is imperative that prehospital providersare prepared to rapidly treat POHCA. To meetthis need, pediatric-specific tools have beendeveloped. This study compares POHCAoutcomes before and after implementationof an age-based resuscitation tool. Methods:Design: retrospective chart review. Setting:county-based ALS service with 87,000 calls peryear, covering a population of 635,000 over2,000 square miles. Subjects: patients <18 yearsold who sustained POHCA with resuscitationattempt without return of spontaneous cir-culation (ROSC) before EMS arrival betweenJanuary 1, 2012 and December 31, 2016. OnJanuary 1, 2014, a commercial tool for POHCA,consisting of age-based medication dosingprotocols, was implemented. Rates of ROSC,survival to hospital admission, and survivalto hospital discharge were calculated andcompared between the pre-implementationand post-implementation groups. Results: Atotal of 132 POHCA patients were identified,of whom 24 were excluded for having ROSCbefore EMS arrival. The remaining 108 patientshad average age of 1.61 years, with similarbaseline characteristics between groups. In thetwo years preceding the tool implementation(control group), there were 37 cardiac arrests.Of these, two had ROSC after EMS arrival andnone survived to hospital admission. In thethree years after implementation (experimentalgroup), there were 71 cardiac arrests. Of these,13 had ROSC after EMS arrival. All patientswith ROSC survived to hospital admission and3 survived to hospital discharge. Between thecontrol and experimental groups, there was a13% difference in ROSC after EMS arrival (5%vs. 18%; 95% CI: −0.01–0.24), 18% differencein hospital admission (0% vs. 18%; 95% CI:0.06–0.29), and 4% difference in overall survivalto discharge (0% vs. 4%; 95% CI: −0.06–0.12).Conclusions: After implementation of an age-based resuscitation tool, there was a statisticallysignificant increase in POHCA survival to hos-pital admission. ROSC rate obtained after EMSarrival and survival to hospital discharge alsoincreased, though failed statistical significance.Based on these results, EMS agencies may con-sider implementing an age-based resuscitationtool as part of a strategy to improve POHCAtreatment.

92. Comparison of CommercialTourniquets in a Pediatric Trauma PatientModel

James Vretis, Center for Tactical MedicineCategory of Submission: Pediatric

Background: Young children and adoles-cents are frequently injured in peacetime andwartime. Reviews of trauma registries at U.S.military medical facilities during the Iraq andAfghanistan conflicts show as the age of a childa child decreases the injury severity and mor-tality increases. Tourniquet use for control ofextremity hemorrhage in adult trauma patientsis associated with increased survival with onlyminimal tourniquet associated morbidity. Useof commercial tourniquets on pediatric patientstreated at US military facilities shows survivalbenefits similar to those seen in the adultpopulation. Hypothesis: We hypothesized thatthere wound be differences in the efficacy ofcommercial tourniquets designed for adultswhen applied to pediatric patients of differentages. Methods: The institutional Ethics ReviewBoard approved the study. The study wasa prospective and non-blinded test of ninecommercial tourniquets on a pediatric armhemorrhage test model using six sized man-nequins to simulate pediatric arms. The StretchWrap And Tuck (SWAT), TacMed K9 (TMK9),and Rapid Application Tourniquet System(RATS) tourniquets apply compressive forcesby the elastic recoil action of the tourniquetstrap. The Combat Application Tourniquet(CAT), Sam XT (SAMXT), Tactical Mechani-cal Tourniquet (TMT), and the SOF TacticalTourniquet – Wide (SOFTTW) use a windlassto increase circumferential compression bydecreasing strap length. The Child RatchetingMedical Tourniquet (CRMT) uses a ratchet andladder mechanism for circumferential compres-sion. The Mechanical Advantage Tourniquet(MAT) has a turnkey apparatus mounted ona fixed length C-shaped housing that pulls aportion of the retaining strap into the housingas a mechanism to increase circumferentialpressure. Results: The SWAT, TMK9 and RATSwere successful stopping the flow of water onall sized mannequins. The CRMT was the onlymechanical advantage tourniquet that was suc-cessful in stopping fluid flow on all mannequinsizes. The TMT and SOFTTW started failing onmannequins with 6.35 cm diameters. The CAT,SAMXT, TMT, and SOFTTW all failed on the5.08 cm diameter mannequin. The MAT failedon the 7.62 and smaller diameter mannequin.Conclusions: We have shown that many com-mercially available tourniquets do not stopfluid flow in our pediatric arm hemorrhage testmodel.

93. Prehospital Blood PressureMeasurement in Major Traumatic BrainInjury: Concordance between EMSProvider Documentation andNon-invasive Monitor Data Tracking

Octavio Perez, Octavio Perez, Eric Helfenbein,Bruce Barnhart, Saeed Babaeizadeh, DawnJorgenson, Chengcheng Hu, Vatsal Chikani,Joshua Gaither, Samuel Keim, Duane Sherrill,Daniel Spaite, University of Arizona Categoryof Submission: Operations, Quality, SafetySystems, Disaster, Disaster

Background: Recent studies have shown thatthe lowest prehospital systolic blood pressure(SBP) is strongly associated with mortalityacross a remarkably wide range (far above90 mmHg) in traumatic brain injury (TBI).Furthermore, in TBI research, case ascertain-ment and risk-adjustment are highly depen-dent upon documentation of prehospital BP.Objective: To identify the concordance betweenthe lowest SBP documented by EMS person-nel in patient care records (PCR) and therecorded non-invasive monitor data in TBI.Methods: A subset of major TBI cases (moder-ate/severe; CDC Barell Matrix 1) in the EPIC



EMS TBI Study (NIH 1R01NS071049) were eval-uated (3/13–3/17). Cases from 6 EMS agenciesthat report continuous monitor data (PhilipsMRxTM) as part of EPIC were included. Allmonitor data available for this post-hoc reviewwere displayed and accessible to the providersduring EMS care. We compared the lowestPCR-documented SBP to the monitor-recordedvalue in each patient. Results: 132 cases wereincluded (median age: 52, 65% male). In 96cases (72.7%), the lowest PCR-documented SBPwas exactly concordant with the lowest mon-itor value. When concordance was defined bythe difference being �5 mmHg, 113 (85.6%)were concordant. Among the 16 patients withguideline-defined hypotension identified bythe monitor (<90 mmHg), only 11 (68.8%) weredocumented in the PCR. Conclusions: Signif-icant disparities were identified between thelowest monitor-recorded SBP and the PCR-documented value. Furthermore, PCRs failedto identify one third of monitor-documentedhypotension. This may be explained, in part,by ongoing care responsibilities and scene dis-tractions that may cause providers to miss BPreadings. Our findings identify a potential hid-den contributor to poor outcomes if hypoten-sion goes unrecognized, and untreated, ratherthan simply not being documented. Further-more, case ascertainment, confounding, andrisk-adjustment in TBI studies may be sub-stantially impacted. Whenever possible, qual-ity improvement and research projects shouldutilize monitor data to identify and evaluatehypotensive TBI patients. Future developmentof monitor-based real-time audiovisual feed-back technology might improve provider iden-tification of hypotension.

94. Evaluating the Gender Gap in EMTSand Paramedics Obtaining National EMSCertification from 2007 to 2016

William Krebs, Remle Crowe, Rebecca Cash,Madison Rivard, Ashley Larrimore, Chris-tine Hamilton, Ashish Panchal, Department ofEM, The Ohio State University Wexner MedicalCenter Category of Submission: Operations,Quality, Safety Systems, Disaster, Disaster

Background: With roots in battlefield medicineand the fire service, the EMS workforce hastraditionally been comprised of mostly maleproviders. As the EMS profession has evolvedin both prominence and function, it is unknownhow the gender composition of the workforcehas changed on a national level. The objectiveof our study was to describe the proportionof females who earned initial National EMSCertification at the EMT and paramedic levelsover a 10-year period (2007–2016). We hypoth-esized that the proportion of female EMTs andparamedics earning certification increased dur-ing this time. Methods: This was a longitudinalassessment of all EMTs and paramedics earn-ing initial National EMS Certification from 2007through 2016. There is no national databaseof all licensed EMS professionals, howeverNational EMS Certification is required to earninitial licensure at one or more provider lev-els in the majority of states. We assessed allEMS professionals who earned initial EMT orparamedic certification between January 1, 2007and December 31, 2016. Descriptive statisticswere calculated. A non-parametric test of trendwas used to assess for increasing or decreas-ing trends in the proportion of females earningcertification during the study period. Results:In 2007, a total of 28.7% of EMTs earning ini-tial certification were female compared to 34.8%in 2016, representing a percent change of 21.3%(p-trend<0.001). An increase was noted in 7 of9 year-to-year comparisons. However, the pro-portion of females earning initial paramediccertification was stagnant during the 10-year

period. While statistically significant (p-trend= 0.03), the overall increase was less than onepercentage point (21.4% in 2007 to 22.1% in2016). A change of less than 2% was noted in8 of 9 year-to-year comparisons. Conclusions:While the proportions of females earning ini-tial National EMS Certification increased forEMTs, the population earning paramedic certi-fication remained relatively stable over the ten-year period. Despite other health care fieldsclosing the gender gap, paramedic certificationhas not followed this trend. Future research isneeded to identify the underlying reasons andbarriers for the lack of change in the paramedicgender composition of those earning NationalEMS Certification.

95. Impact of Community Paramedic HomeVisits on CHF Patients: A Pre-PostAssessment of Heart Failure on Quality ofLife

Sandi Wewerka, Joseph Pasquarella, AnnMajerus, Aaron Burnett, Matthew Simpson,Paula Miller, Regions Hospital Category ofSubmission: Operations, Quality, Safety Sys-tems, Disaster, Disaster

Background: Effective management of conges-tive heart failure (CHF) often requires patientsto make significant lifestyle changes, whichmay be best managed in partnership in thepatient’s home. The objective of this study wasto evaluate the effectiveness of a fire-basedcommunity paramedic (CP) program on CHFmanagement in patients recently dischargedfrom the hospital using the Minnesota Livingwith Heart Failure® Questionnaire (MLHF).We hypothesize that CP visits will contributeto improvement in the patient’s quality of lifeas assessed by the MLHF. Methods: Patientswith a CHF-related hospitalization who pro-vided consent to participate in the CP pro-gram completed the MLHF prior to discharge.The CP program entailed weekly home vis-its from a CP. The MLHF is a validated ques-tionnaire that uses a Likert scale to measurethe effects of CHF symptoms, functional limi-tations and psychological distress. Each symp-tom is rated on a 0–5 scale, with a score of5 corresponding to the greatest detriment toquality of life (QOL). Total MLHF scores rangefrom 0–105. 4–6 weeks after discharge, patientsrepeated the MLHF. Pre/post survey scoreswere analyzed descriptively using means andstandard deviations. Scores were assessed withWilcoxon signed-rank tests in three dimensions:total score, emotional symptoms, and physi-cal symptoms. Results: Twenty-three patientscompleted the pre- and post-tests from March2015 to May 2017. The mean total scores onthe pre-assessment (score = 57.83, SD = 28.09)and post-assessment (score = 45.30, SD = 30.77)were significantly different (p = 0.022). Meanpre-score for physical assessment questions was25.78 (SD = 12.06) while on the post assessmentit was 21.22 (SD = 11.66). Mean of the emotionalscore on the pre-assessment was 12.17 (SD =8.55) while on the post assessment it was 9.96(SD = 8.84). Total scores were significantly dif-ferent between the pre and post assessments(p = 0.0216). Scores for the physical questionsof the assessment were significantly differentbetween the pre and post assessments (p =0.0218). The pre–post difference in emotionalscore was not different (p = 0.21). Conclusions:Using the MLHF, we found significant improve-ment in QOL of CHF patients who completedthe CP program. This study is limited by thesmall sample size but demonstrates encourag-ing improvements to this patient population.

96. Seatbelt Use by Ambulance Personnelin the Patient Compartment Is LowRegardless of Patient Presence, SeatingPosition, or Patient Acuity

Rebecca Cash, Evan Crowe, Remle Crowe,Madison Rivard, Anne Knorr, Ashish Pan-chal, Douglas Kupas, National Registry of Emer-gency Medical Technicians Category of Submis-sion: Operations, Quality, Safety Systems,Disaster, Disaster

Background: Recent crash testing showsEMS professionals are at high risk of injuryor death while riding unrestrained in anambulance, yet seatbelt use is reportedly low.Variation in seatbelt use based on seatinglocation and patient acuity is unknown. Ourobjectives were to describe the prevalence ofseatbelt use by seating location and identifyfactors associated with seatbelt practices. Wehypothesized that seatbelt use would be lowin the patient compartment regardless of pres-ence of a patient, seating position, or patientacuity. Methods: We analyzed a cross-sectionalelectronic questionnaire administered to arandom sample of nationally-certified EMSprofessionals. Respondents reported frequencyof seatbelt usage in the prior 12 months. Inclu-sion criteria consisted of practicing EMTs orhigher in non-military settings who work inambulances. We defined consistent seatbeltuse as reporting frequency of use >50% ofthe time in a seating location. Denominatorsreflect respondents reporting sitting in thespecific seat. Multivariable logistic regression(OR, 95%CI) using an investigator-controlledbackwards selection process was used toassess characteristics associated with wearinga seatbelt on the crew bench while transportingpatients. Results: A total of 1,431 responsesmet inclusion criteria (response rate = 11.4%).Most respondents wore a seatbelt while drivingthe ambulance (97%, n = 1,181/1,221). In thepatient compartment without a patient beingtransported, consistent seatbelt use was poorregardless of seat position (forward-facing seat:60% [n = 49/82], rear-facing airway/jumpseat: 59% [n = 670/1,136], crew bench: 36%[n = 362/997]). During patient transport,consistent seatbelt use on the crew bench wasreported at 23% with stable patients and 11%with critical patients. Factors associated withincreased odds of seatbelt use on the crewbench when transporting a critical patient(lowest seatbelt use) included having a com-pany policy for seatbelt use (6.25,4.06–9.60) andEMT provider level (2.39,1.52–3.78 [referent:AEMT/Paramedic]), controlling for years ofexperience. Conclusions: Seatbelt use by EMSpersonnel in the patient compartment was lowand varied by seat and patient acuity, with usehighest in forward-facing seats. Seatbelt usewas lowest in the patient compartment duringthe potentially more hazardous transport ofcritical patients. Future work should examineways to increase seatbelt use in the patientcompartment.

97. Feasibility of Manual ActiveCompression Decompression CPR in aThirty-Degree Head Up Position

Heather Ellis, David Chase, Ventura City FireDepartment Category of Submission: Cardiac

Background: Manual active compressiondecompression CPR (ACD CPR) with ITD(impedance threshold device) in supineposition has shown improved outcome inout-of-hospital cardiac arrest. Automated ACDCPR with ITD in a thirty-degree head upposition (HUP) has shown improved cerebralperfusion in porcine and human cadaver mod-els. There is controversy regarding the abilityto perform high quality manual ACD CPR inHUP. Hypothesis: High quality manual ACDCPR in HUP to specific standards is feasible.Methods: A recording simulation mannequinwas placed in HUP. After brief instructionand practice using the Zoll ResQCPRTM sys-



tem continuous ACD CPR was started by athree-member first response team. The com-pressor straddled the mannequin. After each200 compressions there was a break to switchcompressors. The CPR feedback from themannequin and the ResQCPRTM system wasrecorded and analyzed looking at depth, rate,and decompression negative pressure (>10kg). 80% beat-to-beat compliance for depth anddecompression and an average rate between75 and 85/minute was considered high qualityCPR. Zoll recommends a rate of 80/minute forthis system. After completion of 15–20 minutesof simulated manual ACD HUP CPR the teammembers were asked to complete a survey toassess the degree of fatigue and muscle strainthey experienced in comparison to standardCPR. Results: 5984 separate compressionswere recorded. Mean (SD; CI95) rate was 78.1(6.9; 75.6–80.6)/minute and mean depth was2.16 (0.07; 2.14–2.19) inches. 30 separate 200compression efforts were analyzed for beat-to-beat compliance for depth and decompression.Mean depth compliance was 78.6% (6.08%;75.8–81.3%). Mean decompression compliancewas 91.4% (1.1%; 88.0–94.8%). 10 of 10 surveyrespondents described manual ACD CPR HUPas more fatiguing than standard CPR and 9 of10 described muscle strain. Discussion: Beat-to-beat % depth compliance fell just short of thebenchmark set. All other defined benchmarkswere met. The authors anticipate that withmore instruction and practice the beat-to-beatdepth compliance of 80% would be achieved.Conclusions: High quality manual ACD HUPCPR can be done; however, it is more fatiguingand causes more muscle strain than standardCPR.

98. Television and Film Depict UnrealisticRates of Cardiac Arrest Survival

Johanna Innes, Brian Clemency, Maxwell Did-dams, Peter Natalzia, Deborah Waldrop, Uni-versity at Buffalo Category of Submission:Cardiac

Background: The media’s portrayal of cardiacarrest management and outcomes may shapepublic perception of a cardiac arrest victim’schance of survival. We sought to determinethe rates of cardiac arrest survival depicted intelevision and film. We hypothesize that thesurvival rates portrayed on television and inmovies were significantly higher than actualcardiac arrest survival rates. Methods: We con-ducted a meta-analysis of existing studies ofcardiac arrest resuscitations depicted on tele-vision and film. A PubMed search was con-ducted using the following search terms: “car-diopulmonary resuscitation and television,” or“resuscitation and television,” or “heart arrestand television.” Two reviewers independentlyreviewed all studies. Studies that included sur-vival data from in hospital and out of hospitalcardiac arrest patients depicted on television orin movies were included in the analysis. Subjectdemographics, rates of return of spontaneouscirculation (ROSC), and survival to dischargewere reviewed and compared to published datafrom the Cardiac Arrest Registry to EnhanceSurvival (CARES) registry. Results: The initialPubMed search yielded 260 unique references.There were 412 resuscitation attempts among532 cardiac arrests, from 8 studies which metthe inclusion criteria. The most common causeof cardiac arrest was trauma (46.2%). All studieshad data on ROSC, which occurred in 203 cases.The average rate of ROSC among the studieswas 49% (range 19% - 79%). Three studies hadno information on survival to discharge. Fivestudies had survival to discharge statistics; 73(25.2%) subjects were lost to follow up. Sur-vival to discharge information was available for217 subjects of which 63 (29.0%) survived to

discharge. This was substantially higher thanthe out of hospital cardiac arrest survival ratereported by the CARES registry (p < 0.001).Conclusions: The media’s depiction of cardiacarrest survival often does not include survivalto discharge information. When television andfilm survival to discharge rates are known,they are significantly greater than actual cardiacarrest survival rates. This may lead to unreal-istic expectations regarding out of hospital car-diac arrest victims’ chances of survival in thegeneral public.

99. Benchmarking the Use of Red Lightsand Sirens in 9-1-1 Systems: A Review of aLarge, National Dataset

Jeffrey Jarvis, Dustin Barton, Lauren Sager,Nick Nudell, Williamson County EMS Cate-gory of Submission: Operations, Quality,Safety Systems, Disaster, Disaster

Background: The use of Red Lights & Sirens(RLS) in responses to and from the scene ofa 9-1-1 call has long been tradition in EMS,although with limited evidence of clinical effi-cacy. There is a growing body of evidence ofthe dangers of RLS response and the effec-tiveness of priority dispatch triage for bettertriage of RLS responses. Little data has beenpublished which defines the prevalence of RLSuse to and from 9-1-1 scenes. We sought todescribe the proportion of RLS responses usinga large national dataset. Methods: Using anelectronic review of 6 1/2 years of data from9-4-1 consenting agencies using ESO’s Elec-tronic Health Record (EHR) system, we identi-fied the transport mode of all responses to andfrom the scene of a 9-1-1 call that resulted intransport to a hospital. The proportion of calls toand from the scene using RLS was determined,along with 95% confidence intervals. Results:There were 7,709,012 9-1-1 calls that resulted ina patient transport. Of these, 5,846,038 (75.8%,75.8–75.9%) involved RLS response to the sceneand 1,494,378 (19.4%, 19.4–19.4%) resulted inRLS response from the scene to the hospital.Conclusions: Using a large national dataset, weprovided baseline information on the preva-lence of the use of RLS to and from 9-1-1 calls.While we are unable to assess the necessity ofsuch response, given the known prevalence ofhigh-acuity 9-1-1 calls, it is possible that the 76%of RLS responses to 9-1-1 scenes could safelybe decreased with appropriate priority dispatchprocesses and ongoing quality improvement.Further efforts using patient outcome shouldassess the necessity of RLS response from thescene.

100. Usefulness of Epinehprine in CardiacArrest

James Hehl, Matthew Wells, Beth Langley,JE Winslow, Cape Fear Valley Mobile IntegratedHealthcare Cumberland County EMS Categoryof Submission: Cardiac

Background: The landscape for treatment ofcardiac arrest is evolving. The importance ofprompt, high quality cardiopulmonary resusci-tation and early defibrillation is receiving moreemphasis. For decades, intravenous (IV) admin-istration of epinephrine every 3–5 minutes hasbeen a component of the standardized proto-col for treatment of cardiac arrest, yet recentstudies suggest that frequency of administra-tion could impede neurological recovery. There-fore, our EMS agency developed a “one doseepinephrine” prehospital protocol for medi-cal cardiac arrest as a quality improvementproject. Hypothesis: Utilizing a “one doseepinephrine” protocol will improve neurolog-ical recovery in survivors of cardiac arrest.Methods: The protocol was revised and imple-mented in February of 2017 to include one

IV dose of epinephrine. All other compo-nents of the cardiac arrest protocol whereunchanged and followed the ACLS algorithm.Each patient was closely followed through aQuality Assurance and Quality Improvementprocess. Data was compared from Februarythrough July 2016, with epinephrine adminis-tered every 3–5 minutes; to February throughJuly 2017, with epinephrine administered once.Evidence of neurological status was obtainedfrom the physician discharge summary in thepatient’s medical record. Results: In the 2016period, 134 cardiac arrest calls were identifiedfrom a total of 27,282 EMS calls. Thirty-threepatients achieved return of spontaneous circu-lation (ROSC) with 10 surviving to discharge.Three of the 33 patients survived to be dis-charged home with no documented neurologi-cal deficit. In the 2017 period, 134 cardiac arrestcalls were identified from a total of 27,572 totalEMS calls. Thirty-nine patients achieved ROSCwith 8 surviving to discharge. Seven of the39 patients survived to be discharged with nodocumented neurological deficits. Outcomes:The number of patients who received the “one-dose epinephrine” protocol and achieved ROSCincreased by 18%. Patient survival to dischargehome with no documented with neurologicaldeficits increased from 30% in 2016 to 87.5%in 2017. Conclusions: Utilization of the “one-dose epinephrine” protocol demonstrated sig-nificant improvement in the percentage of vic-tims who survived a medical cardiac arrest withno documented neurological deficits.

101. Association between Initial BloodGlucose in Out-of-Hospital CardiacArrest and Return of SpontaneousCirculation

Caitlin Howard, Hattie McAviney, DavidWampler, Jeremy Allen, Justin Smith, DavidMiramontes, Joan Polk, United States Army,UTHSCSA Category of Submission: Student,Resident, Fellow

Background: Elevated blood glucose isassociated with poor outcomes in patientsresuscitated from out-of-hospital cardiac arrest(OHCA). In this study, we evaluate whetherinitial blood glucose level in OHCA patients isassociated with return of spontaneous circula-tion (ROSC). Methods: This was a retrospectivereview of a registry containing details ofeach resuscitation attempt by a large, urbanfire-based EMS system where the prevalenceof diabetes is much higher than the nationalaverage (14.2% vs. 9.3%). Data from January 1,2016 through August 15, 2016 was analyzed.Patients were included in the study if thefollowing variables were available: age, gender,initial blood glucose, and outcome (no ROSCvs. ROSC). Patients were excluded if age <17, no age, gender, or initial blood glucoserecorded, multiple blood glucoses crossing200 mg/dl, or no outcome recorded. Only theinitial blood glucose obtained at the onset ofresuscitation was considered. Patients weredivided into two groups: blood glucose < 200mg/dl and blood glucose > 200 mg/dl. At-test was used to analyze continuous variablesand a χ2 test was used to analyze categoricalvariables. Results: 620 patients were includedin this study. Mean age was 64.23 + 17.20 yearswith 385 males (62.10%). 453 patients (73.06%)had an initial blood glucose level < 200 and 167patients (26.94%) had a glucose level > 200. Ofthe patients with glucose < 200, 171 (37.75%)obtained ROSC. Of those with glucose > 200,63 (37.72%) obtained ROSC. There was noassociation between blood glucose levels andachievement of ROSC (P = 0.10). Conclusions:We found no significant association betweeninitial blood glucose levels in OHCA patientsand likelihood of achieving ROSC. The main



limitation to this study was that the patientpopulation was restricted to San Antonio,Texas. Additionally, we only considered theinitial blood glucose obtained during theresuscitation.

102. Implementing a Prehospital Protocolto Treat Behavioral Emergencies withMidazolam Leads to Effective Control ofAgitated Patients

Christopher Richards, Ryan Huebinger, KatieTataris, Joseph Weber, Kenneth Pearlman,Eddie Markul, Matthew Strzalka, Mark Kiely,Leslee Stein-Spencer, Leslie Zun, NorthwesternFeinberg SOM Deptment of Emergency Medicineand Center for Healthcare Studies, Chicago EMSSystem Category of Submission: Medical

Background: Combative patients are com-monly encountered by EMS providers and posechallenges for both patient care and providersafety. Chemical sedation with midazolam iscommonly used in the emergency departmentsetting to treat agitation from psychiatric dis-turbances and intoxication. However, limiteddata exist regarding midazolam use in the pre-hospital setting to treat agitation. We soughtto describe our experience after implementinga protocol for treating patients with behav-ioral emergencies using midazolam in a largeurban EMS system. Hypothesis: We hypothe-size that implementation of a prehospital pro-tocol using midazolam to treat patients havinga behavioral emergency leads to improved clin-ical conditions without causing significant clin-ical deterioration. Methods: We performed aretrospective review of EMS records followingthe implementation of a behavioral emergen-cies protocol in a large urban EMS system fromFebruary 2014 through April 2016. Paramedicswere instructed to administer midazolam 1 mgintravenous (IV) or intraosseous (IO) or 5 mgintramuscular (IM) or intranasal (IN), repeat-ing once as needed, and to record the responseto treatment. Patients receiving midazolam forthe indication of “behavioral emergency” wereincluded, and any patient receiving midazo-lam for “seizure” were excluded. Descriptivestatistics were used to report results, and Spear-man’s rho was calculated to determine cor-relation of dose and route. Results: In total,midazolam was administered in 435 instancesto 390 patients. Median age was 33 (IQR 24–50) years; 69.0% were male, and 53.1% wereAfrican American. Doses administered were1 mg (11.8%), 5 mg (72.3%), and 10 mg (15.1%)via IM (42.2%), IN (41.1%), IV (16.5%), andIO (0.2%) routes. In 37 patients, a seconddose was required, and the same dose (rho =0.84, p < 0.0001) and route (rho = 0.68, p <0.0001) as the first administration was com-mon. Paramedics reported slight or substan-tial improvement in clinical condition in 75.3%of patients, and 24.7% had no clinical change.No paramedic reported clinical deteriorationin a patient’s condition following midazolamadministration. Conclusions: A protocol usingmidazolam in behavioral emergencies can besuccessfully implemented in a large urban EMSsystem. Midazolam successfully treated agita-tion, and paramedics did not feel that patients’clinical conditions worsened after midazolamadministration.

103. Termination of ResuscitationChecklist: Duration and Outcomes ofResuscitation

Katherine Kuefler, Aurora Lybeck, ThomasGrawey, M. Riccardo Colella, Medical College ofWisconsin Category of Submission: Student,Resident, Fellow

Background: Checklists are often used in medi-cal and non-medical fields to aid in error pre-

vention, management of complex processes,and to help produce reliable outcomes. OnApril 1, 2016 a termination of resuscitation(TOR) checklist was implemented for use dur-ing out-of-hospital cardiac arrests (OHCA) byMilwaukee County Emergency Medical Ser-vices (EMS) online medical control (OLMC)physicians concurrently staffing an emergencydepartment. Objective: To evaluate if the useof a TOR checklist by OLMC impacted returnof spontaneous circulation (ROSC) for medi-cal or traumatic OHCA. To compare if check-list use changed the duration of resuscitationsperformed by EMS providers and the durationof the OLMC call. Methods: Medical and trau-matic OHCA data were extracted from the Mil-waukee County EMS database from April 1,2015 to September 30, 2015 (452 medical, 44trauma) and April 1, 2016 to September 30,2016 (482 medical, 71 trauma). Patient outcomeswere measured by occurrence of ROSC duringresuscitation and by presence of ROSC at hospi-tal arrival. Analysis of the data was done usingt-tests. Results: In medical OHCA, incidenceof ROSC during resuscitation increased from41% (185/452) to 46% (220/482) with imple-mentation of the TOR checklist and ROSC athospital arrival increased from 35% (160/452)to 40% (191/482). There was also a significant(p < 0.001) increase in mean duration of resus-citations (26 to 30 minutes) and duration ofOLMC (13 to 15 minutes) after the checklist wasimplemented in cases of medical OHCA. Con-clusions: In medical OHCA the use of a TORchecklist by OLMC significantly increased theduration of both resuscitations and OLMC time.The rates of ROSC during resuscitation and athospital arrival increased after the checklist wasimplemented for medical OHCA. These resultsshow a potential clinical benefit for OLMC useof a TOR checklist for medical OHCA, andalso inform resource utilization in an academicEmergency Department. In traumatic OHCAthere were no significant changes in duration ofresuscitation or OLMC and there was a decreasein ROSC; further study with a larger samplesize may be needed. Neurological outcomesare unknown and further research may providea better understanding of the impact of thesefindings.

104. Qualitative Evaluation of CommunityParamedic Care Transitions InterventionCoach Training

Hunter Lau, Matthew Hollander, JeremyCushman, Amy Kind, Courtney Jones,Michael Lohmeier, Manish Shah, Universityof Wisconsin School of Medicine and Public HealthCategory of Submission: Student, Resident,Fellow

Background: The Care Transitions Intervention(CTI) has potential to improve the emergencydepartment (ED)-to-home transition for olderadults. Community paramedics may functionas the CTI coaches instead of nurses who tradi-tionally serve in that role. To do so requires thatthe community paramedics possess the appro-priate knowledge, skills, and attitudes, whichare not inherently part of traditional EMS edu-cation. The aim of this study is to evaluate anexpert-panel developed training program forcommunity paramedics serving as CTI coacheswho support the ED-to-home transition. Meth-ods: This study is a component of an ongoingtwo-center randomized controlled trial evaluat-ing a community paramedic-implemented CTIto enhance the ED-to-home transition. Com-munity paramedic training covered multipledomains including the CTI program, geriatrics,motivational interviewing, ED discharge, andcommunity paramedicine. One year after start-ing the study, we conducted audio-recordedsemi-structured interviews with community

paramedics in both cities (June–July 2017).After transcribing the interviews verbatim,team members independently performed pre-liminary coding. Ensuing group data analy-sis sessions led to the development of finalcodes and thematic generalizations recurrentin the interviews. Results: All eight partic-ipating community paramedics were inter-viewed. Of the paramedics, five were womenand all were non-Hispanic whites. The meanage was 43. Participants typically had exten-sive backgrounds in healthcare, primarily asEMS providers, but minimal experience withcommunity paramedicine. All reported someprior geriatrics training. Four themes emergedfrom the interviews: (1) certain characteristicsmake coaches more likely succeed in this pro-gram; (2) active rather than passive learningmay achieve the best results for communityparamedic CTI training; (3) training programcomponents require minor refinements; and (4)continuing education should more effectivelyaddress the paramedic coaches’ evolving needs.Conclusions: Paramedics represent a crucialand largely untapped resource for supportingED-to-home care transitions, such as throughthe CTI. Training that leads to the appropri-ate knowledge, skills, and attitudes is criticalfor effective implementation, including choos-ing the optimal candidate coaches, deliveringtraining in the most effective manner for the stu-dents, and delivering content targeted to stu-dent needs.

105. Emergency Medical Services Responseto Mass Shooting and Active ShooterIncidents, United States, 2014–2015

Matthew Sztajnkrycer, Aaron Klassen,Morgan Marshall, Mengtao Dia, N ClayMann, Mayo Clinic Department of EmergencyMedicine Category of Submission: Trauma

Background: According to Federal Bureau ofInvestigation statistics, the number of activeshooter incidents has increased over the pastdecade. The purpose of the current study wasto describe the EMS response and interven-tions to mass shooting and active shooterincidents. Methods: Retrospective analysis of2014 and 2015 National Emergency MedicalServices Information System (NEMSIS) datasets. Date, time, and location for mass shoot-ing incidents were obtained from the opensource Gun Violence Archive and then corre-lated with NEMSIS data set records. Activeshooter incidents were identified through FBIdata. A de-identified database was generatedfor final analysis. Results: A total of 608 massshooting were identified, of which 19 wereclassified as active shooter incidents. Meannumber of injured victims was 4.6 ± 2.5,while mean number of fatalities was 1.2 ±2.2. NEMSIS data identified 652 EMS activa-tions to 226 unique incidents; 5 were activeshooter incidents. 76% of victims were male.80% of victims were African American. Themean age was 27.7 ± 11.1 years. Dispatch com-plaint was reported as not known or unknownproblem/man down in 14.6% of records. Thepredominant response configuration was ALS(78.8%). Volunteer services responded to 7%of events. The most commonly reported inci-dent locations were Street/Highway (38.2%),Home/Residence (32.4%), and Trade/Service(11.5%). Location of wounds included extrem-ities (38%), chest (9%), and head (9%). Tourni-quet use was documented in 6 victims. Gun-shot wound was self-inflicted in 2.3% of vic-tims. When present, cardiac arrest occurredafter EMS arrival in 37.5% of cases. 35.9% ofvictims were transported to the closest facility.Conclusions: Mass shooting and active shooterincidents are prevalent in the United States,with an average of 5.8 victims per incident.



