Air Amb Evidencias

NEW ZEALAND HEALTH TECHNOLOGY ASSESSMENT (NZHTA) Department of Public Health and General Practice Christchurch School of Medicine and Health Sciences Christchurch, New Zealand

Transportation of emergency patients Dr Robert Weir

NZHTA TECHNICAL BRIEF March 2007 Volume 6 Number 4

This report should be referenced as follows: Weir, R. Transportation of emergency patients. NZHTA Technical Brief 2007; 6(4) Titles in this Series can be found on the NZHTA website: http://nzhta.chmeds.ac.nz/ publications 2007 New Zealand Health Technology Assessment (NZHTA) ISBN 978-1-877455-03-2 (Print) ISBN 978-1-877455-04-9 (Web) ISSN 1175-7884

TRANSPORTATION OF EMERGENCY PATIENTS

i

CONTRIBUTIONS BY AUTHORS

It was authored by Dr Robert Weir (Director), who conducted the critical appraisals, prepared the report and coordinated the project.

ACKNOWLEDGEMENTS

This Technical Brief was commissioned by the New Zealand Ministry of Health.

The literature search strategy was developed and undertaken by Susan Bidwell (Information Specialist). Cath Turnbull (Administrator) provided document formatting. Internal peer review was provided by Dr Martin Than.

DISCLAIMER

NZHTA takes great care to ensure the accuracy of the information supplied within the project timeframe, but neither NZHTA nor the University of Otago can accept responsibility for any errors or omissions that may occur. NZHTA and the University of Otago along with their employees accept no liability for any loss of whatever kind, or damage, arising from the reliance in whole or part, by any person, corporate or natural, on the contents of this paper. This document is not intended to be used as personal health advice; people seeking individual medical advice are referred to their physician. The views expressed in this report are those of NZHTA and do not necessarily represent those of the University of Otago, or the New Zealand Ministry of Health.

COPYRIGHT

This work is copyright. Apart from any use as permitted under the Copyright Act 1994 no part may be reproduced by any process without written permission from New Zealand Health Technology Assessment. Requests and inquiries concerning reproduction and rights should be directed to the Director, New Zealand Health Technology Assessment, Christchurch School of Medicine and Health Sciences, P O Box 4345, Christchurch, New Zealand.

CONTACT DETAILS

New Zealand Health Technology Assessment (NZHTA) Department of Public Health and General Practice Christchurch School of Medicine and Health Sciences PO Box 4345 Christchurch New Zealand Tel: +64 3 364 3696 Fax: +64 3 364 3697 Email: [email protected]

Website: http://nzhta.chmeds.ac.nz


ii

LEVEL OF EVIDENCE CONSIDERED IN TECHNICAL BRIEFS

Technical Briefs are rapidly produced assessments of the best available evidence for a topic of highly limited scope. They are less rigorous than systematic reviews. Best evidence is indicated by research designs which are least susceptible to bias according to the National Health and Medical Research Council’s (NHMRC) criteria (see Appendix 2). Where methodologically acceptable and applicable, appraised evidence is limited to systematic reviews, meta-analyses, evidence based clinical practice guidelines, health technology assessments and randomised controlled trials (RCTs). Where not available, poorer quality evidence may be considered.

CONFLICT OF INTEREST

None.


iii

EXECUTIVE SUMMARY

Aim

This technical brief examined four questions:

1. In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on emergency helicopter services improve health outcome when compared with transportation by emergency helicopter without a medical doctor?

2. In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on a road ambulance service improve health outcome when compared with transportation by a road ambulance service without a medical doctor?

3. In adults and children with a medical or trauma related emergency, does the presence of a medical crew able to perform rapid sequence intubation and/or thoracostomy improve health outcome when compared with a medical crew unable to perform rapid sequence intubation and/or tube thoracostomy and/or thoracotomy?

4. In adults and children with a medical or trauma related emergency how does variation in the time from callout to arrival at a medical facility with definitive care influence health outcome?

There are two general strategies about pre-hospital transportation: “scoop and run” and “stay and treat”. Scoop and run consists of short times at the scene with the emphasis being to transport the patient to definitive care as quickly as possible. In contrast, stay and treat involves longer times at the scene in order to start the stabilisation process. The above questions were designed to help address the most appropriate transportation strategy.

Data sources

The literature was searched using the following bibliographic databases: Medline, Embase, Cinahl, Current Contents, Science Citation Index, and Social Science Citation Index. Review databases searched were the Cochrane Database of Systematic Reviews, the Database of Abstracts of Reviews of Effects, the NHS Economic Evaluation Database and the Health Technology Assessment Database. Relevant publications referenced in material obtained in the course of the project were also identified.

Searches were performed between 24 October and 7 November 2006, and were restricted to material in English published from 1980 onwards.

Selection criteria

The selection criteria varied by review question. Selection criteria for each question included:

� primary aim of the study was to evaluate the comparison of interest in each review question

� methods were clearly described

� studies had a relevant control group

� study population included trauma and/or medical emergencies

� minimum sample size of 50

� outcomes included death and days in hospital

� Non-English language articles and publications that had been superseded were excluded.

Criteria were also set for the intervention and comparator based on the review question of interest.


iv

Data extraction

A single reviewer extracted data and appraised the selected studies. Summaries were presented in the form of evidence tables and full text.

Key results and conclusions

Key results were: 1. There was generally more support for the inclusion of doctors on helicopters in the seven

studies appraised in this section. However, there were uncertainties due to study design issues (levels of evidence ranged between III-1 and III-3), lack of consideration about whether non-doctor groups can be trained to perform certain procedures that would improve patient outcome and whether there may be different clinical scenarios that would favour one crew mix over another.

2. Similar considerations applied in the studies examining the use of doctors on board road ambulances. There were four studies in this section with levels of evidence ranging between III-2 and III-3.

3. When considering the outcome in patients who were treated by crews able to perform rapid sequence intubation and/or thoracostomy with other crews who were not able to perform these procedures, the only studies identified that met the study eligibility criteria included doctors amongst those able to perform the procedures of interest. It was therefore not possible to form conclusions about the effectiveness of non-doctor crews able to perform the procedures of interest when compared with crews that included a doctor. There were five studies in this section with levels of evidence ranging between III-1 and III-3.

4. There was inconsistent data on the association between pre-hospital time and patient outcome. However, the general direction was to support improved outcome in association with shorter pre-hospital times. Two studies provided information to consider whether crew mix or rapid transport had a more significant bearing on outcome. The results were conflicting across these two studies. There were 21 studies in this section with levels of evidence all being III-2.

5. Most of the studies included related to trauma rather than medical emergencies.

6. There was insufficient information to consider subgroups based on injury severity or age group.

While the balance of studies support improved outcome associated with doctors on board emergency transportation, the robustness of these studies and the areas of uncertainty that remain (see under research gaps) provide uncertainty about the best approach. The best study supported the use of doctors on board helicopters. The balance of studies supported improved outcome associated with shorter pre-hospital times. The studies identifying improved outcome frequently assessed the linear relationship between pre-hospital outcome and time, meaning that the focus was on any improvement in outcome rather than a set threshold of pre-hospital time to meet in order to achieve improved outcome.

Further research/reviews required

Some general areas of future research that would be helpful include:

1. Is there some form of interaction between pre-hospital time and pre-hospital crew that has impact on patient outcome? Linked to this is whether the same pre-hospital approach (time and crew) results in improved outcome in all emergency patients or whether the best approach is dependent on the clinical situation.

2. Given differences in procedures performed and clinical assessment processes adopted by doctors compared with non-doctor pre-hospital personnel, to what extent would enhanced procedure training for non-doctor groups be helpful?

3. There are cost differences between the “scoop and run” and “stay and treat” approaches, along with the crew mixes used that ideally should be examined in relation to cost effectiveness of different approaches. However, given current uncertainties in


v

effectiveness of the different strategies, incremental cost effectiveness can not be robustly examined at this time.


vi

TABLE OF CONTENTS

CONTRIBUTIONS BY AUTHORS ....................................................................................................................I ACKNOWLEDGEMENTS ...............................................................................................................................I DISCLAIMER ...............................................................................................................................................I COPYRIGHT ................................................................................................................................................I CONTACT DETAILS .....................................................................................................................................I LEVEL OF EVIDENCE CONSIDERED IN TECHNICAL BRIEFS...........................................................................II CONFLICT OF INTEREST .............................................................................................................................II EXECUTIVE SUMMARY .............................................................................................................................III Aim.......................................................................................................................................................... iii Data sources........................................................................................................................................... iii Selection criteria..................................................................................................................................... iii Data extraction........................................................................................................................................ iv Key results and conclusions..................................................................................................................... iv Further research/reviews required.......................................................................................................... iv TABLE OF CONTENTS ............................................................................................................................... VI ABBREVIATIONS ..................................................................................................................................... VII BACKGROUND ...........................................................................................................................................1 International variation in approach to transporting emergency patients.................................................1 Procedures performed by different groups of transport staff ...................................................................2 Dispatch strategies ...................................................................................................................................2 Golden hour..............................................................................................................................................3 Definitions related to transportation times...............................................................................................3 Traumatic versus non-traumatic medical emergencies ............................................................................3 Study types ................................................................................................................................................4 Scope of evaluation...................................................................................................................................6 Review Questions......................................................................................................................................7 SELECTION CRITERIA.................................................................................................................................8 MAIN SEARCH TERMS ..............................................................................................................................11 SEARCH SOURCES ....................................................................................................................................12 Bibliographic databases .........................................................................................................................12 Review databases....................................................................................................................................12 APPRAISAL METHODOLOGY.....................................................................................................................12 RESULTS..................................................................................................................................................13 Doctor versus no doctor on board helicopter.........................................................................................13 Doctor versus no doctor on board road ambulances .............................................................................35 Comparison of outcomes amongst crews that do and do not perform rapid sequence intubation

and/or thoracostomy.............................................................................................................................47 Outcomes by time from ambulance call out to emergency department delivery.....................................63 OVERVIEW ............................................................................................................................................115 Main findings........................................................................................................................................115 Limitations............................................................................................................................................116 Research gaps.......................................................................................................................................116 Conclusions ..........................................................................................................................................116 REFERENCES .........................................................................................................................................117 APPENDIX 1: SEARCH STRATEGY ...................................................................................................122 APPENDIX 2: LEVELS OF EVIDENCE ................................................................................................130 APPENDIX 3: EXCLUDED RETRIEVED PAPERS: DOCTOR VERSUS NO DOCTOR ON HELICOPTERS ......131 APPENDIX 4: EXCLUDED RETRIEVED PAPERS: DOCTOR VERSUS NO DOCTOR ON ROAD

AMBULANCES............................................................................................................134 APPENDIX 5: EXCLUDED RETRIEVED PAPERS: COMPARISON OF OUTCOMES AMONGST CREWS

THAT DO AND DO NOT PERFORM RAPID SEQUENCE INTUBATION AND/OR

THORACOSTOMY .......................................................................................................136 APPENDIX 6: EXCLUDED RETRIEVED PAPERS: OUTCOMES BY TIME FROM AMBULANCE CALL OUT