Despite the fact that extremity wounds werethe most common injury noted, suggesting arole for public access bleeding control, doc-umented EMS tourniquet use was uncom-mon. While mass shooting events pose highrisk for responders, dispatch information waslacking in nearly 15% of records. Respond-ing EMS agencies were diverse and includedBLS providers and volunteers, emphasizing theneed to ensure all EMS providers are preparedto respond to mass shootings.

106. When Dogs Fly: Use of Air MedicalServices to Transport Operational K9sInjured in the Line of Duty

Chelsea Hogan, Chadd Nesbit, Department ofEmergency Medicine, Penn State Milton S. Her-shey Medical Center Category of Submission:Student, Resident, Fellow

Background: Instances of operational K9 airmedical transports have been documented inthe popular press. There have been no stud-ies to look at the prevalence of such trans-ports or to determine what policies flight pro-grams have in place to address this challeng-ing transport issue. We sought to assess theprevalence of operational K9 transports as wellas existence and content of protocols to con-duct such a transport should one be deemednecessary. Methods: We distributed a surveyto air medical programs in the United Statesvia the Research Electronic Data Capture (RED-Cap) program. Programs were identified usingthe Atlas and Database of Air Medical Services(ADAMS ). Programs that could not be reachedvia email were excluded. A survey containingup to 23 questions inquiring about K9 trans-ports, policies and procedures was emailed to295 identified programs. Results: We received147 total survey responses (49.8% response).Twenty-two programs (15%) reported receiv-ing a request to transport a K9 and of those,15 reported flying the K9. Forty-one K9 trans-ports were reported. Smaller numbers of pro-grams reported having any additional train-ing related to care or transport of operationalK9s or a pre-designated emergency veterinar-ian. Six programs reported carrying some typeof equipment for use on K9s and 7 programsreported having some type of protocol in placefor these types of flights. Ninety-five of theprograms reported that they would be able tofly the K9 and handler as well as the normalflight crew. Conclusions: The goal of this sur-vey was to quantify the number of transportsfor injured operational K9s and to identify anypolicies or procedures that programs have inplace to carry out a transport if one is requested.Although supposedly a rare occurrence, 15% ofour respondents have reported such a request.Of those requests the majority of transportswere completed. While some programs maydecide that they will not transport an injuredoperational K9, those programs that will shouldestablish policies and procedures for this type ofmission.

107. Community Paramedic Partnership:Shifting Healthcare Utilization throughPartnership between Municipal Fire/EMSand the Local Level I Trauma Center

Tia Radant, Joseph Pasquarella, Ann Majerus,Matthew Simpson, Paula Miller, SandiWewerka, Aaron Burnett, Regions HospitalEMS Category of Submission: Operations,Quality, Safety Systems, Disaster, Disaster

Background: A partnership between a LevelI Trauma Center and an urban, municipalFire/EMS Department for patients with con-gestive heart failure (CHF) was launched in2014. The program aimed to improve healthcareutilization and reduce readmissions through

a unique Community Paramedic (CP) part-nership. Hypothesis: Patients with congestiveheart failure who receive CP visits for 30 dayspost-discharge have a reduced rate of readmis-sion and an increased use of clinic visits. Meth-ods: Inpatients with CHF were offered visits bya CP for up to 30 days post discharge. Inclusioncriteria included, local residentTia, no home-care services upon discharge, diagnosis of CHF,English speaking, and consent to home visits bya CP. The CP visited the patient in the home1–2 times per week for 4 weeks following dis-charge. At each visit the CP conducted med-ication reconciliation, a physical exam, homesafety evaluation, coordination of follow-upcare and referral to community or healthcareresources as needed. Pre/post comparisonswere analyzed descriptively using means andstandard deviations. Scores were assessed withWilcoxon signed-rank tests. Results: A totalof 64 patients were enrolled between Febru-ary 2015 and July 2017; 32 patients completedthe program with complete data. A compari-son of 90-day healthcare utilization pre- andpost-admission showed that patients who wereprovided CP services had a significant decreasein hospital admissions (68%, p < 0.0001) andED visits (62%, p < 0.0001), and had a 14%increase in clinic visits (ns, p = 0.45). A group ofpatients that met inclusion criteria but declinedconsent to participate was compared to thepatient group that participated in the CP pro-gram. Patients who completed the programhad a significantly higher decrease in admis-sions (p = 0.0145) and ED visits (p = 0.0009)pre- to post-hospitalization than those who didnot enroll (n = 20). There was no significantdifference in change in clinic utilization. Con-clusions: Partnership between fire-based EMSand hospitals for Community Paramedic pro-grams can be successful. CP’s providing post-discharge care results in a shift of healthcareutilization toward reduced admissions/ED vis-its and increased clinic visits. Further researchwith a larger cohort is needed to determine ifutilization patterns would be sustained past 90days.

108. “PDTREE”: Development of a NewPediatric Prehospital TransportDestination EBG

Jennifer Fishe, Kye Fratta, Jennifer Anders,University of Florida COM - Jacksonville, Depart-ment of Emergency Medicine Category of Sub-mission: Pediatric

Background: Prehospital triage should matchpatient needs with hospital service availabil-ity. EBGs guide EMS’ destination choice foradult patients suffering from trauma, MI, andstroke. However, analogous guidelines do notexist for any pediatric condition save trauma.This study’s objective was to create a non-trauma pediatric prehospital transport des-tination EBG. Methods: A systematic liter-ature search identified articles pertinent tonon-trauma pediatric prehospital destinationchoice. Resulting articles were reviewed usingGRADE and compiled into an evidence pro-file. An expert panel (including stakeholdersfrom pediatric EM, EM, EMS medical direc-tors, EMS providers, and patient/family advo-cates) reviewed the evidence profile and datafrom the statewide EMS system where the EBGwould undergo pilot testing. Using a modified-Delphi process with three voting rounds and75% agreement threshold, the panel selecteditems for inclusion, refined terminology, andreached consensus on a pediatric prehospi-tal transport destination EBG. Results: Theliterature search produced 60 articles. AfterGRADE review, 47 articles were included inthe evidence profile. Articles identified specificpediatric populations (ALTE, seizures, special

health care needs) at risk for secondary trans-port or interfacility transport (IFT). IFT deci-sions are made quickly, but patients risk sub-optimal pre-transfer care, and suffer delays indefinitive care and increased morbidity. Quan-titative physiologic data (vital signs, capillaryrefill time, hospital-based scoring systems) inisolation do not accurately or reliably pre-dict the need for pediatric specialty/criticalcare. Combining quantitative and qualitativeprehospital assessments promises more accu-rate, reliable prediction of specialty/criticalcare needs. After reviewing the evidence, theexpert panel’s modified-Delphi process pro-duced a pediatric prehospital destination EBG(“PDTree”). The PDTree is formatted as analgorithm, matching 14 non-trauma condi-tions/risk factors (including ALTE, seizurerequiring EMS-administered benzodiazepine,sepsis, and emergencies related to conditionstreated at a medical home) to three differ-ent levels of pediatric care (specialty, compre-hensive, regional). Conclusions: Existing med-ical literature identifies the need for prehos-pital transport destination guidance for non-trauma pediatric patients. That evidence sup-ported the modified-Delphi process that pro-duced the “PDTree,” a new non-trauma pedi-atric prehospital destination EBG. “PDTree”will be pilot tested by computerized resourcemodeling, prehospital provider simulation, andimplementation in three diverse EMS agencies.

109. Duplicate Procedures and ChargesAssociated with Pediatric Inter-FacilityTransfer from Emergency Departments

Ali Aledhaim, Jon Mark Hirshon, JenniferFishe, Jennifer Anders, University of MarylandDepartment of Emergency Medicine Category ofSubmission: Pediatric

Background: Interfacility Transfer (IFT) ofpatients with emergency conditions from anEmergency Department (ED) delays defini-tive care and burdens the patient with poten-tially harmful duplicate procedures and extracharges. This physical and economic hardshipmay be preventable if patients are taken to adefinitive care facility for their initial destina-tion. Objective: To determine duplicate pro-cedures and charges sustained by pediatricpatients undergoing IFT for inpatient admis-sion after an ED visit to a different facility.Methods: This study utilized three years (2010–2012) of Maryland HCUP ED and inpatient visitdata. A modified probabilistic linkage was per-formed to identify ED patients who were dis-positioned to IFT and admitted to a distantfacility. Included patients were 0–17 years ofage with any of the 20 most common Diag-nosis Categories (DxC) and whose conditionswere classified “emergent” or “urgent”. Afterlinkage, duplicate procedures were identifiedand classified as administrative or clinical. Mul-tiple regression analysis was used to com-pare the average total charges of IFT patients,including duplicate charges, to non-IFT admit-ted patients presenting with the same top 20DxC. Results: Of the 9,447 IFT inpatients iden-tified, 2,254 patients were successfully linked,of which 1713 (76%) had one of the top 20DxC. The most frequent administrative dupli-cate procedure was ER EMTALA emergencymedical screening (1,407). Notable duplicateclinical procedures involving repeat radiationwere chest X-ray (239) and CT scan of head (97)or body (32). IFT patients incurred an averagetotal charge of $11,786.61 including an averageduplicate charge of $1,627.84. In comparison,the average charge incurred by a non-IFT was$8,209.72. Adjusting for the effect of age, gen-der, and race, a weighted regression model esti-mated an average 34% (30.1–37.6%, p < 0.001)increase in total charges for an IFT patient com-



pared to a non-IFT patient. Conclusions: Bothsafety harms (radiation exposure) and signifi-cant economic burden are seen in the subset ofpatients undergoing IFT from an ED for inpa-tient admission to a distant facility. EMS sys-tems can minimize this inefficiency and burdenby transporting patients to definitive care facil-ities whenever feasible.

110. Clinical Events In PrehospitalPatients With St-Elevation MyocardialInfarction Transported To A PCI Center ByBasic Life Support Paramedics In A RuralRegion

Pierre-Alexandre LeBlanc, Sylvain Bussières,François Bégin, Alain Tanguay, Jean-MichelParadis, Denise Hébert, Richard Fleet, Départe-ment de Médecine d’Urgence – Université Laval,Québec, Canada Category of Submission: Stu-dent, Resident, Fellow

Background: Rural areas have limited hospi-tal staff and often rely on basic life support(BLS) paramedics for inter-facility transport.No previous study has established whetherST-segment elevation myocardial infarction(STEMI) patients transported in ambulanceover long distances are at risk of suffering fromclinical events such as bradycardia or hypoten-sion. The objective of this study was to estab-lish clinical events, and to determine if thecomplications occurring in the presence of BLSparamedics are influenced by the transporta-tion time. Methods: In a retrospective cohortstudy, we reviewed 896 consecutive STEMIpatients diverted and transported to the near-est PCI-capable center according to an emer-gency physician interpretation of a 12-lead ECGtransmitted by paramedics. Patients had contin-uous electrocardiogram (ECG) and vital signsmonitoring during transport. A focus groupcomposed of the authors established clinicallyimportant and minor events based on liter-ature search. A multivariate ordinal logisticregression model was used to study the asso-ciation between transportation time (0–14, 15–29 and >30 min) and the occurrence of clin-ical events. Results: Clinically important andminor events were experienced by 18.6 and12.16% of STEMI patients, respectively. Trans-portation time was not associated with higherrisk of suffering from clinical events (p =0.182). The most frequent events were brady-cardia (8.87%), followed by hypotension (6.1%),and ventricular tachycardia / ventricular fib-rillation (VT/VF) (5.13%). All patients suffer-ing from VT/VF were successfully resuscitatedwith defibrillation. No death on arrival at PCIcenter was recorded. Conclusions: Prehospi-tal STEMI diagnosis by transmission of a 12-lead ECG interpreted by emergency physiciansand triage for primary PCI by paramedics with-out advanced care training is a safe option thatcould use less advanced staff in rural areas withlimited resources.

111. Description of Drug-AssistedIntubation in Statewide TreatmentProtocols

Steven Sommerville, Daniel Wilner, DavidSchoenfeld, Beth Israel Deaconess Medical CenterCategory of Submission: Student, Resident,Fellow

Background: Endotracheal intubation in pre-hospital airway management has been a focusof research and debate for decades. Endo-tracheal intubation is performed using drug-assisted intubation (DAI) or without medica-tion. DAI is divided into rapid sequence intu-bation (RSI) where a sedative as well as neu-romuscular blockade is used or sedative-onlyintubation. The extent to which DAI is incor-porated in statewide treatment protocols (STP)has not been described. The majority of states

have STPs that are either mandatory or serveas a guide for medical directors. The purposeof this investigation is to describe the extentto which STPs include DAI and the variabil-ity in pharmacopeia utilized. Methods: Crosssectional study of STP utilizing a standardizedreview of DAI protocols and medications. Pro-tocol revision date was also captured. Results:Thirty one out of fifty states (64%) issue STPs,seven (22%) of which serve as guidelines. RSI isincluded in the STP of 17 states (55%). Sedative-only intubation is included in the STP of 5 states(16%). The most commonly included induc-tion agents are etomidate and midazolam (19STPs each, 61%); other induction agents includeketamine (11 STPs, 35%), fentanyl (2 STPs),and propofol (1 STP). Succinylcholine is themost commonly included paralytic (17 STPs,55%); rocuronium (11 STPs, 35%) and vecuro-nium (7 STPs, 23%) are other approved para-lytic agents. 16 states (52%) permit intubation ofboth adult and pediatric patients while 6 states(19%) only allow DAI of adult patients. All pro-tocols have been revised within the past 5 yearsand 75% of protocols were revised since 2015.Conclusions: The NAEMSP position statementon drug-assisted intubation recommends theuse of a paralytic during DAI, as it increases thelikelihood of first pass success. Just over halfof all STPs allow for DAI, and 16% allow forsedative-only intubation despite the NAEMSPposition statement on DAI. There is significantvariation in both the induction agent as well asthe paralytic utilized for intubation across STPs.There is also variation in the number of statesthat allow for both adult and pediatric intu-bation. Additional research is needed to deter-mine optimal agents and protocols for prehos-pital intubation.

112. Assessment of Intraosseous NeedlePlacement by EMS Providers

Alexandra Petrie, Jeffrey Lubin, Penn State Col-lege of Medicine Category of Submission: Oper-ations, Quality, Safety Systems, Disaster,Disaster

Background: Intraosseous (IO) needle place-ment can be used to provide quick deliveryof various fluids to the patient, particularlyin cases in which venous access is compro-mised; however, if done incorrectly, it can leadto unwanted complications such as extrava-sion of fluid, poor flow, and catheter dislodge-ment (Paxton 2009; Dev 2014; Gluckman 2014).The purpose of this study is to see if EMSproviders can adequately locate the correctlocations for the placement of IO needles inlive models. Methods: We assessed the accu-racy of intraosseous placement by asking EMSproviders from a statewide conference to sim-ulate where they would use an intraosseousneedle on standardized patients. Each partici-pant also filled out a demographic survey thatincluded their experience with intraosseousneedles and a knowledge of acceptable EZIOintraosseous needle landmarks from a list ofoptions. Measurements were established onlive human models using transfer paper withstickers placed in tibial and humeral IO spots,marked so that they easily lined up with themodel via landmarks. The participant wasasked to place a sticker directly on the modelwhere they would insert the EZIO at bothlocations. Afterward, a transfer sheet with thesticker placed at a location correlating withstandard placement was compared against theparticipant-placed sticker. Differences in place-ment were measured with a ruler to the near-est half centimeter. Direction was qualitativelynoted. Numbers were assigned to each par-ticipant so that the demographic survey, loca-tion survey, and sticker location could be linkedto each individual subject (N = 30). Results:Results were analyzed via several 2 sample t

tests using 0 as the standard landmark. Theaverage distance from the landmark on thehumerus was 5.06 cm (95% CI: 4.06–6.06).The average from the tibia was 4.13 cm (95%CI: 3.16–5.10). Both were statistically signifi-cant with a p value of <0.0001. Conclusions:These results show a low accuracy among EMSproviders in identifying correct landmarks forintraosseous needle placement. This suggestsadditional training and skills review may beneeded across the state in order to safely per-form this procedure.

113. Paramedic Recognition andManagement of Anaphylaxis in thePrehospital Setting

Rakesh Gupta, Krystyna Samoraj, SimerpreetSandhanwalia, Matt Kerslake, Luke Ryan,Colleen Shortt, Michelle Welsford, McMasterUniversity Category of Submission: Student,Resident, Fellow

Background: Anaphylaxis is a life-threateningcondition that paramedics are equipped to treateffectively in the field. Current literature sug-gests improvements in paramedic recognitionand treatment of anaphylaxis could be made.The aim of this study was to compare theproportion of cases of anaphylaxis appropri-ately treated with epinephrine by paramedicsbefore and after a targeted educational inter-vention. Methods: This was a retrospectivemedical records review of patients with ana-phylaxis managed by primary or advanced careparamedics in five Emergency Medical Ser-vice areas in Ontario, before and after an edu-cational module was introduced. This mod-ule included education on anaphylaxis diagno-sis, recognition, treatment priorities, and feed-back on the recognition and management fromthe before period. All paramedic call records(PCRs) coded as “local allergic reaction” or“anaphylaxis” during 12-month periods beforeand after the intervention were reviewed bytrained data abstractors to determine if patientsmet an international definition of anaphylaxis.The details of interventions performed by theparamedics were used to determine primaryand secondary outcomes. Results: Of the 600PCRs reviewed, 99/120 PCRs in the before and300/480 in the after period were included. Ofthe charts included, 63/99 (63.6%) in the beforeand 136/300 (45.3%) in the after period met cri-teria for anaphylaxis (p = 0.002). Of the casesmeeting anaphylaxis criteria, 41/63 (65.1%) inthe before and 88/136 (64.7%) in the afterperiod were correctly identified as anaphylaxis(p = 0.96). Epinephrine was administered in37/63 (58.7%) of anaphylaxis cases in the beforeperiod and 76/136 (55.9%) in the after period(p = 0.70). Anaphylactic patients with onlytwo-system involvement received epinephrinein 20/40 (50.0%) cases in the before periodand 45/93 (48.4%) in the after period (p =0.86). Conclusions: There are gaps in paramedicrecognition and management of anaphylaxis,particularly in cases of two-system involve-ment. These gaps persisted after the imple-mentation of an educational intervention. Otherquality interventions and periodic refreshersmay be necessary to improve prehospital treat-ment of anaphylaxis. Limitations include anincrease in overall cases and decrease in rateof true anaphylaxis in the after period, whichmay relate to better case identification afterelectronic PCR implementation and changes inparamedic recognition.

114. National Description of PatientRefusals Following PrehospitalAdministration of Naloxone

Mirinda Gormley, Juan Lu, Virginia Common-wealth University Category of Submission:Medical



Background: Emergency medical services(EMS) personnel deliver Naloxone to reversedeadly opioid overdoses. However, EMS per-sonnel may experience challenges with patientcare, including being unable to convince apatient to be transported to the hospital.Without accessing appropriate follow-up care,these patients could overdose again. Objec-tive: Identify characteristics associated withpatients who received Naloxone from EMS butrefused to be transported to hospital. Methods:Data came from the 2015 National EmergencyMedical Services Information System. Theincident/patient disposition was used tocreate a binary outcome (“transported” or“refused”), where “treated, transferred care,”“treated, transported by EMS,” and “treated,transported by Law Enforcement” made up“transported,” and “no treatment required,”“patient refused care,” “treated and released,”and “treated, transported by private vehicle”comprised “refused.” Characteristics includedage, gender, race, prior aid, location, U.S.census region, and urbanicity. Descriptive andmultivariable logistic regression were utilized.Results: In 2015, EMS agencies reported 585,108Naloxone administrations by personnel of atransport unit during a 9-1-1 response. Aftertreatment, 1.6% of patients refused transport.These patients were primarily male (65.0%),white (76.8%), and had a median age of 48 (IQR= 32–61). Compared to transported patients,those who refused were more likely to be foundin a residence (75.6% vs. 68.0%), or receive aidprior to EMS arrival (60.3% vs. 23.9%). Largerproportions of patients went to the hospital iffound in a public location (19.7% vs. 13.5%),or a rural/wilderness area (10.4% vs. 6.6%).Patients had nearly double the odds of beingtransported from a public location rather thana residence (OR = 1.704, 95% CI = 1.58–1.84),and patients in rural/wilderness locations were1.5 times more likely to be transported thanurban patients (OR = 1.58, 95% CI = 1.44–1.73).Patients who did not receive aid prior to EMSarrival were nearly twice as likely to go to thehospital (OR = 1.71, 95% CI = 1.61—1.81).Conclusions: While effective at reversingfatal overdoses, prehospital administration ofNaloxone is not sufficient to address addiction,whereas those transported to hospital couldaccess treatment. EMS agencies should worktogether with public safety partners to planhow to work with patients most at risk ofrefusing transport following initial treatment.

115. EMS Compass Benchmarks Using aNational EMS Dataset: Status Epilepticusand Hypoglycemia Performance Measures

Jeffrey Jarvis, Dustin Barton, Lauren Sager,Nick Nudell, Williamson County EMS Cate-gory of Submission: Medical

Background: Status epilepticus and hypo-glycemia are emergent conditions, both ofwhich can be effectively treated by EMS. It isunclear how often these assessments and treat-ments are given. EMS Compass is a nationalorganization that has developed several clini-cal measures. No work has been done to bench-mark these measures against large, nationaldatasets. This is necessary for quality improve-ment efforts and refinement of the measuresthemselves. We aimed to describe compliancewith these measures using a large, nationalcohort. Methods: Using a 6 ½-year sampleof 9-4-1 consenting EMS agencies using theESO electronic health record (EHR), we calcu-lated compliance rates among transported 9-1-1 patients for the following measures: (1) sometype of glucose given to those with blood glu-cose under 60, (2) a blood glucose documentedfor those felt to be in status epilepticus, and(3) a benzodiazepine given for those in felt to

be in status epilepticus. For measures requir-ing administration of a medication, only ALSproviders were included. For each measure, arate and 95% Confidence Interval were calcu-lated. Results: A total of 147,238 patients hada documented blood glucose <60. Of these,117,358 (79.7%, 95% CI 79.5–79.9%) receivedsome type of glucose. Of 11,148 patients witha status epilepticus, 8,072 (72.4%, 71.6–73.2%)had a blood glucose documented and 6,250(56.1%, 55.1–56.0%) had some type of ben-zodiazepine given by ALS agencies. Conclu-sions: We describe the compliance rates on sev-eral EMS Compass measures using a nationalcohort. We found a low rate of benzodiazepineuse for status epilepticus. It is possible thatthis is a function of poor, non-standard doc-umentation, imprecise measure definitions, orpoor clinical performances. In any case, theseresults identify opportunities for important sys-tem improvement.

116. Analysis of Medication StorageTemperatures in a Modern EMS Fleet:Preliminary Results from the Analysis ofMedication Storage Temperatures Trial(AFIRE)

Timothy Burns, Alan Butsch, CristopherTouzeau, Roger Stone, Barry Reid, Mont-gomery County (MD) Fire And Rescue ServiceCategory of Submission: Professional

Background: EMS operational programsdeploy medications using a variety of meansunder all kinds of conditions. Because of thisdeployment versatility, medications that wereonce limited to somewhat controlled clinicalsettings are now deployed on vehicles whoseclimate control is more difficult. Purpose:To explore whether or not EMS medicationsdeployed in modern fire and EMS vehiclesexperience temperatures that are outsidestorage temperature ranges from the US Phar-macopeia. Hypothesis: Medications will beexposed to temperatures outside the guidelinesin all types of our apparatus. Methods: Werecorded ambient temperatures on two of ourparamedic engines and in two of our transportunits during two summer months in 2017 usingtemperature data loggers. Once downloadedinto a database, these measurements createda continuous stream of temperature data forthe entire study period. Results: Data fromthe paramedic engine location reveals that theambient temperature was above the definitionof “extreme heat” from the USP (104°F) for1,350 minutes (1.4%) of the 94620-minute studyperiod, in the range “warm” (86–104°F) for60168 minutes (64%) of the study period, andout of the “controlled room temperature”range 89229 minutes (94%.) Neither position inthe transport unit was subjected to “extremeheat,” but they were in the “warm” rangefor 5759 minutes (6%) and 12092 minutes(13%) respectively, during the study period.Transport unit temperatures were outside ofthe “controlled room temperature” range for51138 minutes (54%) and 67131 minutes (71%).Conclusions: Temperature is much morecontrolled in the transport units than in theParamedic Engines. Medications deployed onour paramedic engines experience a significant“extreme heat” exposure. During the summermonths of the study period, all environmentstested temperatures in the “warm” range andall were out of “controlled room temperature”range a majority of the time. Limitations: Theseobservations only define the temperatures towhich the medications were exposed. Furtherstudy would need to be conducted on theeffects of this exposure.

117. Epidemiology And Outcomes OfAnaphylaxis-Associated Out-Of-HospitalCardiac Arrest

Seung Chul Lee, Sun Young Lee, Sang DoShin, Jeong Ho Park, Dongguk University IlsanHospital Category of Submission: Cardiac

Background: Understanding the epidemiologiccharacteristics of anaphylaxis-associated out-of-hospital cardiac arrests (OHCAs) is thefirst step for developing preventative strate-gies and optimizing care systems. We aimedto describe and compare the epidemiologicfeatures and outcomes among patients withanaphylaxis-associated OHCAs according tocausative agents group. Methods: We identi-fied emergency medical service (EMS)-treatedanaphylaxis-associated OHCA patients from anationwide OHCA registry between 2008 and2015. We compared epidemiologic character-istics and outcomes according to the causalagents and evaluated temporal variability inanaphylaxis-associated OHCA incidence. Therate of survival to discharge was comparedamong causative agents groups using mul-tivariate logistic regression analysis. Results:During the study period (8 years), a total of 224anaphylaxis-associated OHCAs were includedin the analysis. Natural agents group includ-ing insect sting and foods were 192 (85.6%)and iatrogenic agents group were 32 (14.3%).There was significant variability in the fre-quency of anaphylaxis-associated OHCA byhour of the day (p value < 0.01) and sea-son of the year (p value < 0.01), with thehighest incidence occurring during the day-time (7:01 am to 3 pm; 64.6%) and in summer(June to August, 48.7%). Compared with nat-ural agents, the adjusted odds ratios (AORs)for survival to discharge in iatrogenic agentswere statistically insignificant (AORs 3.61, 95%CIs 0.86 to 15.06). Conclusions: There was sig-nificant temporal variability in the incidenceof anaphylaxis-associated OHCA, with its peakduring the summer. Anaphylaxis-associatedOHCA by natural agents accounted for thegreater proportion of anaphylaxis than iatro-genic agents but there was no difference in sur-vival to discharge between the two groups.

118. Relationship between Adult BodyMass Index and Anticipated Failure Rateof Needle Decompression Using a 5cmNeedle for Tension Pneumothorax

John Lyng, Kristin Pokorney-Colling,Michaela West, Gregory Beilman, NorthMemorial Health Ambulance and Air Care Cate-gory of Submission: Trauma

Background: Tension pneumothorax is trau-matic injury that can lead to rapid circulatorycollapse and death. Emergent needle thoracos-tomy can quickly treat tension pneumothorax,but the best anatomic location and catheterlength necessary to perform the interventionhas been questioned in the recent years giventhe increasing rates of obesity in our popula-tion. Methods: We conducted a retrospectivereview of a convenience sample of all traumapatients admitted to our level 1 trauma centerin Minneapolis, MN that underwent chest com-puted tomography (CT) during their admis-sion between 2011 and 2012. Using these CTradiographs, chest wall thickness was mea-sured bilaterally at the 2nd intercostal space(ICS) at the midclavicular line, and at the 4thand 5th intercostal spaces at the anterior axillaryline. Baseline demographic data including age,sex, BMI, ISS and associated chest wall traumawere collected from medical chart review. Nee-dle thoracostomy failure was defined as chestwall thickness (CWT) of > 5cm, based on thelength of commonly used needle decompres-sion needles. Results: Atotal of 141 patients thatmet all inclusion criteria were identified. Therewere no significant differences in mean CWTat any of the anatomic sites. CWT was simi-lar between males and females. BMI > 30 was



associated with an adjusted odds ratio of 13.8(Confidence interval 4.8–39.8) for failure witha standard 5cm catheter needle decompression.Conclusions: In the increasingly obese generalpopulation, needle thoracostomy with a stan-dard 5cm needle may be more prone to fail-ure. Adult BMI > 30 is a significant risk fac-tor for anticipated failure of needle tube decom-pression. Alternative anatomic sites for nee-dle decompression did not appear increase theanticipated success of the intervention.

119. Evaluating the Incorporation of aJournal Club Series into Paramedic InitialEducation

Lauren Maloney, Paul Werfel, Robert Mar-shall, Scott Johnson, Stony Brook UniversityDept of Emergency Medicine Category of Sub-mission: Student, Resident, Fellow

Background: Given Paramedic National Stan-dard Curriculum cognitive objectives, wedeveloped an 8-hour curriculum that guideseducators and paramedic students (PS) throughthe scientific process and offers a simple wayto find and evaluate research articles. We thenevaluated its effect on PS perception of findingand evaluating research articles, and theirinterest in participating in future prehospitalresearch studies. Methods: PS participated infour 2-hour long journal club sessions. First, theeducator provided PS with four types of articlesand highlighted differences between formats.Next, PS used search engines to fact checkreferences of a free open access article. Third,PS sent articles on a topic selected by the classto the educator, who facilitated a discussion ofseveral articles after a short statistics lecture.Finally, PS found an article on a topic of theirchoice and verbally presented as it as if tellingtheir partner about it between calls. Before andafter the module, PS completed a survey withdemographic questions and a series of affectivedomain questions, with surveys linked usingunique identifiers. PS will receive a follow-upsurvey in one year. Results: A total of 21 PSparticipated. 81% were male, with an averageage of 24. 43% were college graduates. Beforethe module, 76% of PS could identify a researcharticle, 29% had a journal subscription, andmany read articles several times a month (38%)or year (33%). Affective survey questions hadfive-point Likert scale responses that were con-verted to numeric responses (strongly disagree= 1, strongly agree = 5) and analyzed usinga paired t-test, with p < .05 for significance.After the module, PS had significantly moreagreement that they could find research articles(p < .001) and are interested in attending ajournal club for their continuing education (p= .02). PS significantly disagreed more thatpatient care decisions should be based onpersonal experience instead of research basedevidence (p = .01). All PS agreed the modulewas a productive use and time and wouldrecommend it to others. Conclusions: Thiscohort of PS demonstrated their ability to find,interpret, and communicate important findingsin research articles, and had overall positivetrends in opinions about evidence-basedmedicine.

120. Double-Sequential Defibrillation:Efficacy and Risk of Defibrillator DamageAre Highly Dependent on the Choice ofShock Timing and Shock Vectors

Tyson Taylor, Sharon Melnick, Fred Chapman,Gregory Walcott, Physio-Control, Inc. Categoryof Submission: Cardiac

Background: Double-sequential defibrillation(DSD) is the use of two defibrillators for deliv-ery of two near-simultaneous shocks in anattempt to terminate refractory VF. We hypoth-

esized that the efficacy of DSD compared tocontrol shocks (one vector, one device, shocksize the same as each of the two DSD shocks)would depend on the time between the twoshocks. Furthermore, we hypothesized that thepotential for damaging a defibrillator duringDSD would depend on the choice of shock vec-tors. Methods: To assess shock efficacy, defib-rillation pads were applied in lateral-lateral(LL) and anterior-posterior positions in 10 anes-thetized pigs. Episodes of electrically-inducedVF were treated with a shock of a block-randomized therapy. Shock energy was chosento yield approximately 25% success for a singleLL shock. We compared LL stacked shocks (i.e.,a failed LL shock was repeated) and seven DSDshock intervals (Overlapping; 10, 50, 100, 200,500, 1000 ms apart), with n = 81 VF episodesper therapy. To assess the potential for damag-ing a defibrillator, two sets of pads were appliedin six different configurations (either approx-imately parallel or approximately orthogonaldefibrillation vectors). Ten 360 J shocks weredelivered from one set of pads while the volt-age across the second set of pads was measured.We compared the voltage coupling ratio (VCR):ratio of the measured voltage to the deliv-ered voltage. Results: Compared to stacked LLshocks, DSD shocks that were Overlapping, 10,and 100 ms apart significantly increased effi-cacy (p < 0.05), DSD shocks that were 50 msapart significantly decreased efficacy (p < 0.05),and DSD shocks 200, 500, and 1000ms apartwere not different. During DSD potential dam-age assessment, voltage of delivered shocks was1833±5 V and voltage across the second set ofpads ranged from 1.2 to 503 V; parallel vec-tors resulted in significantly higher VCR com-pared to orthogonal vectors (15.2 ± 0.6% vs.1.6±0.2%, p < 0.0001). Conclusions: The effi-cacy of orthogonal-vector DSD is highly depen-dent on time between shocks and can increase,decrease, or not change compared to stackedshocks on a single vector. Potential for defibril-lator damage during DSD can likely be mini-mized by choosing near-orthogonal defibrilla-tion vectors.