TO EMERGENCY DEPARTMENT DELIVERY..................................................................142 APPENDIX 7: INCLUDED PAPERS ....................................................................................................151


vii

LIST OF TABLES

Table 1. Inclusion/exclusion criteria for the effectiveness of including medical doctors on emergency helicopter transportation.................................................................................. 8 Table 2. Inclusion/exclusion criteria for the effectiveness of including medical doctors on road ambulances ........................................................................................................................ 9 Table 3. Inclusion/exclusion criteria for the effectiveness of including a crew able to perform rapid sequence intubation and/or thoracostomy on pre-hospital emergency transportation ................................................................................................................... 10 Table 4 Inclusion/exclusion criteria for the ideal time between call out and delivery of emergency patients to the emergency department ........................................................... 11 Table 5 Reasons for exclusion of studies before retrieval in full text: doctor versus no doctor on board helicopter ............................................................................................................... 13 Table 6 Reasons for exclusion of studies retrieved in full text: doctor versus no doctor on board helicopter ........................................................................................................................ 13 Table 7 Evidence tables of studies comparing the outcome or patients transported by helicopter with and without a medical doctor on board.................................................................... 18 Table 8 Reasons for exclusion of studies before retrieval in full text: doctor versus no doctor on board road ambulances .................................................................................................... 36 Table 9 Reasons for exclusion of studies retrieved in full text: doctor versus no doctor on board road ambulances .............................................................................................................. 36 Table 10 Evidence tables of studies comparing the outcome of patients transported by road ambulance with and without a medical doctor on board.................................................. 39 Table 11 Reasons for exclusion of studies before retrieval in full text (from additional search): patients transported by crews that do and do not have the ability to perform rapid sequence intubation and/or thoracostomy........................................................................ 47 Table 12 Reasons for exclusion of studies retrieved in full text: patients transported by crews that do and do not have the ability to perform rapid sequence intubation and/or thoracostomy................................................................................................................... 47 Table 13 Evidence tables of studies comparing crews that do and do not perform rapid sequence intubation and/or thoracostomy ....................................................................................... 52 Table 14 Reasons for exclusion of studies before retrieval in full text: time from callout to emergency department delivery ....................................................................................... 63 Table 15 Reasons for exclusion of studies retrieved in full text: time from callout to emergency department delivery ......................................................................................................... 63 Table 16 Evidence tables of studies examining time from ambulance callout to emergency department arrival ............................................................................................................ 73 Table 17 Key results for studies examining time from ambulance callout to emergency department arrival that found an association between prolonged pre-hospital time and poor outcome......................................................................................................................... 112 Table 18 Key results for studies examining time from ambulance callout to emergency department arrival in studies that did not directly compare pre-hospital time with outcome ........... 113 Table 19 Key results for studies examining time from ambulance callout to emergency department arrival in studies that did not find an association between pre-hospital time and outcome.......................................................................................................................... 114


viii

ABBREVIATIONS

AIS abbreviated injury scale

ALS advanced life support

APACHE acute physiology and chronic health evaluation

ASCOT A severity characterisation of trauma

AUC area under the curve

CI confidence interval

CPR cardiopulmonary resuscitation

DHB District Health Board

ED emergency department

GCS Glasgow Coma Scale

HEMS helicopter emergency medical service

ICU intensive care unit

ISS injury severity score

IV intravenous

MECU mobile emergency care unit

MI myocardial infarction

MTOS Major Trauma Outcome Study

NSW New South Wales

OR odds ratio

Ps probability of survival

RTS revised trauma score

TISS Therapeutic Intervention Scoring System

TRISS Trauma and Injury Severity Score

TS trauma score

UK United Kingdom

USA United States of America


1

BACKGROUND

This technical brief was requested by Paul Howard, DHB Funding and Performance Directorate, Ministry of Health, New Zealand Government.

International variation in approach to transporting emergency patients

Variations are seen in the approach to transportation of emergency patients internationally. Some use helicopters with doctors on board, others use helicopters that include nurses or paramedics. Some regions use road ambulance when others would use helicopters and some have used a combination of helicopters supported by ground services (Frankema et al. 2004). At longer distances, there is also variation between fixed wing and rotary wing aircraft. The most appropriate approaches are controversial, particularly in relation to:

1. Staffing of emergency helicopter services

2. Road versus air ambulance transportation.

There is also debate about time spent on the trauma scene (Feero et al. 1995; Sampalis et al. 1993), with Europe tending towards placing more importance on field stabilisation while the North American philosophy tends to be that field stabilisation only contributes to reducing mortality if on-scene time is not prolonged or if the patient would die without immediate intervention (Osterwalder 2002). In some regions, such as Europe and Quebec, field stabilisation is usually achieved by physicians whereas in others paramedics with access to voice control by physicians are used in the field (Sampalis et al. 1993).

The optimal staffing for helicopter transport is controversial (Bartolacci et al. 1998; Cameron 1999) (Cameron and Zalstein 1998; Gisvold 2002; Rhee et al. 1986). Proponents of the inclusion of a physician on helicopter transport suggest improved outcomes can be expected from improved assessment and better use of advanced life support (ALS) skills (Rhee et al. 1986). However, the major advantage from the use of helicopters may be a result of speed (Matsumoto et al. 2006). Under those circumstances, the use of a limited range of ALS skills would be more appropriate, thus obviating the need for physician skills. Some suggest ALS may be more beneficial in certain circumstances, such as blunt trauma (Cameron 1999).

A range of studies exist that compare road with air ambulance. These studies reflect ongoing controversy about the most appropriate use of these two forms of transport (Cameron and Zalstein 1998; Kerr et al. 1999; Nicholl et al. 1995; Thomas et al. 2002).

In the UK the first helicopter ambulance service started in 1987. Since that time air ambulance services operating in the UK have expanded and, interestingly, bear little resemblance to one another, further emphasising the uncertainty about the most appropriate transportation methods in different circumstances. A study set in London reported crew arrangements at the time. The service, which operated from Royal London Hospital, was crewed by two pilots, a registrar and a paramedic (Nicholl et al. 1995). In another study, set in Italy, the helicopter was crewed by an anaesthetist, registered nurse with ICU or pre-hospital emergency experience, a flight co-ordinator with the same skills as the registered nurse and the pilot (Sanson et al. 1999). A national air ambulance service implemented in Norway in 1988 also included anaesthetists on all helicopter services (Nielsen et al. 2002). A study set in Michigan used helicopters crewed by a pilot, flight nurse (with emergency of intensive care experience) and a physician (resident or fellow), (Rhee et al. 1986). Japan has recently started using helicopters with physicians on board to transport emergency patients (Matsumoto et al. 2006).

Some air ambulance services are developed to overcome issues of remoteness. For example, an emergency medical retrieval service was set up in the Argyll and Clyde Health Board, Scotland to support rural community hospitals in the area (Corfield et al. 2006). Transferring patients via helicopter resulted in highly significant differences in transfer times when compared with road/ferry combinations in the regions covered by this service.


2

Procedures performed by different groups of transport staff

Procedures performed by different groups vary by region. For example, paramedics in New South Wales intubate without paralysis or sedation. Therefore, only unconscious patients could be intubated. In other regions, paramedics can use paralysing agents and sedation (Garner et al. 1999; Murphy-Macabobby et al. 1992). In Quebec, emergency medical technicians are prohibited from performing any invasive procedures or administering medications. Physicians are dispatched to the trauma scene based on severity of the trauma. Only physicians can intubate, insert intravenous (IV) lines, and administer IV fluids and medications. Emergency technicians are restricted to basic life support procedures including extrication, wound dressing, head and spine immobilisation, oxygen administration, fracture splinting and cardiopulmonary resuscitation (CPR) (Sampalis et al. 1992).

Thoracostomy may be differentially performed by staff groups. There is some controversy concerning relative outcomes of thoracostomy pre-hospital versus in the emergency department setting. Spanjersberg et al. (2005) compared thoracostomies in these two settings and found no difference in complication rates between the two groups.

Intubation frequency and difficulty in the pre-hospital setting has been examined. A study set in the USA found there were 11,951 intubations from 1,544,791 patient care reports (0.77%), (Wang et al. 2005). In a study set in Germany, pre-hospital intubation was performed in 342 of 3669 (9.3%) patients treated by helicopter emergency medical service (HEMS), (Helm et al. 2006). The first attempt was successful in 87.4% of the 342 intubations.

Some studies have raised uncertainties about the use of IV fluids pre-hospital. For example, Sampalis et al. (1997) found that IV fluid replacement was associated with an increase in mortality risk and this association was exacerbated by increased pre-hospital times. A health technology assessment of the issue found no evidence that pre-hospital IV fluid resuscitation was beneficial and some evidence was identified that it may be harmful. However, they commented that this evidence was not conclusive (Dretzke et al. 2004).

Further differences in procedures performed are documented in the studies selected for appraisal in this review.

Dispatch strategies

Dispatch strategies vary internationally and within countries (Garner et al. 1999). In some countries (and studies) there is no consistency in approach between systems. For example, a study comparing a physician with a paramedic crew appeared to use the physician crew on a selective basis (Garner et al. 1999), making interpretation of the study results difficult.

Some have stated that helicopters have a role in transporting critically ill trauma patients over distances greater than 50 km or 30 minutes by road (Cameron and Zalstein 1998; Garner et al. 1999; Ministerial Taskforce on Trauma and Emergency Services 1999).

The following section provides examples of selected dispatch strategies to illustrate the variation in approaches.

The dispatch of CareFlight, a medically staffed helicopter service operating in New South Wales (NSW), is at the discretion of the NSW Ambulance Service. Dispatch is based on injury severity, entrapment, remote location or difficult hoist (needing a rescue hoist for extraction), (Bartolacci et al. 1998).

In Rotterdam, the following criteria are used for the primary deployment of a helicopter transported medical team for trauma patients (Frankema et al. 2004):

� place difficult to reach for ambulances

� in the professional opinion of the dispatcher, the helicopter service provides additional value

� motor vehicle crashes with estimated speed > 30 km/hr


3

� frontal collisions outside the built up area of town

� fall from > 6 metres or third floor

� entrapment in vehicle

� death of other occupant

� ejected from vehicle

� explosions

� near drowning or diving accidents

� exposure to toxic chemicals

� inhalation trauma or severe burns

� penetrating injuries to head, neck or trunk

� pelvic, spinal or femur fracture

� comatose (Glasgow Coma Scale ≤ 8)

� systolic blood pressure < 95 mm Hg or pulse > 120 per min

� major estimated blood loss (> 1 litre)

� respiratory distress.

Some centres have a paramedic stationed in ambulance service control to help identify calls that would benefit from helicopter retrieval (Nicholl et al. 1995).

Golden hour

Trunkey classified deaths as immediate, early and late. Immediate deaths were defined as occurring instantaneously or within one hour of the time of injury. Such deaths resulted from severe injuries to the brain, major blood vessels, heart or spinal cord. Early deaths were defined as occurring between one hour and one week following injury. These deaths resulted from major haemorrhage, multiple brain injuries and severe brain damage. These injuries should not result in death if definitive care can be given within one hour (the golden hour) of the time of injury. Late deaths occur more than one week after injury and result from later complications or infections (Trunkey 1983). The data used to support this classification was largely based on wartime findings. There is debate whether the golden hour applies to civilian settings (Lerner and Moscati 2001).

Definitions related to transportation times

There are five phases in the transportation of emergency patients to the emergency department:

1. Activation.

2. Dispatch.

3. Travel.

4. Patient preparation and treatment.

5. Travel.

Activation, dispatch and travel to the scene represent the response time. The retrieval time encompasses all five phases (Ministerial Taskforce on Trauma and Emergency Services 1999).

Traumatic versus non-traumatic medical emergencies

The scope of this review included both traumatic and non-traumatic medical emergencies. The studies examining the staff mix on helicopters tended to focus on study populations resulting from trauma as opposed to medical emergencies. Trauma patients are thought to be good subjects as the predicted mortality can be estimated objectively and compared with the actual mortality (Osterwalder 2003). The


4

great majority of studies eligible for the comparison of outcomes by pre-hospital time were also conducted in trauma populations. Reasons for this are explained in more detail in the relevant results section.