121. Biometric Analysis of ThoracolumbarMovement during Ambulance Transport

David Wampler, Ronald Stewart, Rena Sum-mers, Lawrence Roakes, Mike Shown, CraigCooley, Chetan Kharod, Tasia Long, BrianEastridge, The University of Texas Health ScienceCenter at San Antonio Category of Submission:Trauma

Background: Within the community of traumasurgeons, emergency medicine physicians andemergency medical services (EMS) providersresponsible for the care of injured patients, thereis mounting concern that the long spine board(LSB) does little to reduce spinal motion, andthat risk outweighs benefit. The purpose of thisstudy was to evaluate the movement of thethorocolumbar spine during ambulance trans-port, comparing different patient positions withand without LSB. We hypothesized that trans-port on a mattress with the head of the bedelevated would limit thoracolumbar movementmore effectively than a LSB. Methods: Thiswas a randomized 10-treatment adult healthyvolunteer crossover trial. Real-time 3D motionanalysis of the thoracolumbar region was mea-sured using a wireless motion tracking system.Positions analyzed included: on LSB at zero andten degree incline, and on EMS stretcher withhead elevated to 10, 30, 45, and 60 degrees.All subjects were fitted with a rigid cervicalcollar (c-collar) and headblocks when on LSB.Subjects on stretcher without LSB were fittedwith a c-collar and were transported with andwithout foam headblocks. Each subject under-went simulated ambulance transport over a city

street course at, or below, posted speed lim-its. The driver was blinded to the subject posi-tion. Composite volume of motion was mea-sured at the T12-L1 body area. Statistical sig-nificance was determined using t-test. Results:Nine healthy subjects participated, 66% weremale. Comparing movement between LSB andno LSB respectively, there was no statistical dif-ference in three-dimensional volumetric move-ment of the thorocolumbar spine (2 ± 0.6 mm3LSB vs. 4.7 ± 5 mm3 no LSB). The two posi-tions that allowed the lowest mean volume ofspinal movement were: head of the bed ele-vated to 10 degrees and 30 degrees with head-blocks adhered to the stretcher mattress (1.2 ±1.5 mm3 and 0.9 ± 0.5 mm3, respectively). Con-clusions: In healthy volunteers thorocolumbarspinal motion was limited in all groups and notcontingent upon use of LSB. These data sup-port the assertion that the long spine board isnot superior for immobilization, and that moreinvestigation should be performed to evaluateoptimal thoracolumbar immobilization.

122. Supraglottic Airway Utilization vsEndotracheal Intubation Pre/PostDeployment of the I-Gel LMA in a LargeGround and Air-Based EMS Service

John Lyng, Michael Perlmutter, Alex Tremb-ley II, Marc Conterato, Michaela West, NorthMemorial Health Ambulance and Air Care Cat-egory of Submission: Operations, Quality,Safety Systems, Disaster

Background: Identify changes in invasive air-way management using supraglottic airways(SGA) and endotracheal intubation (ETI) asprimary and secondary interventions follow-ing transition from the King LTS-D to the i-gel LMA in an EMS setting. Methods: This isa retrospective observational study performedin a US-based ground/air EMS performing86,000 transports annually. Charts document-ing an attempt at invasive airway placementover a 12 month period were abstracted for age,gender, airway indication, type(s) of invasiveairway device(s) attempted, number of place-ment attempts, and placement success. Twocohorts were defined: cohort “K” represent-ing King LTD and cohort “I” representing i-gel LMA. ETI was continuously available. Pri-mary endpoint was number of airways success-fully managed using an SGA. Secondary end-points included rate of use of invasive devicesbased on clinical indication and use of devicesas primary or secondary interventions. Descrip-tive statistics were utilized. Results: A total of660 charts were abstracted, 259 cohort K and401 cohort I. Age (57.5 +/− 21.9y), and gen-der (63.5% male) were consistent across cohorts(p = 0.07 and 0.81, respectively). Acuity wassimilar across cohorts. SGAs were the primarydevice in 1.9% of cohort K and 37.9% of cohortI, and the secondary device in 10.4% of cohortK and 10.2% of cohort I. Success for first devicewas ETI 84.0% and SGA 40% in cohort K, andETI 80.1% and SGA 92.7% in cohort I. Finalsuccessful device in cohort K was ETI 87.3%,SGA 11.1%, and in cohort I was ETI 54.6%and SGA 44.7%. Successful airway manage-ment was achieved using any invasive deviceat 94.2% in cohort K and at 98% in cohort I (p= 0.015). Conclusions: Deployment of the i-gelLMA improved invasive airway managementin this EMS service, achieving a 4% increase insuccess, and a final 98% overall success rate.Introduction of the i-gel resulted in an increasein use of SGAs as a primary device, and neu-tral effect on use of SGAs as a secondary device.Despite that successful invasive airway man-agement by any device improved following i-gel deployment, erosion of ETI skills is identi-fied as a potential collateral effect that requiressurveillance.



123. I Love My Community Paramedic:Patients Report OverwhelmingSatisfaction with Community ParamedicProgram

Tia Radant, Paula Miller, Joseph Pasquarella,Ann Majerus, Jennifer Murphree, StephenBloomstrand, Aaron Burnett, Regions HospitalEMS Category of Submission: Operations,Quality, Safety Systems, Disaster

Background: Patient satisfaction is a keyindicator of healthcare quality. CommunityParamedic (CP) is an emerging professionand as such is there is limited data on patientsatisfaction with CP programs. Hypothesis:Patients enrolled in a 30-day post-dischargecommunity paramedic program report highsatisfaction with both the program and the careprovided by the CP. Methods: Inpatients with adiagnosis of CHF were offered post-dischargehome visits by a CP for up to 30 days afterdischarge. Inclusion criteria required that thepatient was a local resident, not eligible forhome-health services upon discharge, diag-nosis of CHF, English speaking, and written,informed consent to home visits by a CP. TheCP visited the patient in the home 1–2 timesper week for 4 weeks following discharge.At the final visit the patient was surveyedto assess their satisfaction with the program.Scores ranged 1–4, 1 being “very dissatisfied”and 4 being “very satisfied.” Mean scoresfrom the Likert scale were analyzed and arereported descriptively. Results: A total of 59patients completed surveys regarding theirsatisfaction with the program. Mean scores foreach question were as follows: willingness tolisten carefully to the patient (4.0), time takento answer patient questions (4.0), amount oftime spent with the patient (4.0), explainingthings in a way the patient could understand(3.95), instructions regarding medication andfollow-up care (3.97), thoroughness of theexamination (4.0), advice given on ways to stayhealthy (3.94), and overall satisfaction level(4.0). 100% of patients responded they wouldrecommend the community paramedic serviceto others. Conclusions: Patients providedoverwhelmingly positive feedback on the CPprogram. Patient’s open responses included: “Iwas glad that they were here the first day thatI got out of the hospital.” “When I got out ofthe hospital I was just so messed up, I had allthese drugs and stuff, and she went throughthem and got everything worked out. It madea big difference; I was so overwhelmed at thattime.” This study is limited by the small samplesize. We hope to maintain these results as theprogram continues.

124. Evaluation of Educational Methodsfor Prehospital Medical Command (PMC)Training for Emergency MedicineResidents (EMRS)

Jeffrey Luk, Cristina Carpintero, StephanieGaines, Amy Pound, University Hospitals Cleve-land Medical Center/CWRU School of MedicineCategory of Submission: Operations, Qual-ity, Safety Systems, Disaster

Background: PMC is a crucial part of EMRtraining. This skill can be difficult to teachin predominantly off-line prehospital systems,and training for PMC may not be standardizedacross programs. The purpose of this study wasto evaluate a phased comprehensive PMC cur-riculum for EMRs. Methods: Setting: Tertiaryacademic medical center. Participants: EMRs.Design: Subjects were taught PMC in phasesconsisting of (1) lecture; (2) review of PMC calls;and (3) simulated PMC calls. A survey wasgiven pre-training and after each phase usinga Likert Scale to assess comfort with medicalcommand (MC), refusal of medical assistance

(RMA), and field termination (FT) along withfamiliarity of protocols (FP) and phase useful-ness (PU). Mean and median Likert scores forthese categories were compared among phasesusing the t-test and Mann-Whitney test, respec-tively, with statistical significance set a pri-ori at < 0.05. Participants indicated the mostuseful and instructional phase. Pre- and post-tests were given to evaluate changes in knowl-edge. Mean and median test scores were sim-ilarly compared. The IRB deemed this studyexempt. Results: Statistically significant differ-ences were found in all comparisons, except forFT, FP, and PU from phase 1 to 2 and fromphase 2 to 3, and for MC and RMA from phase2 to 3. For the former group, statistically signif-icant differences were found from phase 1 to 3.A statistically significant increase was found intest scores (mean 50% to 65%, median 40% to67%). Participants found phase 2 the most use-ful and simulated calls the best way to learnPMC. Conclusions: A statistically significantincrease in Likert scores was found in all cat-egories from pre-training to completion of allphases. The study found a possible cumulativeeffect of phases 2 and 3 for FT and FP, sug-gesting a benefit from the addition of simulatedcalls to review of PMC calls alone. The statisti-cally significant increase in test scores demon-strated an increase in PMC knowledge from thetraining. One limitation was the lack of a consis-tent population due to EMR schedules. Furtherresearch should provide the training over oneday to ensure consistency.

125. Effect of Transport Time Interval onNeurological Recovery afterOut-of-Hospital Cardiac Arrest inPatients without a Prehospital Return ofSpontaneous Circulation

Jeong Ho Park, Yu Jin Kim, Young Sun Ro,Sola Kim, Sang Do Shin, Kyung Jun Song, SoYeon Kong, Ki Jeong Hong, Sun Young Lee,Department of Emergency Medicine, Seoul NationalUniversity Hospital Category of Submission:Cardiac

Background: Longer transport can adverselyaffect the outcomes of out-of-hospital cardiacarrest (OHCA) patients without return of spon-taneous circulation (ROSC), and those effectscan be aggravated when resuscitation effortsat the scene are insufficient. The aim of thisstudy was to determine the association betweenthe transport time interval (TTI) and neurologicoutcomes in OHCA patients without ROSC.Methods: We analyzed 57,902 adult OHCApatients with presumed cardiac etiology andwithout prehospital ROSC. The primary expo-sure was TTI, which was categorized as short(1–5 min), intermediate (6–10 min), and long(�11 min). The primary outcome was good neu-rological recovery at discharge (cerebral perfor-mance category 1 or 2). Multiple logistic regres-sion analysis was used, and the final modelincluded an interaction term between TTI andscene time interval (STI). Results: Among thepatients, 40%, 36%, and 24% were classifiedas short, intermediate, and long TTI, respec-tively. Good neurological recovery occurred in1.0%, 0.6%, and 0.3% of the short, intermedi-ate, and long TTI groups, respectively. Refer-encing the short TTI group, the adjusted oddsratios (aORs) [95% confidence interval (CI)] ofTTI for good neurological recovery was 0.58(0.47–0.73) for intermediate TTIs and 0.30 (0.21–0.41) for long TTIs. In the interaction model, theaOR of TTI for good neurological recovery wassmaller in the 1- to 5-min STI group than in the�6-min STI group. Conclusions: A longer TTIadversely affected the likelihood of good neu-rologic recovery among OHCA patients with-out prehospital ROSC. This negative effect wasintensified when the STI was short.

126. Community Paramedic Point of CareBlood Analysis: Validity and UsabilityTesting of Two Commercially AvailableDevices

Ian Blanchard, Ryan Kozicky, Dana Dal-garno, Stacy Goulder, Suzanne Snozyk,Karen Leaman, Susan Biesbrook, LenorePage, Lyle Redman, Keith Spackman,Tyler Williamson, Eddy Lang, Christo-pher Doig, Gerald Lazarenko, Alberta HealthServices/University of Calgary Category ofSubmission: Professional

Background: Community Paramedics (CPs)require access to timely blood analysis in thefield to guide treatment. Point of care test-ing (POCT), as opposed to traditional labora-tory analysis, may offer a solution, but lim-ited research exists on CP POCT. Purpose: Inthe CP setting, to assess the validity of twodevices (Abbott i-STAT and Alere epoc) andcontrast their usability. Methods: In a CP pro-gramme responding to 6,000 annual patientcare events, a split sample validation of POCTagainst traditional laboratory analysis for sevenanalytes (sodium, potassium, chloride, creati-nine, hemoglobin, hematocrit, and glucose) wasconducted on a consecutive sample of patientsrequiring blood analysis. The difference of pro-portion of discrepant results between POCTand laboratory was compared using a twosample proportion test. Usability was anal-ysed by survey of CP experience, linear mixedeffects model of Systems Usability Scale (SUS)adjusted for experience, expert heuristic eval-uation of devices, device-logged errors, andcoded observations of quality control testing.Results: Of 1,649 study period patient careevents, 174 had a blood draw, with 108 events(62.1%) enrolled from 73 participants. Partici-pants had a mean age of 58.7 years (SD16.3);49% were female. In 4 of 646 (0.6%) individ-ual comparisons, POCT reported a critical valuebut the laboratory did not; occurring moreoften in i-STAT (0.9%; 95%CI: 0.0%,1.9%) com-pared to epoc (0.3%; 95%CI: 0.0%, 0.9%; p =0.323). There were no instances of the labo-ratory reporting a critical value when POCTdid not. In 88 of 1,046 (8.4%) individual com-parisons the a priori defined acceptable differ-ence between POCT and the laboratory wasexceeded; occurring more often in epoc (10.7%;95%CI: 8.1%, 13.3%) compared to i-STAT (6.1%;95%CI:4.1%, 8.2%; p = 0.007). Eighteen of 19CP surveys were returned, with 11/18 (61.1%)preferring i-STAT over epoc. The i-STAT had ahigher mean SUS score compared to the epoc(84.0/100 vs. 59.6/100; p < 0.011). Fewer fieldblood analysis device-logged errors occurred ini-STAT (7.8%; 95%CI: 2.9%,12.7%) compared toepoc (15.5%; 95%CI: 9.3%, 21.7%; p = 0.063).A possible explanation may relate to usabilityissues with the epoc cartridge and test menus.Conclusions: CP programs can expect validresults from POCT in most instances, howeveran important discrepancy between traditionallaboratory did occur. Usability assessment sug-gests a preference for i-STAT.

127. Characteristics of ParamedicGraduates Who Retest after anUnsuccessful Attempt on a NationalCognitive Examination

Ashley Larrimore, Rebecca Cash, RemleCrowe, Madison Rivard, William Krebs,Jeremy Miller, Ashish Panchal, Department ofEmergency Medicine, The Ohio State UniversityWexner Medical Center Category of Submis-sion: Operations, Quality, Safety Systems,Disaster

Background: Paramedic program graduatesinvest significant time and effort in complet-ing their training. However, some graduates are



unsuccessful on the national paramedic certifi-cation examination on their first attempt. Theproportion of paramedic graduates who do notretest, despite available attempts, is unknown.The objective was to describe paramedic grad-uates who do not retest and their associatedcharacteristics. We hypothesized that few grad-uates chose not to retest and retesting was notassociated with specific candidate characteris-tics. Methods: We conducted a cross-sectionalevaluation of the national paramedic certifica-tion cognitive examination results for the classof 2013. This computer adaptive test terminateswhen the 95% confidence interval surroundingthe estimate of the candidate’s ability is entirelyabove or below the passing standard. Testlength ranged from a minimum of 80 to a max-imum of 150 questions. Unsuccessful testerswere defined as candidates who had a grade offail or incomplete (did not finish the examina-tion) on their first examination attempt. Grad-uates of military only training programs wereexcluded. Chi-square tests, Wilcoxon Rank Sumtest, and two tailed independent t-test wereused to compare demographics and individ-ual performance on the examination betweensuccessful and unsuccessful testers. Results: In2013, 11,090 paramedic graduates attempted thenational paramedic cognitive examination forthe first time with an overall pass rate of 73%.Paramedic graduates who failed were morelikely to be maximum length testers (38%, N= 1,148) than minimum length testers (29%,N = 892). Most graduates who were unsuc-cessful chose to retest (89%, N = 2,697). Therewas no clinically significant difference in themedian age (28 vs. 29 years, p = 0.0156) orrace/ethnicity (white, non-Hispanic 88% vs.minority 89%; p = 0.706) of students who choseto retest. Female students (86%, N = 734) wereless likely to retest than male students (90%, N =1,911, p = 0.001). Conclusions: The majority ofgraduates who were unsuccessful on their firstattempt retested on the national paramedic cog-nitive examination with female graduates hav-ing lower retest rates. This study was limited bythe lack of graduate specific information con-cerning their reasons for retesting. Future stud-ies will need to focus on the individual charac-teristics which affect whether graduates chosenot to retest.

128. Interaction Effect between BystanderCardiopulmonary Resuscitation andCommunity Urbanization Level onOutcomes after Out-of-Hospital CardiacArrest

Jeong Ho Park, Young Sun Ro, Sang DoShin, Kyung Jun Song, Ki Jeong Hong, SoYeon Kong, Dong Sun Choi, Gwan Jin Park,Department of Emergency Medicine, Seoul NationalUniversity Hospital Category of Submission:Cardiac

Background: Positive association betweenbystander cardiopulmonary resuscitation andoutcomes of out-of-hospital cardiac arrest(OHCA) are reported. There are various dif-ferences of sociodemographic and EMS factorsbetween rural areas and urban areas. The aim ofthis study was to investigate whether the effectof bystander CPR on outcomes after OHCAdiffered by urbanization level of community.Methods: This study was a cross-sectionalstudy using a nationwide EMS-based OHCAregistry in Korea. We included adult witnessedOHCA patients with presumed cardiac etiol-ogy from 2013 to 2015. The primary exposurewas bystander CPR categorized into 3 groups:bystander CPR with dispatcher assistance,bystander CPR without dispatcher assistance,and no bystander CPR. The endpoint wasgood neurologic recovery at discharge. Wecompared outcomes between bystander CPR

group using multivariable logistic regressionwith an interaction term between bystanderCPR and community urbanization level (Ruralvs Urban). Results: Among 108,253 patients,53,528 patients were included. 49.1% receivedbystander CPR (12.8% without dispatcherassistance and 36.3% with dispatcher assis-tance), and 50.9% received no bystander CPR.Good neurological recovery rate was 5.0% inbystander CPR with dispatcher assistance,5.5% in bystander CPR without dispatcherassistance, and 2.2% in no bystander CPRgroup. In the interaction model, the adjustedOR of bystander CPR for good neurologicalrecovery was different in urban areas [AOR(95% CI): 1.38 (1.17–1.63) without dispatcherassistance and 1.64 (1.44–1.86) with dispatcherassistance] and rural areas [AOR (95% CI): 2.80(1.33–5.92) without dispatcher assistance and4.46 (2.28–8.74) with dispatcher assistance].Conclusions: The effect of Bystander CPR andDA-CPR was more prominent in rural areasthan urban areas.

129. A National Description of the Use ofContinuous Positive Airway Pressure inthe Prehospital Setting

Rebecca Cash, Remle Crowe, Jeremiah Kins-man, Madison Rivard, Dave Bryson, GamunuWijetunge, Ashish Panchal, National Registryof Emergency Medical Technicians Category ofSubmission: Medical

Background: The use of continuous and bilevelpositive airway pressure (CPAP/BiPAP) is lim-ited to paramedics under the 2007 NationalEMS Scope of Practice Model. However, stateand local practices may vary and currentnational trends of CPAP/BiPAP use by otherEMS licensure levels is unknown. Our objec-tive was to describe use and outcomes ofCPAP/BiPAP by EMS licensure level nationally.We hypothesized that basic life support (BLS)only agencies use CPAP/BiPAP with similarpatient outcomes compared to agencies withadvanced life support (ALS) capability. Meth-ods: Using the 2012–2015 National EmergencyMedical Services Information Systems (NEM-SIS) datasets, we evaluated all records withCPAP/BiPAP use documented by EMS pro-fessionals in agencies with BLS-only responseversus a response with a combination of BLSand ALS (ALS-BLS). Only 911 responses wereincluded. Variables assessed included patientand response characteristics, additional pro-cedures performed, and cardiac arrest occur-rences. Chi-square tests were used to evalu-ate differences between BLS-only and ALS-BLSresponders. Results: There were 259,099 casesof CPAP/BiPAP use documented during thestudy period. Of these, 253,728 (98%) were per-formed by services with ALS-BLS responders.Most patients were 70 years or older (78%)and 49% were male. The most common inci-dent locations were residences (65%) and healthcare facilities (20%). The proportion of patientstreated by BLS-only responders who sufferedcardiac arrest after EMS arrival was signifi-cantly greater (4% vs. 0.5% for ALS-BLS respon-ders, p < 0.001) with a concomitant increasein the provision of chest compressions (BLS-only: 4%, ALS-BLS: 1%, p <0.001). BLS-onlyresponse agencies more frequently upgradedto lights and sirens during transport (7%) thanALS-BLS responders (2%, p < 0.001). Conclu-sions: Use of CPAP/BiPAP by EMS agencieswith BLS-only response occurred in 2% of cases.BLS-only responders documented more cardiacarrest events after EMS arrival than ALS-BLSresponders, although the reasons for this find-ing require further evaluation beyond the scopeof this dataset including geographical locationand patient population served. This evaluationlikely underestimates the use of CPAP/BiPAP

by BLS practitioners since the dataset is unableto separate combined BLS-only and ALS-BLSresponse agencies. Further work is needed tounderstand the trends of CPAP/BiPAP use byBLS EMS professionals.

130. Association Between BMI and Returnof Spontaneous Circulation inOut-of-Hospital Cardiac Arrest

Caitlin Howard, Jeremy Allen, DavidWampler, Hattie McAviney, Justin Smith,David Miramontes, Joan Polk, United StatesArmy and UTHSCSA Category of Submission:Student, Resident, Fellow

Background: Sudden cardiac arrest (SCA) con-tinues to be the leading cause of death in theU.S. Current studies suggest that there is nostrong correlation between BMI and resusci-tation rates. The objective of this study wasto evaluate what effect BMI has on the rateof return of spontaneous circulation (ROSC).Methods: This was a retrospective review ofan in-house cardiac arrest registry containingdetails of each resuscitation attempted by alarge, urban fire-based EMS system. Data wasanalyzed from January 1, 2016 through August15, 2016. The BMI recorded was a subjec-tive measurement obtained from the paramedicat the time of data collection. Patients wereincluded in the study if the following variableswere available: age, gender, BMI, and outcome(no ROSC vs. ROSC). Patients were excludedif age < 17, no age or gender recorded, noBMI data available, or no outcome available.Patients were divided into four groups basedon the recorded BMI (under, normal, over, mor-bid). An ANOVA test was utilized to analyzecontinuous variables and a χ2 test was usedto analyze categorical variables. Results: Therewere a total of 771 possible patients. 516 patientswere included in the analysis. The mean ageof the subjects was 63.08 + 17.96 years with319 males (61.82%). 64 (12.40%) patients wereunderweight, 224 (43.41%) patients were nor-mal weight, 168 (32.56%) patients were over-weight, and 60 (11.63%) patients were mor-bidly obese. There was no statistically signifi-cant difference in outcome (no ROSC vs ROSC)between the BMI categories (P = 0.37). Con-clusions: BMI did not have an association withrates of ROSC in this study. Our study did havelimitations. First, the BMI was a subjective mea-surement and not calculated. Second, the data isfrom a single system cardiac arrest registry andmay not be extrapolated to other systems.

131. Paramedics Providing Palliative Careat Home: Management of Pain andBreathlessness

Brianne Robinson, Alix Carter, Judah Gold-stein, Michelle Harrison, Marianne Arab, Dal-housie University Category of Submission:Student, Resident, Fellow

Background: Palliative care is aimed at alleviat-ing pain and distressing symptoms while offer-ing support. Paramedics routinely respond topalliative patients and can assist with symp-tom relief. In Nova Scotia, a novel clinicalpractice guideline was implemented enablingparamedics to assist families with home med-ications, collaborate with on-scene home careteams, or to administer opiates through anexpanded EMS formulary with the goal to treatat home if the patient desired. Paramedics com-fort with the dose and range of opiates forpalliative care is increasing. Our objective wasto describe paramedic medication administra-tion practices for the management of pain andbreathlessness. Methods: We conducted a ret-rospective review of 100 consecutive palliativecare responses from February 1, 2016 to June30, 2016. An electronic query would fail to cap-



ture assistance with home medications; a man-ual chart review including standard medicationadministration fields and the free-text narrativewas conducted to fully capture the care pro-vided. Descriptive analysis was conducted andresults were reported with n and % or mean andstandard deviation. Results: Study populationincluded 94 unique patients; 6 patients had 2–4 calls and the remaining had one. Paramedicsadministered medication to 58 (58%) patients,and of those 42 (72.4%) remained at home com-pared to 17/42 (40.5%) with no medication.Most common CC was pain; despite this, only36 (80%) pain patients received treatment and 6(13.3%) had both pre- and post-treatment painscores. Only 12 (44.4%) breathlessness patientsreceived medication. Paramedics assisted withhome medication 10 (17.2%), administered fromdrug kit 45 (77.6%) and both 3 (5.2%). Meanoral morphine equivalent was 13 ± 7.5 mg.Contact with an OnLine Medical Physician(OLMP) occurred during 57 encounters, andwas increased when medication was adminis-tered 46 (79.3%) compared to no medication 11(26.2%). Conclusions: Medication administra-tion would be underestimated in an electronicquery alone. Even with inclusion of assistancewith home medications, management of painand breathlessness may not be optimized. Pre-and particularly post-medication pain scoreswould confirm symptom control. Contact withOLMP when paramedics were not going toadminister medication should increase admin-istration and non-transport through increasedcomfort and confidence.

132. Ketamine Indications in StatewideTreatment Protocols

Christie Fritz, Christina Loporcaro, DavidSchoenfeld, Beth Israel Deaconess Medical Cen-ter/Harvard Medical School Category of Submis-sion: Student, Resident, Fellow

Background: Ketamine was discovered in the1960s, and since that time has been usedfor multiple indications including pain con-trol, procedural sedation, induction, depres-sion, and excited delirium/behavioral distur-bances. Ketamine has a more favorable hemo-dynamic profile than many of its alternativesfor the same indications. It can be adminis-tered through the intravenous, intraosseous orintramuscular routes. The purpose of this inves-tigation is to describe the overall prevalenceof ketamine in STPs and the indications forwhich it can be utilized. Methods: Cross sec-tional study of STPs for inclusion of ketaminein any protocols. Protocol revision date was alsocaptured. Results: Thirty-one out of fifty (62%)states issue ALS STPs, seven of which serve asguidelines. 48% of STPs include ketamine asan approved medication in their pharmacopeia.Ten states (32%) include ketamine for induc-tion during rapid sequence intubation, andfive states (16%) allow ketamine for proceduralsedation. Six states (19%) include ketamine intheir pain control protocols. Eight states (26%)have excited delirium protocols which includethe use of ketamine. One state also includesketamine as an agent for shivering. 60% of stateswhich include ketamine in their protocols onlyallow its use for one indication. 75% of protocolshave been revised since 2015 and all have beenrevised within the past 5 years. Conclusions:Ketamine is a versatile medication with a vari-ety of applications in prehospital care. Despitethis, less than half of STPs include ketaminein their pharmacopoeia, and the majority ofthose that include it have limited indications.Ketamine is a hemodynamically stable optionfor pain control or induction for RSI, but aminority of states with STPs include ketaminefor these indications. Ketamine has had a recentresurgence in emergency medicine, although as

most protocols have been revised in the last3 years, it is unlikely that protocol revision tim-ing has been a barrier to ketamine adoption intoSTPs. Further study is needed to examine thebarriers to introduction and indication expan-sion of ketamine in STPs.

133. Manual Syringe Aspiration andAdministration of Epinephrine byEmergency Medical Technicians forPrehospital Treatment of Anaphylaxis

Andrew Latimer, Sofia Husain, JonathanNolan, Vinod Doreswamy, Thomas Rea,Michael Sayre, Mickey Eisenberg, Universityof Washington Department of Emergency MedicineCategory of Submission: Student, Resident,Fellow

Background: In recent years, the costs ofepinephrine autoinjectors (EAIs) in the UnitedStates have risen substantially. In 2014,emergency medical services within a largeurban/suburban county in the United Statesimplemented the “Check and Inject” pro-gram to replace EAIs by teaching emergencymedical technicians (EMTs) to manually aspi-rate epinephrine from a single-use 1 mg/mLepinephrine vial using a needle and syringefollowed by prehospital intramuscular admin-istration of the correct adult or pediatric dose ofepinephrine for anaphylaxis or serious allergicreaction. Treatment was guided by an EMTprotocol that required a trigger and symptoms.We sought to determine if the “Check andInject” program was safely implemented byEMTs treating presumed prehospital anaphy-laxis or serious allergic reaction. Methods: Weconducted a prospective investigation of allcases treated as part of the “Check and Inject”program from July 2014 through December2016 in the suburban aspects of the County andJanuary 2016 through December 2016 withinthe major American city located within thecounty. All cases were prospectively collectedusing a custom quality improvement dataform completed by the first responding EMTs.Two physicians completed a structured reviewof each EMS medical record to determine ifthe EMTs followed the “Check and Inject”protocol and determine if epinephrine wasclinically-indicated based on physician review.Results: Of the 411 cases eligible for analysis,EMTs followed the protocol appropriately in367 (89.3%) cases. In the remaining 44 (10.7%)cases, the EMS incident report form failedto document either a clear inciting allergictrigger or an appropriate symptom from theprotocol list. Physician review determined thatepinephrine was clinically indicated in 36 of the44 cases. Among the remaining 8 cases (1.9%)that did not meet protocol criteria and were notclinically-indicated based on physician review,none had a documented adverse reaction to theepinephrine. Conclusions: We observed thatEMTs successfully implemented the manual“Check and Inject” program for severe allergicreactions and anaphylaxis in a manner thattypically agreed with physician review andwithout any overt identified safety issues.

134. Timely Treatment of Tiny Tummies: TheUse of Oral Ondansetron in thePrehospital Environment

Kelly Meehan-Coussee, Abhijit Srun-gavarapu, John Martel, Michael Bohanske, J.Matthew Sholl, Tania Strout, Maine MedicalCenter Emergency Medicine Division of EMS, TuftsUniversity Category of Submission: Student,Resident, Fellow

Background: Nausea and vomiting are com-mon emergency department (ED) complaints.While oral rehydration therapy is the preferred

treatment modality for dehydration, emesis is atherapeutic barrier. In 2013, Maine’s statewideEmergency Medical Services (EMS) proto-cols added oral ondansetron for paramedicadministration to children with nausea andvomiting, as unnecessary prehospital intra-venous (IV) catheter placement is associatedwith discomfort, prolonged scene time andincreased cost. Prehospital oral ondansetronadministration has not previously been eval-uated for clinical endpoints. Our objectivewas to evaluate the impact of prehospitaloral ondansetron administration to pediatricpatients on frequency of use, additional inter-ventions, ED length of stay, rate of hospitaladmission and ED recidivism. Methods: Weconducted a simple interrupted time-seriesanalysis to assess the effect of oral ondansetronavailability on study endpoints. Pediatricpatients transported via EMS to our tertiarycare pediatric referral center ED who receivedeither oral or IV ondansetron in the prehospitalsetting for nausea or vomiting from 2011–2015were included. Pre- and post-oral ondansetronprotocol implementation groups were com-pared using chi-square, Fisher’s exact or t-testas appropriate. Results: A total of 48 patientsmet inclusion criteria with a greater numbertreated in the post-protocol implementationperiod (34 vs. 14). A statistically significantincrease in the proportion of patients receivingoral ondansetron in the prehospital setting wasnoted following protocol implementation (0%vs. 47%, p = 0.002). This was associated with asignificant decline in the proportion receivingprehospital IVs (100% vs. 65%, p = 0.010) andprehospital IV ondansetron (100% vs. 53%, p =0.002). Significant changes in other prehospital(p = 0.521) or ED interventions (p = 0.741),length of stay (p = 0.253), hospital admissionrates (p = 0.161), or 48-hour ED return visits(p = 0.254) were not observed. Conclusions:The results of this study suggest that theavailability of prehospital oral ondansetronincreases the frequency of antiemetic use,decreasing the need for vascular access andimproving patient comfort. An increase in otherinterventions, hospital admissions, or returnED visits was not observed. Despite concernthat ondansetron may mask a medical or sur-gical emergency, this study suggests that pedi-atric patients treated with oral ondansetron pre-hospitally are not at increased risk of symptom-masking and subsequent return ED visits.

135. Use of a Community ParamedicineProgram to Address High Utilizers of the9-1-1 System

Thomas Grawey, Mario Colella, Steven Riegg,Michael Wright, Medical College of WisconsinCategory of Submission: Student, Resident,Fellow

Background: The role of communityparamedics (CP) has been expanding overrecent years. Many programs exist across thecountry, attempting to meet the unique needsof the local community. The Milwaukee FireDepartment (MFD) has created a CP programwhich addresses high utilizers of the 9-1-1system, attempting to decrease unnecessaryuse of resources and improve patient quality oflife. Objective: To determine if enrolling highutilizers of the 9-1-1 system in a one monthcommunity paramedicine program decreasedsystem usage. Methods: This is a retrospectivechart review. Data from MFD’s program in2016 was reviewed. 9 out of 12 months hadpatients enrolled in the program, varying from2–8 patients per month. Data was availableand analyzed based on month of enrollment inthe program. The number of 9-1-1 calls fromthe patients enrolled were reviewed with the6 months prior to participation compared to



6 months after the program was completed.Total hours of community paramedic contacttime was also monitored; 30 patients wereexcluded because they dropped out of theprogram prior to completion. Results: In all9 months of implementation there was a dropin number of 9-1-1 use during the 6 monthsafter completion of the program. August par-ticipants saw the least change, where 5 patientswho required a total of 54.1 CP work hours sawa decrease from 18 to 13 runs over a 6 monthperiod (−28%). In July there was the largestdrop in 9-1-1 usage of 77% (48 to 11), duringwhich time 4 patients were enrolled and 43.8contact hours were performed. In 2016 therewere a total of 47 patients enrolled in the pro-gram accounting for 337 calls pre-interventionand 149 calls (−56%) after 419.8 total hours ofCP care were performed. For every 27 minutesof care provided, one less 9-1-1 call occurred.Conclusions: Participation in a communityparamedicine program established to decrease9-1-1 utilization cut use by 56%. Limitationsinclude lack of information about nature of9-1-1 calls including which calls requiredhospital transport. A future study could look atcost savings provided by the program.

136. Understanding How TransactionalStress Relates to Stress Reactions andSafety Outcomes

Elizabeth Donnelly, Paul Bradford, CathieHedges, Matthew Davis, Doug Socha, PeterMorassutti, University of Windsor Category ofSubmission: Operations, Quality, Safety Sys-tems, Disaster

Background: Increasing attention is being paidto the impact of stress and fatigue on safety inparamedicine. Specifically, empirical linkageshave been established between fatigue, chronicwork stress, critical incident stress, and safetyoutcomes. However, the relationship betweentransactional stresses, stress reactions like post-traumatic stress, fatigue, and safety outcomes(safety compromising behaviors, medicationerrors and adverse events, and injuries or expo-sures) have not been assessed. There are twotypes of transactional stress. Internal transac-tional stresses are associated with the day today provision of service (e.g., offload delays,being placed on standby, dealing with dispatch,inappropriate use of EMS, mandatory over-time, and dealing with frequent service users).External transactional stress involves interact-ing with allied professions (e.g., law enforce-ment, Base Hospital, ER Charge nurses, ERPhysicians, ER primary nurses, and fire fight-ers). The purpose of this study was to see ifthere was significant variation in levels of trans-actional stress in paramedics that endorsedhigh levels of posttraumatic stress, fatigue, orreported negative safety outcomes. Methods:An online survey was conducted with ten Cana-dian paramedic services with a 40.5% responserate (n = 717). T-tests were used to assess for sig-nificant differences. Results: Analyses revealedhigh levels of internal and external transac-tional stress in those paramedics with high lev-els of posttraumatic stress (p < .001), thosewho reported being fatigued (p < .001), thosewho reported injuries or exposures at work[internal ambulance stress (p < .05), externalambulance stress (p < .001), safety compro-mising behaviors (p < .001), and medicationerrors (p < .05)]. Conclusions: These resultsindicate that there are higher levels of transac-tional stresses between paramedics that reportpathological levels of posttraumatic stress, sig-nificant fatigue, and negative safety outcomes.These exploratory analyses indicate that trans-actional stresses may influence the wellbeing ofparamedics. The ability to further break downand focus on the specific factors may offer

opportunities addressing posttraumatic stressand negative safety outcomes.

137. Prehospital Availability and Use ofMedications for Managing HazmatEmergencies

Kubwimana Mhayamaguru, Amber Bel-lafiore, Eric Lederer, Carl Youngs, RobertFrench, Joshua Gaither, Kristina Waters,Frank Walter, The University of Arizona Cat-egory of Submission: Student, Resident,Fellow

Background: A minimal amount is knownabout prehospital availability and use of med-ications to treat hazardous materials (hazmat)emergencies. The purpose of this study wasto identify the availability and use of hazmatmedications among paramedics with advancedhazmat training, practicing in prehospital set-tings in the United States (U.S.). Methods:An email Qualtrics® survey was sent to U.S.paramedics who completed the Advanced Haz-mat Life Support (AHLS®) Provider Coursefrom 1999–2017. The survey asked what spe-cific hazmat medications were available to eachrespondent, how they were carried, and howfrequently they had been used. For analysis,responses were grouped into those medica-tions with hazmat indications only and thosewith multiple uses. Availability and use of eachhazmat medication is reported using simpledescriptive statistics, including number (n) andpercent (%). Hazmat medications were con-sidered to have been used if the surveyedparamedic gave them anytime in the last fiveyears. Results: Of the 4,360 surveys sent, 784(18.0%) were completed. Of the completed sur-veys, 279 (35.6%) paramedics had dedicatedhazmat medication kits and 505 (64.4%) hadhazmat medications carried with other medica-tions. For those hazmat medications with haz-mat uses only, availability/use was: cyanideantidotes 463 (59.1%) / 36 (4.6%), atropine+ pralidoxime auto-injectors 376 (48.0%) /5 (0.6%), pralidoxime multi-dose vials 122 /(15.6%) / 3 (0.4%), and methylene blue 103(13.1%) / 5 (0.6%). The availability/use of haz-mat medications with other uses was: atropine513 (65.4%) / 63 (8.0%), calcium chloride 540(68.9%) / 83 (10.6%), calcium gluconate 247(31.5%) / 26 (3.3%), diazepam 498 (63.5%) /49 (6.3%), lorazepam 262 (33.4%) / 18 (2.3%),midazolam 619 (79.0%) / 29 (3.7%), ophthalmictopical anesthetics 254 (32.4%) / 50 (6.4%), andtopical lubricating jelly 462 (58.9%) / 28 (3.6%).Conclusions: Among paramedics with AHLS®Provider training there is limited availabilityand use of hazmat medications. Although localscope of practice, financial, and other geo-graphical considerations likely contribute tothese results, further work is needed to iden-tify which medications should be available toparamedics to optimize the cost benefit ratio ofstocking and using hazmat medications.

138. Validation of a Prehospital Falls RiskAssessment Tool

Allison Infinger, Meghan Wally, Rachel Sey-mour, Jonathan Studnek, Mecklenburg EMSAgency Category of Submission: Trauma

Background: Every 15 seconds an older adultwill present to the emergency room with afall related injury. Prevention programs havedemonstrated efficacy; however, health careproviders must be able to identify at riskpatients. This study aimed to develop a con-tent valid and reliable assessment of environ-mental fall risk performed in the prehospitalsetting. Methods: First, we identified validateditems for screening extrinsic factors from the lit-erature. Then, a multidisciplinary expert panelcompleted two rounds of assessment using con-

tent validity index (CVI) scores to eliminateitems. The remaining items were revised forprehospital use and rated by EMS profession-als for clarity, relevance, and feasibility. Thedraft assessment tool was deployed for fieldtesting with two paramedics to determine thefeasibility and frequency of item identification.Following descriptive analysis and structuredinterviews, a second field test was conductedwith a revised tool. Paired crews completedthe assessment independently on low acuitypatients whose home they entered. Pair agree-ment on the final tool was measured usingCohen’s kappa. Results: Atotal of 87 items mea-suring extrinsic factors were identified. Roundone of content validity testing eliminated 33items (CVI � 0.76); 22 items were condensed orremoved due to redundancy. Round two elim-inated another 6 items (CVI � 0.70). Twenty-eight items were included in the initial EMSassessment and items with CVI scores � 0.70(n = 4) were eliminated. Twenty-two itemswere deployed for field testing. Round one offield testing (n = 12) revealed paramedics infre-quently accessing the kitchen (41.6%), bath-room (0.0%), or bedroom (25%) and excludedroom-specific items. Five crews completed 57paired assessments in round two using a nine-item tool. One item (κ = 0.8721) returned ahigh level of agreement, whereas the remain-ing items showed low to moderate agreement(κ = 0.3322–0.5369). Conclusions: A nine-item,content-valid, prehospital falls risk assessmenttool was created using a standardized process.After two rounds of field testing, the tool is notyet highly reliable. It is hypothesized that thelow agreement is due to the variation in priori-ties of providers on scene. Future efforts shouldtest the accuracy of extrinsic factor identifica-tion among secondary care providers only.

139. Development of a Hypoxic AsphyxialModel of Pseudo-Pulseless ElectricalActivity in Swine

Norman Paradis, Sarah Crockett, JeffreyGould, Christopher Kaufman, Karen Moodie,Dartmouth-Hitchcock Medical Center Categoryof Submission: Cardiac

Background: Pulseless electrical activity (PEA)is an increasingly prevalent initial rhythm incardiac arrest, particularly with in-hospital res-piratory arrests. Pseudo-PEA (p-PEA), whichoften precedes true PEA, is characterized by alow-flow state in which cardiac contraction pro-duces a non-palpable blood pressure, and is dif-ficult to treat. We set out to develop a repro-ducible, stable, and clinically relevant animalmodel of p-PEA for testing novel treatments.Hypothesis: Rapid induction of a hypoxicasphyxial state will result in a reproducible p-PEA state sufficient for study of pathophysiol-ogy and therapy. Methods: A state of p-PEAwas induced via progressive hypoxia in twelvedomestic swine ∼32 kg with standard physio-logical monitoring. Blood flow was measuredin the common carotid artery and jugular vein.FiO2 was reduced to 6% by increasing the frac-tion of nitrogen in inspired gas. A target systolicblood pressure (SBP) of 40 mmHg was usedto mimic p-PEA. After resuscitation, the ani-mal was stabilized. This cycle of hypoxic p-PEAand resuscitation was repeated until return ofspontaneous circulation could not be achieved.Results: p-PEA could be reliably created byhypoxic asphyxiation. In this model, p-PEAwascharacterized by a mean heart rate of 77 ±16 bpm, mean aortic blood pressure of 23 ±6 mmHg, mean right atrial pressure of 14 ±2 mmHg, mean carotid flow of 48 ± 16 mL/min,mean jugular flow of 10±5 mL/min, and meanintracranial pressure of 24 ± 3 mmHg. Time toachieve target systolic blood pressure was sig-nificantly less in the second round, however



the physiological responses were similar forboth rounds. Conclusions: A reproducible, sta-ble and clinically relevant porcine model of p-PEA via hypoxic asphyxiation was developed.Time to induction was reduced after multipleinsults. This model offers an improved methodfor testing innovative therapies for p-PEA.

140. Characteristics of Acute MyocardialInfarction Cases Coded as Low-Acuity atDispatch

Marie Gardett, Greg Scott, Chris Olola,Meghan Broadbent, International Academies ofEmergency Dispatch Category of Submission:Cardiac

Background: Identification of acute myocar-dial infarction (AMI) can be complicated bythe wide variety of symptomologies or pre-sentations. While the most common symp-tom of AMI is chest pain, so-called “atypi-cal” presentations are in fact quite commonand extremely variable, and AMI sometimespresents with very mild-seeming symptomssuch as flu-like chills and nausea, abdominalpain, or lightheadedness. Correctly identify-ing mild-seeming presentations that actuallyturn out to be AMIs can help ensure appropri-ate response and treatment. This study iden-tified hospital-confirmed AMI cases coded aslow-acuity at dispatch to determine whetherany common characteristics could help identifythese cases in the future. Methods: This was aretrospective study utilizing emergency med-ical dispatch (EMD), emergency medical ser-vices (EMS), and hospital discharge datasets.The study sample included all cases that arrivedto the hospital via EMS. Primary outcome mea-sures were the numbers of hospital-diagnosedAMIs categorized by patient age and gen-der, Chief Complaint Protocol, and dispatchdeterminant code; secondary measures werecomparisons between EMD- and EMS-recordedsymptoms. Descriptive statistics were used tocharacterize the distributions of all ALPHA-level cases and of ALPHA-level AMIs, cate-gorization by hospital discharge destinations,and Chief Complaint. Results: A total of 8,007ALPHA priority-level cases with correspond-ing hospital records were identified. Of these,40 (0.50%) were identified as AMIs. TheseALPHA-level AMI cases fell into only five ChiefComplaint Protocols (Sick Person, Falls, Uncon-scious/Fainting, Abdominal Pain/Problems,and Hemorrhage/Lacerations). Older age anddischarge to medical facility (rather than tohome or self-care) were identified with AMIcases. The most commonly reported symptomwas a fall, especially ground-level fall in anolder-age patient. Certain “sick person” char-acteristics were also somewhat associated withAMI diagnosis. Conclusions: Overall, the num-ber of AMI cases assigned to the ALPHA pri-ority level is very low and is confined tovery few Chief Complaint Protocols. In general,the ALPHA-coded AMIs in this study showedcharacteristics consistent with missed or silentAMIs widely described in other healthcare set-tings.

141. Heat Index Is the Main FactorInfluencing Rates of Patient Presentationat East Carolina University FootballGames

An Truong, Stephen Taylor, Roberto Portela,Kori Brewer, Brody School of Medicine at EastCarolina University Category of Submission:Student, Resident, Fellow

Background: Mass gathering events are largegatherings of greater than 1000 people whereaccess to patients is difficult and responseby emergency medical services (EMS) may bedelayed. Current literature suggests that mul-

tiple factors can influence patient presenta-tion rates during these events. Local emer-gency physicians and EMS provide medicalcare at East Carolina University (ECU) foot-ball games with a stadium capacity of 51,082.ECU football games are typically staffed by sixEMS units placed around the field’s perime-ter, one field-dedicated EMS unit, and 2 Med-ical Treatment Areas staffed with four physi-cians. Cooling tents are used as needed basedon weather forecasts for the game. Objective:This study aimed to quantify patient presenta-tion rates and factors influencing patient pre-sentation during ECU football games between2008 and 2016. Methods: A retrospective reviewof EMS field records and 9-1-1 incident numbersoriginating from the stadium on the dates andtimes of home football games from 2008–2016was conducted. JMP Version 13 (Cary, NC) wasused to conduct a bivariate correlation analysison the cumulative data set to determine rela-tionships between external factors and patientpresentation as well as emergency department(ED) transport rates per 10,000 attendees. Heatindex, attendance, and kickoff times were themain factors evaluated. RESULTS: Data from47 home football games with attendance rang-ing from 33,048 to 51,082 were included. Theheat index during the games ranged from 37.8to 89.6 °F. Kickoff times ranged from 1200 to2000 hours. Bivariate correlation analysis ofheat index and patient presentation was cal-culated as 0.432 (p < .05). This result sug-gests a positive correlation between heat indexand patient presentation rates. The correlationbetween heat index and rates of ED transportwas moderately positive at 0.316 (p < .05).The bivariate analysis of attendance and kickofftimes with patient presentation and ED trans-port rates showed little to no correlation withno statistical significance. Conclusions: Heatindex values were shown to have a moderatelystrong correlation with rates of patient presen-tation at ECU football games. There was nocorrelation between attendance at the footballgames, kickoff times, and patient presentationrates.

142. Reducing 9-1-1 Over-Utilizationthrough a Targeted Community ParamedicHospice Referral Program

Peter Antevy, Kenneth Scheppke, Juan Car-dona, Susan Toolan, Sharon Maraj, FrankBabinec, Julie Corona, Paul Pepe, MemorialHealthcare System Category of Submission:Medical

Background: Over-utilization of 9-1-1 systemsis a nationwide problem that overburdens EMSagencies and often results in hospital transportsbetter suited for other dispositions. For exam-ple, EMS professionals often are called to attendand transport patients who likely require out-of-hospital end-of-life care, yet still have unmethealthcare needs. The purpose of this studywas to evaluate if a community paramedic (CP)could successfully refer appropriate patientsto local hospice partners and thereby dimin-ish EMS responses for those patients. Meth-ods: Between April 1, 2015 and December 31,2016, front-line EMS crews, guided by estab-lished criteria, referred potential hospice can-didates to a single designated CP who visitedthose patients at their residence then referredthose meeting specified hospice criteria to ahospice partner (VITAS Healthcare) for enroll-ment. Demographics, diagnoses, length of stay(LOS), and outcomes were collected for patientsenrolled. The associated 9-1-1 utilization, beforeand after enrollment, was tracked and mea-sured. Results: The CP attended 320 poten-tial hospice patients over the 21-month period.Of the 136 patients seen in 2015, 42 (30.9%)were enrolled in hospice and, similarly, 64 of

184 (34.8%) seen in 2016 were also enrolled.Of those 106 total patients enrolled, 58 weremen and 48 were women. While ranging in agefrom 3 to 86 years, 95.2% (n = 101) were over68 and the main diagnoses involved includedCOPD, CHF, dementia and cancer. The aver-age combined LOS with hospice services was71 days and 23.5% (n = 25) of the 106 patientsused their full 6-month hospice benefit. Another11.3% (n = 12) are still enrolled. The total num-ber of 9-1-1 responses for this cohort (priorto hospice enrollment) had been 439. This fellto 17 after enrollment (a 96.1% reduction inrelated EMS utilization). Conclusions: Based onthis experience, it is concluded that communityparamedic programs can play a very importantrole in facilitating the care of hospice-eligiblepatients and thus help to avoid unneeded EMSsystem utilization for such patients. Appro-priate education of front-line EMS profession-als, working in synchrony with a designatedCP, can reduce unneeded 9-1-1 utilization, but,more importantly, facilitate the most appropri-ate and expert care through hospice-partnerresources.

143. Prehospital Provider Year of HireCorrelates to Time Spent On-Scene inEmergent Trauma

Clark Smith, Steven Hulac, Spencer Knierim,Zachary McDade, David Edwards, DenverHealth and Hospital Authority Category of Sub-mission: Trauma

Background: The definitive prehospital man-agement of critically-injured blunt or penetrat-ing trauma patients is rapid transport to atrauma center. Retrospective studies of traumaregistry data have indicated that prolonged on-scene times may worsen mortality in the mostcritically-injured patients. The preponderanceof available research suggests that optimal man-agement of these patients is the provision ofbasic stabilization measures while minimizingtime spent on-scene. The objective of our studywas to investigate if prehospital provider dateof hire was associated with time spent on-scene in patients transported emergently withtraumatic injuries. Methods: We conducted adata analysis of emergent transports of traumapatients by paramedics hired by our EMSagency during the years 2006 through 2015. Weexamined the on-scene times for these calls asrecorded through the agency’s computer-aideddispatch system, from January 2011 to June2017. We compared the mean on-scene timesfor paramedics over this period, aggregated byyear of hire. We excluded calls in which theprovider indicated a specific delay or barrierto care in the electronic patient care report.Results: During the study period, paramedicsfrom the included years of hire transported atotal of 2,910 emergent trauma patients. Thenumber of emergent trauma transports forparamedics from each year of hire range from179 to 380. Paramedics with earlier years of hirehave lower average on-scene times than thosehired later. Paramedics hired in 2006 average7.16 minutes on scene, while paramedics hiredin 2015 average 9.14 minutes on scene. Linearregression of this data yielded an R-squaredvalue of 0.82. Utilizing a one-way between sub-jects ANOVA, there was a significant effect ofyear of hire on average on-scene time at the p <0.05 level [F(2,2900) = 4.713, p < 0.001]. Conclu-sions: There was a distinct association betweenparamedic year of hire and on-scene times inemergent transports of trauma patients. This isthe first study comparing providers’ years inservice to their on-scene times with criticallyinjured patients. Further research is needed todetermine if this trend is seen in other similaragencies and to investigate its impact on patientoutcomes.



144. Benchmarking EMS Compass TraumaScene Times and Traumatic PainManagement Performance Measures Usinga National Dataset

Jeffrey Jarvis, Dustin Barton, Lauren Sager,Nick Nudell, Williamson County EMS Cate-gory of Submission: Trauma

Background: Minimizing scene times forpatients with critical trauma has long beenrecommended. Additionally, pain from trau-matic injuries is very common. Assessmentand management of this pain has been iden-tified as a key clinical performance measureby the EMS Compass initiative. There hasbeen little work done using national datato benchmark these measures. We sought todescribe the performance on these measuresusing a large commercial dataset. Methods:Using anonymous data from 9-4-1 consentingagencies, we analyzed 6 ½-years of data fromESO Solution’s electronic health record (HER)to calculate benchmarks for: (1) the percentageof patients with trauma alert criteria as definedby the CDC trauma triage criteria for transportto a trauma center who have a scene timeunder 10 minutes, and (2) of patients with anytraumatic injury, the proportion with at leastone pain scale documented. For those withan initial pain score >5, the proportion witha second score reassessing pain. Of patientsfrom ALS agencies who had an initial score>5, the proportion with decreased pain fromthe first to last pain score. We calculated boththe proportion and 95% Confidence Interval aswell as average, median and interquartile range(IQR) for time-based measures. Results: Of the66,414 critical trauma patients, 16,162 (24.3%,24.0–24.7%) had a scene time less than 10 min-utes. The average scene time was 16.4 min, IQR14.7(10.2, 20.2). Of 2,166,680 trauma patients,1,053,747 (48.6%, 48.6–48.7%) had a pain scoredocumented. Of 503,656 patients with initialscores of >5, 305,493 (60.7%, 60.5–60.8%) hada reassessment. Of the 310,737 patients of ALSagencies with a score >5, 64,076 (20.6%, 20.5–20.8%) had an improvement in pain scores.Conclusions: We provide the first benchmarkson critical trauma scene times and pain man-agement using a large national dataset. Theresults indicate additional efforts are needed,both for assessing/documenting traumaticpain and in addressing it. Additionally, scenetimes on critical patients are rarely under the“platinum” 10 minutes, indicating either needfor improvement or a more realistic goal.

145. Stop the Bleed: The Effect ofHemorrhage Control Education onLaypersons’ Willingness to RespondDuring a Traumatic Medical Emergency

Derek Brown, Elliot Ross, Theodore Redman,Julian Mapp, Kaori Tanaka, Chetan Kharod,Craig Cooley, David Wampler, SAUSHEC Mil-itary EMS & Disaster Medicine Fellowship Cat-egory of Submission: Student, Resident,Fellow

Background: The “Stop the Bleed” campaignadvocates for non-medical personnel to betrained in basic hemorrhage control. How-ever, it is not clear what type of educationor the duration of instruction that is requiredto meet that condition. The objective of thisstudy was to determine the impact of a briefhemorrhage control educational curriculum onthe willingness of laypersons to respond dur-ing a traumatic emergency. Methods: Thiseducation initiative was conducted betweenSEP 2016 and MAR 2017, and subjects wererecruited from multiple community groups ina large metropolitan area. Individuals with for-mal medical certification were excluded. Par-ticipants completed a pre- and post-education

questionnaire assessing personal comfort lev-els and their knowledge and attitudes abouttourniquets and responding to traumatic emer-gencies. Each training course included 20 min-utes of didactic instruction on hemorrhagecontrol techniques, encompassing indicationsfor tourniquets, and hands-on instruction withtourniquet application on both adult and pedi-atric mannequins. The primary outcome waswillingness to use a tourniquet in response toa traumatic medical emergency. Results: Of 236participants, 218 met eligibility criteria. Wheninitially asked if they would use a tourniquetin real life 64% (140/218) responded “Yes”. Fol-lowing training, 96% (194/203) of participantsresponded that they would use a tourniquet inreal life. Of participants who initially responded“No” (2%, 6/218), all responded “Yes” follow-ing training. Before training, men were statis-tically more likely to respond “Yes” to usingtourniquets than women (80.9% vs. 57.1%, p= 0.003), but that difference resolved follow-ing training. When participants were askedabout their comfort level with using a tourni-quet in real life, there was a statistically sig-nificant improvement between their initial andpost-training response (2.5 vs. 4.0, based on 5-point Likert scale, p < 0.001). Conclusions: Inthis hemorrhage control education study wefound that a short educational intervention canimprove layperson’s self-efficacy and reportedwillingness to use a tourniquet in an emergency.Significant gender differences exist in the statedwillingness to respond in emergencies. Identi-fied barriers to act should be addressed whendesigning future hemorrhage control publichealth education campaigns. Community edu-cation should continue to be a priority of the“Stop the Bleed” campaign.

146. Can Prehospital Providers CorrectlyTriage Patients to FreestandingEmergency Departments?

Charles Hwang, Desmond Fitzpatrick, JasonJones, University of Florida Department of Emer-gency Medicine Category of Submission:Student, Resident, Fellow

Background: Freestanding emergency depart-ments (FSEDs) are equipped to care for mostemergencies but do not have all the resourcesthat hospital-based emergency departments(EDs) offer. Emergency medical services (EMS)must routinely determine whether a FSEDis an appropriate destination. Inappropriatetriage may increase morbidity and mortalitydue to delay in definitive care. We soughtto evaluate paramedics’ ability in determin-ing whether a FSED is the most appropriatedestination. Methods: We conducted a retro-spective study of two county EMS agenciesand two FSEDs over more than 2 years. BothEMS agencies allow paramedic discretion indetermining transport destination; both proto-cols read, “Any patient potentially requiringadmission in the paramedic’s best judgment(Ex. elderly, weakness, dizziness, dialysis, etc.)will be EXCLUDED and not considered eligi-ble for transport to a FSED.” The primary out-come was whether paramedics can correctlyidentify patients that can be cared fully at aFSED without additional resources. We soughtto identify the percentage of patients broughtby EMS to FSEDs that were discharged with-out additional hospital-based services. Results:Between January 1, 2015 and February 6, 2017,1,247 EMS patients had a selected destinationof FSED. We excluded patients that did notarrive at their selected FSED destination, leftbefore FSED disposition, or were transferredfrom the FSED to unaffiliated hospitals. A totalof 1,184 patients were included for analysis, and885 (74.7%) did not require additional hospi-tal resources. Comparing the two EMS agen-

cies yielded similar results. Of note, multipleEMS narratives revealed that paramedics trans-ported patients to a hospital-based ED insteadof a FSED because the main hospital had moreresources. Conclusions: The primary goal oftriage is “determining how best to get the rightperson to the right place at the right time usingthe right amount of resources”. The burgeon-ing of FSEDs highlights the significance of thiscritical concept. As FSEDs become more popu-lar, a burden is frequently placed on paramedicsto determine which patients are appropriatefor specific emergency departments. Our studydemonstrated that paramedics have a reason-able ability to appropriately triage patients toFSEDs and to predict the need for hospitalresources.

147. Outcome Impacts of CommunityBystander Defibrillation VersusDispatcher-Assisted CPR (DA-CPR) inOut-of-Hospital Cardiac Arrest at PublicLocations

Patrick Chow-In Ko, Shih-Chieh Huang, Yu-Wen Chen, Hong-Yi Hsiao, Matthew Huei-Ming Ma, Chung-Liang Shih, National TaiwanUniversity, College of Medicine, Department ofEmergency Medicine Category of Submission:Cardiac

Background: We compared the outcomesbetween a community-wide bystander defib-rillation program and a DA-CPR program inpatients after out-of-hospital cardiac arrest atpublic sites. Methods: A prospective 2-yearcommunity-wide observational database col-lected from a metropolitan OHCA e-Registrywas studied, after a citywide bystander defibril-lation rescue program had been launched thatstrategically provided publicly accessed AEDs(automated external defibrillators) in desig-nated locations that were also e-registered; anda DA-CPR program had been run. The survivaloutcomes of OHCA at pubic locations betweenthe two program interventions were compared.Outcomes included 2-hour sustained ROSC(return of spontaneous circulation) at hospital,survival to hospital discharge, and good CPC(Cerebral Performance Category Scale 1 or2). All patient prehospital characteristics andoutcome relations were evaluated and adjustedby regression analysis. Results: The density ofpublic AEDs distribution increased from 3.96to 6.24 per square kilometers in the studied2 years. Among a total of 6,356 OHCA, 627patients occurred at public locations, including28 patients (male for 82%, witnessed arrestfor 79%) received bystander aid by publicAEDs plus CPR rescue and 243 patients (malefor 64%, witnessed arrest for 61%) receiveddispatcher-assisted CPR. For these 28 patients,53.6% (15/28) achieved prehospital ROSCat scene or during transport, 71.4% (20/28)achieved sustained ROSC after resuscitationat hospital, 57.1% (16/28) achieved survival-to-discharge and noticeably all those 16 (100%,16/16) survival-to-discharge patients achievedexcellent neurological outcome of CPC 1 (CPCScale 1). Their outcomes were significantlybetter [71.4 vs. 43.6%, OR: 3.2 (95%CI: 1.4–7.6)for sustained ROSC; 57.1 vs. 25.9%, OR: 3.8(95%CI: 1.7–8.5) for survival of discharge;57.1 vs 16.9%, OR: 6.6 (95%CI: 2.9–14.9) forgood CPC; and 100 vs. 65.1% for good CPCamong survival-to-discharge] compared withthose 243 patients by dispatcher-assisted CPRrescue. In 28 patients by bystander defibrilla-tion rescue only one man without prehospitalROSC still achieved survival-to-discharge andgood CPC. Conclusions: For OHCA patients atpublic locations, we found that a community-wide bystander defibrillation program wereassociated with excellent neurological outcomeof CPC 1 and survival to hospital discharge that



were significantly higher than those associatedwith dispatcher-assisted CPR program.

148. Randomized Trial of a ShearReduction Surface in AmbulanceTransport

Kathleen Berns, Ann Tescher, Lucas Myers,Patrick Koehler, Kip Salzwedel, HeatherMcCormack, Marianne Russon, Josh Bur-ton, Christine Lohse, Jay Mandrekar, EvanCall, Scott Zietlow, Mayo Clinic Category ofSubmissions: Operations, Quality, SafetySystems

Background: Shear is a known risk factor inpressure injury development such as decubitusulcers. The purpose of this study is to examinethe effectiveness of an anti-shear mattress over-lay (ASMO) in reducing shear/pressure andincreasing comfort on an ambulance stretcher.Methods: This was a randomized, cross-overdesign. Thirty adult volunteers in 3 BMI cat-egories served as their own controls. PRE-DIA shear/pressure sensors were applied tothe sacrum, ischial tuberosity (IT), and heel.The stretcher was placed in sequential 0°, 15°,and 30° elevations, with and without ASMO.The ambulance travelled over a closed courseachieving 30 mph, with 5 complete stops at eachhead of bed elevation for a total of 900 trials.Subjects rated discomfort on a 0–10 scale aftereach series of 5 runs. Results: Peak shear dif-ference between surfaces was −0.89, indicatingthat after adjusting for elevation, sensor loca-tion, BMI, starting pause peak shear levels were0.89 Newtons (N) lower for ASMO comparedwith standard surface (p = 0.057). Comparedwith 0°, elevations of 15° and 30° increasedthese levels by 2.41N (p < 0.001) and 3.44N (p <0.001), respectively. Using the sacrum as the ref-erence, IT and heel had increased shear levels of2.54N (p < 0.001) and 1.01N (p = 0.079), respec-tively. Peak pressure difference between sur-faces was −1.69, indicating pre-run peak pres-sure levels were 1.69 mmHg lower for ASMOcompared with standard surface (p = 0.070).Discomfort was lower on ASMO than standardsurface at 0° and 30° (p = 0.004, p = 0.014).Both surfaces had increased discomfort mov-ing from 0° to 30° (p = 0.005 and 0.039, respec-tively). Conclusions: ASMO reduced levels ofshear, pressure and discomfort. During trans-port, attention should particularly be given tothe heels and head of bed elevation.

149. Sleep Disorders Are Common RiskFactors for Occupational Injury

Matthew Weaver, Jason Sullivan, ConorO’Brien, Salim Qadri, Charles Czeisler, LauraBarger, Brigham and Women’s Hospital andHarvard Medical School Category of Submis-sion: Operations, Quality, Safety Systems,Disaster

Background: The rate of occupational injury inEMS is high and crashes are common. Fatiguehas been identified as an important risk fac-tor. Sleep disorders are common, often over-looked contributors to fatigue. We sought toexamine the prevalence of common sleep dis-orders and their impact on occupational safety.Methods: A nationwide cross-sectional studycollected data from 66 fire departments acrossthe US who participated in a workplace-basedsleep disorders screening and education pro-gram. Participants were screened for commonsleep disorders using reliable and valid screen-ing questionnaires and asked a series of ques-tions about adverse safety outcomes whichoccurred in the past month. The cooperationrate was 58.6%. For this secondary analysis,the dataset was limited to participants whoreported their primary responsibility as med-ical care and listed an EMT-Basic or higher

certification. The prevalence of common sleepdisorders is reported using descriptive statis-tics. The association between sleep disorderscreening result and safety outcomes was testedusing multi-level mixed effects logistic regres-sion models which accounted for clusteredresponses. Models controlled for individualand agency-level risk factors, including age,gender, body mass index, exercise frequency,years of experience, shift schedule, work at mul-tiple jobs, and annual call volume. Results:Responses were obtained from 2,992 fire-basedEMS providers employed at 65 departments.Most were male (93%), full-time employees(99%), who worked 24 hour shifts (77.2%). Onein three was obese (33.2%). Nearly half (45.1%),screened positive for at least one sleep disorder.Over 1/3 (33.9%) screened positive for obstruc-tive sleep apnea, 7.5% screened positive forinsomnia, and 10.1% screened positive for shiftwork disorder. More than 2/3 (71.6%) reportedsleeping less than 7 hours per night on aver-age and 33.8% had excessive daytime sleepi-ness. After controlling for potentially confound-ing variables, positive sleep disorder screeningwas independently associated with more thantwice the odds of an occupational injury (OR2.04; 95% CI 1.48–2.81), motor vehicle crash (OR2.10; 95% CI 1.12–3.93), and near-crash (OR 2.27;95% CI 1.94–2.66). Conclusions: Sleep disordersare highly prevalent among EMS providers.Sleep disorder screening may help to identifyproviders who are vulnerable to adverse safetyoutcomes.