Study types

Judgment of skills required for specific trauma incidents

In some studies the judgment and skills shown by physicians were used to evaluate their role on helicopter transport (Dalton et al. 1992; Rhee et al. 1986). These studies can be expected to be somewhat subjective and the added value proposed for physicians may be dependent on who is doing the evaluation. In some cases the added value of the physician may be underestimated and in others it may be overestimated, depending on the study design. It is difficult to extrapolate the findings of these studies to precise estimates of improvement in health outcome. If a delay in transportation to the base hospital occurs as a result of the physician involvement, this may adversely affect outcome and such factors are unlikely to be considered in this type of study. Therefore, this type of study was omitted from this review, in favour of studies that included relevant health outcomes. A variant on the above study design was also identified. In this design, records were retrospectively reviewed and classified into groups indicating whether a physician was required (Nielsen et al. 2002; Snow et al. 1986). Similar considerations apply to those outlined above, so these studies were also excluded. Gries et al. (2006) examined the frequency of defined procedures and conditions that may require physicians to adequately manage emergency situations. In this study it was estimated that patients with life threatening conditions such as acute coronary syndrome, stroke, head trauma, and multiple trauma only occurred once every 0.4-14.5 months and CPR and intubation was carried out once every 0.5-1.5 months. The ranges represent time periods before encountering each specific outcome across both helicopter and ground transportation. Chest tubes were inserted every six months to six years. This was in the context of a service with 82,002 scene calls registered for ground crews during a 54 month period and 47,184 calls for air rescue services over 24 months.

Scientific methods used to compare the predicted mortality with actual mortality

A common method used in the studies selected for this review made use of a comparison between predicted and actual mortality. Most of these studies made use of the Trauma and Injury Severity Score (TRISS). Another, less frequently used comparison was with the ASCOT (A Severity Characterization of Trauma) model. Both these approaches are detailed below.

Study methods involving the TRISS approach evolved for two reasons: understanding the limitations of a retrospective evaluation of helicopter use in relation to patient outcome and the practical difficulties associated with conducting a randomised controlled trial. TRISS incorporates physiologic (trauma score), anatomic (injury severity score) and age (55 years as cut-off) independent variables into a logistic regression model. Predicted mortality can then be compared with actual mortality (Boyd et al. 1987).

The trauma score (TS) includes five components:

1. Systolic blood pressure.

2. Capillary refill.

3. Respiratory rate.

4. Respiratory expansion.

5. Glasgow Coma Scale (GCS).

Boyd et al. (1987) considered the trauma score had a sensitivity of 80% (meaning 20% of patients with severe injury will not be identified with this score) and specificity of 75% (meaning overestimation of severity will occur when physiologic changes are related to factors other than the consequences of hypovolaemia, cerebral oedema or hypoxia). The predictive value is greatly improved when combined with an injury severity score (ISS).

The ISS is based on the abbreviated injury scale (AIS). It was first proposed by Baker et al. (1974) as a method of using the AIS but adjusting for multiple injuries. The AIS is a list of several hundred injuries


5

each with a score that can range from 1 (minor injuries) to 6 (nearly always fatal). The ISS takes values from 1-75. If the patient has any AIS 6 injury, the ISS score is automatically 75. Otherwise, the highest AIS severity score in each of six body regions is identified, and the squares of the largest three are added to obtain the ISS.

Age greater than 55 years was shown to be associated with significantly increased mortality given comparable levels of physiologic derangement and anatomic injury severity in the Major Trauma Outcome Study (MTOS), (Boyd et al. 1987).

The probability of survival can be estimated from:

Ps = 1/(1 + e-b)

Where b= b0 + b1(TS) + b2(ISS) + b3(A).

b0….3 are regression coefficients that were initially derived from the patients included in the MTOS. As improvements in trauma care result in decreased mortality, these MTOS coefficients can be expected to change. Different sets of coefficients are used for blunt and penetrating trauma.

A revised trauma score (RTS) evolved out of a critical analysis of patients whose outcome was not predicted by the TRISS methods. Three parameters were used: GCS, systolic blood pressure and respiratory rate. When using the RTS the coefficients are different from those originally proposed. The RTS classification rather than the TS classification was used in the MTOS results published in 1990 (Champion et al. 1990b). The MTOS norms were obtained for adults (15+ years) with either blunt injuries (n=15,754) or penetrating injuries (n=7,423).

Various statistics are calculated using the TRISS methodology. The Z statistic compares outcome in two population subsets (Flora 1978). It quantitates the difference in the actual number of deaths in the test subset with the predicted number of deaths. Z values can be affected by the injury severity match between the study and baseline patient sets. The M statistic is a measure of that match. Values for M range from zero to one. The closer the value is to one, the better is the match of injury severity. Z

values associated with lower values of M (< 0.88) should be viewed with sceptism.

Younge et al. (1997) discussed the use of the W statistic in order to compare trauma survival rates between different institutions and reference databases hampered by different injury severity mixes. The W statistic estimates excess survivors per 100 patients that would be achieved if the study centre treated patients with the same distribution of injury severity as the reference database (e.g. MTOS). It represents the number of excess survivors per 100 patients attending a particular centre that would be achieved if that centre received patients with the same distribution of injury severity as the reference database. The standardising process places undue emphasis on patients with a good probability of survival. In lower strata of survival, which are based on smaller patient numbers in the reference database, the linear regression model over-predicts survival. Overall, the TRISS model tends to over-predict survival when using the MTOS database, meaning the performance of individual centres with a poor prognosis casemix will appear to be poor. This problem is reduced if the W statistic for each probability of survival interval at the study centre is compared with the W statistic for the equivalent interval in the reference database.

As a general comment, the analytic complexity of TRISS based studies is one aspect that lacks appeal to a wide readership. Stratification by severity marker is a method of overcoming this limitation but requires large patient numbers to achieve adequate study power. Other limitations include:

� unmeasured factors (e.g. pre-existing medical conditions, mixture of injury types, injury mechanism, time between injury and assessment of RTS) may account for differences between predicted and actual outcome

� distribution of probability of survival (Ps) may differ within comparative Ps intervals

� choosing different Ps intervals may alter the result.

The MTOS was a retrospective study of injury severity that initially aimed to develop national norms for trauma care that could be used for quality assurance. The study was co-ordinated through the


6

American College of Surgeons’ Committee on Trauma and data collection started in 1982. Over 140 hospitals from the USA, Canada, Australia and the UK submitted demographic, aetiologic, injury severity and outcome data. At the time of publishing results in 1990 more than 120,000 trauma patients had been treated at the participating hospitals, including 80,544 trauma patients from 139 USA and Canadian institutions for the period October 1982-1987. These 80,544 patients were the focus of a publication by Champion et al. (1990b). It should be noted that MTOS was not population based, participation was voluntary, participating centres may have tended to have increased trauma care expertise and may have been biased towards more severe trauma.

The ASCOT model was developed by Champion et al. (1990a). This approach combines emergency department admission values of GCS, systolic blood pressure, respiratory rate, patient age and AIS anatomic injury scores in a way that was designed to overcome ISS shortcomings. In his original description, Champion et al. suggested, based on Hosmer-Lemeshow statistics, that ASCOT reliably predicted patient outcome in penetrating injuries and “nearly so” for blunt injured patients. He commented that statistically reliable predictions were not achieved by TRISS in either patient group.

Other more recent studies have compared TRISS, ASCOT and other models (Frankema et al. 2005; Gabbe et al. 2005). Gabbe et al. (2005), estimated the sensitivity, specificity and the area under the curve (AUC) of receiver operating characteristic curves (which is a discriminating estimate of measure performance). The sensitivity, specificity and AUC of TRISS were 19%, 98% and 0.87 respectively. Likewise, the sensitivity, specificity and AUC of ASCOT were 75%, 58% and 0.78 respectively. In contrast in Frankema et al. (2005), the AUC for TRISS was 0.940 and for ASCOT was 0.956. Thus, on the basis of these two studies, it was not clear which of the two measures were better. What is clear is that neither is perfect, indicating potential problems with the estimation of excess unexpected deaths and excess unexpected survivors in the studies that use this approach in this review.

Other measures have been used to compare severity of injury at baseline in some of the included studies. One measure was the APACHE (Acute physiology and chronic health evaluation) II score. Some studies have been conducted evaluating the performance of APACHE II. One study estimated an AUC in receiver operating characteristic curves with mortality as the outcome of interest of 0.84 (95% CI 0.83-0.85), (Suistomaa et al. 2002). Another study estimated an AUC of 0.787 and also noted there were significant differences between observed and predicted mortality (p<0.001), (Moreno and Morais 1997). Vassar et al. (1999) compared TRISS, APACHE II and APACHE III. TRISS and APACHE were described as having poor performance characteristics. In relation to predicting mortality both had poor goodness of fit characteristics (Hosmer-Lemeshow testing). The AUC for APACHE II was 0.87 and for TRISS was 0.82. APACHE had better goodness of fit characteristics and the AUC was 0.89. Similar performance characteristics for APACHE II were found by Muckart et al. (1997). In this study, goodness of fit was poor compared with a new model the authors developed. The AUC for APACHE II was 0.78.

Stratified/multivariate studies

As previously mentioned, there are drawbacks to studies comparing predicted with actual mortality. One method that has been used to overcome some of these limitations is stratification by severity marker. However, this design requires large sample sizes. Multivariate analysis can help with this.

Scope of evaluation

The scope of the systematic review included two broad areas:

1. The usefulness of including medical doctors on emergency transportation.

2. Ideal times between receipt of call out and delivery to base hospital for emergency patients.

The second aspect of the scope was designed to assist with the process of determining the most appropriate form of transport (air or road ambulance) in different locations of New Zealand. The literature was restricted to 1980 onwards.


7

Review Questions

Question 1

In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on emergency helicopter services improve health outcome when compared with transportation by emergency helicopter without a medical doctor? The health outcomes of interest are death in-transit and in-hospital and time to hospital discharge to the community.

Explanatory notes:

1. The cut off between childhood and adulthood was 16 years.

2. Neonates were excluded.

3. An inpatient stay in a local secondary hospital after discharge from the “definitive care” hospital was included in ‘length of stay’ where possible.

4. Subanalysis for Injury Severity Score, ISS >15.

Question 2

In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on a road ambulance service improve health outcome when compared with transportation by a road ambulance service without a medical doctor? The health outcomes of interest are death in-transit and in-hospital and time to hospital discharge to the community.

Explanatory notes:





Question 3

In adults and children with a medical or trauma related emergency, does the presence of a medical crew able to perform rapid sequence intubation and/or thoracostomy improve health outcome when compared with a medical crew unable to perform rapid sequence intubation and/or tube thoracostomy and/or thoracotomy? The health outcomes of interest are death in-transit and in-hospital and time to hospital discharge to the community.

Explanatory notes:





Question 4

In adults and children with a medical or trauma related emergency how does variation in the time from callout to arrival at a medical facility with definitive care influence health outcome? The health outcomes of interest are death in-transit and in-hospital and time to hospital discharge to the community.


8

Explanatory notes:





SELECTION CRITERIA

Question 1

The selection criteria for question 1 are set out in Table 1.

Table 1. Inclusion/exclusion criteria for the effectiveness of including medical doctors on

emergency helicopter transportation

Characteristic Criteria

Inclusion criteria

Aim A primary aim of the study was to evaluate the

effectiveness of including medical doctors on emergency

helicopter transportation

Methods The methods were clearly described

Publication type Randomised controlled trials, cohort studies, case control

studies, interrupted time series and systematic reviews of

the above publication types

Population

Adults or children (excluding neonates) who present to

the emergency department after helicopter

transportation

Medical emergencies or trauma related health

emergencies will be included

Sample size At least 50 human patients

Intervention/test Inclusion of a medical doctor on the helicopter

transporting the emergency patient

Comparator

The absence of a medical doctor for the helicopter


Outcome Death: in transit, death in hospital, days to hospital

discharge

Exclusion criteria

Publication type Non-systematic reviews, case series, letters, editorials,

expert opinion articles, conference proceedings,

comments and articles published in abstract form.