150. Effectiveness of Manual Ventilationin Intubated Helicopter EMS TransportedTrauma Patients

Timothy Lenz, Brett McLachlan, Craig Bil-brey, Keith Mausner, Medical College of Wiscon-sin Category of Submission: Trauma

Background: Helicopter EMS agencies are fre-quently called to prehospital settings to trans-port intubated patients to definitive care at atrauma center. There is no current evidenceto inform the decision of ventilation in thispopulation. Current practice varies by groupfrom hand-operated bag-valve-mask (BVM) tomechanical ventilation. Our goal was to evalu-ate the effectiveness of manual BVM ventilatorysupport in our population of severely injuredtrauma patients. We hypothesized that man-ual control of ventilation will provide adequatesupport to maintain a physiologic end-tidal car-bon dioxide (ETCO2). Methods: This researchrepresents a prospective, observational, proofof concept study. Over a seven month periodof data collection (June 2015 to December 2015)and across the three distinct bases of our flightprogram, twenty patients were enrolled. Inclu-sion criteria for the study was limited to trau-matic mechanisms and patients endotracheallyintubated on scene and transported by heli-copter. Excluded were any interfacility trans-ports, non-scene calls, and any patient intu-bated with a supraglottic device. ETCO2 mon-itoring was accomplished with a ZOLL ProPacprogrammed to collect data at 30 second inter-vals for the duration of the flight. Additionalinformation on demographics and mechanismwas also collected. As a descriptive pilot study,there were no considerations of power; weenrolled all patients during the study periodwho met the inclusion criteria. Results: Thesubject group of 20 trauma patients was usedto collect data for over 500 cumulative minutesof manual ventilation. The percentage of cumu-lative time spent with adequate oxygen satura-tions (�90% Sp02) was 83.6%. The percentageof cumulative total time spent with adequateETCO2 (35–45 mmHg) was 48.7%, with 34.6%of time spent under this range and 16.7% abovethis range. Conclusions: Manual control of ven-

tilation via BVM was able to maintain a physio-logic ETCO2 only 48.7% of the time. There wassignificant variability, which resulted in inter-mittent hypoxia, as well as significant hyper-ventilation. Prior research has linked theseevents to increased morbidity and mortality.Further studies to compare similar data againstmechanically ventilated patients is warrantedbefore changes to practice can be made.

151. Development and Validation ofReality-Based Training ScenariosSimulating Violent EMS Encounters

Mallory DeLuca, Donald Garner, Jr., RemleCrowe, Rebecca Cash, Madison Rivard, Jeffer-son Williams, Ashish Panchal, Jose Cabanas,Wake County EMS Category of Submission:Professional

Background: Emergency Medical Services(EMS) providers are often exposed to violenceduring patient encounters. Traditional EMStraining may not adequately address appro-priate responses to potentially threateningsituations. Our objective was to develop andvalidate scenarios to evaluate EMS providers’response to threatening situations. We hypoth-esized that provider recognition and perceptionof threatening situations would not differ givendifferent patient presentations or aggressors.Methods: Using an iterative process, EMSphysicians, EMS educators and law enforce-ment training staff developed four simulationscenarios to assess provider responses tothreatening situations. Each scenario involvedpatient presentations and distractors that sim-ulated common high-stress EMS encounters.The scenarios were standardized for timing(8 minutes) and distinct phases of escalation(e.g., entrance of distractor, physical contactwith patient, physical contact with crew),with the same 51 data elements collected. Thescenarios used actors in an immersive, realistic,video-recorded environment. Role playersand evaluators attended a week-long courseto standardize simulation performance andassessment. Providers were told that they wereparticipating in a “patient care scenario” butotherwise blinded to the purpose of the simu-lation. Each provider participated in a singlescenario as a member of a two-person team. Theevaluator to participant ratio was 1:1. Charac-teristics were compared using chi-square tests.Results: A total of 272 EMS providers wereevaluated across the four scenarios: domesticabuse (n = 94, 35%), possible overdose (n =44, 16%), deceased mother (n = 68, 25%), andintoxicated homeless person (n = 66, 24%), with<3% missing data across elements. There wereno differences in participant characteristicsby scenario: certification levels (p = 0.96), sex(p = 0.28), and years of EMS experience (p =0.86). Most providers felt their scenario wasrealistic (n = 219/265, 83%) and this rating didnot differ across scenarios (p = 0.08). Overall,63% (n = 170/269) of providers stated thatif this scenario had occurred in real life, theywould have felt threatened, with no differenceacross scenarios (p = 0.31). Conclusions: Wecreated and validated four realistic scenariosfor prehospital providers that simulated threat-ening patient encounters with standardizedphases of escalation and data collection points.Future research should focus on evaluatingthe characteristics of threatening encounterphases that alert providers to the potential forviolence.

152. Paramedics Providing Palliative Careat Home: An Evaluation of ParamedicComfort and Confidence in ProvidingPalliative Support

Alix Carter, Judah Goldstein, Marianne Arab,Michelle Harrison, Wilma Crowell, Katherine



Houde, Jan Jensen, Mireille Lecours, JamesSullivan, Carolyn Villard, Kathryn Downer,Dalhousie University Category of Submission:Operations, Quality, Safety Systems, Disas-ter

Background: Paramedics are called for crisisand symptom management for patients receiv-ing palliative care. To address the mismatchbetween EMS current care and the patient’sgoals of care, a new program was implementedin two provincial EMS systems. Prior to pro-gram launch, all paramedics were trained inthe Learning Essentials Approach to PalliativeCare (LEAP) Mini for Paramedics. We eval-uated paramedic comfort and confidence todeliver palliative or end of life care. Methods:A prospective, cross-sectional electronic sur-vey was delivered before and 18 months aftertraining and program launch. A total of 1,255paramedics received an email invitation. Partic-ipants scored questions on comfort and confi-dence on a 4-point Likert scale, and attitudeson a 7-point Likert scale. Scores are reportedas Median (IQR). Wilcoxon ranked sum testedbefore and after differences. Open-ended ques-tions were thematically analyzed by one author.Results: Pre-launch, 235 (18.9%) responded;105were primary care paramedics (PCP) (44.7%).Post-launch, 267 responded (21.3%), 118 by(44.2%) PCPs. Paramedic comfort to providepalliative care scores improved: pre = 3 (IQR 1)to post = 3 (IQR 1) (p = 0.00009), where 4 = verycomfortable. Comfort to provide palliative carewithout transport increased: pre = 3 (IQR 1) vs.post = 3 (IQR 1), p =< 0.000001). Confidence inhaving the right interventions tools to deliverpalliative care increased from: pre = 2 (IQR 1)to post = 3 (IQR 0) (p =<0.000001); for carewithout transport to hospital: pre = 2 (IQR 1) topost = 3 (IQR 1), p =<0.000001). Respondentsstrongly agreed that all paramedics should beable to provide good basic palliative care: 7[IQR 6, 7]) and that a patient with an incur-able illness should receive palliative care: 6 [IQR4, 7]). Thematic analysis revealed paramedicsfeel delivering palliative care is rewarding,although additional experiential training, con-tinued expansion of the role of PCPs and addi-tional medications were recommended. Con-clusions: The palliative care training and addi-tional resources resulted in improved comfortand confidence. Paramedics strongly agree withparamedic administration of palliative care, citepalliative care as an important and rewardingpart of their job, and identified recommenda-tions for further training and scope.

153. Complications with Use of aTransport Ventilator with a King-LTDBased on Peak Airway Pressure

Leonard Weiss, Gabriel Diamond, ThomasSegerson, Justin Talarico, Francis Guyette,Christian Martin-Gill, Department of Emer-gency Medicine, University of Pittsburgh School ofMedicine Category of Submission: Medical

Background: Our prior pilot data demon-strated that mechanical ventilation during crit-ical care transport using the King Laryn-geal Tube Disposable airway (King-LTD) wasassociated with peak inspiratory pressures(PIP) above the manufacturer recommended 30cmH2O in almost half of cases. In the currentstudy, we aimed to determine prehospital andin-hospital complications associated with useof King-LTD when PIP with mechanical venti-lation is above or below 30 cmH2O. Methods:We retrospectively reviewed all King-LTD useswith mechanical ventilation in a large multi-state critical care transport service from Decem-ber, 2006 through November, 2015. Cases of dis-continuation of ventilatory efforts with King-LTD or missing PIP data were excluded. Pri-mary outcomes were the incidence of prehos-

pital complications (cardiac arrest, oxygena-tion or ventilation failure, emesis, and docu-mented air leaks) and where hospital outcomedata were available, the incidence of aspira-tion on radiologic studies (compared with chisquare tests). We secondarily compared prehos-pital oxygenation and ventilation parameters,in-hospital ventilator days, ICU days, hospitaldays, and in-hospital death, using descriptivestatistics. Results: Of 137 cases meeting inclu-sion, N = 93 (68%) were male, and average agewas 50 years (+/−18). Median initial PIP was30 cmH2O (IQR 24–40). In patients with PIP �30 cmH2O at any time (N = 74, 54%) vs. PIP<30 cmH2O (N = 63, 46%), final prehospitalvital parameters were SpO2 99.5 (IQR 96–100)vs. 98.5 (IQR 95–100) and ETCO2 35 (IQR 32–41)vs. 33.5 (IQR 29–38). Prehospital complicationsoccurred in 11 (8%) vs. 10 (7%) (p = 0.68). Of87 patients with in-hospital data with PIP � 30cmH2O (N = 46) or <30 cmH2O (N = 41), inci-dence of aspiration was N = 11 (23.9%) vs. N= 5 (12.2%) (p = 0.16). Median ventilator dayswere 4 (IQR 1–10) vs. 3 (IQR 1–11.5), ICU dayswere 5.5 (IQR 2–16) vs. 3 (IQR 2–19), and hos-pital days were 8 (IQR 3–22) vs. 7.5 (IQR 2–27).N = 19 (40.4%) vs. N = 13 (31.7%) died. Conclu-sions: Although confounders such as aspirationprior to airway placement may exist, these datasuggest that patients receiving mechanical ven-tilation via the King-LTD with PIP � 30 cmH2Ohave similar incidence of prehospital and in-hospital complications.

154. The Effect of IV vs. IO Access inPrehospital Cardiac Arrest ROSC Rates

Colby Redfield, Stephen Suarez, JessicaDaniels, Cristina Sanchez, Heidi Siples,Kim Landry, Leon County EMS Category ofSubmission: Cardiac

Background: The prevailing standard of care inprehospital emergency medical services is thateither intravenous (IV) or intraosseous (IO) areacceptable routes for obtaining vascular accessand delivery of resuscitation medications andvolume expanders in cardiac arrest patients.Our local EMS agency’s current cardiac arrestprotocol allows for either IV or IO access to beplaced without preference. Objective: To eval-uate the effectiveness of IV access versus IOaccess, in terms of Return of Spontaneous Cir-culation (ROSC), for patients suffering fromcardiac arrest. Methods: Quality Improvementretrospective review project examining cardiacarrest data with a single ACLS EMS agency withaverage call volume of 37,000 calls annually. Weexamined a four year period from 2013 to 2016.Cardiac Arrest patients were identified from aQuality Assurance Database. Exclusion criteriaincluded trauma arrest, pediatrics, pregnancy,and obvious signs of death. Method of vascularaccess was determined by reviewing the reportand placed into an excel spreadsheet along withROSC determination. Results: A total of 1,028patient care reports were examined from Jan-uary 1, 2013 to December 31, 2016. There were230 patients where resuscitation was not initi-ated due to obvious signs of death. A total of 46patients were excluded as trauma related car-diac arrests and 31 patients excluded due toage less than 18 years. A total of 721 patientsremained after applying the exclusion criteria.A total of 361 cardiac arrest patients had an IVplaced with a ROSC in 148 (41.1%). Atotal of 360cardiac arrest patients had an IO placed with aROSC in 80 (22.2%). IV use during cardiac arresthad improved ROSC when compared to IO use(p < 0.001). Conclusions: In this small retro-spective review, there is a correlation betweenhigher ROSC rates and IV access versus IOaccess. Limitations include small sample size,single EMS agency and retrospective nature ofstudy. Future studies should further evaluate

the effectiveness of IO vs IV access in cardiacarrest and other low perfusion states such asshock in a prospective manner.

155. Type of Airway Device Does Not AffectPhysiologic Markers In PatientsUndergoing Mechanical CPR: ThePrehospital Airway and Mechanical CPREvaluation Study

Torben Becker, Arjun Prabhu, Aric Bern-ing, Clifton Callaway, Francis Guyette, Chris-tian Martin-Gill, Torben Becker, University ofFlorida Category of Submission: Cardiac

Background: Mechanical chest compression(MCPR) devices and manual chest compres-sions achieve similar survival for patients without-of-hospital cardiac arrest (OHCA). How-ever, recent data suggest supraglottic airwaydevices (SGA) during MCPR may impair ven-tilation compared with an endotracheal tube(ETT). In this study, we tested whether mark-ers of oxygenation, ventilation, and perfu-sion differed between OHCA patients receivingMCPR with SGA and OHCA patients receivingMCPR with ETT. Methods: We retrospectivelyreviewed prehospital and in-hospital electronichealth records from three Emergency Medi-cal Services (EMS) agencies from January 1,2014 to December 21, 2016. We included allpatients with OHCA who underwent MCPRand who had their airway managed with anETT or SGA. The primary outcome was intra-arrest end-tidal carbon dioxide (etCO2) mea-surements. We also examined ventilation rates,vital signs upon return of spontaneous cir-culation (ROSC), as well as vital signs, lacticacid values, and venous or arterial blood gasresults in the emergency department (ED). Wealso recorded rates of ROSC and survival at24 hours, 30 days, and 90 days. Results: Of140 patients who received MCPR, valid datawere available for 126 patients. Of includedpatients, 84 (66.7%) had an ETT placed, and 42(33.3%) had a SGAplaced. Twenty-eight (22.6%)achieved ROSC. In-hospital data were availablefor 13 (10.3%) patients. There were no groupdifferences in etCO2 values during arrest, vitalsigns upon ROSC or ED arrival, or arterial orvenous partial pressure of oxygen, partial pres-sure of carbon dioxide, pH and lactic acid lev-els in the ED. There were no group differencesin ROSC or survival at 24 hours, 30 days, or90 days. Conclusions: We detected no differ-ence in markers of oxygenation, ventilation orperfusion and no differences in survival forOHCA patient managed with either an ETT orSGA in combination with MCPR.

156. The Use of Airway SimulationScenarios to Augment Systemic QualityImprovement Initiatives in a Fire-BasedEMS Agency

Eric Cortez, Tyler Smith, Andrew Little, RichLatham, William Krebs, James Davis, DavidKeseg, Ohio Health Doctors Hospital Categoryof Submission: Operations, Quality, SafetySystems, Disaster

Background: Airway simulation for prehospi-tal providers has several benefits, includingprovider exposure to low-frequency proceduresand identification of systemic quality improve-ment concerns. The objective of this study wasto analyze two airway simulation scenariosduring a two-hour paramedic airway course.We hypothesized that the simulation scenarioswould identify areas of focus for future qualityimprovement initiatives. Methods: This was aprospective evaluation of paramedics in an alladvanced life support (ALS) fire-based emer-gency medical services (EMS) system duringtwo simulated airway scenarios in a hospital-based simulation center. During each session,



teams of paramedics (4–6 individuals) man-aged one trauma patient and one acute decom-pensated heart failure patient. Trained EMSagency instructors and simulation center per-sonnel using a standard scoring sheet withpredefined data points evaluated teams. Theprimary outcome was successful endotrachealintubation. Secondary outcomes included sev-eral pre-intubation and post-intubation assess-ment and management steps. Descriptive statis-tics were reported as medians with interquar-tile ranges (IQR) and proportions. Results:A total of 375 paramedics participated in 61trauma scenarios and 74 heart failure scenarios.The median number of self-reported successfulintubations in the previous six months was 1(IQR 0–2). Successful intubation was achievedin 59 (97%) of the trauma scenarios and 73(99%) of the heart failure scenarios. End-tidalcapnography confirmation was performed in 60(98%) of the trauma scenarios and 73 (99%) ofthe heart failure scenarios. Preoxygenation wasperformed in 60 (98%) of the trauma scenar-ios and 72 (97%) of the heart failure scenarios.Basic airway maneuvers (repositioning, suc-tioning) were performed in 13 (21%) of traumascenarios and 31 (42%) of heart failure scenar-ios. In the heart failure scenario, allergies werereviewed in 10 (13.5%) encounters, and endo-tracheal tube dislodgement was recognized in57 (77%) encounters. Conclusions: This studyfound high intubation success rates during thesimulated scenarios, while other tasks, such asbasic airway maneuvers and reviewing aller-gies, were performed at lower than expectedrates. Developing quality improvement initia-tives is challenging for low-frequency proce-dures. This study exemplifies the utility ofairway simulation in helping to help guidequality improvement initiatives for large EMSagencies.

157. Incidences of Adverse ReactionsSecondary to the Administration ofHydroxocobalamin for Suspected CyanidePoisoning in the Prehospital Setting

Albert Arslan, Doug Isaacs, Pamela Lai,Matthew Melamed, Glenn Asaeda, DavidPrezant, Fire Department City of New York andNorthwell Health EMS Fellowship Category ofSubmission: Student, Resident, Fellow

Background: The objective of this study wasto assess the incidences of adverse reactionssecondary to the prehospital administration ofhydroxocobalamin to patients with suspectedcyanide poisoning after exposure to smokeinhalation. Exposure to fires involves a highmorbidity and mortality, in part by the cellu-lar asphyxiant cyanide - a byproduct of thecombustion of synthetic materials. Hydroxo-cobalamin, one of the most common antidotes,combines with cyanide to form a nontoxicmetabolite. Since 2009, our department hasadministered hydroxocobalamin in 239 cases,creating one of the largest prehospital caseseries for a single agency. Methods: This is aretrospective analysis of adverse reactions inpatients who were administered hydroxocobal-amin in the setting of suspected cyanide poi-soning by review of patient care reports as wellas hydroxocobalamin-specific questionnaires.Patients were separated into two study popu-lations: those in cardiac arrest, and those thatwere experiencing respiratory failure, alteredmental status, seizures, coma, or hypotensionof unknown etiology. Patients received 1.25–5 grams of hydroxocobalamin intravascularly.Adverse reactions recorded included erythema,nausea, seizures, headaches, allergic reactions,or increased blood pressures. Results: A totalof 239 patients, with ages ranging from 1–99 years and a median age of 52 years, of whom58% were male, were administered hydroxo-

cobalamin. Patients in cardiac arrest comprised36.8% of the patients studied and were excludedfrom subjective adverse reactions. For theremaining patients, one was observed to havenausea and another with post-administrationseizure. An increase in blood pressure wasnoted in 42.4% of the patients, with a change insystolic measurements between 1–106 mmHgwith a mean change of 13.9 mmHg (median7 mmHg, SD = 17.6 mmHg) and change in dias-tolic measurements between 1–77 mmHg witha mean change of 19 mmHg (median 10 mmHg,SD = 24.8 mmHg). Of these patients, 7.9% expe-rienced a clinically significant increase in bloodpressure that resulted with a value greater than180/110 mmHg. Conclusions: The administra-tion of hydroxocobalamin was associated witha low incidence of previously reported adversereactions when given in the prehospital settingfor the treatment of suspected cyanide toxicity.Limitations for this study include its retrospec-tive nature and its lack of hospital patient out-comes.

158. BIS: Bispectral Index Monitoring forPatients During Out-of-Hospital CardiacArrest

Ralph Frascone, Jeffrey Anderson, JosephPasquarella, Nicholas Loken, Sandi Wewerka,Regions Hospital EMS Category of Submission:Cardiac

Background: Progress in the treatment ofOHCA has resulted in a need to rapidly deter-mine the likelihood of neurological viability inpatients during CPR. End tidal (Et) CO2 lev-els have been used as a measure of circula-tion during CPR, however, EtCO2 is not predic-tive of neurological recovery. Based upon stud-ies in our animal laboratory, we hypothesizethat measuring processed electroencephalogra-phy (EEG) during CPR can be used togetherwith EtCO2 to determine if there are signs ofbrain electrical activity that may predict neuro-logically intact recovery from a cardiac arrest.The primary research question was to deter-mine if EEG activity alone or in combina-tion with another non-invasive measurement,EtCO2, could be used to predict the returnof spontaneous circulation (ROSC). Methods:This is a multi-agency, prospective, proof-of-concept, prehospital, cohort study to deter-mine the relationship between EtCO2 and BIS.Paramedics from three agencies were trained inthe application of BIS. Sensors were applied asearly as possible during resuscitation. BIS wasrecorded until the patient achieved ROSC orwas pronounced dead. The BIS monitor trans-forms the EEG waveform into a dimensionlesspercent ranging from 0 (complete cerebral sup-pression) to 100 (fully awake and alert). Datawas analyzed using descriptive statistics andunadjusted logistic regression. Results: Forty-two patients with BIS measures were enrolled.(ROSC) was achieved in 13 patients (31%). Nei-ther BIS at initiation of CPR (p = 0.513) orBIS nadir (0.975) was significantly associatedwith ROSC. 29/40 (73%) died prior to or dur-ing transfer to the ED. BIS measures at initi-ation of CPR (p = 0.973) or at nadir (0.285)were not significantly associated with mortal-ity. 2/11 patients who survived the ED transferhad BIS measures that fell below 5%. Similarly,among 40 patients with available data, ETCO2at initiation of CPR or at nadir did not sig-nificantly predict ROSC outcomes (p0 = 0.995;pnadir = 0.416) or mortality (p0=0.772; pnadir= 0.532). Using ETCO2 <5% as a stopping ruleonly would have achieved 91% sensitivity forsurvival, as one patient who survived ED trans-fer had ETCO2 readings <5 during monitoring.Conclusions: In this small study, neither BISnor ETCO2 monitoring are predictive of ROSCor survival through ED transfer.

159. Descriptive Analysis of PatientsAdministered Naloxone by PrehospitalProviders

Eric Cortez, Kaitlin Bowers, Judd Shelton,Andrew Little, Robert Lowe, Sam Kotran, OhioHealth Doctors Hospital Category of Submis-sion: Medical

Background: Emergency medical services(EMS) providers are administrating naloxonemore frequently and at higher doses. Theobjective of this study was to analyze patientsthat received naloxone by EMS providers. Wehypothesized that a proportion of prehospi-tal patients were administered naloxone inthe absence of apnea. Methods: This was aretrospective study of patients that receivedprehospital naloxone between October 1, 2015and March 31, 2016. All patients adminis-tered naloxone and transported to emergencydepartments (EDs) within the study’s health-care system were included. Patients wereexcluded if they were transported to EDsoutside of the healthcare system. The primaryoutcome was the presence of prehospital apneabefore naloxone administration. Secondaryoutcomes included the proportion of patientsdiagnosed with opioid overdose in the ED, andthe presence of prehospital unresponsiveness,miosis and hypoxia (< 94% pulse oximetry).Data points were defined a priori and a stan-dardized data sheet was utilized. Data werereported as percentages, and medians withinterquartile ranges (IQR). Results: A totalof 350 patients were included. The medianage was 45 years (IQR 31–56), and 61% weremales. The most common naloxone doses were2 mg (54%), 4 mg (26%), and 6 mg (7.2%).Of 347 patients with available prehospitalphysical exam findings, apnea was present in27%, unresponsiveness in 56%, miosis in 51%,and hypoxia in 17%. Final ED diagnosis wasavailable for 284 patients, and 128 (45%) werediagnosed with opioid overdose. Conclusions:In this study, a proportion of EMS patientsreceived naloxone in the absence of apnea andother signs of opioid toxicity. Furthermore,over half of the patients were not diagnosedwith opioid overdose in the ED. This datahighlights several important considerationsfor EMS naloxone administration: indications(obvious opioid toxicity vs. undifferentiatedoverdose), endpoints of therapy (reversalof apnea vs. confusion), need for re-dosing(potent opioid toxicity vs. non-opioid overdosewith partial response), and effects of naloxoneadministration in the setting of medical ortraumatic emergencies. Limitations include ahigh number of missing ED diagnoses, andexclusion of patients transported to otherhospitals.

160. Can Grip Technique and Bag SizeImprove Volume Delivered with aBag-Valve-Mask by EMS Providers?

Melissa Kroll, Jyotirmoy Das, Jeffrey Siel-ger, Washington University/ Barnes-Jewish Hospi-tal Category of Submission: Medical

Background: Emergency Medical Services(EMS) professionals rely on the bag-valve-mask(BVM) to provide life-saving positive pressureventilation in the prehospital setting. Multipleemergency medicine and critical care studieshave shown that lung-protective ventilationprotocols reduce morbidity and mortality. Arecent study has shown that the volumes typi-cally delivered by EMS professionals with theadult BVM are often higher than recommendedby lung-protective ventilation protocols. Ourprimary objective was to determine if a groupof EMS professionals could reduce the volumedelivered by adjusting the way the BVM washeld. Secondary objectives included (1) if the



adjusted grip allowed for volumes more consis-tent with lung-protection ventilation strategiesand (2) comparing volumes to similar gripstrategies used with a smaller BVM. Methods:A patient simulator of a head and thorax wasused to record respiratory rate, tidal volume,peak pressure, and minute volume deliveredby participants for 1 minute each across sixdifferent scenarios: three different grips (usingthe thumb and either three fingers, two fingers,or one finger) with two different sized BVMs(adult and pediatric). Trials were randomizedby blindly selecting a paper with the scenariolisted. A convenience sample of EMS providerswas used based on EMS provider and researchstaff availability. Results: We enrolled 50providers from a large, busy, urban hospital-based EMS agency a mean 8.60 (SD = 9.76)years of experience. Median volumes for eachscenario were 836.0mL, 834.5mL, 794mL for theadult BMV (p = 0.003) and 576.0mL, 571.5mL,547.0mL for the pediatric BVM (p < 0.001).Across all three grips, the pediatric BVM pro-vided more breaths within the recommendedvolume range for a 70kg patient (46.4% vs 0.4%;p < 0.001) with only a 1.1% of breaths belowthe recommended tidal volume. Conclusions:The study suggests that it is possible to alter thevolume provided by the BVM by altering thegrip on the BVM. The tidal volumes recordedwith the pediatric BVM were more consistentwith lung-protective ventilation volumes.

161. Retrospective Refinement andValidation of a Hypoglycemia DecisionTool for Paramedics

Julie Sinclair, Michael Austin, ShannonLeduc, Zachary Cantor, Richard Dionne,Penny Price, Justin Maloney, Andrew Reed,Andrew Willmore, Valerie Charbonneau,Christian Vaillancourt, Regional ParamedicProgram for Eastern Ontario Category ofSubmission: Medical

Background: Hypoglycemia symptoms areoften treated by paramedics in the prehospitalenvironment. Some evidence suggests thatnot all patients require transport to hospitalfollowing successful reversal of symptoms. Wesought to refine and validate a decision toolderived to identify patients that could safelybe assessed, treated, and not transported tohospital following paramedic care for hypo-glycemia. Methods: We conducted a healthrecord review of paramedic call reports andemergency department (ED) health recordsover a 6-month period (July 1, 2015–December31, 2015). Prehospital records were queried toidentify all adult patients with a prehospitalreading of <72 mg/dl (4.0 mmol/L) excludingcardiac arrests and terminally ill patients.We used standardized case report forms tocollect data. We defined short-term adverseevents as admission to hospital, repeat accessto paramedics/ED care, or death, occurringwithin 72 hrs of the initial prehospital hypo-glycemic event. The hypoglycemia decisiontool incorporates the following variables: oninsulin, not on corticosteroid/oral diabeticagent, no seizure disorder or cardiovasculardisease, and given CHO/protein. We per-formed descriptive, logistic regression analysisand test characteristics of the decision tool.Results: There were 392 included patientswith the following characteristics: mean age57.5 [range 18–97], male 55.9%, diabetic 72.5%,on insulin 60.2%, oral diabetic agents 10.7%,>1 paramedic encounter 18.6%; 247 (63.0%)were transported to hospital and 57 (14.5%)were admitted; 34 (8.7%) had repeat access toparamedic/ED care. A significant associationwas found between these patient characteris-tics and short-term events: renal disease, liverdisease, homelessness and on chemotherapy

agent; 60 (15.3%) patients met the revisedhypoglycemia decision tool for non-transport.Of these, 8.3% were transported to hospitaland all were discharged from ED with noadditional management for hypoglycemia;6.7% had repeat access to paramedics/ED carefor hypoglycemia and none were admitted.The sensitivity of the hypoglycemia decisiontool was 93.3%, specificity 17.8%, PPV 25.0%,NPV 90.0%. Conclusions: Demonstrating highsensitivity and NPV, this tool is potentiallysafe to rule out transport to hospital followingparamedic care for prehospital hypoglycemia.Further research is needed to prospectivelyvalidate the tool and evaluate its impact onprehospital and healthcare systems.

162. Fall Risk Inventory by ParamedicsPredicts Future Hospitalization and EDUtilization by Elders

Ryan Carter, Joanne McGovern, James Dziura,Fangyong Li, Geliang Gan, David Cone,Sandy Bogucki, Yale University Category ofSubmission: Medical

Background: One-third of community-dwelling elders fall each year. Previouswork showed that more than half of elderswho fall and activate EMS for “lift assists”without transport will activate EMS againwithin 30 days. Objective. To evaluate whetherseveral parameters assessed by a researchparamedic at a scheduled home visit predictED visits and hospitalizations within 90 daysin elders at risk for falls. Methods: For thisprospective study, informed consent to trackfuture healthcare utilization was obtained, andparticipants were enrolled via three pathways:9-1-1 activation for lift assist, ED visit, orself-referral. Participants had scheduled homevisits by research paramedics, who assessedhome safety and fall risk (a 15-item survey ofyes/no questions adapted for field use froma previously validated instrument), balance,and medical disability, and by a visiting nurse,who evaluated home health needs. Subsequenthealthcare utilization within 90 days after thevisiting nurse evaluation was identified byquerying electronic hospital records. A mul-tivariate analysis was performed, includingseveral of the research paramedics’ assess-ments plus race, sex, medication count, historyof prior healthcare utilization, and enrollmentpathway with the dependent variable beingED or hospital admission within 90 days.Results: Of 2,265 participants, 1,512 com-pleted their research paramedic and visitingnurse appointments, with at least 90 daysof subsequent observation. The median agewas 77, with 69% female, 19% black, and 11%Hispanic. 390 (25.8%) had an ED or hospitaladmission within the 90-day time period. In themultivariate analysis, significant independentpredictors of 90-day healthcare encountersincluded history of prior encounter (adjustedOR 2.94, p-value <0.0001), medication count(1.06, 0.0001), and fall risk (0.91, 0.0002). Inan analysis using the same variables withthe single outcome of 90-day hospitalization,these factors remained significant independentpredictors, with similar adjusted odds ratios.Conclusons: This study demonstrates thatthe fall risk inventory, along with medicationcount and history of previous encounter, is anindependent predictor of future healthcare uti-lization and hospitalization within 90 days. Thefield-adapted fall risk inventory is a simple toolfor paramedics to enhance the EMS assessmentof patients at risk of falls.

163. Factors Associated with a GoodOutcome Following PediatricOut-of-Hospital Cardiac Arrest in theYears Following the 2010 ResuscitationGuidelines

Paul Banerjee, Paul Pepe, Amninder Singh,Latha Ganti, Polk County Fire Rescue, PolkCounty, FL Category of Submission: Pedi-atric

Background: To determine which factors hadthe strongest association with good outcomesafter pediatric out-of-hospital cardiac arrest(POHCA) since 2010 when clinical practiceguidelines became more aligned with thoseused for adults. Methods: Conducted in alarge EMS urban/surburban jurisdiction thatuses a comprehensive Utstein-style database,all POHCA cases encountered over 5 calen-dar years (January 1, 2012 through December31, 2016) were analyzed for associated out-come correlations following full implementa-tion of the latest (2010) international guide-lines for childhood basic and advanced life sup-port. The analysis was used to identify currentpredictors for return of spontaneous circula-tion (ROSC), hospital admission (HA) and sur-vival to successful hospital discharge (SURV).Logistic regression models of traditional predic-tors were performed using JMP 12.0 for Mac.Results: Of 133 consecutive POCHA cases stud-ied, the interquartile range (IQR) for responseintervals was 16 to 47 minutes (range: 0–490) and the majority presented with asys-tole. As traditionally predicted, shorter timesfrom arrest to EMS arrival were associated (sig-nificantly) with ROSC, HA and SURV (all p< 0.0001) whereas bystander-witnessed arrestcases (only 13%) were not (p = NS). Still, in95% of cases, the arrest was identified by abystander prior to EMS arrival and, contraryto previous studies (with lower reported fre-quencies of bystander CPR), chest compres-sions were performed by bystanders in 59% ofcases. The earlier CPR was provided by EMSpersonnel was itself (significantly) associatedwith ROSC, HA and SURV (all P < 0.0001), butsome form of treatment before EMS arrival wasprovided in 54% of cases and such actions werestrongly associated with ROSC, HA and SURV(p < 0.0001 for all 3 outcomes) whereas AEDplacement (50% of cases) was not. Conclusions:Although “witnessed arrest” cases and AEDplacement were not identified as contribut-ing factors in this subpopulation of cardiacarrests (likely reflecting infrequent ventriculardysrhythmia etiologies), as expected, shorterelapsed intervals from the moment of arrest toEMS arrival; performance of CPR prior to EMSarrival; and, most importantly, any treatmentprovided before EMS arrival, all resulted in sig-nificantly higher rates of ROSC, hospital admis-sion and survival beyond hospital discharge.

164. Adherence to Quality CPR PrinciplesDuring the EMS to ED Handoff inSimulated Pediatric Cardiac Arrest

Ariel Cohen, Jen Anders, Jordan Duval-Arnould, UCSD Category of Submission:Pediatric

Background: The aim of this study is to quanti-tatively evaluate adherence to 2015 AHA guide-lines for quality CPR during the transition ofpatient care from EMS to ED. We hypothe-sized that quality would be compromised dur-ing this complicated period; as measured bypauses in chest compressions. Methods: Wesimulated the handoff and resuscitation of apediatric patient in a tertiary pediatric EDusing EMS and hospital volunteers. This wasa pilot study conducted over two, four hoursessions, where as many simulations as pos-sible were run. Simulation began with entryof the prehospital gurney in the ER hallwayand continued through first 10 seconds of dedi-cated compressions from ER staff on emergencydepartment bed. CPR recording defibrillatorscollected CPR data (chest compression pauses(sec), rate (cc/min), depth (in) and CC frac-



tion (CCF, %) throughout the scenario. Quali-tative assessment was performed using videorecording and post-simulation participant sur-veys. The primary outcome was number ofpauses in chest compression longer than 10 sec-onds. Secondary outcomes include analysis ofdepth and rate of compressions and qualita-tive feedback from participants about poten-tial for errors. Results: A total of 16 simulatedresuscitations were analyzed, with a total of16 minutes of CPR. Only two simulations, eachhad a total of one pause longer than 10 sec-onds. Average depth of compressions rangedfrom 0.5–1.2 in. Average rate ranged from 107–146 cc/min, with the majority of compressionsbeing above 120cc/min. Conclusions: Simu-lated CPR during EMS to ED handoff did nothave an issue with prolonged pauses. How-ever, the majority of the resuscitation did notmeet quality goals/2015 Pediatric BLS Guide-lines for depth and rate of compressions. Limi-tations include that this was a simulated resus-citation scenario and only one size mannequinwas used. Future studies observing real-timeresuscitation should evaluate the validity of thispilot study findings to possibly guide efforts toimprove resuscitation quality.

165. Factors Associated with PediatricInterfacility Transfer from EmergencyDepartments

Ali Aledhaim, Jon Mark Hirshon, JenniferFishe, Jennifer Anders, University of MarylandDepartment of Emergency Medicine Category ofSubmission: Pediatric

Background: In regionalized health systems,pediatric patients often require interfacilitytransfer (IFT) from an initial emergency care toa second acute care facility to reach definitivecare. IFT is associated with patient safety risks,delays in definitive care, and increased cost.EMS triage tools to guide pediatric destinationchoice should be developed to reduce the needfor IFT. Objective: To determine factors associ-ated with the likelihood of pediatric ED patientsundergoing Interfacility Transfer (IFT). Meth-ods: This study encompassed 3 years (2010–2012) of Maryland HCUP ED visit data. Weincluded patients 0–17 years of age with a dis-charge or transfer disposition. The analysis waslimited to visits classified as “emergent” and the20 most common Diagnosis Categories (DxC)associated with IFT. Factors assessed includedDxC, age, gender, race, and insurance type. Thelikelihood of IFT from the ED was evaluated byweighted multivariate logistic regression mod-eling design. Results: For the three-year period,146,995 pediatric ED patients were diagnosedwith one of the top 20 DxC emergent conditions;10,143 underwent IFT. All factors assessed werestatistically significant with varying effect sizes.The largest difference was seen between the top20 DxC medical (11.5% IFT) vs. trauma con-ditions (3.2% IFT, p < 0.001). Age was associ-ated with incremental increases in transfer rate.Compared to 0–4 yo, the ORs of IFT were 1.35,2.48, and 3.54 for 5–9 yo, 10–14 yo, and 15–17 yo,respectively (p < 0.001). In the adjusted logis-tic model, pediatric patients with medical con-ditions were 4.6 (4.41–4.85) times more likely tobe transferred than patients with trauma condi-tions (p < 0.001). African-Americans were 22%less likely to undergo IFT than Caucasians. Pri-vate insurance and self-pay had a higher ORof transfer than Medicaid, 1.08 (p = 0.002),and 1.51 (p < 0.001), respectively. Conclusions:For pediatric medical patients, IFT from EDto another acute facility for admission is morecommon than for trauma. EMS triage tools toguide destination choice for pediatric medi-cal patients may help reduce this discrepancy.Additionally, racial and insurance disparitiesexist for pediatric IFT from EDs.

166. Paramedics’ Perceptions of FocusedPoint of Care Cardiac Ultrasound

John Reynolds, Juan March, Roberto Portela,Steven Taylor, Bryan Kitch, Department of Emer-gency Medicine, Division of EMS, Brody Schoolof Medicine, East Carolina Category of Submis-sion: Student, Resident, Fellow

Background: Focused point of care cardiacultrasound (FOCUS) has been used success-fully in screening for many life threateningemergencies such as cardiac standstill, pericar-dial effusion, and others. There has been lim-ited research on paramedics’ ability to per-form FOCUS, but none looking at their per-ceptions. The goal of this study was to evalu-ate paramedics’ perceptions of FOCUS beforeand after an educational intervention. Meth-ods: A prospective study was performed in asuburban/urban setting with a population of180,000 and 26,000 EMS calls annually. Overa six month period a convenience sampleof fire-based paramedics were recruited. Theparamedics attended a 60 minute ultrasoundlecture and practicum. An emergency medicinetrained physician educated in basic ultrasoundskills delivered the educational intervention tothe paramedics. The paramedics completed a 5question survey both before and after the edu-cation, regarding their perceptions of prehos-pital ultrasound. A Chi-square test or FischerExact test was used to determine statistical sig-nificance. Results: All 27 (100%) paramedicscompleted the pre-survey, education interven-tion, and the post-survey. Pre-survey only 2 of27 paramedics felt they had a significant knowl-edge regarding FOCUS, while in the post-survey that number increased to 13 of 27, p <0.001. Pre-survey 4 of 27 paramedics felt com-fortable performing and reading a FOCUS dur-ing a cardiac arrest compared to 23 of 27 post-survey, p < 0.001. Pre-survey 8 of 27 paramedicsagreed that the cost of FOCUS justifies the ben-efits as compared to 21 of 27 post-survey, OR =8.3, 95% CI: 2.4–28.4. Almost half (13 of 27) ofthe paramedics thought that FOCUS performedby paramedics during cardiac arrest would beeasy to perform based on the pre-survey, com-pared to 24 of 27 on the post-survey, OR=8.6, 95% CI: 2.1–35.6. Pre-survey the majorityof paramedics (19 of 27) already believed thatthey should have access to prehospital ultra-sound, and in the post-survey that numberincreased to 25 of 27, p < 0.08. Conclusions:This study suggests that without previous edu-cation paramedics were not comfortable usingultrasound and believed FOCUS was not costeffective. Yet, after a brief educational inter-vention, paramedics’ perceptions significantlychanged.

167. Paramedics Can Successfully PerformCardiac Ultrasonography Utilizing theParasternal Long Axis Approach

John Reynolds, Juan March, Roberto Portela,Steven Taylor, Bryan Kitch, Department of Emer-gency Medicine, Division of EMS, Brody Schoolof Medicine, East Carolina Category of Submis-sion: Student, Resident, Fellow

Background: Focused point of care cardiacultrasound (FOCUS) has been used success-fully in screening for many life-threateningemergencies such as cardiac standstill, peri-cardial effusion, and others. There has beenlimited research on paramedics’ ability toperform FOCUS, but none specifically com-paring paramedics’ ability to utilize differentcardiac views. This study aimed to determineif paramedics can perform FOCUS utilizingtwo different views. Methods: A prospectivestudy was performed in a suburban/urbansetting with a population of 180,000 and 26,000EMS calls annually. Twenty-seven fire-based

paramedics were recruited for this study. Theparamedics completed a 60-minute educationalintervention on FOCUS which included alecture followed by a hands-on practicumconcentrated on using the parasternal longaxis and subxiphoid views only. An emer-gency medicine trained physician educatedin basic ultrasound skills delivered the edu-cational intervention to the paramedics. Theparamedics were given a brief overview andtraining of ultrasound knobology, and thenasked to perform FOCUS using only theparasternal long axis and subxiphoid views.Participants were then graded using the Car-diac Ultrasound Structural Assessment Scale(CUSAS). CUSAS is a 6-point graded scale thatevaluates visualization of the cardiac structure.A CUSAS score of 6 is given when multiplechambers are visualized. A CUSAS score of 3 isgiven when there is only partial visualizationof the ventricle. A CUSAS score of 1 is givenwhen no chambers are visualized. Results:All 27 paramedics were able to view the heartduring the practicum. When performing theparasternal long axis view 27 of 27 paramedics(100%) received a CUSAS score of 6 (multi-ple chambers visualized). In contrast, whenperforming the subxiphoid view 0 of 27 (0%)received a CUSAS score of 5 or 6, p < 0.001, 5 of27 (19%) paramedics received a CUSAS score of4 (multiple partial chambers including one ven-tricle) and 22 of 27 (81%) paramedics receiveda CUSAS score of 3. Conclusions: Our pilotstudy suggests paramedics with only limitededucation can be taught to successfully performa FOCUS using the parasternal long axis view,but have difficulty using the subxiphoid view.

168. Increasing Cardiac Arrest Survivalthrough a Novel Dispatcher CPRInstruction Program

Brittany Farrell, E. Brooke Lerner, M. Ric-cardo Colella, Kenneth Sternig, Lesley Simley,Christine Westrich, Charles Cady, Medical Col-lege of Wisconsin Category of Submission: Car-diac

Background: Out-of-hospital cardiac arrest(OHCA) survival rates remain low as dobystander CPR rates. When a dispatcher pro-vides CPR instructions to a bystander whoperforms compressions the odds of survivalincrease. However, many communities donot provide this lifesaving intervention, oftenciting the barriers of limited personnel andfunding. Objective: To describe the implemen-tation of a novel centralized dispatcher CPRinstruction program that serves seven PSAPsin a single county and compare bystanderCPR rates before and after implementation.Methods: As of April 22, 2016, seven munic-ipal public safety answering points (PSAPs)that did not previously provide dispatcherinstructions implemented this novel pro-gram. Using a simple 30-minute self-directedvideo, 84 PSAP dispatchers were trained toutilize a two-question protocol to identify andtransfer suspected OHCA cases to a centralcommunication center. At this center, a trainedcommunicator delivered CPR instructionsto the caller. Training of the 26 central com-municators was accomplished with a 2-hourin-person didactic session followed by a 2-hourpractice session. We compared pre and postcountywide EMS medical record data throughDecember 2016 using descriptive statistics. Wealso collected and analyzed data from record-ings of communicator-to-caller interactions.Results: 169 calls were transferred to the cen-tral dispatch center. Of those, 106 needed CPRinstructions. Of those, 56 callers performedcompressions before EMS arrival (52%). The 63non-OHCA calls were for a variety of ailmentsranging from severe to mild and the number



of non-OHCA calls decreased over time (May44%; Dec 29%). 11 victims survived to hospitaldischarge, for a 19% survival rate; previously,the countywide survival rate was 10%. Thecountywide bystander CPR rate increasedfrom 19% to 24%. Approximately 109 OHCAcalls were not transferred for instructions,work continues to increase the rate of OHCAcalls transferred. Conclusions: Implementinga novel centralized dispatcher CPR programincreased the rate of bystander CPR. Using acentral communication center for instructionsallowed us to train and maintain a smallergroup of communicators, leading to less costand more experience for those communica-tors, while limiting the burden on the PSAPdispatchers.

169. Qualitative Study of EmergencyMedical Technician and PatientPerspectives on the Transport PlusProgram

Hayley Neher, Ksenia Gorbenko, NadirTan, Diana Grigoriou, Hugh Chapin, LynneRichardson, Ula Hwang, Kevin Munjal, IcahnSchool of Medicine at Mount Sinai Category ofSubmission: Student, Resident, Fellow

Background: “Transport PLUS” is an educa-tional intervention in which Emergency Med-ical Technicians (EMTs) are trained to use achecklist to perform discharge instruction com-prehension assessments and home fall safetyassessments for older adult patients trans-ported home following hospitalizations. Pre-viously reported preliminary findings demon-strated high rates of patient acceptance andremoving fall hazards following the interven-tion. In this qualitative study, we endeavoredto identify potential barriers to success andrefine the existing checklist and other modifi-able aspects of the program in order to max-imize its effectiveness. Methods: This qual-itative study consisted of two homogenousfocus groups led by an experienced facilita-tor with Transport PLUS trained EMTs andpotential older adult patients to assess bar-riers and opportunities for improving theprogram. Three independent analysts codedanonymous transcripts for themes, comparedfor consistency, and resolved disagreementsthrough discussion. Results: Trained EMTs andpotential patients found the program valuablebut uncovered a number of potential barri-ers to success and suggested improvements.Themes identified by both groups includedconcerns for patient privacy and the impor-tance of obtaining buy-in from both patientsand providers. Trained EMTs also suggestedimproving phrasing of items on the checklistand optimizing delivery of educational infor-mation. Patient focus group suggested ways toenhance comprehension. Suggested improve-ments included emphasis on situational aware-ness during EMT training, building rapport,question order, normalizing safety measures,and explaining the reasoning behind includ-ing and excluding specific items on the check-list. Conclusions: The Transport PLUS programwas well received by both EMTs and patients.We found a high degree of agreement betweenthe two groups in identified barriers to suc-cess. Adjustments in EMT training and support-ive materials, including checklists and hand-outs, were made based on suggestions obtainedduring the focus group interviews. Trainingwas specifically enhanced to emphasize qual-ity, consistency, communication skills and tech-nique. Documents were enhanced to be morevisually appealing, easier to understand, andpromote better flow throughout the encounter.A randomized controlled trial to assess effec-tiveness of the program is already underway.If successful, our program will reduce the bur-

den of preventable injuries and readmissions onfrontline providers and health care systems.

170. Am I awake? Lack of SedationProtocols for Intubated Patients duringTransport in Statewide TreatmentProtocols

Christina Loporcaro, David Schoenfeld, BethIsrael Deaconess Medical Center/Harvard MedicalSchool Category of Submission: Student, Res-ident, Fellow

Background: In our constantly evolving health-care system the transfer of intubated patientsbetween facilities is an ever more commonoccurrence. While there is a paucity of litera-ture regarding the impact of adequate sedationin the out of hospital environment, intensivecare unit (ICU) studies have shown significantoutcome measures such as ICU length of stayis associated with inadequate patient sedation.The purpose of this study was to describe cur-rent protocols for sedation of intubated patientsduring interfacility transfer (IFT), as well asthe use of standardized sedation assessmentscoring to guide sedative medication admin-istration. Methods: Cross sectional study ofSTPs utilizing a standardized review to eval-uate sedation protocols for intubated patientsand the use of standardized sedation assess-ment scores. Protocol revision date was alsocaptured. Results: Thirty-one out of fifty states(62%) issue ALS STPs. Of those thirty-onestates, only one (3%) has a protocol for seda-tion of intubated patients. No STP incorpo-rates or references any sedation scoring tool tohelp guide sedative administration or aid inpatient assessment. 75% of protocols have beenrevised since 2015 and all have been revisedwithin the past 5 years. Conclusions: Althoughthere is little in the prehospital literature regard-ing patient outcomes with respect to inade-quately sedated patients, self-extubation, exces-sive agitation on hospital arrival and vital signabnormalities are complications well known toproviders. This study demonstrates that cur-rent STPs do not provide paramedics with thetools to optimally assess and sedate intubatedpatients in the out of hospital environment.While sedation plans may be developed withmedical control prior to transfer, a protocolizedapproach to sedation scoring and medicationadministration may be beneficial. This repre-sents a serious deficiency in our ability to pro-vide high quality care to intubated patients inthe out of hospital environment. In the future,we hope to develop and validate a prehospitalsedation scoring model and associated protocolfor the management of intubated patients in theout of hospital environment.

171. Prevalence of Recurring PatientEncounters that Require Administrationof Prehospital Naloxone: A RetrospectiveChart Review

Thomas Dykstra, Jen Knapp, Patrick Dugan,Rhees Nickel, City of Fort Wayne, EMS Founda-tion Chair Category of Submission: Student,Resident, Fellow

Background: A significant proportion ofpatients responded to by EMS personnel foropioid overdose will continue to abuse opioidsafter treatment and resuscitation, leading tosubsequent overdoses in the future that requireadditional treatment. The aim of this study is toidentify the prevalence of recurrent encountersthat require the administration of Naloxone toreverse opioid overdose. The trend of opioidabuse within the United States has continuedto increase despite efforts to decrease theiraccessibility. To deter this issue, stricter guide-lines regarding the prescription of medicinalopioids has led many individuals with addic-

tion to seek illicit substances. The major healthconcern of opioid abuse, respiratory depres-sion, is treated mainly with Naloxone, whichcounteracts opioids at the receptor level. It isthought by many EMS personnel that peopleexperiencing resuscitation with Naloxone willcontinue to abuse opioids. Methods: Thisretrospective chart review examined electronicpatient care reports provided by the ThreeRivers Ambulance Authority (TRAA). Allencounters in which Naloxone was adminis-tered between November 1, 2010 and October31, 2016 by TRAA or other bystanders wereexamined. The number of encounters eachpatient had during this date range was used toanalyze a general recurrence rate of opioid use.Results: The increase in number of individualsexperiencing more than one Naloxone relatedevent annually did not differ significantly fromwhat was expected over the 6-year range, χ2 =9.81, p = 0.08. However, the number of patientsfalling into this category increased more thantriple throughout the study. Conclusions: Theresults of this study suggest that the numberof recurrent patient encounters involving theadministration of Naloxone has increased.While the extent of the increase of recurrenceis much less than initially believed by EMSpersonnel, additional future studies to cor-rectly identify the impact of recurrent patientencounters may show significant results toassist combating addiction.

172. Pharmacologic Opioid Alternativesfor Pain Control in Statewide TreatmentProtocols

Christie Fritz, Christina Loporcaro, DavidSchoenfeld, Beth Israel Deaconess Medical Cen-ter/Harvard Medical School Category of Submis-sion: Student, Resident, Fellow

Background: There has been an increasingfocus on reducing opioid use across health-care in light of the opioid epidemic. Thereare multiple pharmacologic options for treat-ing pain in the prehospital setting includ-ing ketamine, nitrous oxide, acetaminophen,ibuprofen, ketorolac and aspirin. The major-ity of states issue statewide treatment proto-cols (STPs) that are either mandatory, or serveas a guide for medical directors. The pur-pose of this investigation is to describe theextent to which STPs include alternatives toopioids for pain control. Methods: Cross sec-tional study of STPs, utilizing a standardizedreview of pharmacopeia in pain control proto-cols. Protocol revision date was also captured.Results: Thirty-two of fifty states (64%) issueSTPs; 78% are mandatory; 38% of STPs limitpain management to opioid medications only;and 62% of STPs provide for pharmacologicalternatives to opioids for pain management.Pharmacologic alternatives for pain control arevariable across STPs and include Nitrous oxide(50%), ketamine (19%), Tylenol (25%), ketoro-lac (25%), Ibuprofen (16%), and aspirin (6%).A total of 75% of protocols have been revisedsince 2015 and all have been revised withinthe past five years. All ALS statewide treat-ment protocols have explicit orders for opiatesin their pain control protocols. Conclusions:The opiate epidemic in the U.S. has led to anincreased focus on the use of alternatives to nar-cotic medications in healthcare. Pain manage-ment is an important part of prehospital care,however many states do not provide pharmaco-logic alternatives to narcotic medications. Whileno studies have identified prehospital narcoticadministration as a cause of or contributor tothe opiate epidemic, we should strive to reducethe use of narcotics when appropriate alterna-tives exist. Despite the majority of STPs under-going protocol revisions within the last twoyears which is during the ongoing opiate epi-



demic, STPs have not fully incorporated alter-natives to opiates for pain control. This repre-sents a significant opportunity to improve ourSTPs to include alternatives to narcotic medi-cation for the management of pain, and do oursmall part to help combat the opiate epidemic.Further study is needed to better understandthe barriers to adoption of non-opiate phar-macologic treatment or adjuncts for pain treat-ment.

173. The Heavy Lift: Impact of a RegionalBariatric Transport Program

Gerald Wydro, Larry Loose, Alvin Wang,Aria Jefferson, Health Department of EmergencyMedicine Category of Submission: Opera-tions, Quality, Safety Systems, Disaster

Background: Obesity is an epidemic in thisnation and provides serious challenges to EMSfor care and transport. Many systems haveidentified the problem, but few provide a solu-tion to their providers. Alternatively, EMS sys-tems should create a solution that is deploy-able, cost effective, and provides safe digni-fied transport. We describe the characteristics ofa regional Bariatric Support Unit (BSU) trans-port system dispatched via the 9-1-1 system forbariatric patients. Methods: Descriptive anal-ysis of a regional BSU transport system inour suburban EMS system served by 17 agen-cies covering an area of 622 sq. miles with apopulation of over 620,000. Requests for EMSservice exceed 53,000 annually and are han-dled via a single 9-1-1 center. The BSU trans-port system utilizes three specially equippedambulances (bariatric stretchers, lifts, ramps,and winches) strategically located throughoutthe county. The BSU ambulances rendezvouswith the on scene EMS unit and assist withtransport of the patient and crew to the hos-pital. Results: There were 121 requests forBSU transport during the 12 month periodof review with 108 (89%) ending in transportto the hospital. The average weight of trans-ported patients was 419 lbs. Of BSU requests,66 (55%) were dispatched ALS, with less thanhalf receiving an ALS intervention. The mostcommon complaint type was Acute ExtremityPain (19%). Twenty Four patients (20%) usedthe system more than once. Average on-scenetime increased by 150% for patients transportedvia BSU (30 minutes) compared to our systemaverage on-scene time (12 minutes). Patient andEMS crew satisfaction was high with the BSUsystem and there were no reported injuries topatients or EMS providers during the reviewperiod. Conclusions: A regional BSU transportsystem provides a cost effective, safe and dig-nified means of transport of bariatric patientsduring EMS response. While more than half ofcases were dispatched ALS, the most commoncomplaint was Extremity Pain. No practitionersused unconventional modes of transportationfor transporting a patient to the hospital dur-ing this period; 20% of patients utilized the sys-tem more than once. On-scene times were sig-nificantly increased however no adverse eventswere reported.

174. Nationwide Quality E-Registry forDispatcher-Assisted CardiopulmonaryResuscitation (DACPR) of Out-of-HospitalCardiac Arrest (OHCA) – The Design forStructured Measurement

Patrick Chow-In Ko, Mei-Fen Yang,Kah--Meng Chong, Hui-Chih Wang, Chien-Hsin Lu, Chih-Hao Lin, Yen-Bing Chen,Yen-Ho Yang, Ming--Shian Lee, Wen-ChihChou, Chih-Chiang Cheng, Wen-Long Chen,National Taiwan University, College of Medicine,Department of Emergency Medicine Categoryof Submission: Operations, Quality, SafetySystems, Disaster

Background: Following the guidelines ofdispatch-assisted CPR (DACPR) may enhancebystander CPR rate after OHCA. Registry ofquality measurement for DACPR has neverbeen explored. We designed a nationwidequality registry for DACPR performance andinnovated a structured format of measurement.Methods: A nationwide Google Forms basedonline e-registry system covering over twentyadministrative regions and more than twentymillions of population was designed andlaunched for DACPR performance and qual-ity measurement at individual case level fornon-traumatic OHCA patient. Audio records ofindividual EMS call were reviewed for perfor-mance rating. System data inputted could beimmediately retrieved as feedback to each cor-responding administrative region. Recognitionof cardiac arrest by call communication, CPRInstructions upon the recognized OHCA, andchest compression upon the recognized OHCAwere the three major categorical performanceindicators, and each operational time intervalof call-to-recognition, call-to-instruction, andcall-to-compression were evaluated. Eachcategorical performance indicator (Y-axis) waspaired with its operational time interval (X-axis) as a set of quality index for diagrammaticcomparison in our design. We used regressionanalysis for statistical analysis. Results: Atotal of 6,078 audio records for OHCA EMScalls across 18 regions were centralized intothe nationwide DACPR Quality Registryin 6 months (minimal 40 to maximal 1,625cases/region according to its population).Regional recognition rate significantly variedfrom 10.0% to 88.1% (p < 0.01; averaged60.4%, SD 21.2%). Instruction rate varied from41.3% to 93.1%% (p < 0.01; averaged 77.4%,SD 14.9%). Compression rate varied from45.2% to 88.4% (p < 0.01; averaged 75.3%, SD12.8%). Averaged regional call-to-recognitiontime, call-to-instruction time, and call-to-compression time were 58 (SD 21), 92 (SD48), and 174 (SD 71) seconds. The designateddiagrammatic comparisons may indicate theadministrative regions of better performancelocated at the upward and leftward dimension,and the ones of unsatisfied performance locatedat the downward and rightward dimension(diagrams will be illustrated). Conclusions:We successfully innovated and launcheda nationwide DACPR quality e-registryshowing a wide variety of regional perfor-mance needing improvement. The designateddiagram may easily indicate and comparethe individual performance across the jointregions.

175. Pilot Randomized Control Trial ofPelvic Binder Compared to Standard Carein Prehospital Patients with a SuspectedPelvic Fracture

Jonathan Studnek, Allison Infinger, MeghanWally, Sarah Pierrie, Malcolm Leirmoe,Joseph Hsu, Rachel Seymour, MecklenburgEMS Agency Category of Submission: Trauma

Background: Pelvic ring fractures are associ-ated with high morbidity and mortality, how-ever, pelvic stabilization and hemorrhage con-trol has not been rigorously tested. The primaryobjective of this study was to determine the fea-sibility of conducting a randomized controlledtrial comparing a commercially available pelvicbinder to standard care in prehospital patientswith a suspected pelvic fracture. Methods: Thisprospective study collected data from an EMSagency - which serves a population of nearly1 million and transports approximately 114,000patients per year - and a level 1 trauma center.Eligible patients were those �18 years with ahigh-energy mechanism of injury and prehospi-tal suspicion of pelvic fracture. Exclusion crite-

ria were low-energy mechanism of injury, pen-etrating pelvic injuries, pregnancy, and inabil-ity to secure binder due to patient size. Com-munity consultation to meet the requirementsof Exception from Informed Consent was con-ducted prior to study implementation. EMScrews were randomly assigned a sealed kit atthe start of each shift containing either a com-mercial binder (intervention) or towels (com-parison), indicating standard of care. Prehospi-tal providers were blinded to the contents untilit was opened after identifying a patient meet-ing inclusion criteria. Outcomes included skincomplications, hospital and ICU admissions,angioembolization, surgical control of bleeding,transfusions, and 30-day readmission and mor-tality. Results: A total of 30 patients with sus-pected pelvic fractures were enrolled from June2016 to April 2017. Fourteen (46.7%) patientswere randomized to the binder group; six(42.9%) had pelvic fracture diagnoses comparedto four (25.0%) in the comparison group. Thepatient population was majority male (n=19,63.3%) and averaged 31.5 years. Of the twenty-three patients (74.2%) admitted to the hospital,11 (47.8%) were admitted to the ICU. Only 3(9.6%) patients required angioembolization, 2(6.5%) surgical control of bleeding, and 3 (9.6%)a transfusion. One (3.2%) patient required read-mission and died within 30 days. There were noserious adverse events. Conclusions: This pilottest demonstrates that prehospital providersare able to implement a randomized trial,including identification of eligible patients,maintaining the randomization scheme andassignment to treatment, and handoff to theclinical and research teams at the receivinghospital.

176. Descriptive Analysis of PatientsTransported via Ground and Air CriticalCare Teams on Extracorporeal MembraneOxygenation (ECMO)

Matthew Sztajnkrycer, Ryan Sherden,Meghan Lamp, Kathleen Berns, David Clay-pool, Mayo Clinic Category of Submission:Medical

Background: Despite improved portability andease of cannulation, few U.S.-based medicaltransport services currently transfer patientson ECMO. The purpose of the current studywas to perform a descriptive analysis of acohort of patients transported via air or groundwhile on ECMO. Methods: Retrospective caseseries of patients transported by a single crit-ical care transport provider to a single ter-tiary care facility between January 1, 2014 andMay 31, 2017. Patients were included if trans-ported while on ECMO. T-test and Fisher’sExact Test were performed for statistical analy-ses. Results: Twenty-five patients met inclusioncriteria, of which 16 (64%) were male. Mean agewas 43.4 ± 17.6 years (range 1–68 years). Six-teen patients were transported on VA-ECMO,while 9 were transported on VV-ECMO. Threepatients were transported by ground criticalcare team, while 9 were transported by rotorwing and 13 were transported by fixed wing.Mean transport time was 60.8 ± 28.4 minutes.The most common indications for ECMO wererespiratory failure/acute lung injury (48%) andcardiogenic shock (28%). Four patients receivedECMO as extracorporeal life support (ECLS) forrefractory cardiac arrest. No patient died dur-ing ECMO initiation or transport. Two patientsrequired fluid boluses for low blood flow, while5 received blood transfusion for cannulation-related blood loss. The most common interven-tions in transit were sedation, muscle relax-ation, and heparinization. Survival to hospi-tal discharge was 48%, with improved survivalamongst younger patients (p = 0.52). Mortalityfor patients on VA-ECMO was 62.5%, compared



with 33.3% for those on VV-ECMO (p = 0.35).In patients receiving ECLS, 50% survived todischarge; both had refractory VF/VT arrests.No difference in survival was noted basedupon early (40%) versus late (50%, p = 0.70)ECMO initiation. Conclusions: In our patientcohort, transport on ECMO was not associatedwith significant adverse event or mortality. VAECMO for cardiopulmonary support was asso-ciated with worse final outcome. ECLS sec-ondary to VF/VT arrest was associated withbetter survival to discharge compared withother dysrhythmias. The current data suggestthat transportation of ECMO patients is safe,and initiation of ECMO need not be delayedpending transfer.

177. Does Prehospital Mode of ArrivalInfluence Women’s Decisions toParticipate in Research?

Madeline Karafanda, Martina Anto-Ocrah,Vivian Lewis, Todd Jusko, Jeff Bazarian,Edwin van Wijngaarden, Courtney Jones,Department of Environmental Health Categoryof Submission: Medical

Background: Advances in medicine requirevoluntary participation in research. Thisrequirement however, may compromise studygeneralizability, as it is often unclear howrefusals and participants differ. Further com-plicating the matter is the National Institutesof Health (NIH) requirement, that proposedresearch studies address any possible dispar-ities in gender. Investigators have exploredthe barriers and facilitators for research par-ticipation. Few, however, have focused onhow prehospital factors, specifically modeof Emergency Department (ED) arrival, mayinfluence the recruitment of female subjects inresearch studies. We explored how prehospitalmode of arrival (ambulance vs non-ambulance)affects research participation, sampling femalesubjects only. We hypothesize that womenarriving via ambulance will be less willingto participate in research, compared to thosearriving as walk-in/ambulatory patients.Methods: From January 9 to July 8 2017, wecollected data on 373 women presenting forcare in the ED of a Level 1 Trauma Center.All subjects were required to have GCS �13and/or deemed capable of providing informedconsent (following standard protocol and/orprovider approval). Refusals were comparedto participants based on variables abstractedfrom the medical record. Comparisons weremade between groups using 2-tailed inde-pendent t tests or χ2 tests, as appropriate.Results: Ambulance users comprised a thirdof the sample (33.5%, n = 125) and non-users represented 66.5% (n = 248). The meanage was 28.5(+/−7.9), with no statisticaldifferences between ambulance users andnon-users (p = 0.4). Compared to non-users,a significant proportion of ambulance users(41.6% vs. 14.1%, p < 0.001), were involved inmotor vehicle crashes, and sought care within24 hours of their injury. Over a quarter ofambulance users (versus non-users) refusedto participate in research (28.8% vs.19.4%, p= 0.039). Ambulance use was associated witha 69% increased odds of refusal to participatein research (95% CI: 1.02, 2.78). Reasons forrefusing included “disinterest in research”,time constraints” and “discomfort with con-sent process/nature of study”. Conclusions:As hypothesized, ambulance transport tothe ED is associated with increased oddsof research refusal. Future studies shouldexplore if this finding is unique to femalesonly, and tailor ED-based research recruitmentefforts, with ambulance mode of arrival as aconsideration.