Population Restriction to publication of incidents during

transportation

Comparator Expected outcome based on expert panel

Publication superseded Publication superseded by a later publication with longer

follow-up data and overlap in the patient population

Language Non-English language articles will be excluded


9

Question 2


Table 2. Inclusion/exclusion criteria for the effectiveness of including medical doctors on road

ambulances


Inclusion criteria


effectiveness of including medical doctors on road

ambulances


Publication type Randomised controlled trials, pseudorandomised

controlled trials, cohort studies, case control studies,

interrupted time series and systematic reviews of the

above publication types

Population


the emergency department after road ambulance

transportation




Intervention/test Inclusion of a medical doctor on the road ambulance


Comparator

The absence of a medical doctor on the road

ambulance transporting the emergency patient


discharge to community

Exclusion criteria





transportation






10

Question 3


Table 3. Inclusion/exclusion criteria for the effectiveness of including a crew able to perform

rapid sequence intubation and/or thoracostomy on pre-hospital emergency

transportation


Inclusion criteria


effectiveness of different crew configurations on health

outcome


Publication type Randomised controlled trials, pseudorandomised

controlled trials, cohort studies, case control studies,

interrupted time series and systematic reviews of the

above publication types

Population


the emergency department after emergency

transportation




Intervention/test Inclusion of a crew able to perform rapid sequence

intubation using muscle relaxants and/or tube

thoracostomy and/or thoracotomy

Comparator

Inclusion of a crew unable to perform rapid sequence

intubation and/or thoracostomy


discharge to community

Exclusion criteria





transportation






11

Question 4


Table 4 Inclusion/exclusion criteria for the ideal time between call out and delivery of

emergency patients to the emergency department


Inclusion criteria

Aim A primary aim of the study was to evaluate the effect of

different transportation times (from call out to arrival at an

emergency department) on health outcome


Publication type Study that includes a control group

Population


the emergency department after helicopter

transportation




Comparison Comparison of at least two different categories of time

from call out to time of delivery of patients to the

emergency department

Comparison of different methods of transport that have

different mean times of transport


discharge

Exclusion criteria

Publication type Non-systematic reviews, letters, editorials, expert opinion

articles, conference proceedings, comments and articles

published in abstract form.




MAIN SEARCH TERMS

Details of the search strategies are presented in Appendix 1.

Medline Subject Headings (MeSH headings): air ambulances, aircraft, “personnel staffing and scheduling”, personnel selection, patient care team, physician’s role, exp physicians, allied health personnel, nurse’s role, nurses, manpower[as floated subheading], time facts, survival analysis, treatment outcome, patient discharge, length of stay, morbidity, mortality, “outcome assessment (health care)”, emergencies, exp emergency medical services, emergency service-hospital, transportation of patients, ambulances

Additional keywords: helicopter$, medivac, medivac, casivac, casevac, evac, aeromedic$, air ambulanc$ , flight ambulanc$, medical practitioner$, medic$ adj qualif$, paramedic$, medic, medics, flight nurse$, doctor$, staff$, physician$, personnel, time adj3 delay$, ((pre-hospital or, pre-hospital) adj (time or care or treatment)), ((call-out or callout) and (arrival or admit$ or hospital or medical


12

facility or definitive care or emergency department or ED)), scene time, “out of hospital time”, transport adj time$, transfer$ adj time$, survival, outcome, golden hour, golden minute$

SEARCH SOURCES

Bibliographic databases

Medline Embase Cinahl Current Contents Science/Social Science Citation Index PubMed (last 90 days)

Review databases

Cochrane Database of Systematic Reviews Database of Abstracts of Reviews of Effectiveness (DARE) NHS Economic Evaluation Database Health Technology Assessment Database

Articles published in English language only were considered.

The search was restricted to literature published from 1980 onwards. Searching was undertaken between 24 October and 7 November, 2006.

APPRAISAL METHODOLOGY

Summaries of appraisal results are shown in tabular form (known as Evidence Tables) which detail study design, study setting, sample, methods, results, and reported conclusions.

The evidence presented in the selected studies were assessed and classified according to the NHMRC’s revised hierarchy of evidence (Appendix 2).


13

RESULTS

Doctor versus no doctor on board helicopter

From the search strategy for question one (comparing helicopter transportation with and without medical doctor staffing) we identified, 1068 potentially relevant articles/abstracts of which 58 were retrieved. Of these retrieved articles, 51 were excluded. These excluded papers are presented in Appendix 3. Two additional articles were identified from reference lists. Both these studies were excluded: one was not relevant to the review question and the other used an incorrect comparator. Reasons for exclusion of studies before retrieval in full text are outlined in Table 5. Reasons for exclusion of studies retrieved in full text are detailed in Table 6.

Table 5 Reasons for exclusion of studies before retrieval in full text: doctor versus no doctor

on board helicopter

Reason for exclusion Number

Not relevant to review question aim 894

Methods were not clearly described 0

Wrong publication type 98

Incorrect population 1

Sample size less than 50 4

Incorrect comparator 7

Incorrect outcomes 3

Publication superseded 0

Non-English language 0

Neonatal study 3

Total 1010

Table 6 Reasons for exclusion of studies retrieved in full text: doctor versus no doctor on

board helicopter







Incorrect intervention group 3





Neonatal study 0

Patient transfer 1

Article unable to be obtained 1

Total 51

Seven retrieved articles were appraised and are listed in the references and Appendix 7. Included papers are presented in the evidence table below. Included studies ranged from level III-1 to III-3 and above according to NHMRC’s hierarchy of evidence, including one pseudorandomised controlled trial, four cohort studies and two before and after studies. Evidence tables for the included studies are found at the end of this section (Table 7). Two of the seven studies were of marginal relevance given the high proportion of patients transferred from a hospital rather than the scene (Burney et al. 1992; Burney et al. 1995). These are presented separately from the remaining five studies.


14

Baxt et al. 1987

Baxt et al. (1987) conducted a pseudorandomised controlled trial (Level III-1 evidence) comparing mortality in a group of consecutive patients with blunt trauma transported by helicopter with physician and nurse on board with another group with paramedic and nurse on board. There were 574 participants (316 in the physician group and 258 in the comparator group). Actual mortality was compared with predicted mortality using TRISS methodology. The trauma score scale rather than the revised trauma score scale was included in this calculation. Differences in procedures that the different groups could perform were noted and are detailed in Table 7. Actual mortality was not statistically significantly different from predicted mortality in the paramedic group while, in the physician group, actual mortality was significantly lower than predicted (P<0.05). There was a statistically significant difference in the Z statistic between the two groups, supporting reduced mortality in the physician group compared with the paramedic group.

There were potential sources of confounding and bias that should be considered when interpreting this study. Key issues included:

� the study was not truly randomised so is susceptible to confounding. However, significant baseline differences were not observed for transport time, trauma score, ISS, GCS, predicted survival or patient age.

� the key difference between the groups related to the category of patients who survived but were expected to die. There were only 22 patients who were expected to die in the physician group (five survived) and 16 who were expected to die in the paramedic group (none survived).

� the TRISS methodology used in this review did not assess the degree of match in injury severity between the two groups, or adjust for different casemix in the study groups.

� the study did not use all-cause mortality as the outcome (the focus was on mortality due to trauma), resulting in the potential for misclassification of outcome.

Despite the limitations this was the strongest study considering the effectiveness of including doctors on board helicopters.

Garner (2004) noted that the non-doctor team in this study was “considerably more procedurally capable than most Australian paramedics” and observed that despite this, a better outcome was observed in the physician treatment group.

Hamman et al. 1991

Hamman et al. (1991) conducted a before and after study (Level III-3 evidence) comparing mortality in a group of consecutive patients transported by helicopter crewed by physician and nurse (before phase) with another group crewed by paramedic and nurse or two nurses (after phase). There were 145 in the physician group and 114 in the non-physician group. Actual mortality was compared with predicted mortality using TRISS methodology. Differences in procedures that the different groups could perform were noted and are detailed in Table 7. Comparison of actual with predicted mortality showed a lower than predicted mortality in both groups. However, there was no overall difference in patient outcomes between the two groups.

Despite the above results, this study should be viewed cautiously, given a number of significant limitations:

� before and after design is a weak method as factors other than the factor of interest may contribute to the estimated measure of effect

� the study lacked power to detect a difference between the two groups as there were only 32 patients in total who were expected to die based on estimates of the probability of survival

� comparison of injury severity between the two groups is consistent with a poor match in severity between the groups (M statistic 0.87)

� lack of documentation about the timing of deaths and timing of estimation of RTS added to difficulties interpreting the results.


15

Schmidt et al. 1992

Schmidt et al. (1992) reported on a registry based study (level III-2) that compared mortality in two centres with differing crew configurations on their emergency helicopter services. In the German centre a trauma surgeon was included on all flights, whereas in the USA centre the trauma surgeon was replaced with either a flight nurse or a paramedic. There were 221 participants in the setting with a surgeon and 186 in the setting without a surgeon. Actual mortality was compared with predicted mortality using TRISS methodology. Differences in procedures that the different groups could perform were noted and are detailed in Table 7. The Z statistic was calculated for the comparison between actual and predicted outcome at both centres. This statistic was consistent with improved survival compared with that predicted in the surgeon centre. This improvement was consistent with an additional 1.35 survivors per 100 patients when compared with the MTOS reference population. No such improvement was noted in the non-surgeon centre. However, there were reservations about the degree of match in injury severity between the actual data and the reference population in the non-surgeon setting. There were a lower proportion of deaths in the first six hours in the surgeon present setting.

There were limitations to this study:

� registry based study that resulted in the omission of 37 patients in the surgeon setting due to missing charts or data. This is likely to have resulted in selection bias, in that it seems likely the excluded patients were not representative of the general population, given no deaths occurred amongst these 37 patients.

� the measurement of RTS, a key component of the projected survival populations, occurred at different times in the two centres. This results in variation in the accuracy of the comparison with the reference population, and therefore in the Z statistic.

� there was variation in key time intervals relating to transport. In particular, there was a significantly shorter time between the incident and arrival at the scene in the surgeon setting. This may have resulted in improved prognosis in this setting.

� management was more aggressive in the setting with a surgeon and this increased aggressiveness may have produced the more favourable results, rather than the presence of a surgeon per se. The procedures adopted that reflected this increased aggressiveness could be performed by flight nurses and paramedics as well as surgeons.

More detail about this study is presented in Table 7.

Garner et al. 1999

Garner et al. (1999) reported on a retrospective study set in Australia (level III-2). They compared the outcome (mortality) between groups transported via helicopter with a physician on board versus patients transported with a paramedic on board. All patients were transported directly from the scene to the relevant hospital (hospitals varied by the crew mix on the helicopter). There were 67 patients in the physician group and 140 in the paramedic group. The patients were restricted to those with blunt trauma and an ISS score greater than 10. Actual mortality was compared with predicted mortality using TRISS methodology (MTOS as the reference population) and the adjusted W statistic was also used to directly compare the physician and paramedic groups. The degree of match on injury severity between the MTOS population and the physician and paramedic groups was poor, therefore the adjusted W statistic was appropriately presented. When comparing the physician group with the MTOS population it was estimated that 9.48 (95% CI 3.84-15.12) extra lives per 100 population were saved in the physician group. There was no significant difference in mortality between the MTOS population and the paramedic group. Direct comparison between the paramedic and physician groups suggested 13.44 (95% CI 7.80-19.08) extra lives per 100 population were saved in the physician group. There were differences in the procedures performed between the two groups. These are detailed in Table 7. In general, the physician group treated patients more aggressively and also conducted a number of rapid sequence intubations and tube thoracostomies (the paramedic group did not).


16

The study had limitations:

� it was a retrospective study

� there were sources of selection bias – with differences in baseline measures between the two study groups (the physician group appeared to manage a more severely injured group) and seven patients who died were excluded from the paramedic group due to missing case sheets

� confounding was a potential problem between the two groups although use of the W statistic should have partially controlled confounding (based on injury severity)

� the degree of match in injury severity with the MTOS study was poor and there were variations in methods adopted in the MTOS study and this study (most notably related to the timing of RTS measurement) which limits the usefulness of the MTOS cohort as a reference population.