178. Pediatric Bypass: Characteristics andEffects on EMS Resources

Jennifer Fishe, Kevin Psoter, Kyle Fratta, CarlaTilchin, Jennifer Anders, University of FloridaCollege of Medicine - Jacksonville Category ofSubmission: Pediatric

Background: Regionalization of pediatriccare decreases pediatric service availabilityat community hospitals. However, pediatricregionalization’s effects on EMS operationsare unknown. This study describes pediatrictransport characteristics, focusing on bypasspatients. Methods: This retrospective studyexamined all transports ages 0–17 years fromthree geographically diverse EMS agencies(urban, suburban, and rural) over a 12-monthperiod. Those agencies only pediatric destina-tion protocol is the CDC Trauma Triage Tree.Scene response, destination facility, and sur-rounding facility locations were geocoded, andeach facility assigned a category denoting pedi-atric service availability. Bypass was defined astransport to any facility other than the nearest.Results: The three agencies transported 12,223pediatric patients during the study period,and 8,039 (66%) bypassed the nearest facility(80% to a higher level of pediatric care, and20% to an equivalent or lower care level).Over half of urban (71%), suburban (60%), andrural (59%) agency pediatric transports werebypasses. The majority of children bypassedto a higher level of care were transported topediatric trauma/specialty facilities (55%),followed by regional pediatric facilities (24%),and comprehensive pediatric facilities (21%).The top five EMS clinical impressions werepain, other, seizure, traumatic injury, and noapparent illness/injury. Patients bypassed tothe same or lower care level were transportedto community (51%), trauma/specialty (30%),comprehensive (12%), and regional facilities(7%). The top five impressions were pain, other,no apparent illness/injury, traumatic injury,and asthma. For bypass patients, median EMStransport distance was 6.2 km (25–75th per-centiles: 3.2–10.8) and median driving time was15.8 minutes (25–75th percentiles: 10.8–21.9),representing an additional median 2.9 km(25–75th percentiles: 1–6.4) and 6.8 minutes(25–75th percentiles: 3.1–12.6) from the closestfacility. Median transport distance was 2.8 km(25–75th percentiles: 1.6–4.6) and median driv-ing time was 8.6 (25–75th percentiles: 6.2–11.9)for patients transported to the closest facility.Conclusions: This study demonstrates highpediatric bypass rates, which coupled withincreased transport distances and times, affectEMS resource allocation by occupying vehiclesand crews for longer runs. Future work willdetermine each bypass’ appropriateness toinform both EMS operations and pediatricdestination decisions.

179. Removal of Left Bundle Branch Blockfrom Prehospital ST-Elevation CriteriaDecreases Number of Unnecessary CathLab Activations

Rachel Semmons, Elizabeth Mannion,Andrew Thomas, Quinn Frier, Jason Wilson,Cory Thomas, Tampa Fire Rescue, Universityof South Florida Category of Submission:Cardiac

Background: Prehospital identification ofSTEMI allows decreased time to PCI. Falsepositive prehospital STEMI Alerts may wasteresources through unnecessary cath lab acti-vation as well as pose risks to patients. Ourcurrent prehospital STEMI Alert Criteriaincludes ST-segment elevation >1mm in two ormore contiguous leads and/or presumed newleft bundle branch block (LBBB) in the presence

of anginal symptoms. LBBB was removed fromSTEMI criteria in the most recent AHA/ACCSTEMI guidelines as a result of low specificity.We hypothesize that LBBB has led to a highnumber of false positive activations in our sys-tem and can safely be removed from our STEMIcriteria. Methods: We conducted a one-yearretrospective analysis of prehospital STEMIalerts. Hospital records were reviewed for EDphysician interpretation of EKG findings ofST-elevation, LBBB, or neither of these criteria(nondiagnostic). Primary outcomes were cathlab activation and intervention. Secondaryoutcomes analyzed were presence of initialelevated troponin. We excluded patients withmissing records. Results: A total of 107 STEMIAlerts were transported over the study period,with records available for 102 patients. Ofpatients identified as EMS STEMI Alerts, 45.1%went to cath, and 36.3% received coronaryintervention. Rates of cath lab activation andcoronary intervention were significantly higherin patients with an ED physician interpretationof ST-elevation compared to those with LBBB(71% vs. 9%, OR 22.03, CI 9.77–49.68, p <0.0001). One patient with LBBB received emer-gent cath with stenting after testing revealedelevated troponin. Secondary analysis of thispatient’s EKG showed that he did not haveSgarbossa criteria. Conclusions: The majorityof EMS STEMI alerts did not require emergentcardiac catheterization. More false positivealerts were due to nondiagnostic EKGs ratherthan LBBB. It appears that removal of LBBBas criteria for STEMI activation can safelylower STEMI alert numbers. Future protocolswill direct EMS to transport patients withLBBB and anginal symptoms to a PCI-capablecenter without designating the patient as aSTEMI alert. Further efforts will be aimed atdecreasing the number of false positive alertsthrough EMS education.

180. Is Door-to-needle Time Reduced forEMS Transported Stroke Patients RoutedDirectly to the CT Scanner on ED Arrival?

Bryan Sloane, Nichole Bosson, Jeffrey Saver,Nerses Sanossian, Marianne Gausche-Hill,Harbor-UCLA Medical Center Category of Sub-mission: Medical

Background: To evaluate if a protocol to routeEMS-transported stroke patients directly to theCT scanner on ED arrival reduces door-to-needle time (DTN). We hypothesized a reducedDTN compared to initial routing to an EDbed. Methods: This is a retrospective analy-sis from a large regionalized stroke system.EMS utilize the modified Los Angeles Prehos-pital Stroke Screen (mLAPSS) and transport allsuspected acute stroke patients to one of 46Approved Stroke Centers (ASC). Some ASCroute EMS directly to CT. ASC report patienttreatment and outcomes to a registry, fromwhich data were abstracted from May 2015through April 2016. Adult patients transportedby EMS and treated with intravenous throm-bolytic therapy (IV tPA) were included. The pri-mary outcome was median DTN at hospitalswith CT routing protocols compared to hos-pitals with ED routing. Secondary outcomeswere door-to-imaging time, hospital length ofstay, and modified Rankin Scale at discharge.A subgroup analysis of patients with positivemLAPSS was planned a priori. Outcomes werecompared with Hodges-Lehmann’s median dif-ference. Results: EMS transported 6315 patientsfor suspected stroke and 797 (13%) were treatedwith IV tPA, 143 at hospitals with CT routingand 654 at hospitals with ED routing. Patientcharacteristics were similar between groups;overall 420 (53%) were male, 500 (62%) Whiterace, and 189 (24%) Hispanic ethnicity. MedianNIHSS was 12 (IQR 8–19) in the CT rout-



ing group and 11 (IQR 5–19) in the ED rout-ing group. Positive mLAPSS and EMS notifi-cation occurred respectively in 63% and 96%in the CT routing group and 66% and 86%in the ED routing group. DTN was not dif-ferent between groups, median DTN 59 min-utes (IQR 45–78) for CT routing and 54 (IQR40–73) for ED routing, median difference 4.5(IQR 0–9). There were no differences betweenthe groups in terms of secondary outcomes orwithin the mLAPSS-positive subgroup. Con-clusions: In this regional stroke system, hospi-tals with protocols for routing EMS-transportedstroke patients directly to CT did not havereduced DTN compared to hospitals withoutsuch protocols. These results are limited by thefact that the actual routing of each patient is notknown.

181. At Risk Referral: Characteristics ofPatients Identified by Paramedics As AtRisk of Becoming Frequent 9-1-1 Callers

Nancy Newall, Neil McDonald, Tara Stewart,Brandon University Category of Submission:Medical

Background: Paramedics often encounterpatients who are frequent users of emergencyservices for non-emergent conditions. In somejurisdictions, these patients are treated bycommunity paramedics or mobile integratedhealthcare teams. After establishing a programof home visits by community paramedics forfrequent callers, our service sought to identifypatients who might benefit from this model ofcare prior to their becoming high users. In 2014,we integrated a semi-structured assessmenttool into our electronic patient care report(ePCR) that allows front-line paramedics toflag patients they encounter as being at riskof harm or of becoming frequent users. ThisAt Risk Referral (ARR) evaluates patientsover a range of risk factors and forwardstheir information to a community paramedicfor follow-up. The purpose of the presentstudy is to describe the ARR and to examineincidents that were flagged as involving at-riskpatients over a 2.5-year period. Methods:All ePCRs that contained an ARR betweenJanuary 1, 2014 and July 20, 2016 were iden-tified out of the total number of incidents(N = 169,810). Data extracted for the presentstudy included: date, age, sex, and at-riskcharacteristics (lack of social supports, mobilityissues, cognitive impairment, communicationimpairment, safety concerns, and others).Results: On average, paramedics identified42.2 incidents/month or 1.4 incidents/dayas involving at-risk patients, representing0.8% of total incidents over 2.5 years. Overall1,309 ARRs were completed for 943 patients,19% of which had more than one incidentover 2.5 years. Fifty-five percent of incidentsinvolved female patients; and the average agewas 68 years (range 13–99 years). Of the 1,309incidents, the two most common reportedat-risk categories were lack of social supports(71.9%) and safety concerns (70.3%), with 51%of incidents categorized by both of these riskfactors. Conclusions: This study demonstratesthat front-line paramedics regularly use anARR to identify patients deemed at risk ofharm or of becoming common callers. Incidentswere generally characterized by more thanone at-risk factor. Future studies will trackoutcomes among identified patients and screenfor other patients not identified who becamefrequent callers.

182. Association between BMI andPrehospital Selection of Advanced Airwayin Out-of-Hospital Cardiac Arrest

Caitlin Howard, David Wampler, JeremyAllen, Hattie McAviney, Justin Smith, David

Miramontes, Joan Polk, United States Army andUTHSCSA Category of Submission: Student,Resident, Fellow

Background: Obesity is associated with diffi-cult prehospital endotracheal intubation. Theobjective of this study was to examine the asso-ciation between patient BMI and the selectionof advanced airway by prehospital providersduring out-of-hospital cardiac arrest (OHCA).Methods: This was a retrospective review ofan in-house cardiac arrest registry containingdetails of each resuscitation attempted by alarge, urban fire-based EMS system. Advancedairway selection was at the discretion of theresuscitation team. The BMI recorded wasa subjective measurement obtained from theparamedic at the time of data collection. Datawas analyzed from January 1, 2016 throughAugust 15, 2016. Patients were included in thestudy if the following variables were available:age, gender, BMI, and initial airway attempted(supraglottic vs ETT). Patients were excludedif age < 17, no age, gender, or BMI recorded,or an airway other than supraglottic or ETTwas used. Patients were divided into 4 groupsbased on the BMI (under, normal, over, mor-bid). A subcategory analysis of endotrachealintubation method (direct laryngoscopy (DL)vs video laryngoscopy (VL)) was also exam-ined. ANOVA was utilized to analyze con-tinuous variables and a χ2 test was used toanalyze categorical variables. Results: A totalof 474 patients were included. Mean age forthe population was 63.56 + 17.65 years with293 males (61.81%). Most patients were classi-fied as normal BMI (209 patients, 44.09%) orover BMI (156 patients, 32.91%). The ETT wasmore frequently utilized as the initial airwayof choice in under BMI vs morbid BMI (P =0.03). Compared to normal BMI, more over BMIand more morbid BMI had a supraglottic air-way selected as the initial airway (P = 0.03and P = 0.009, respectively). Subgroup analysisof laryngoscopy method used for endotrachealintubation was not significant between the BMIgroups (P = 0.80). Conclusions: We found thatparamedics tended to favor endotracheal intu-bation with lower BMI patients. There was nodifference noted between BMI and DL vs VL.Limitations included that the BMI was not cal-culated and we only looked at the initial airwayattempt, which may not have been the conclu-sive airway.

183. Development of Modified Trauma andInjury Severity Score Model to PredictDisability for Acute Trauma Patients

Ki Jeong Hong, Sang Do Shin, Kyoung JunSong, Young Sun Ro, So Yeon Kong, Tae HanKim, Jeong Ho Park, Department of EmergencyMedicine, Seoul National University Boramae Med-ical Center Category of Submission: Trauma

Background: Trauma and Injury Severity Score(TRISS) has been used to predict mortal-ity of trauma patients and to perform qual-ity improvement of trauma care system. Inadvanced countries, functional outcome includ-ing disability is recently emphasized as a qual-ity indicator for trauma care system. The goalof this investigation is to develop modifiedmodel of Trauma Related Injury Severity Scoreto predict Disability (TRISS-D) for acute traumapatients. Methods: We used emergency med-ical services based severe trauma database ofthe Korea Centers for Disease Control. Weenrolled severe trauma cases transported byfire department from January to December 2013in 10 provinces across Korea. We calculatedrevised trauma score (RTS) and injury sever-ity score (ISS) for enrolled cases. We devel-oped modified TRISS model predicting severedisability and worsening disability using ageindex (0–14, 15–54, 55– years), RTS and ISS.

TRISS-D model 1 added injury mechanism cat-egory divided by blunt or penetrating injury.TRISS-D model 2 added presence of severehead injury when abbreviated injury scale (AIS)of head is from 3 to 6. We developed coeffi-cients of each TRISS-D model for severe dis-ability and worsening disability. Severe dis-ability was defined when Glasgow outcomescale (GOS) at hospital discharge was 1,2,3.If the difference of GOS at hospital dischargeand GOS before trauma incident is 1 point ormore, we defined the case as worsening dis-ability. We assessed discriminative power ofeach model by Area Under the ROC Curve(AUC) value. Results: A total of 14,791 patientswere enrolled. 3,757 cases were severe disabil-ity and 6,018 cases were worsening disability.AUC value of TRISS-D model 1 and model 2for severe disability was 0.948(95% ConfidenceInterval (CI): 0.944–0.952) and 0.950(95% CI:0.946–0.954), respectively. AUC value of TRISS-D model 1 and model 2 for worsening disabilitywas 0.810(95% Confidence Interval (CI): 0.803–0.817) and 0.816(95% CI: 0.809–0.823) respec-tively. Conclusions: We developed modifiedTRISS model for functional outcome like severedisability and worsening disability of acutetrauma patients. TRISS-D model for severe dis-ability showed excellent discriminative powerwith AUC value higher than 0.9. AUC valueof TRISS-D model for worsening disability washigher than 0.8.

184. Most Civilian Prehospital CareProviders Require Additional WoundPacking Training and Protocols

Mark Liao, Daniel O’Donnell, Thomas Lar-daro, Indiana University Category of Submis-sion: Student, Resident, Fellow

Background: Civilian EMS agencies are increas-ingly interested in adopting hemostatic dress-ings and wound packing for treating difficult-to-control or junctional hemorrhage. However,baseline civilian provider training of hemo-static dressings and wound packing is vari-able. We conducted a survey of prehospitalcare providers in a large metropolitan EMSagency that did not equip hemostatic dressingsor provide wound packing training to evalu-ate baseline provider knowledge and comfortlevel of these techniques. Methods: A total of178 prehospital providers (68% of whom wereparamedics) completed an online survey. Thissurvey queried their prior training, understand-ing of local EMS protocol, knowledge of hemo-static dressing efficacy vs. regular gauze forwound packing and comfort level performingwound packing. Results: Only 27 civilian pre-hospital providers (15%) identified themselvesas having prior military/Tactical Combat Casu-alty Care/law enforcement training that hadfamiliarized them with wound packing andhemostatic dressings. 81.5% of all respondents(n = 145) did not believe that wound packingwas authorized to perform as part of local orState EMS protocols, when in fact the procedurewas not mentioned in any such regulations. 64%(n = 114) providers believed that hemostaticdressings would be more effective than a reg-ular gauze roll when used for wound pack-ing; those with prior familiarization were morelikely to think this was true. (92.5% for thosewith prior familiarization vs. 58.9% without. χ2statistic = 11.69, p = 0.002 ). On a ranked scaleof 0 to 100 indicating comfort level using hemo-static dressings/wound packing, individualswith prior familiarization were more likely torate a higher comfort level (mean score of 89 forthose with prior military/LEO/TCCC trainingvs. 64 without, t-value 5.9, p < 0.00001). Conclu-sions: Current civilian EMTs and Paramedicsare interested and motivated to utilize hemo-static dressings and wound packing techniques,



but most require additional training to increasecomfort with these interventions. Prior military,tactical combat casualty care or law enforce-ment training appears effective in increasingconfidence. Civilian EMS protocols may needto explicitly reference wound packing to ensureproviders are aware that they can utilize thisskill.

185. Transport Determinates forContinuing Care Residents Assessed by anEMS Urgent Response Team: ARetrospective Observational Study

Kevin Lobay, Robyn Palmer, Lorissa Mews,Robert Sharman, Brian Boswell, Priya Jaggi,University of Alberta Department of EmergencyMedicine Category of Submission: Opera-tions, Quality, Safety Systems, Disaster

Background: Alberta Health Services (AHS)Emergency Medical Services (EMS) in theCity of Edmonton recently introduced an“EMS Continuing Care Urgent Response Team”(ECCURT) to support continuing care resi-dents by providing urgent care on-site, therebyreducing unnecessary patient transfers to emer-gency departments. ECCURT is comprised ofAdvanced Care Paramedics and Nurse Prac-titioners, and is dispatched via a dedicatedconsult line and/or 9-1-1. Various patientcharacteristics are tracked within our inter-nal database including age, diagnosis, Goalsof Care Designation (GCD), and CanadianTriage Acuity Scale (CTAS) score. Objectives:This study will provide an analysis of vari-ous ECCURT patient characteristics, and deter-mine whether age, GCD and CTAS score arecorrelated with frequency of transport to hos-pital. Methods: This is a six-month retrospec-tive, observational study of patient data. Allnew patients assessed between January 1, 2016and June 30, 2016 were included. Multipleregression analysis was performed to deter-mine whether a statistically significant correla-tion exists comparing age, GCD and CTAS scorewith transport frequency. Results: A total of 471(83%) of 567 new patients assessed by ECCURTduring the study period had established GCDsin place; 521 (92%) of our patients had a CTASscore assigned; 131 (23%) of our patients weretransported to hospital. All patients with a GCDof C2 (specifically requesting no transfer to hos-pital) were managed by our team on-site. Mul-tiple regression analysis reveals a statisticallysignificant correlation of age, GCD, and CTASscore with transport frequency (F statistic = 3.26E-11). P-values for each variable are: age = 0.92;GCD = 0.05; CTAS = 5.08 E-12. Conclusions:Although patient age appears not to be stronglycorrelated with transport frequency indepen-dently, GCD and CTAS score may be quite use-ful predictors for Community Care EMS Teamswhen selecting patients who can be managedon-site without transport to hospital.

186. Early Double Sequence DefibrillationImproves Outcomes in RefractoryVentricular Fibrillation

Matthew Harris, Ronald Klebacher, JoshuaSchwarzbaum, Andrew Parrish, Michael Carr,Andrew Torres, Navin Ariyaprakai, AmundepTagore, Eric Wasserman, Bauter Robert, MarkMerlin, Newark Beth Israel Medical Center Cate-gory of Submission: Cardiac

Background: Refractory ventricular fibrilla-tion (RVF) has been defined as VF that per-sists after 5 standard attempts at defibrillation(SD), though no uniform definition exists. Itsincidence has been estimated at 0.5–0.6 per100,000 population. Double sequence defibrilla-tion (DSD) has emerged as a possible treatmentfor RVF to improve rates of ROSC. Methods:A retrospective chart review of patients greater

than 17 years with RVF, defined as VF persis-tent after 3 standard defibrillations (SD), afterthe implementation of a quality project allow-ing paramedics to perform DSD in patientswith RVF. After 3 SDs, 2 sets of defibrilla-tion pads were placed on the patients withRVF. Two rapid sequence defibrillations at 360Joules were delivered. No limit was placed onthe number of DSD shocks provided. We com-pared patients who received DSD to those whoreceived SD. Our primary outcome was ROSC.We performed descriptive statistics, and asso-ciation and correlation between variables withANOVA and Chai-squared. Results: We identi-fied 280 patients with RVF, 229 (82%) receivedSD only and 51 (18%) received DSD. Compar-ing the SD group vs. DSD group: Mean Age67.7 years vs. 66.8 years (p < 0.001), Male gen-der 76.2 % vs. 72.5% (p < 0.001), and meanweight (kg) 89.3 vs. 90.1 (p < 0.001). There werefewer witnessed arrests in the control group(61.6% vs. 80.4%). Time to 1st shock was theidentical (14.7min ± 9.5 vs 14.7min ± 10.1), andin those who received DSD, mean time to firstDSD was 33.6 minutes. The rate of ROSC washigher in the control arm compared to therapyarm, though this was not statistically significant(31.4% vs. 23.5%) (p value = 0.26). Of the 32/51patients with ROSC in DSD arm, average timeto 1st DSD was lower (32.7 min vs. 35.01). Thesepatients had similar numbers of primary shocks(4.42 vs. 4.78) but required fewer DSD (2.8 vs.3.47). Conclusions: The management of RVFremains challenging. While the achievement ofROSC was higher in the non-DSD group, thedifference did not meet statistical significance.Those who received DSD earlier had higherrates of ROSC than those with more delay, andrequired fewer DSD attempts.

187. Prehospital Online MedicalOversight (Promo) an Analysis of theInteraction between Emergency RoomPhysicians and Paramedics

Jason Prpic, Alicia Violin, Sylvie Michaud,Nicole Sykes, Paul Myre, Health Sciences NorthCentre for Prehospital Care Category of Submis-sion: Operations, Quality, Safety Systems,Disaster

Background: In Ontario, paramedics operatemainly under off-line medical direction, theyuse online medical control when it is mandatoryaccording to provincial medical directives or ifa patient presents with a condition that does notfit into their protocols. Literature that encom-passes the interaction that occurs between over-sight physicians and paramedics is limited eventhough this interaction is critical to ensurepatients receive appropriate prehospital care.Objective: The objective was to describe thequality of online medical control in a Cana-dian EMS system and use the study findings todevelop a quality improvement program whichwill enhance the outcome of online medical con-trol. Methods: A retrospective review of writ-ten and audio records of online medical con-trol interactions from April 1, 2016 to March31, 2017. Audio recordings were assessed by asingle reviewer to evaluate predetermined cri-teria which gauged the efficiency of commu-nication that occurred during each interaction.Results: There were 454 online interactions inthe fiscal year, 14 cases were excluded as audiowas unavailable and 27 could not be retrieveddue to technology failure at the dispatch level.Therefore 413 cases were assessed. Three hun-dred thirty-eight patches (81.8%) were manda-tory provincial patch points with 289 (85.5%)regarding patients in cardiac arrest. Analge-sia administration made up 30.7% of the non-mandatory calls, and all resulted in medica-tion orders. In 100% of patches additional infor-mation was requested by the oversight physi-

cian and in 131 (31.7%) patches no request wasmade by the paramedic. The average lengthof patch was 0:02:03 (SD = 0:01:07) and theparamedic had to wait on average 0:01:11 (SD= 0:00:44) before talking to an oversight physi-cian. Conclusions: Implementing standardiza-tion of information handover will allow forpatch calls to be more efficient and ensure allpivotal information is communicated. This willallow oversight physicians to make informedclinical decisions optimizing the care providedto patients. To further enhance the medical con-trol provided by oversight physicians it wouldbe beneficial to determine the most effectiveway to provide EMS training to these physi-cians. Also, as all requests for analgesia weregranted, implementing a medical directive withincreased paramedic autonomy for pain controlwould be warranted.

188. Quantifying EMS Resource Allocationfor Pediatric Transports

Jennifer Anders, Jennifer Fishe, Kevin Psoter,Carla Tilchin, Kyle Fratta, Johns Hopkins Uni-versity School of Medicine Category of Submis-sion: Pediatric

Background: Regionalization of pediatric caredecreases available pediatric services at com-munity hospitals. Therefore, some childrenshould bypass closer hospitals for direct trans-port to pediatric specialty facilities. Futuretools assisting EMS with transport destinationchoices must balance EMS resource allocationwith direct transport’s benefits. To do so, thecurrent burden of pediatric transport on EMSagencies must be quantified to provide a bench-mark for future systems changes. Objective:The objective of this study was to describe thebaseline EMS services utilization for pediatrictransport in three geographically diverse juris-dictions (urban, suburban, and rural). Methods:This study examined a 12-month retrospectivecohort of pediatric (0–17 years) EMS transportsfrom three Maryland counties. All agencies usethe same patient care protocols, EMR, and Heli-copter EMS (HEMS) system. Each patient trans-port location, actual transport times, demo-graphics, and clinical variables were abstractedfrom the EMR. The response scene and destina-tion hospital locations were geocoded to calcu-late road driving distance. Each agency’s base-line EMS utilization for pediatric transport wasthen estimated using transport miles and min-utes. Results: The three counties transporteda total of 12,223 pediatric patients during the12-month period (urban n = 6,033, suburbann = 5,987, and rural n = 243). Total EMS uti-lization for pediatric transport was 63,631 min-utes and 27,613 miles in the urban jurisdic-tion; 91,002 minutes and 77,831 miles in thesuburban jurisdiction; and 5,248 minutes and7,605 miles in the rural jurisdiction. HEMS usefor pediatric transport was zero in the urbancounty, 0.1% in the suburban county and 4.8%in the rural county (p < 0.001). The mean trans-port time per patient varied significantly at10.6, 15.2, and 21.6 minutes, respectively (p <0.001). Mean road transport miles per patientwas 4.6, 13.0, and 31.3 miles, respectively (p< 0.001). On a population basis, EMS utiliza-tion for pediatric transport was 0.493, 0.494, and0.445 minutes per pediatric citizen and 0.214,0.716, and 0.381 road miles per pediatric citi-zen per year, respectively (p < 0.001). Conclu-sions: EMS resource use for pediatric transportsis noteworthy and varies significantly betweenurban, suburban, and rural jurisdictions. Thisstudy provides essential benchmarks for futuredevelopment of pediatric direct transport pro-tocols.

189. Assessment of Emergency MedicalServices Provider Research Literacy andInvolvement



Lauren Maloney, Robert Marshall, HenryThode Jr, Adam Singer, Scot Johnson, StonyBrook University Department of EmergencyMedicine Category of Submission: Student,Resident, Fellow

Background: For a needs assessment for futurecontinuing medical education classes androllout of prehospital clinical research, a surveywas developed to gather data on providerattitudes towards evidence-based medicine(EBM), participating in clinical research, andinformed consent. Methods: A 35 questionsurvey was distributed to 71 employees ofa university-based EMS system. Surveysincluded demographic and experience items.Responses to various statements were gradedon a 5-point Likert scale from “strongly dis-agree” to “strongly agree” and analyzed withChi square tests. Results: Of 54 analyzablesurveys, 81.5% respondents were paramedicsand 18.5% were EMT-Bs. 78% of respondentswere male. Mean age was 39 with an average18 years of EMS experience. 61% held collegedegrees, 48% subscribed to medical journals,and read articles a couple times a week (20%),month (32%), or year (35%). At least 95% ofproviders agreed about the importance of pre-hospital EBM and their responsibility to staycurrent with medical advances. Paramedicswere more likely than EMT-Bs to disagree thatEMS protocols are updated promptly. 37%agree that patient care decisions should bebased on research evidence and not personalexperience, (45% males vs. 8% females). 65%of those surveyed disagreed with limiting therights of an individual to better the care of alarge group, and disagreement was higher infemales than males (92% vs. 57%), respectively.A total of 96% agreed with an option to readresearch articles for CME; those without acollege degree were more likely to disagree. Nosignificant relationship between age, educa-tion, provider level, or experience existed withfrequency of reading research articles. A totalof 65% disagreed that spending an additional5 minutes after a call to complete clinical trialpaperwork would be a burden; those who readarticles a couple times a year or never weremore likely to agree. A total of 44% disagreedwith enrolling a critical patient in a trial ifdelayed consent is obtained, with a significantrelationship to age; younger respondents wereless likely to disagree than other age groups.Conclusions: In this cohort of prehospitalpersonnel, evaluating medical research andinvolvement in future prehospital clinical trialswas overall well received.

190. Can Heart Rate Variability RiskStratify Patients with UndifferentiatedNon-Traumatic Chest Pain?

Juan March, Carmon Russoniello, NicholasMurray, Walter Robey, East Carolina Univer-sity Department of Emergency Medicine Division ofEMS Category of Submission: Cardiac

Background: Previous research suggests thatheart rate variability (HRV), also known as R toR variability, can be used to risk stratify patientswith known acute coronary syndromes. TheHRV spectrum contains two major components.One component of HRV is the high frequency(0.18–0.4 Hz) component, which is synchronouswith respiration and is identical to respira-tory sinus arrhythmia. The second is a low fre-quency (0.04–0.15 Hz) component that appearsto be mediated by both the vagus and car-diac sympathetic nerves. This study examinedwhether heart rate variability can be used to riskstratify patients presenting with undifferenti-ated non traumatic chest pain. Methods: Thisexploratory study was performed at a percu-taneous coronary intervention capable tertiaryteaching hospital with 900 beds and an Emer-

gency Department (ED) with an annual cen-sus of 120,000. A convenience sample of adultpatients presenting to the Emergency Depart-ment with a chief complaint of non-traumaticchest pain were enrolled. HRV was capturedusing a physiological status monitor (PSM)affixed to the chest for a 5–10 minute periodduring the patient’s ED stay. High risk patientswere identified by either a positive troponin,positive stress test, positive cardiac catheteri-zation, ST elevation on EKG, or death within30 days. A low frequency/high frequency ratioof less than 1.0 was used as the cutoff. Data anal-ysis was performed with a Fischer Exact test.Results: A total of 26 patients were enrolled.All six patients identified as high risk had aLF/HF ratio of less than 1.0; sensitivity = 100%.Furthermore, all 20 patients who were deter-mined to be low risk had an HF/LF ratio >1.0; specificity = 100%, p< 0.0001. Conclusions:This pilot study suggests that heart rate vari-ability with a LF/HF ratio < 1.0 may be usedto rapidly risk stratify patients with undifferen-tiated non traumatic chest pain. Further stud-ies in the prehospital environment with a largersample size are needed to determine if HRV canbe used by EMS to rapidly risk stratify patientswith undifferentiated non traumatic chest pain.

191. Correlation of EEG-Based BrainResuscitation Index and End Tidal Co2 inPorcine Cardiac Arrest Model

Dongsun Choi, Hee Jin Kim, Taehan Kim,Ki Jeong Hong, Young Sun Ro, Kyoung JunSong, Hee Chan Kim, Shin Sang Do, SeoulNational University Hospital, Department of Emer-gency Medicine Category of Submission: Car-diac

Background: Evaluation and monitoring ofbrain viability is important during resusci-tation of cardiac arrest. We developed non-invasive EEG-based brain resuscitation index(EBRI) and evaluated correlation EBRI and end-tidal CO2(ETCO2). Methods: A crossover ani-mal experimental study using porcine cardiacarrest model was designed. After 1 minuteof untreated ventricular fibrillation, alterna-tion of high quality CPR (compression depth5 cm and compression rate 100/min) and lowquality CPR (compression depth 3 cm andcompression rate 60/min) was performed forevery 50 seconds in 10 phases. EBRI wascalculated from selected single EEG channelwhich have the lowest noise. Mixed modelanalysis was conducted to compare the dif-ferences of hemodynamic parameters, ETCO2and EBRI between high quality CPR periodand low quality CPR period. Pearson’s correla-tion coefficient was calculated to assess correla-tion between EBRI and ETCO2. Results: Exper-iment was performed in five female porcine(44.6 ± 2.8kg). EBRI and EtCO2 was obtainedaccording to quality of CPR received. DeltaEBRI obtained during high quality CPR wassignificantly higher than delta EBRI of lowerquality CPR (HQ: Median 0.17, (0.04–0.30), LQ:Median −0.18 (−0.05–−0.32), p =< 0.01). EBRIhad statistically moderate positive correlationwith ETCO2 (r = 0.56). Conclusions: In porcinecardiac arrest model, EEG-based Brain Resusci-tation Index was successfully obtained duringresuscitation and had statistically moderate cor-relation with ETCO2.