Cameron et al. 2005

A retrospective chart review was conducted in Australia (Cameron et al. 2005). This study used a before and after design (Level III-3 evidence) to compare outcome in a period where helicopters included emergency physicians with a subsequent period where the helicopter did not include an emergency physician (intensive care paramedics were used). Given the nature of the data recorded measures of injury severity were restricted to the RTS. Chart abstraction was primarily performed by one person but a 10% sample was validated by another abstractor. There was an excellent level of agreement between the two abstractors. There were 163 patients in the physician group and 211 in the paramedic group. Mortality was measured at 30 days. There were 10 deaths in total and no significant difference was detected between the study groups. There was also no significant difference in the length of stay, although the mean length of stay was only two days and one day respectively in the physician and paramedic groups.

There were significant limitations to this study:

� the effects of a retrospective design were apparent. As the authors documented, they were unable to extract data that would have provided a better indication of injury/illness severity.

� there was no control over potential confounders. Given baseline differences suggestive of a more severe casemix in the physician group this may have led to bias in the comparison.

� the outcome of the group discharged from ED was not obtained.

� the study power was low, particularly for mortality, so the lack of a significant difference in outcome was not surprising.

Studies that included predominantly transfer patients

Burney et al. 1992

Burney et al. (1992) reported on a retrospective cohort study (Level III-2 evidence) that compared outcome following the helicopter transportation of patients by physician/nurse with nurse/nurse. Both groups were studied concurrently during 1987-1988. There were 659 participants (418 in the P/N team and 241 in the N/N team). Most of the transports originated in an emergency department (ED), followed by inpatient units with only 5.5% being from the scene. There was no overall difference in mortality or number of hospital days. No statistically significant differences in these outcomes were observed on subgroup analysis (by clinical category or point of origin of the transport). However, the latter analyses had low study power.

This study had a number of limitations:

� measures of injury severity were not typical of those used in the studies conducted to investigate this review question

� lack of documentation about the procedures available to the two groups, dispatch criteria for the two groups, and number of receiving hospitals involved


17

� there was variation in baseline characteristics between the two groups in the following variables: age, clinical category and origin of patient

� there was no adequate method of controlling for confounding used (the stratified analyses resulted in loss of study power)

� low proportion of participants were transported from the scene, thus having little relevance to the New Zealand setting.

Details of this study are provided in Table 7.

Burney et al. 1995

Burney et al. (1995) followed their retrospective cohort study with a prospective cohort study (Level III-2 evidence). The same study comparisons were made and the same data collection tools were used (although severity measures were measured more frequently). There were 255 patients in the P/N group and 914 in the N/N group. As previously, the proportion of transports originating from the scene was low (7.6%). There were differences in criteria for the selection of who was to attend the patient particularly in the second year of the study when the physician group attended patients with more complicated injuries. Overall, there was no significant difference in mortality between the two groups (25% in physician group, 21% in nurse group, P=0.12). There was also no difference in the group transported from the scene (mortality of 16% in both groups). Duration of hospital days was assessed separately across each study year. The nurse group was associated with a significantly shorter stay in the second year but this wasn’t surprising given the more complex cases attended by the physician group in that study period.

This study had similar limitations to that conducted by Burney et al. in 1992, except it had the advantage of a prospective approach. If the study was to have 80% power to detect a 5% difference in mortality at the 5% significance level a sample of 2,000 patients would have been required (there were 1,169 patients). The increased complexity of the patients transported by physicians in the second year makes it very difficult to interpret the results of this study.


18

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ba

xt

an

d M

oo

dy 1

987

)

USA

Pse

ud

o-

ran

do

mis

ed

co

ntr

olle

d t

rial

Leve

l III-

1

Stu

dy s

ett

ing

.

Co

nse

cu

tive

pa

tie

nts

with

blu

nt

tra

um

a

ove

r a

24 m

on

th p

erio

d.

All

pa

tie

nts

we

re t

ran

spo

rte

d t

o t

he

Un

ive

rsity o

f

Ca

lifo

rnia

, Sa

n D

ieg

o,

Me

dic

al T

rau

ma

Ce

nte

r w

he

re t

he

y w

ere

ca

red

fo

r u

ntil

dis

ch

arg

e.

Pa

rtic

ipa

nts

:

Tota

l sa

mp

le 5

74

Inte

rve

ntio

n.

He

lico

pte

r st

aff

ed

by f

ligh

t n

urs

e a

nd

ph

ysi

cia

n (

n=

316

)

Pro

ce

du

res:

ora

l/n

asa

l* e

nd

otr

ac

he

al

intu

ba

tio

n

Ne

ed

le/t

ub

e t

ho

rac

ost

om

y p

lac

em

en

t*

Pe

rip

he

ral/

ce

ntr

al I

V li

ne

pla

ce

me

nt

Pe

ric

ard

ioc

en

tesi

s*

Cric

oth

yro

ido

tom

y p

lac

em

en

t*

Exp

an

de

d m

ed

ica

tio

ns*

Co

mp

ara

tor.

He

lico

pte

r st

aff

ed

by f

ligh

t n

urs

e a

nd

pa

ram

ed

ic (

n=

25

8)

Pro

ce

du

res:

ora

l en

do

tra

ch

ea

l

intu

ba

tio

n

IV li

ne

pla

ce

me

nt

Lim

ite

d m

ed

ica

tio

ns

Pn

eu

ma

tic

an

tish

oc

k g

arm

en

t

pla

ce

me

nt*

Inc

lu/e

xcl c

rite

ria

.

Pa

tie

nts

with

blu

nt

tra

um

a

Exc

lud

ed

pa

tie

nts

wh

o d

id n

ot

ha

ve

an

y r

esu

scita

tive

pro

ce

du

res

in t

he

fie

ld

Da

ta c

olle

ctio

n

Tra

um

a s

co

re c

alc

ula

ted

on

pa

tie

nt

co

nta

ct

by t

he

cre

w.

Inju

ry s

eve

rity

sc

ore

ca

lcu

late

d f

rom

pa

tie

nt

rec

ord

s a

nd

au

top

sy

rep

ort

s.

Ou

tco

me

me

asu

res

Mo

rta

lity: d

efin

ed

as

de

ath

du

e t

o

the

in

itia

l in

jurie

s o

r c

om

plic

atio

ns

of

the

in

jurie

s

Follo

w-u

p in

terv

al

Min

imu

m o

f si

x m

on

ths

An

aly

sis

Pre

dic

ted

mo

rta

lity e

stim

ate

d u

sin

g

the

TR

ISS m

eth

od

olo

gy, u

tilis

ing

th

e

mo

st r

ec

en

t c

oe

ffic

ien

ts a

va

ilab

le

at

the

tim

e.

An

aly

tic

me

tho

ds

inc

lud

ed

usi

ng

χ2,

two

ta

iled

Stu

de

nt

t te

st,

Ma

nte

l-

He

an

sze

l te

st a

nd

th

e Z

sta

tist

ic o

f

co

mp

ariso

n b

etw

ee

n p

red

icte

d

an

d a

ctu

al s

urv

iva

l.

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Pa

ram

ed

ic g

rou

p:

Pre

dic

ted

= 1

9.5

Ac

tua

l=1

9

Z s

tatist

ic 0

.208 (

P>

0.0

5)

Ph

ysi

cia

n g

rou

p:

Pre

dic

ted

16.9

Ac

tua

l 11

Z s

tatist

ic 2

.284 (

P<

0.0

5)

Diffe

ren

ce

in Z

sta

tist

ic b

etw

ee

n t

he

two

gro

up

s: 2

.076 (

P<

0.0

5)

Ind

ica

tin

g a

sta

tist

ica

lly s

ign

ific

an

t

imp

rove

d o

utc

om

e in

th

e p

hysi

cia

n

gro

up

.

Dis

trib

utio

n o

f p

atie

nts

by

pro

ba

bili

ty o

f su

rviv

al

Exp

ec

ted

to

die

bu

t liv

ed

:

Pa

ram

ed

ic: 0

Ph

ysi

cia

n: 5

Exp

ec

ted

to

live

bu

t d

ied

:

Pa

ram

ed

ic: 5

Ph

ysi

cia

n: 3

Exp

ec

ted

to

die

an

d d

ied

:

Pa

ram

ed

ic: 1

4

Ph

ysi

cia

n: 8

Lim

ita

tio

ns

�

No

t tr

uly

ra

nd

om

ise

d: d

isp

atc

h

de

pe

nd

ed

on

ro

tatio

n o

f c

alls

or

wh

ich

he

lico

pte

r w

as

clo

ser

to t

he

sce

ne

at

tim

e o

f d

isp

atc

h.

�

Ce

ntr

al c

om

mu

nic

atio

n d

id n

ot

ap

pe

ar

to b

e a

va

ilab

le f

or

eith

er

sta

ff

gro

up

.

�

The

leve

l of

exp

erie

nc

e o

f b

oth

sta

ff

gro

up

s w

as

un

cle

ar.

�

TRIS

S m

eth

od

olo

gy in

clu

de

d t

he

tra

um

a s

co

re (

TS)

rath

er

tha

n t

he

revis

ed

tra

um

a s

co

re (

RTS

). In

late

r

ye

ars

th

e R

TS s

co

re w

as

co

nsi

de

red

to

be

mo

re a

cc

ura

te t

ha

n t

he

TS s

co

re.

Su

bse

qu

en

t d

eve

lop

me

nts

in T

RIS

S

me

tho

do

log

y m

ad

e u

se o

f th

e M

sta

tist

ic w

hic

h a

llow

s a

n a

sse

ssm

en

t o

f

the

in

jury

se

ve

rity

mix

be

twe

en

stu

dy

gro

up

s a

nd

, if a

pp

rop

ria

te,

the

W

sta

tist

ic, w

hic

h d

ea

ls w

ith

diffe

ren

t

ca

se m

ixe

s a

cro

ss s

am

ple

s.

�

All-

ca

use

mo

rta

lity n

ot

use

d. M

ay

ha

ve

lea

d t

o o

utc

om

e

mis

cla

ssific

atio

n.

�

No

do

cu

me

nta

tio

n o

f b

lind

ing

in t

he

ass

ess

me

nt

of

ca

use

of

de

ath

.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps

(alth

ou

gh

all

rec

ord

ed

de

ath

s o

cc

urr

ed

with

in 4

8 h

ou

rs o

f

ad

mis

sio

n).


19

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ba

xt

an

d M

oo

dy 1

987

)

USA

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

.

Me

dia

n p

rob

ab

ility

of

surv

iva

l:

Pa

ram

ed

ic c

rew

: 0.9

5-1

.0

Ph

ysi

cia

n c

rew

: 0.9

5-1

.0

Me

dia

n G

lasg

ow

Co

ma

Sc

ore

of

seve

re

bra

in in

jure

d p

atie

nts

(G

CS≤8

)

Pa

ram

ed

ic c

rew

(n

=44,

17%

of

tota

l): 4

Ph

ysi

cia

n c

rew

(n

=5

4, 1

7%

of

tota

l):

5

Me

an

Tra

um

a S

co

re:

Pa

ram

ed

ic g

rou

p:

14.2

Ph

ysi

cia

n g

rou

p:

14.3

Me

an

In

jury

se

ve

rity

sc

ore

:

Pa

ram

ed

ic g

rou

p:

13.6

Ph

ysi

cia

n g

rou

p:1

3.9

Me

an

pre

dic

ted

su

rviv

al

Pa

ram

ed

ic g

rou

p:

0.9

24

Ph

ysi

cia

n g

rou

p:

0.9

47

Me

an

ag

e

Pa

ram

ed

ic g

rou

p:

27.6

ye

ars

Ph

ysi

cia

n g

rou

p:

27.8

ye

ars

Tra

nsp

ort

tim

es:

Me

an

re

spo

nse

tim

e:

Pa

ram

ed

ic g

rou

p 1

5.5

min

ute

s

Ph

ysi

cia

n g

rou

p:

14.9

min

ute

s

Me

an

sc

en

e t

ime

:

Pa

ram

ed

ic g

rou

p:

18.6

min

ute

s

Ph

ysi

cia

n g

rou

p:

19.1

min

ute

s

Me

an

de

live

ry t

ime

:

Pa

ram

ed

ic g

rou

p:

16.9

min

ute

s

Ph

ysi

cia

n g

rou

p:

16.1

min

ute

s.