192. Social Connectedness and CopingStyles in EMS Workers and TheirAssociation with Burnout and PerceivedStress

Lori Boland, Pamela Mink, Jonathan Kamrud,Jessica Jeruzal, Russell Myers, Charles Lick,Andrew Stevens, Allina Health Emergency Med-ical Services Category of Submission: Profes-sional

Background: To assess social connectednessand coping styles among emergency medi-cal services (EMS) providers and explore theirassociation with occupational burnout and per-ceived stress. Methods: A 167-item electronicsurvey was distributed to employees of a largeambulance service that provides 9-1-1 responsein Minnesota. The survey included the MaslachBurnout Inventory (MBI), Cohen’s 4-item Per-ceived Stress Scale (PSS), the Brief COPE Inven-tory, and the Berkman-Syme Social NetworkIndex (SNI). Burnout was defined as a highscore on the emotional exhaustion (�27) ordepersonalization (�13) subscales of the MBI.The COPE inventory assesses an individual’stendency to use 14 coping styles in responseto stressful situations, with scores rangingfrom 2 (low use) to 8 (high use). Results:Responses were received from 217 providers(54% response); the mean age was 40, 60%were male, and 55% had an EMS tenure of10+ years. The prevalence of burnout was 18%and the mean PSS score was 4.8 (SD=3.2). TheSNI characterized respondents as socially iso-lated (15%), moderately isolated (33%), moder-ately integrated (29%), and socially integrated(24%), and the prevalence of burnout in eachgroup was 38%, 19%, 16%, and 7%, respec-tively. After adjustment for age, gender, EMStenure and marital status, providers charac-terized as socially isolated were more likelyto experience burnout than those who weresocially integrated (OR = 6.4; 95%CI = 1.3–32.2). Decreased social connectedness was asso-ciated with increased mean PSS score: sociallyintegrated = 3.8, moderately integrated= 5.0,moderately isolated = 4.8, and socially isolated= 6.0 (p for trend = 0.03). Commonly usedcoping strategies included planning, positivereframing, and active coping, while substanceabuse, behavioral disengagement, and denialwere infrequently used. Higher use (scores =6,7,8) of religion, use of emotional support, anduse of instrumental support to cope were asso-ciated with a lower prevalence of burnout. Con-versely, higher scores on the coping subscalesof self-blame, food, and substance abuse wereassociated with increased burnout and werecorrelated positively with PSS scores (all p <0.002). Conclusions: EMS providers who aresocially isolated or who frequently use the cop-ing strategies of self-blame, food, or substanceabuse may be at increased risk of burnout andstress, however the temporality of these charac-teristics remains unclear.

193. Impact of the Implementation of aCritically Ill Patient Bundle of Care onthe Performance of Key MedicalInterventions for Respiratory DistressPatients by Paramedics in the Field

Mark Pinchalk, Mark Tomassi, Roth Ronald,Jeffery Reim Jr., James Dlutowski, SimonTaxel, Thomas Goode, City of Pittsburgh EMSCategory of Submission: Professional

Background: Medical intervention patient carebundles have been advocated as a process basedsystem to improve patient care and outcomesusing evidence based guidelines. We soughtto evaluate the effect of the implementation ofa Prehospital “Crashing Patient” Critical CareBundle on the performance of key prehospitalintervention for patients presenting with respi-ratory distress. Methods: A “Crashing Patient”bundle of care addressing key interventions forcritically ill patients was implemented in anurban Advanced Life Support (ALS) EMS sys-tem from 2012–2014. After full implementationof the care bundle, retrospective Patient CareReport (PCR) review was conducted of PCRswith a chief complaint of “Respiratory Distress”for the first calendar quarter after implemen-tation (July–September 2014) and compared to



PCRs for the most recent quarter (April–June2017). Rates of EKG & end tidal carbon diox-ide (EtCO2), vascular access and CPAP appli-cation were compared for all respiratory dis-tress cases. For the subset of patients whoreceived Albuterol for bronchospasm, the ratesof administration of Methylprednisolone, Mag-nesium Sulfate and 1:1000 Epinephrine werecompared between the two time intervals.Results: There were 905 respiratory distressPCRs in the 2014 interval and 885 in 2017. In2017 there were improvements in EKG monitor-ing from 32.6% to 45.9% (p < 0.0001) of cases,EtCO2 monitoring from 7.1% to 17.3% (p <0.0001), vascular access from 37.2 to 45% (p =0.0009) & CPAP use from 6.5% to 10.8% (p =0.0013). 408 of the patients received Albuterolfor bronchospasm in 2014 compared to 306 in2017. In this subset there were improvements inthe administration of Methylprednisolone from24.4% in 2014 to 52% (p < 0.0001), Magne-sium Sulfate from 12.5% to 19.9% (p = 0.0091)& 1:1000 Epinephrine from 3.2% to 6.8% (p =0.0318). These care improvements were asso-ciated with a decrease of the rate of cardiacarrest after EMS contact for medical patientsin the system from 12% to 9.1% (NS). Conclu-sions: The implementation of a prehospital crit-ical (“crashing”) patient bundle of care resultedin a significant performance improvements inaccomplishing key interventions for respiratorydistress patients. Patient care bundles may havesignificant utility to improve patient care andsafety in the prehospital setting.

194. Tracking Violations of NewlyImplemented Behavioral EmergencyTreatment Protocol

Timothy Lynch, Christie Fritz, DavidSchoenfeld, Beth Israel Deaconess MedicalCenter/Harvard Medical School Category ofSubmission: Student, Resident, Fellow

Background: In September 2014, Massachusettsstatewide EMS protocols authorized the use ofhaloperidol and/or a benzodiazepine for man-agement of behavioral emergencies. The newlyadopted protocol allows for medication admin-istration with contraindications of age <18, his-tory of seizures, or prolonged QT interval. Geri-atric dosing was reduced by 50%. The new pro-tocol was implemented following a standardtraining module. The purpose of this investiga-tion is to describe the frequency and type of pro-tocol violations observed during the implemen-tation of a new protocol, with the goal of help-ing to better understand the types of errors, soas to improve implementation of future treat-ment protocols. This will help to determinewhat further training if any is needed and planfor future protocol roll out difficulties. Meth-ods: Retrospective chart review of calls occur-ring between October 1, 2014 and June 30, 2015,in which the new behavioral emergencies pro-tocol was utilized. Cases were reviewed forprotocol violations and the type of violationwas recorded. Results: There were a total of 56calls during the study period that utilized thenew behavioral emergencies protocol includ-ing the administration of haloperidol. Proto-col deviations were identified in 29% (95%CI18–42%) of cases. The most common error wasprotocol violations at 17%(95%CI 9–26%), with13% (95%CI 6–24%) having a seizure historyof or reported seizure and 4%(95%CI 1–13%)pediatric administrations. 9%(95%CI 4–20%) ofhaloperidol administrations were not reducedfor geriatric use. While not required by the pro-tocol, OLMC was contacted in 14% (95%CI 7–27%). Conclusions: Standard Treatment Proto-cols allow for rapid implementation of care byprehospital providers, without the need to con-tact OLMC. Little is known about the type andfrequency of errors observed when adopting a

new protocol and this analysis can provide use-ful insight to help better tailor training for newprotocol implementation. Additionally, unnec-essary calls to OLMC were observed, suggest-ing a lack of familiarity or confidence with thenew protocol. This investigation demonstratespotential risks in new protocol implementationand we recommend further study to developbest practices for training and implementationof new clinical protocols.

195. Emergency Physicain TelehealthDispositions of Low-Acuity 9-1-1 Patients

Michael Gonzalez, David Persse, Guy Gleis-berg, Karen DuPont, Andrew Kincannon,Houston Fire Department Category of Submis-sion: Medical

Background: Every day within the UnitedStates low-acuity patients are transported toemergency departments (ED) for primary care.American College of Emergency Physicians andNational Association of EMS Physicians believenot all patients require ALS care and in thesecircumstances, alternate transport and desti-nation may be appropriate. EMS patient dis-positions are traditionally determined by themedic assessment along with off-line medicaldirection. At present, literature regarding pre-hospital physician telehealth patient disposi-tions are limited. The aim of this study wasto measure and report prehospital EmergencyTelehealth and Navigation (ETHAN) mobile-integrated patient dispositions for alternatetransportation and/or destination. Methods:This retrospective study was conducted on con-secutive EMS patients triaged by telehealthemergency physicians in a major metropoli-tan urban fire-based EMS system from Decem-ber 2014 through May 2017. Once on scene,EMS completes a patient assessment togetherwith ETHAN inclusion/exclusion criteria. Ifeligible, the medic transfers the ePCR andcontacts the Physician, who interviews thepatient via real-time video/voice conferenc-ing and determines the appropriate disposi-tion. Those cases where the ETHAN proto-col was employed were abstracted from theePCR system. Descriptive statistics describestudy characteristics and a 95% confidenceinterval was calculated for telehealth disposi-tions. Results: During the study period 10,042patients met the ETHAN criteria. Among thisgroup of telehealth dispositions; alternate trans-port and/or destination 77% (95% CI 76–78%)(5942/Referred to ED by ETHAN MD-Cab orSelf Transport, 639/Referred to ETHAN Clinic-Cab or Self Transport, 839/Patient DeclinedClinic Referred wants ED Visit-Cab or SelfTransport, 340/Referred to Patient Care Physi-cian/Alternate Clinic-Cab or Self Transport),traditional transport 15% (1497/Referred forEMS Transport to ED-Ambulance) and miscel-laneous 8% (274/Patient Declined to speak withETHAN MD, 130/Unable to Complete Due toTechnical Issues, 230/Referred for Home CareInstructions Only, 31/Patient Refused EMSTransportation, 120/Other). The mean studyage was 44 years (range 1–99 years), 54% werefemale and no patient adverse events werereported. Conclusions: In this system’s popula-tion, telehealth alternate transport and/or des-tination dispositions significantly reduced low-acuity ambulance transports and ED visits. Fur-ther studies are warranted to develop guide-lines for uniform reporting of prehospital caredepositions based on the Physician telehealthmodel.

196. Novel Measure to CaptureTransactional Stress in ParamedicServices

Elizabeth Donnelly, Paul Bradford, CathieHedges, Matthew Davis, Doug Socha, Peter

Morassutti, University of Windsor Category ofSubmission: Operations, Quality, Safety Sys-tems, Disaster

Background: In the past few years, there hasbeen an increase in awareness of the challengeof managing work related stress in EMS. Extantresearch has liked different types of chronicand critical incident stress to stress reactionslike posttraumatic stress. However, there is notool to capture the transactional stresses that areassociated with the day to day provision of ser-vice (e.g., dealing with offload delays or manda-tory overtime) and interacting with allied pro-fessions (e.g., emergency department staff) orallied agencies (e.g., law enforcement). The pur-pose of this study was to develop and validate ameasure which captured transactional stressesin paramedics. Methods: An online survey wasconducted with ten Canadian Paramedic Ser-vices with a 40.5% response rate (n = 717). Fac-tor analysis was used to identify variation inresponses related to the latent factor of trans-actional stress. The scale was validated usingboth exploratory and confirmatory factor analy-ses. Results: The sample of transactional stressquestions was split to allow for multiple anal-yses (EFA n = 360/ CFA n = 357). In theexploratory factor analysis, principal axis fac-toring with an oblique rotation revealed a two-factor, twelve item solution, (KMO = .832, x2= 1440.19, df = 66, p < .001). Confirmatoryfactor analysis also endorsed a two factor, 12item solution, (x2 = 130.39, df = 51, p < .001,CFI = .95, TLI = .93, RMSEA = .07, SRMR= .06). Results supported two groups of six-item factors that captured transactional stressin the provision of service. The factors, clearlyaligned with transactional stress issues internalto the ambulance and transactional stress rela-tionships external to the ambulance. Both sub-scales demonstrated good internal reliability (α= .843/α = .768) and were correlated (p � .01)with a convergent validity measure. Conclu-sions: This study successfully validated a two-factor scale which captures stress associatedwith the day to day provision of EMS and theinteraction with allied professions. The devel-opment of this measure of transactional stressesfurther expands the potential that paramedics,Paramedic Services, employers, and prehospi-tal physicians may understand the dynamicsthat influence provider health and safety. Asa result, there may be greater opportunitiesto intervene holistically to improve paramedichealth and well-being.

197. Review of Emergency Medical Services(EMS) Transports to a FreestandingEmergency Department (FSED)

Matthew Chinn, Brittany Farrell, M. RiccardoColella, Medical College of Wisconsin Categoryof Submission: Operations, Quality, SafetySystems, Disaster

Background: Freestanding emergency depart-ments (FSED) are an area of expansion inhealthcare. Despite rapid growth, there is a min-imal amount of literature regarding the appro-priate triage of patients to these facilities byemergency medical services (EMS) providers.Purpose: The study seeks to review and developa list of objective markers for improving EMSfield triage to a FSED through evidence-basedrecommendations. Methods: Patient data wasretrospectively reviewed from the EPIC elec-tronic medical record system of all patientsbrought in to a single FSED by ambulance dur-ing a six month convenience period. A reportwas generated to abstract patient demograph-ics, medical information, and disposition. Miss-ing data fields were then manually entered.Ambulance services were all previously givena list of FSED capabilities and guidance onbypass for major trauma, STEMI, and stroke



care, amongst others, in accordance with bestpractice guidelines upon facility opening. Datawas analyzed using an unpaired t-test. Results:There were 138 patients brought to the FSEDfor the six month period of September 2016–February 2017 by 12 ambulance services. A totalof 105 patients were discharged home directlyfrom the FSED and 20 were transferred to a full-service hospital for admission or specialty care;7 were admitted to a psychiatric facility; 6 wereadmitted to a skilled nursing facility. Therewas a statistically significant difference in agebetween patients discharged home and thosetransferred to a full-service hospital (52.69 yearsvs. 71.75 years; p = 0.0011). There was no statis-tically significant difference between these twogroups in initial FSED pulse rate, respiratoryrate, systolic blood pressure, or temperature.There was a trend towards a longer length ofstay in the FSED for patients transferred to afull-service hospital (183.51 min vs. 236.25min;p = 0.0865). Conclusions: The data reflects thatolder patients are more likely to require transferand possibly admission after initial treatment ata FSED. The FSED initial vitals were not goodpredictors of the need for transfer; the use of ini-tial vitals as a surrogate for prehospital vitals isa limitation. Further research is needed to eval-uate other possible indicators that may be usedto triage patients to the most appropriate emer-gency department.

198. Variability of California Local EMSAgencies’ Pediatric Respiratory DistressProtocols and Their Corresponding Levelof Evidence

Jennifer Farah, J. Joelle Donofrio, NicholasAldridge, University of California, San DiegoCategory of Submission: Student, Resident,Fellow

Background: We sought to compare Califor-nia local EMS agencies’ (LEMSA) protocolsand review evidence-based guidelines on thetreatment of three main pediatric respiratorycomplaints by presentation: asthma (wheez-ing), bronchiolitis (wheezing <24 months), andcroup (stridor). Methods: In 2016, publiclyavailable protocols from 33 California LEMSAswere itemized and reviewed in the followingcategories: wheezing, wheezing <24 months,and stridor. Descriptive statistics were used tocompare these protocols. Literature reviews,including the American Academy of Pedi-atrics’ (AAP) current treatment guidelines,were used to create level of evidence (LOE)tables for asthma, bronchiolitis, and croup. Ofnote, steroids were included only in the litera-ture review, as California LEMSAs do not cur-rently use steroids prehospital. The evidence-based tables were compared to California localEMS agency protocols. Results: Among the33 LEMSAs, wheezing protocols had the leastamount of variability with only two of thesix treatments, ipratropium (15/33) and neb-ulized epinephrine (3/33), having >2 LEM-SAs with variability. The most common wheez-ing treatments included albuterol (33/33) andIV/IM epinephrine (33/33). The least commontreatments included nebulized epinephrine andmagnesium (2/33). Current evidence stronglysupports the use of albuterol, ipratropium,epinephrine, magnesium, steroids, and nonin-vasive positive pressure ventilation (NIPPV) inthe asthmatic child. Only three agencies dif-ferentiated wheezing in children <1 year ofage, referencing this as possible bronchioli-tis. All three included albuterol and NIPPVas their recommended treatments but did notinclude nebulized hypertonic saline, nebulizedepinephrine, steroids or suctioning. For chil-dren <24 months, albuterol and steroids areno longer strongly recommended based onnew AAP guidelines. Stridor had the highest

protocol variability, with no treatment havinguniform use among agencies. The most com-mon treatments included IV/IM epinephrine(24/33), NIPPV (29/33), and humidified mist(18/33). The least common treatments werenebulized epinephrine (12/33) and suctioning(4/33). For stridor, evidence supports the effi-cacy of all formulations of epinephrine. Conclu-sions: There is wide variation among CaliforniaLEMSAs in their management of pediatric res-piratory distress. Recent changes to treatmentguidelines have likely created the discordancebetween current treatment practices and LOEtables. Timely evidence-based updates willlikely benefit prehospital agencies’ treatmentprotocols.

199. Early Impact of an Emerging MihProgram for 9-1-1 High Utilizers

Jon Ehrenfeld, Ashley Clayton, CatherineCounts, Michael Sayre, Seattle Fire DepartmentCategory of Submission: Operations, Qual-ity, Safety Systems, Disaster

Background: Vulnerable, medically complexpatients comprise a disproportionate share ofresponses in an urban, fire-based EMS sys-tem. A social worker-based Mobile IntegratedHealthcare (MIH) program was designedto intervene using either direct engagement(DE) or indirect care coordination (CC). Wehypothesized that sustained outreach wouldreduce 9-1-1 activations and engage moreappropriate services. Methods: We usedcomputer-aided dispatch and electronic healthrecords to identify housed individuals with�3 EMS responses in the previous quarter.The social worker then assigned enrollees tothe DE or CC cohort based on previous callvolume, current services, vulnerability, andcase management history. We recorded medicaland social services in place, existing casemanagers or providers, services and referralsinitiated, ongoing 9-1-1 utilization, and reasonfor disenrollment when applicable. Groupswere compared by chi-squared and t-tests.Results: During the baseline quarter, EMSresponded 389 times to 45 patients. Twenty-eight were female, the median age was 64(IQR 56–71), 29 were Caucasian, and 12 wereAfrican American. All were medically andsocially complex, with a mean of >4 medical orsocial comorbidities per patient. Nineteen wereassigned to DE and 26 to CC. In the baselinequarter the cohorts had a similar number ofresponses (DE 9.5 ± 7.2, CC8.4 ± 4.7, p = 0.54).More patients in the DE cohort received multi-disciplinary care conferences (37% vs. 8%, p =0.02) and primary care linkage (53% vs. 15%, p= 0.008), while case management staffing alonewas more prevalent in the CC cohort (77%vs.. 47%, p = 0.041). Quarterly EMS responsesdeclined to 6 ± 5.7 after 3 months, 6.4 ± 6.6after 6 months, and 3.9 ± 4.5 after 9 months ofenrollment. Clients in the third quarter aver-aged a six call decrease compared to baseline(1.8–10.2, p = 0.011). Nine were disenrolleddue to death, relocation, or reduction in EMScalls. Conclusions: These preliminary findingsindicate that MIH direct engagement and carecoordination yield a reduction in 9-1-1 utiliza-tion. This study was limited by a small samplesize and lack of randomization, but stronglyindicates that additional investigation iswarranted.

200. Urban Law Enforcement NaloxoneDeployment for Treatment of SuspectedOut-of-Hospital Opioid Overdoses: A PilotProgram

Eric Cortez, David Keseg, James Davis, Ken-neth Kuebler, Ashish Panchal, Ohio Health Doc-tors Hospital Category of Submission: Opera-tions, Quality, Safety Systems, Disaster

Background: Law enforcement (LE) nalox-one programs aimed at early recognition andtreatment of opioid overdoses have increased.Implementation is often challenged by emer-gency medical services (EMS) engagement,which may impact adoption and overall suc-cess. The objective of this study was to ana-lyze the implementation of a naloxone pilot pro-gram at a large urban LE agency supportedby local EMS providers. We hypothesized thatwith direct training and interaction with EMSproviders, LE adoption would be high. Meth-ods: This prospective pilot program was con-ducted between May 2016 and December 2016.LE officers, investigative personnel, and sup-port personnel underwent training by the city’sfire-based EMS providers. LE training includedidentifying the symptoms of opiate overdose,and administration of naloxone if opioid over-dose was suspected and respiratory depressionwas present. LE personnel were deployed with2 mg naloxone doses administered intranasallywith a mucosal atomizer device. At the endof the study period, LE personnel completeda survey concerning their training and expe-rience with naloxone administration. LE clini-cal performance was monitored for each nalox-one administration. Outcomes included patientsurvival at the time of EMS arrival, and theresults of the post program survey. Results:A total of 124 LE officers underwent naloxonetraining with 31 (25%) LE officers administer-ing naloxone to 58 suspected overdose patients.Thirteen (42%) administered naloxone to morethan one patient. Fifty-six (97%) of the patientsreceived a single 2 mg dose of naloxone, and 2(3%) of the patients received two 2 mg doses ofnaloxone. Of the treated patients, 98% (57/58)patients survived to EMS arrival. The post pro-gram survey demonstrated that 82% of LE offi-cers felt they received adequate naloxone train-ing, 90% felt that the program promoted timelyand safe use of naloxone, and 90% felt pre-pared to handle issues on scene. Conclusions:This study suggests that urban LE agenciespartnered with EMS may successfully imple-ment naloxone administration programs forsuspected opioid overdoses. Limitations to thisstudy include the lack of patient-centered out-comes, and the significant number of LE officersthat did not administer naloxone.

201. Association of Case Volume PerAmbulance Station with Outcome ofOut-of-Hospital Cardiac Arrest (OHCA)

Tae Han Kim, Sang Do Shin, Kyoung JunSong, Ki Jeong Hong, Young Sun Ro, SoYeon Kong, Seoul National University Hospi-tal,Department of Emergency Medicine Categoryof Submission: Operations, Quality, SafetySystems, Disaster

Background: Sufficient case volume for emer-gency medical service may be important forretention of resuscitation skills and proce-dures during prehospital management of Out-of-Hospital Cardiac Arrest (OHCA). We eval-uated association of case volume per ambu-lance station with outcome of OHCA. Meth-ods: Nationwide data of all adult OHCA dur-ing 2013 to 2014 was retrospectively analyzed.All ambulance station was stratified in to 4groups according to annual average numberof OHCA treated by EMS teams dispatchedfrom each ambulance station. Multivariablelogistic regression model was conducted toevaluate effect of increased case volume peran ambulance station on survival outcome ofOHCA. Results: From 2013 to 2014, total of47,637 OHCAs were treated and transportedby EMS teams from 1,205 ambulance stationsnationwide. Mean annual number of OHCAdispatched from each ambulance station was19.8 cases. Overall survival to discharge rate



was 5.5% with 2.9% of discharge with favor-able neurological outcome. Survival was high-est in groups with largest case volume (7.2% ingroup 4(largest case volume) vs. 3.3% in group1(smallest case volume)). Adjusted odds ratio oflargest case volume per ambulance station forpredicting survival was 1.46(95% CI 1.26 – 1.70).Conclusions: Case volume of OHCA per ambu-lance station might be associated to survivaloutcome of EMS treated OHCA. Appropriateprehospital EMS dispatching strategy accord-ing to case volume should be further studied.

202. Resource Utilization and ClinicalOutcomes of Older Adult EMS Patientswith Traumatic Brain Injury Who WereTransferred to a Level I Trauma Center

Courtney Jones, Vasisht Srinivasan, JeremyCushman, Julius Cheng, Timmy Li, SuzanneGillespie, Martina Anto-Ocrah, Nancy Wood,Heather Lenhardt, Ann Dozier, Jeffrey Bazar-ian, Manish Shah, University of Rochester, Schoolof Medicine and Dentistry Category of Submis-sion: Trauma

Background: Traumatic brain injury (TBI) isa substantial source of death, disability, andhealthcare utilization among older adults.Older patients are frequently under-triagedby EMS to community hospitals and requiresubsequent transfer to a trauma center forfurther care. However, a minimal amountis known regarding the provision of careand patient outcomes at the final receivinghospital. We aimed to describe trauma centercare among geriatric transfer patients withTBI. Methods: We conducted a secondaryanalysis on a sub-cohort from a prospectivemulti-center study focusing on ambulance andemergency department (ED) care of injuredolder adults transported via ambulance. Thecurrent analysis focused on patients trans-ferred to the region’s Level I trauma centerfrom another hospital. The trauma centerfor the present study serves a nine countycatchment area of over one million people.Transfer paperwork from the originatinghospital was reviewed and a detailed medicalrecord abstraction was conducted, includingcomputed tomography (CT) findings, proce-dures, length of stay (LOS), and ED disposition.We used descriptive statistics to characterizethe study sample including proportions andconfidence intervals. Results: There were 205patients transported by EMS to a communityhospital who were subsequently transferredto the Level I trauma center. Thirty had con-firmed abnormalities on head CT (14.6%). Themean age was 78 years (range: 55–91), 57%female, and the most frequent mechanism ofinjury was falls (93%). Median length of stayat the trauma center was 13.5 days (range:0–230), with 8 patients staying one day orless. CT findings included subdural hematoma(60%), subarachnoid hemorrhage (50%), andintraparenchymal hemorrhage (36.7%). Fivepatients required neurosurgical intervention(17%), eight required ICU admission (27%),two were discharged from the ED (7%), andtwo transitioned to inpatient hospice (7%).Conclusions: In our sample, geriatric patientswith TBI who were subsequently transferred toa trauma center were overwhelmingly injuredvia falls and had variable resource utilizationand clinical outcomes. Additional ways forresponding EMS providers to identify geriatricfall patients who are at high risk for TBI arewarranted.

203. Relationships between Right Atrialand Aortic Pressures and Jugular andCarotid Flows Respectively in a SwineModel of Asphyxial Pseudo-PulselessElectrical Activity

Norman Paradis, Karen Moodie, Sarah Crock-ett, Jeffrey Gould, Christopher Kaufman,Dartmouth-Hitchcock Medical Center Categoryof Submission: Cardiac

Background: The initial cardiac rhythms foundduring in-hospital respiratory arrests are typi-cally either pulseless electrical activity (PEA) orasystole. Pseudo-PEA (p-PEA) often precedestrue PEA and is characterized by a low-flowstate in which cardiac contraction producesa non-palpable blood pressure. The purposeof the study was to characterize the relation-ships between venous and arterial pressuresand the flows that drive brain perfusion in ahypoxic asphyxial model of p-PEA. Hypoth-esis: We hypothesized that during CPR rightatrial pressure (RAP) would be related to jugu-lar venous flow (JVF), and that aortic pressure(AOP) would be related to carotid flow, and thatthese relationships might change with time dur-ing p-PEA. Methods: Pseudo-PEA was inducedvia hypoxic asphyxiation in 12 domestic swine(∼32 kg) with standard physiological monitor-ing. AOP and RAP were measured with solidstate transducers placed in the thoracic aortaand right atrium. Blood flow was measured inthe common carotid artery and jugular veinwith ultrasonic flow probes. FiO2 was reducedto 6% by increasing the fraction of nitrogen. Atarget systolic blood pressure (SBP) of 40 mmHgwas used to define p-PEA. The relationshipbetween pressures and flows was determinedwith a Pearson correlation coefficient. Results:Overall, RAP was significantly negatively cor-related with JVF (r = −0.51, p < 0.05), however,the relationship varied over time during p-PEA(Figure). AOP was significantly positively cor-related with carotid flow (r = 0.85, p < 0.05), butdid not show the same time dependence as seenwith RAP and JVF. Conclusions: In an asphyx-ial model of p-PEA, venous blood pressuresand flows were negatively associated and therelationship varied as a function of time. Arte-rial pressures and flow were positively associ-ated and the relationship varied less over time.These findings have implications for how andwhen chest compressions or other interventionsshould be applied when treating p-PEA.

204. Change in the Utilization ofEmergency Care after Establishment ofEmergency Centre in Yaoundé, Cameroon:A Before and After Cross-SectionalAnalysis

So Yeon Kong, Sang Do Shin, Young Sun Ro,Yun Jeong Kim, Joong Sik Jeong, Dae Han Wi,Seoul National University Hospital Category ofSubmission: Medical

Background: In effort to address the shortage ofemergency medical care in Cameroon, YaoundéMedico-Surgical Emergency Center (CURY)was established in June, 2015 in Yaoundé,Cameroon. To evaluate its impact on the com-munities of Yaoundé, we assessed the changesin utilizations of emergency medical care sincethe establishment of CURY. Methods: In 2014the first survey was conducted on randomlyselected 619 households (3,358 individuals) liv-ing in six health districts of Yaoundé. In 2017the second quantitative survey was conductedon 634 households (3,466 individuals) usingthe same survey methods as the first survey.In both surveys, data on demographic infor-mation, socioeconomic status, and utilizationof healthcare, including emergency care in thepast year were collected on every member ofthe households via face-to-face interview. Dataon two surveys were compared and emergencyunit utilization by the distance from CURY wasexamined. Results: Participants in the both sur-veys had similar age and gender distributionwith mean age of 24 and 54% being male.In 2014 survey, healthcare utilization rates for

outpatient, emergency unit, and hospitalizationwere 37.2%, 4.5%, and 9.6%, respectively. In2017 survey, corresponding rates were 32.4%,5.7%, and 8.7%, respectively. The increase inthe utilization of emergency unit between twosurveys were statistically significantly (p=0.01).When the emergency unit utilization rates wereexamined by 3 km radius from CURY, therewas decrease in the utilization of emergencycare among residents living near CURY (27.3%in 2014 to 22.8% in 2017). Conclusions: Afterthe establishment of emergency medical center(CURY) in Yaoundé, Cameroon, the utilizationof emergency care was significantly increased.This increase was regardless of the distancefrom the patients’ residential places to the emer-gency medical center, suggesting that the estab-lishment of an emergency medical center mayhave impacted the utilization of emergencycare throughout the entire communities ofYaoundé.

205. Prehospital Push Dose Epinephrine inHypotension

Mark Merlin, Navin Ariyaprakai,Ammundeep Tagore, Matthew Harris,Andrew Parrish, Josh Schwarzbaum, AlexTorres, Michael Carr, Susmith Koneru, NewarkBeth Israel Medical Center/RWJBarnabas-MONOCCategory of Submission: Medical

Background: Hypotension is commonlyencountered in the prehospital arena andoccurs in the setting of illness, trauma or maybe iatrogenic during rapid sequence intubation(RSI). The mainstay of prehospital treatmenthas been intravenous (IV) fluids; however, thismethod is not always effective. Push dosesof epinephrine or phenylephrine, so called“push-dose pressors,” have long been used byanesthesiologists for acute hypotension in theoperating room. Push dose epinephrine (PDE)offers another tool to advanced life support(ALS) providers to combat hypotension. Meth-ods: A retrospective review of data collected forthe administration of PDE for the managementof acute hypotension in the prehospital setting.We included patients >17 years old withsystolic blood pressures <90 mmHg during theperi intubation period. Primary outcome wascardiac arrest. Secondary outcomes includedchanges in vital signs and shock index (SI).We performed descriptive statistics on demo-graphics, biometrics and derived the mean,median and standard deviations for continuousvariables of both the interventional and con-trol group. RESULTS: PDE was administered75 times in the two-year study period. 22 ofthose were peri-intubation (treatment group).Mean age in PDE was 69 years vs. 72.4 yearsin control group (P = 0.23). When comparingpre- and post-intubation vital signs of patientsreceiving PDE, we found significant increasesin mean HR, SBP, DBP, MAP, and SI (P <0.001). In the control group SBP, DBP, MAP,SI, and RR all achieved a statistical significantdecrease of the mean (P < 0.001). The meandose of epinephrine was 10 micrograms (range10–80mcg); 19.7% of peri-intubation patientsin the control group went into cardiac arrest.Only 4.5% of patients in the treatment groupwent into cardiac arrest. This did not reachstatistical significance. Conclusions: PDEused in the management of peri-intubationhypotension in the prehospital setting resultedin statistically significant improvements in SBP,DBP, MAP and SI. The control group showedstatistically significant worsening of vital signsafter intubation. Overall, fewer patients wentinto peri-intubation cardiac arrest after receiv-ing PDE. Readily available, easily composedand rapidly effective, PDE is a useful tool tocombat acute hypotension in the prehospitalarena.



206. Accuracy of Stroke Dispatch by aLarge Urban EMS Dispatch System

Thomas Lardaro, Dustin Holland, TomArkins, Dan O’Donnell, Indiana UniversitySchool of Medicine Category of Submission:Medical

Background: Stroke is a time sensitive emer-gency that requires appropriate triage in EMStransport planning. The existence of hospitalswith varying stroke-care capabilities and morerecently mobile stroke units (MSUs) neces-sitates early recognition of stroke symptomsand accurate triage of patients to appropriateresources. This study investigates the accuracyof the EMS dispatch system in a major U.S.metropolitan area in predicting whether or not a

patient is having a stroke. Objective: The objec-tive of this study was to evaluate the accuracyof stroke recognition by a large urban EMS dis-patch system in the United States. Methods: Weperformed a retrospective cohort study look-ing at the initial dispatch for stroke within alarge urban-area EMS system. We then com-pared these patients to a stroke registry froma large urban tertiary hospital in the same cityover a two-year period (2015–2016). Results:Over the study period, a total of 33,910 patientswere transported to the tertiary care hospital forany complaint, including 778 patients with aninitial dispatch code for stroke. Of the patientswith initial dispatch coded as stroke, 133 werethen confirmed as truly having a stroke basedon stroke registry data. Dispatch for stroke had

a sensitivity of 43.2% (95% CI 37.6–48.9), speci-ficity of 98.1% (95% CI 97.9–98.2), positive pre-dictive value of 17.1% (95%CI 15.1–19.3), andnegative predictive value of 99.5% (95% CI 99.4–99.5). Conclusions: These findings imply EMSdispatch alone is not sufficient to rule-in stroke.In the case of MSUs, dispatch alone may leadto patients being inappropriately triaged to thisresource due to the 82.9% false positive rate. Theauthors conclude that (1) triage tools beyonddispatch are required to ensure appropriatetriage of potential stroke patients for interceptby a MSU or transport to a stroke center and (2)EMS systems need triage tools to prevent inap-propriate triage of non-stroke patients to suchresources such as MSUs to ensure patient safetyand to prevent delays in definitive care.

Date post:	05-Nov-2020
Category:	Documents
Upload:	others
View:	2 times
Download:	0 times