B

lind

ed

ch

art

an

aly

sis

ass

ess

ing

ad

he

ren

ce

to

writt

en

me

dic

al

tre

atm

en

t p

roto

co

ls.

1.

Pa

tie

nts

wh

o s

urv

ive

d

Pa

ram

ed

ic g

rou

p18 o

f 23

9 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

Ph

ysi

cia

n g

rou

p:

2 o

f 305

pa

tie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

.

2.

Pa

tie

nts

wh

o d

ied

Pa

ram

ed

ic g

rou

p:

9 o

f 19 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

Ph

ysi

cia

n g

rou

p:

0 o

f 11 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

.

No

te in

co

nsi

ste

nc

ies

inc

lud

ed

failu

re t

o c

on

du

ct

the

ind

ica

ted

pro

ce

du

re o

r m

ed

ica

l pro

ce

du

re

no

t fo

llow

ed

.

�

Po

ten

tia

l fo

r c

on

fou

nd

ing

, a

lth

ou

gh

ba

selin

e a

na

lyse

s in

dic

ate

litt

le

diffe

ren

ce

in t

ran

spo

rt t

ime

, tr

au

ma

sco

re, in

jury

se

ve

rity

sc

ore

, G

CS,

pre

dic

ted

su

rviv

al a

nd

pa

tie

nt

ag

e

be

twe

en

gro

up

s.

�

Ke

y r

esu

lts

are

ba

sed

on

a s

ma

ll

nu

mb

er

of

pa

tie

nts

wh

o s

urv

ive

d b

ut

we

re e

xpe

cte

d t

o d

ie (

5 o

f 22 in

th

e

ph

ysi

cia

n g

rou

p a

nd

0 o

f 1

6 in

th

e

pa

ram

ed

ic g

rou

p).

�

Un

cle

ar

if R

SI w

as

ava

ilab

le t

o t

he

co

mp

ara

tor

gro

up

Co

mm

en

ts

�

All

pa

tie

nts

att

en

de

d a

sin

gle

tra

um

a

ce

ntr

e.

�

The

TR

ISS m

eth

od

olo

gy w

as

ap

plie

d t

o

the

tw

o s

tud

y g

rou

ps

usi

ng

da

ta t

ha

t

we

re c

olle

cte

d a

t th

e s

am

e t

ime

be

twe

en

th

e t

wo

gro

up

s.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

�

Thre

e p

atie

nts

we

re e

xclu

de

d in

ea

ch

gro

up

du

e t

o t

he

lac

k o

f re

susc

ita

tiv

e

me

asu

res

in t

he

fie

ld.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

A s

tatist

ica

lly s

ign

ific

an

t re

du

ctio

n in

th

e

mo

rta

lity o

f p

atie

nts

with

blu

nt

tra

um

a

tre

ate

d b

y a

me

dic

al h

elic

op

ter

em

erg

en

cy

ca

re s

erv

ice

sta

ffe

d b

y a

nu

rse

/ph

ysi

cia

n

co

mb

ina

tio

n c

ou

ld b

e d

em

on

stra

ted

co

mp

are

d w

ith

th

at

sta

ffe

d b

y a

nu

rse

/pa

ram

ed

ic c

om

bin

atio

n.

* D

iffer

ence

in p

roce

dure

s av

aila

ble

to th

e tw

o st

affin

g gr

oups


20

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ha

mm

an

et

al.

1991

)

USA

Be

fore

an

d

aft

er

de

sig

n

Leve

l III-

3.

Stu

dy s

ett

ing

.

All

pa

tie

nts

tra

nsp

ort

ed

fro

m a

cc

ide

nt

sce

ne

to

Hu

ma

na

Ho

spita

l U

niv

ers

ity

in

a 1

0 m

on

th p

erio

d in

198

5 (

with

ph

ysi

cia

n)

an

d c

om

pa

red

with

a g

rou

p

tra

nsp

ort

ed

du

rin

g a

10 m

on

th p

erio

d in

1987 (

with

ou

t p

hysi

cia

n).

Pa

rtic

ipa

nts

:

Tota

l sa

mp

le 2

59

Inte

rve

ntio

n (

n=

11

4).

Sto

pp

ed

th

e p

rese

nc

e o

f a

me

dic

al

do

cto

r o

n e

me

rge

nc

y h

elic

op

ter

(tw

o

nu

rse

s o

r n

urs

e +

pa

ram

ed

ic p

rese

nt)

Co

mp

ara

tor

(n=

14

5).

Me

dic

al d

oc

tor

(Fa

cu

lty m

em

be

r o

r a

sec

on

d o

r th

ird

ye

ar

resi

de

nt

ph

ysi

cia

n)

pre

sen

t o

n h

elic

op

ter

(be

fore

ph

ase

of

the

stu

dy).

Nu

rse

als

o o

n b

oa

rd.

Ph

ysi

cia

ns

ab

le t

o p

erf

orm

th

e f

ollo

win

g

ad

ditio

na

l pro

ce

du

res:

cric

oth

yro

ido

tom

ies

an

d t

ub

e

tho

rac

ost

om

ies.

Bo

th in

terv

en

tio

n a

nd

co

mp

ara

tor

gro

up

s in

tub

ate

d, o

bta

ine

d IV

ac

ce

ss,

pe

rfo

rme

d a

dva

nc

ed

CP

R, p

lac

ed

pn

eu

ma

tic

an

tish

oc

k g

arm

en

t,

pe

rfo

rme

d n

ee

dle

th

ora

co

sto

my a

nd

pe

rica

rdio

ce

nte

sis,

im

mo

bili

sed

th

e

ne

ck, sp

linte

d f

rac

ture

s a

nd

dre

sse

d

op

en

wo

un

ds.

Inc

lu/e

xcl c

rite

ria

.

Co

nse

cu

tive

ad

ults

du

rin

g t

he

tw

o

stu

dy p

erio

ds.

Exc

lud

ed

pa

tie

nts

with

an

initia

l an

d

sub

seq

ue

ntly u

nc

ha

ng

ed

RTS

of

0

an

d b

urn

vic

tim

s.

Da

ta c

olle

ctio

n

Dis

tan

ce

tra

nsp

ort

ed

Tim

e a

t sc

en

e

Pro

ce

du

res

pe

rfo

rme

d

RTS

ISS

Ou

tco

me

me

asu

res

Mo

rta

lity

Follo

w-u

p in

terv

al

No

t st

ate

d

An

aly

sis

Me

an

co

mp

ariso

ns

use

d e

ith

er

χ2,

Be

hre

ns-

Fis

he

r t’

, o

r th

e W

elc

h d

f’

test

to

de

term

ine

sim

ilarity

.

Re

gre

ssio

n c

on

sta

nts

we

re d

erive

d

fro

m M

TOS.

The

Z a

nd

M s

tatist

ics

we

re c

alc

ula

ted

.

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Ph

ysi

cia

n p

rese

nt

gro

up

Pre

dic

ted

: 17

Ac

tua

l: 1

2

% r

ed

uc

tio

n:

30

Z s

tatist

ic -

2.0

3

Ph

ysi

cia

n a

bse

nt

gro

up

Pre

dic

ted

: 15

Ac

tua

l: 8

% r

ed

uc

tio

n:

47

Z s

tatist

ic -

3.1

1

Fre

qu

en

cy o

f p

ote

ntia

lly li

fe

thre

ate

nin

g in

jurie

s n

ot

ad

dre

sse

d

be

fore

arr

iva

l at

ED

Ph

ysi

cia

n p

rese

nt

gro

up

: 12%

Ph

ysi

cia

n a

bse

nt

gro

up

: 6

%

P>

0.0

5

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

Be

fore

an

d a

fte

r d

esi

gn

ha

ve

sig

nific

an

t lim

ita

tio

ns

du

e t

o t

he

po

ssib

le c

ha

ng

e o

f fa

cto

rs o

the

r th

an

the

in

terv

en

tio

n.

�

Len

gth

of

follo

w-u

p o

f p

art

icip

an

ts

wa

s n

ot

cle

ar.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps.

�

Hig

h p

ote

ntia

l fo

r c

on

fou

nd

ing

.

�

Sig

nific

an

t d

iffe

ren

ce

in a

ge

be

twe

en

the

tw

o g

rou

ps

with

th

e p

hysi

cia

n

ab

sen

t g

rou

p b

ein

g o

f yo

un

ge

r

ove

rall

ag

e a

nd

th

ere

fore

imp

rove

d

ou

tco

me

is f

avo

ure

d in

th

is g

rou

p.

�

Ke

y r

esu

lts

are

ba

sed

on

a s

ma

ll

nu

mb

er

of

pa

tie

nts

wh

o w

ere

exp

ec

ted

to

die

(32 in

to

tal)

th

us

un

de

rpo

we

red

to

de

tec

t a

diffe

ren

ce

in o

utc

om

e b

etw

ee

n t

he

tw

o s

tud

y

gro

up

s.

�

M s

tatist

ic w

as

0.8

7 w

hic

h is

less

th

an

the

ad

vis

ed

cu

t o

ff o

f 0.8

8 a

nd

th

us

ind

ica

tes

a p

oo

r m

atc

h in

inju

ry

seve

rity

be

twe

en

th

e t

wo

gro

up

s. W

sta

tist

ic w

as

no

t c

alc

ula

ted

to

ad

just

for

the

va

ria

tio

n in

ca

sem

ix.

�

Tim

ing

of

de

ath

s w

as

no

t

do

cu

me

nte

d.

�

Un

cle

ar

if R

TS w

as

est

ima

ted

at

the

sam

e t

ime

po

st in

jury

in b

oth

gro

up

s.

RTS

is s

usc

ep

tib

le t

o c

ha

ng

e o

ve

r

tim

e.


21

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ha

mm

an

et

al.

1991

)

USA

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

.

Ave

rag

e a

ge

(ye

ars

):

Ph

ysi

cia

n p

rese

nt:

34

Ph

ysi

cia

n a

bse

nt:

30

P<

0.0

5

Dis

tan

ce

tra

nsp

ort

ed

(m

iles)

Ph

ysi

cia

n p

rese

nt:

21

Ph

ysi

cia

n a

bse

nt:

23

Initia

l tra

um

a s

co

re

Ph

ysi

cia

n p

rese

nt:

7

Ph

ysi

cia

n a

bse

nt:

7

Nu

mb

er

of

org

an

s in

jure

d:

Ph

ysi

cia

n p

rese

nt:

2

Ph

ysi

cia

n a

bse

nt:

2

ISS

Ph

ysi

cia

n p

rese

nt:

15

Ph

ysi

cia

n a

bse

nt:

15

Tim

e a

t sc

en

e (

min

ute

s)

Ph

ysi

cia

n p

rese

nt:

15

Ph

ysi

cia

n a

bse

nt:

15

Nu

mb

er

of

sce

ne

pro

ce

du

res

(% p

er

att

en

da

nc

e)

Ph

ysi

cia

n p

rese

nt:

221

Ph

ysi

cia

n a

bse

nt:

25

6

Nu

mb

er

of

in-f

ligh

t p

roc

ed

ure

s (%

pe

r

att

en

da

nc

e)

Ph

ysi

cia

n p

rese

nt:

8

Ph

ysi

cia

n a

bse

nt:

4

Co

mm

en

ts

�

All

pa

tie

nts

att

en

de

d a

sin

gle

tra

um

a

ce

ntr

e.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

TRIS

S m

eth

od

olo

gy a

pp

rop

ria

tely

use

d

RTS

sc

ore

.

�

TRIS

S a

na

lysi

s a

pp

rop

riate

ly in

clu

de

d

est

ima

tio

n o

f th

e M

sta

tist

ic.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

It a

pp

ea

rs t

ha

t e

xpe

rie

nc

ed

nu

rse

s a

nd

pa

ram

ed

ics,

op

era

tin

g w

ith

we

ll-e

sta

blis

he

d

pro

toc

ols

, c

an

pro

vid

e a

gg

ress

ive

ca

re t

ha

t

yie

lds

eq

ua

l ou

tco

me

re

sults

co

mp

are

d w

ith

tho

se o

f a

flig

ht

tea

m t

ha

t in

clu

de

s a

ph

ysi

cia

n.


22

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ha

mm

an

et

al.

1991

)

USA

co

ntin

ue

d

N

um

be

r o

f e

me

rge

nc

y r

oo

m

pro

ce

du

res

(% p

er

att

en

da

nc

e)

Ph

ysi

cia

n p

rese

nt:

26

Ph

ysi

cia

n a

bse

nt:

20

Blu

nt

ve

rsu

s p

en

etr

atin

g in

jurie

s (%

)

Ph

ysi

cia

n p

rese

nt:

95:5

Ph

ysi

cia

n a

bse

nt:

93

:7


23

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hm

idt

et

al.

1992

)

USA

an

d G

erm

an

y

Re

gis

try b

ase

d

stu

dy

co

mp

arin

g

mo

rta

lity in

two

ce

ntr

es

Leve

l III-

2.

Stu

dy s

ett

ing

.

Two

se

ttin

gs:

USA

an

d G

erm

an

y. Th

e

USA

se

ttin

g c

on

sist

ed

of

he

lico

pte

rs

cre

we

d b

y f

ligh

t n

urs

e a

nd

pa

ram

ed

ic

(Ju

ly 1

98

8-J

un

e 1

989

). T

he

Ge

rma

n

sett

ing

co

nsi

ste

d o

f h

elic

op

ters

cre

we

d

by t

rau

ma

su

rge

on

an

d p

ara

me

dic

(No

ve

mb

er

198

8-O

cto

be

r 19

89

).

Pa

rtic

ipa

nts

:

Tota

l 407

Inte

rve

ntio

n (

n=

22

1)

Ge

rma

n s

ett

ing

(w

ith

su

rge

on

)

Co

mp

ara

tor

(n=

18

6).

USA

se

ttin

g (

with

ou

t su

rge

on

)

Pre

-ho

spita

l A

LS in

clu

de

d

IV f

luid

s, E

T in

tub

atio

n,

tub

e

tho

rac

ost

om

y/n

ee

dle

de

co

mp

ress

ion

.

Hig

h u

se o

f IV

an

ae

sth

esi

a in

Ge

rma

ny.

Inc

lu/e

xcl c

rite

ria

.

All

pa

tie

nts

with

mu

ltip

le in

jurie

s

tra

nsp

ort

ed

to

th

e t

wo

re

spe

ctive

tra

um

a c

en

tre

s d

urin

g a

on

e y

ea

r

pe

riod

.

Pa

tie

nts

we

re t

ran

spo

rte

d d

ire

ctly

to t

he

tra

um

a c

en

tre

ra

the

r th

an

via

an

inte

rme

dia

te f

ac

ility

.

Exc

lud

ed

pa

tie

nts

with

inc

om

ple

te

or

mis

sin

g c

ha

rts.

Da

ta c

olle

ctio

n

Da

ta r

etr

ieve

d f

rom

tra

um

a r

eg

istr

y

rec

ord

s a

nd

flig

ht

log

s.

Ext

rac

ted

pa

tie

nt

ag

e,

me

ch

an

ism

of

inju

ry, flig

ht

tim

es

fro

m t

he

sc

en

e

to t

he

tra

um

a c

en

tre

, R

TS, IS

S,

AIS

-

85,

mo

rta

lity,

tim

e o

f d

ea

th f

rom

ho

spita

l ad

mis

sio

n a

nd

pre

-ho

spita

l

ALS

re

ce

ive

d.

RTS

me

asu

red

in t

he

fie

ld in

Ge

rma

ny a

nd

up

on

arr

iva

l

at

the

tra

um

a c

en

tre

in U

SA

.

Ou

tco

me

me

asu

res

Mo

rta

lity

Follo

w-u

p in

terv

al

No

t st

ate

d

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Su

rge

on

gro

up

Z s

tatist

ic +

2.4

59

P<

0.0

25

M s

tatist

ic 0

.89

W s

tatist

ic in

dic

ate

s 1.3

5 a

dd

itio

na

l

surv

ivo

rs f

rom

100

pa

tie

nts

an

aly

sed

co

mp

are

d w

ith

th

e M

TOS

po

pu

latio

n.

No

n-s

urg

eo

n g

rou

p

Z s

tatist

ic +

1.0

49

P>

0.0

5

No

te h

ow

eve

r, M

st

atist

ic o

f 0.8

74

ind

ica

tes

less

th

an

ide

al d

eg

ree

of

ma

tch

in in

jury

se

ve

rity

(a

dvis

ed

cu

t

off

0.8

8)

De

ath

with

in 6

ho

urs

Su

rge

on

pre

sen

t: 4

of

21

de

ath

s

Su

rge

on

ab

sen

t: 1

2 o

f 21 d

ea

ths

P<

0.0

1

Lim

ita

tio

ns

�

Re

tro

spe

ctive

, re

gis

try b

ase

d s

tud

y.

�

Exc

lud

ed

37 p

atie

nts

fro

m t

he

Ge

rma

n s

ett

ing

du

e t

o in

co

mp

lete

or

mis

sin

g c

ha

rts.

Th

is is

a lik

ely

so

urc

e o

f

sele

ctio

n b

ias

giv

en

th

ere

we

re n

o

de

ath

s in

th

is g

rou

p.

�

Va

riatio

n in

tim

ing

of

RTS

est

ima

tio

n

will

re

sult in

va

ria

tio

n in

th

e p

red

icte

d

surv

iva

l in

th

e t

wo

gro

up

s. T

he

refo

re,

co

nc

ern

s e

xis

t a

bo

ut

the

dire

ct

co

mp

ara

bili

ty o

f th

e p

atie

nt

surv

iva

l

da

ta b

etw

ee

n t

he

tw

o g

rou

ps.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps.

�

Hig

h p

ote

ntia

l fo

r c

on

fou

nd

ing

.

�

Sig

nific

an

tly s

ho

rte

r tim

e f

rom

inc

ide

nt

to la

un

ch

in t

he

su

rge

on

gro

up

po

ten

tia

lly im

pro

vin

g p

rog

no

sis.

�

Sig

nific

an

tly lo

ng

er

sce

ne

tim

e in

th

e

surg

eo

n g

rou

p.


24

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hm

idt

et

al.

1992

)

USA

an

d G

erm

an

y

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

.

Me

an

ag

e (

ye

ars

)

Su

rge

on

pre

sen

t: 36

(ra

ng

e 1

-90)

Su

rge

on

ab

sen

t: 2

9.5

(ra

ng

e 1

-86)

Blu

nt

me

ch

an

ism

of

inju

ry (

%)

Su

rge

on

pre

sen

t: 10

0

Su

rge

on

ab

sen

t: 8

9.2

Me

an

ISS s

co

re

Su

rge

on

pre

sen

t: 1

8

Su

rge

on

ab

sen

t: 1

9.8

Tota

l de

ath

s:

Su

rge

on

pre

sen

t: 2

1 (

9.5

%)

Su

rge

on

ab

sen

t: 2

1 (

11.3

%)

Me

an

tim

e f

rom

in

cid

en

t to

he

lico

pte

r

lau

nc

h (

min

ute

s)

Su

rge

on

pre

sen

t: 5

.5

Su

rge

on

ab

sen

t: 1

7.5

P<

0.0

5

Me

an

sc

en

e t

ime

(m

inu

tes)

Su

rge

on

pre

sen

t: 1

9.3

Su

rge

on

ab

sen

t: 1

0

P<

0.0

5

Tota

l mis

sio

n t

ime

fro

m in

cid

en

t to

ho

spita

l arr

iva

l (m

inu

tes)

Su

rge

on

pre

sen

t: 4

2.5

Su

rge

on

ab

sen

t: 5

4.5

P>

0.0

5

ISS s

co

res

ha

d s

imila

r d

istr

ibu

tio

ns

at

bo

th c

en

tre

s. S

imila

r d

istr

ibu

tio

ns

of

ag

e

an

d ISS w

ere

als

o o

bse

rve

d a

t b

oth

site

s. U

ne

qu

al d

istr

ibu

tio

n o

f se

ve

re

inju

rie

s to

on

e b

od

y r

eg

ion

wa

s a

lso

no

t

de

mo

nst

rate

d.

An

aly

sis

Me

an

s fo

r a

ge

, IS

S a

nd

flig

ht

tim

es

ca

lcu

late

d.

Dis

trib

utio

ns

of

ISS c

on

stru

cte

d.

Co

mp

ara

bili

ty o

f b

od

y r

eg

ion

s

inju

red

co

mp

are

d u

sin

g t

he

AIS

sco

re.

TRIS

S m

eth

od

s u

sed

(re

fere

nc

e

MTO

S)

to c

om

pa

re b

oth

ce

ntr

es.

Pro

ba

bili

ty o

f su

rviv

al u

sed

to

ca

lcu

late

th

e Z

sta

tist

ic.

The

M a

nd

W s

tatist

ics

we

re c

alc

ula

ted

fo

r

ea

ch

inst

itu

tio

n.

Ad

va

nc

ed

pre

-ho

spita

l ca

re

ET

intu

ba

tio

n

Su

rge

on

pre

sen

t: 3

7.1

%

Su

rge

on

ab

sen

t: 1

3.4

%

P<

0001

Tho

rac

ic c

av

ity d

ec

om

pre

ssio

n

Su

rge

on

pre

sen

t: 9

.1%

Su

rge

on

ab

sen

t: 0

.5%

P<

0.0

01

IV f

luid

infu

sio

n (

me

an

mls

)

Su

rge

on

pre

sen

t: 1

800

Su

rge

on

ab

sen

t: 8

25

P<

0.0

5

Co

mm

en

ts

�

All

pa

tie

nts

att

en

de

d a

sin

gle

tra

um

a

ce

ntr

e in

ea

ch

co

un

try.

�

TRIS

S m

eth

od

olo

gy a

pp

rop

ria

tely

use

d

RTS

sc

ore

.

�

TRIS

S a

na

lysi

s a

pp

rop

riate

ly in

clu

de

d

est

ima

tio

n o

f th

e M

an

d W

sta

tist

ics.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

�

Me

dic

al c

on

tro

l ava

ilab

le b

y r

ad

io in

USA

.

�

Re

sults

ma

y r

efle

ct

mo

re a

gg

ress

ive

the

rap

y in

Ge

rma

ny r

ath

er

tha

n t

he

spe

cific

cre

w c

on

fig

ura

tio

n s

inc

e k

ey

diffe

ren

ce

s in

th

e p

roc

ed

ure

s u

sed

ca

n b

e c

on

du

cte

d b

y p

ara

me

dic

s

an

d f

ligh

t n

urs

es

as

we

ll a

s d

oc

tors

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

With

on

sc

en

e e

xpe

rie

nc

e a

nd

jud

ge

me

nt

of

a t

rau

ma

su

rge

on

, th

e G

erm

an

ae

rom

ed

ica

l syst

em

pro

vid

ed

mo

re

ag

gre

ssiv

e p

re-h

osp

ita

l re

susc

ita

tio

n,

pa

rtic

ula

rly in

th

e a

rea

s o

f a

irw

ay a

nd

ve

ntila

tio

n m

an

ag

em

en

t.


25

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ga

rne

r e

t a

l. 19

99

)

Au

stra

lia

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

He

lico

pte

r tr

an

spo

rta

tio

n o

f p

atie

nts

fro

m t

he

sc

en

e o

ve

r a

28 m

on

th p

erio

d.

Pa

rtic

ipa

nts

(n

=2

07

):

Inte

rve

ntio

n (

n=

67

).

Ph

ysi

cia

n s

taff

ed

NR

MA

Ca

reFl

igh

t

he

lico

pte

r

Co

mp

ara

tor

(n=

14

0).

Pa

ram

ed

ic s

taff

ed

We

stp

ac

Hu

nte

r

reg

ion

he

lico

pte

r.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

dia

n a

ge

(ye

ars

)

Ph

ysi

cia

n g

rou

p:

31 (

13

-70

)

Pa

ram

ed

ic g

rou

p:

33 (

2-8

9)

Me

ch

an

ism

of

inju

ry

Mo

tor

ve

hic

le o

cc

up

an

t

Ph

ysi

cia

n g

rou

p:

63%

Pa

ram

ed

ic g

rou

p:

63

%

Mo

tor

bik

e r

ide

r

Ph

ysi

cia

n g

rou

p:

9%

Pa

ram

ed

ic g

rou

p:

11

%

Pe

da

l cyc

list

or

pe

de

stria

n

Ph

ysi

cia

n g

rou

p:

3%

Pa

ram

ed

ic g

rou

p:

5%

Inc

lu/e

xcl c

rite

ria

.

Blu

nt

tra

um

a

ISS ≥

10

Tra

nsp

ort

ed

dire

ctly f

rom

th

e

inc

ide

nt

sce

ne

Inc

ide

nt

oc

cu

rre

d b

etw

ee

n

Jan

ua

ry 1

996 a

nd

Ap

ril 1

998.

Da

ta c

olle

ctio

n

Pa

ram

ed

ic g

rou

p r

etr

osp

ec

tive

ly

ide

ntifie

d f

rom

th

e t

rau

ma

re

gis

try

of

Joh

n H

un

ter

Ho

spita

l, N

ew

ca

stle

.

Ph

ysi

cia

n g

rou

p id

en

tifie

d f

rom

th

e

me

dic

al d

ata

ba

se o

f N

RM

A

Ca

reFl

igh

t/N

SW

Me

dic

al r

etr

ieva

l

Se

rvic

e w

ho

we

re t

ran

spo

rte

d t

o

We

stm

ea

d o

r N

ep

ea

n h

osp

ita

ls in

Syd

ne

y.

Pre

-ho

spita

l ca

se s

he

ets

we

re

exa

min

ed

to

allo

w t

he

ca

lcu

latio

n

of

the

RTS

(firs

t re

co

rde

d d

ata

use

d).

Oth

er

da

ta c

olle

cte

d

inc

lud

ed

de

mo

gra

ph

ics,

me

ch

an

ism

of

inju

ry, re

spo

nse

,

sce

ne

an

d t

ran

spo

rt t

ime

s,

en

tra

pm

en

t a

t th

e s

ce

ne

,

req

uire

me

nt

for

win

ch

extr

ac

tio

n,

flu

ids

ad

min

iste

red

an

d p

roc

ed

ure

s

pe

rfo

rme

d a

t th

e s

ce

ne

or

in t

ran

sit.

Ou

tco

me

me

asu

res

Mo

rta

lity. C

om

pa

red

with

TR

ISS

me

tho

do

log

y u

sin

g c

oe

ffic

ien

ts

de

rive

d f

rom

th

e M

TOS u

sin

g t

he

1990 a

bb

revia

ted

inju

ry s

ca

le.

Co

mp

aris

on

be

twe

en

ob

serv

ed

an

d p

red

icte

d m

ort

alit

y m

ad

e a

t

ho

spita

l dis

ch

arg

e.

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Ph

ysi

cia

n g

rou

p

Z s

tatist

ic +

2.7

2

P<

0.0

1

M s

tatist

ic 0

.62

Ad

just

ed

W s

tatist

ic 9

.48 (

95%

CI

3.8

4-1

5.1

2)

co

mp

are

d w

ith

th

e

MTO

S p

op

ula

tio

n.

Pa

ram

ed

ic g

rou

p

Z s

tatist

ic -

1.1

6

P=

0.2

5

M s

tatist

ic 0

.68

Ad

just

ed

W s

tatist

ic -

2.3

7 (

95%

CI -

6.8

1 t

o 2

.07)

co

mp

are

d w

ith

th

e

MTO

S p

op

ula

tio

n.

Dire

ct

co

mp

ariso

n b

etw

ee

n

ph

ysi

cia

n a

nd

pa

ram

ed

ic g

rou

p

Ad

just

ed

W s

tatist

ic 1

3.4

4 (

95%

CI

7.8

0-1

9.0

8)

sug

ge

stin

g a

n a

dd

itio

na

l

13 s

urv

ivo

rs p

er

10

0 p

atie

nts

tre

ate

d

in t

he

ph

ysi

cia

n g

rou

p c

om

pa

red

with

th

e p

ara

me

dic

gro

up

.

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

Gro

up

s tr

ea

ted

in d

iffe

ren

t h

osp

ita

ls –

pa

ram

ed

ic g

rou

p t

rea

ted

in a

leve

l 6

ho

spita

l, p

hysi

cia

n g

rou

p in

a le

ve

l 5

or

leve

l 6 h

osp

ita

l.

�

Sta

tist

ica

lly s

ign

ific

an

t d

iffe

ren

ce

in

ba

selin

e G

CS a

nd

ISS s

co

res.

co

nsi

ste

nt

with

in

cre

ase

d s

eve

rity

in

the

ph

ysi

cia

n g

rou

p.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g a

lth

ou

gh

use

of

the

W

sta

tist

ic h

elp

s a

dju

st f

or

TRIS

S v

aria

ble

s

be

twe

en

th

e s

tud

y p

op

ula

tio

ns.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

RTS

sc

ore

in t

he

MTO

S s

tud

y w

as

ca

lcu

late

d a

t a

dm

issi

on

ra

the

r th

an

at

the

sc

en

e a

s p

erf

orm

ed

in t

his

stu

dy. Th

e a

pp

roa

ch

use

d in

th

is s

tud

y

ha

s th

e a

dva

nta

ge

of

co

llec

tin

g R

TS

da

ta b

efo

re in

tub

atio

n a

nd

co

mp

arin

g R

TS in

bo

th g

rou

ps

at

ap

pro

xim

ate

ly t

he

sa

me

tim

e.

Ho

we

ve

r, c

om

pa

riso

n w

ith

MTO

S is

no

t so

va

lid g

ive

n t

he

diffe

ren

ce

in

tim

ing

.

�

Se

ve

n p

atie

nts

we

re e

xclu

de

d d

ue

to

mis

sin

g c

ase

sh

ee

ts.

All

7 w

ere

in t

he

pa

ram

ed

ic g

rou

p p

rod

uc

ing

a

sele

ctio

n b

ias.


26

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ga

rne

r e

t a

l. 19

99

)

Au

stra

lia

co

ntin

ue

d

Fa

lls

Ph

ysi

cia

n g

rou

p:

10%

Pa

ram

ed

ic g

rou

p:

6%

Me

dia

n t

ime

inte

rva

ls (

min

ute

s)

Ca

ll to

sc

en

e a

rriv

al

Ph

ysi

cia

n g

rou

p:

29

Pa

ram

ed

ic g

rou

p:

26

Sc

en

e t

ime

(e

xclu

din

g t

rap

pe

d a

nd

win

ch

ed

pa

tie

nts

)

Ph

ysi

cia

n g

rou

p:

33

Pa

ram

ed

ic g

rou

p:

34

Tra

nsp

ort

tim

e t

o h

osp

ita

l

Ph

ysi

cia

n g

rou

p:

15

Pa

ram

ed

ic g

rou

p:

12

Tota

l pre

-ho

spita

l tim

e

Ph

ysi

cia

n g

rou

p:

86

Pa

ram

ed

ic g

rou

p:

82

Me

dia

n R

TS

Ph

ysi

cia

n g

rou

p:

6.9

0

Pa

ram

ed

ic g

rou

p:

7

.55

Me

dia

n G

CS

Ph

ysi

cia

n g

rou

p: 1

3

Pa

ram

ed

ic g

rou

p:

14

P=

0.0

5

Me

dia

n ISS

Ph

ysi

cia

n g

rou

p:

25

Pa

ram

ed

ic g

rou

p:

18

P=

0.0

5

Follo

w-u

p in

terv

al

No

t st

ate

d

An

aly

sis

Ca

teg

oric

al v

aria

ble

s: χ

2 o

r Fi

she

r’s

exa

ct

test

, a

s a

pp

rop

ria

te.

Co

ntin

uo

us

va

ria

ble

s: M

an

n-

Wh

itn

ey U

te

st.

Co

mp

aris

on

be

twe

en

pre

dic

ted

an

d o

bse

rve

d m

ort

alit

y u

sin

g Z

, W

an

d M

sta

tist

ics.

An

ad

just

ed

W

sta

tist

ic w

as

ca

lcu

late

d b

y t

he

me

tho

d o

f Y

ou

ng

e w

he

n t

he

M

sta

tist

ic in

dic

ate

d a

po

or

ma

tch

with

th

e M

TOS c

oh

ort

an

d t

o

dire

ctly c

om

pa

re t

he

pa

ram

ed

ic

an

d p

hysi

cia

n t

rea

ted

gro

up

s.

Pro

ce

du

res

at

sce

ne

Me

dia

n v

olu

me

of

flu

id in

fuse

d in

pa

tie

nts

wh

o r

ec

eiv

ed

> 5

0m

L

Ph

ysi

cia

n g

rou

p:

250

0

Pa

ram

ed

ic g

rou

p:

825

P<

0.0

01

Me

dia

n v

olu

me

of

flu

id (

mL)

infu

sed

in p

atie

nts

with

initia

l hyp

ote

nsi

on

(syst

olic

BP

<90

mm

Hg

)

Ph

ysi

cia

n g

rou

p:

503

5

Pa

ram

ed

ic g

rou

p:

147

5

P<

0.0

01

Nu

mb

er

of

pa

tie

nts

intu

ba

ted

Ph

ysi

cia

n g

rou

p:

34/6

7 (

1

cric

oth

yro

ido

tom

y, m

usc

le r

ela

xan

t

dru

gs

use

d in

28 o

f th

e 3

4

intu

ba

tio

ns)

Pa

ram

ed

ic g

rou

p:

14

/14

0

P<

0.0

01

Pro

po

rtio

n o

f p

atie

nts

with

GC

S<

9

intu

ba

ted

Ph

ysi

cia

n g

rou

p:

23/2

3

Pa

ram

ed

ic g

rou

p:

14

/36

P<

0.0

01

Tho

rac

ic d

ec

om

pre

ssio

ns

Ph

ysi

cia

n g

rou

p:

8/6

7 (

6 t

ub

e, 2

ne

ed

le)

Pa

ram

ed

ic g

rou

p:

2/1

40 (

bo

th

ne

ed

le)

P<

0.0

1

Co

mm

en

ts

�

TRIS

S m

eth

od

olo

gy a

pp

rop

ria

tely

use

d

RTS

sc

ore

.

�

TRIS

S a

na

lysi

s a

pp

rop

riate

ly in

clu

de

d

est

ima

tio

n o

f th

e M

an

d W

sta

tist

ics.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

�

Diffe

ren

ce

s in

pro

ce

du

res

pe

rfo

rme

d

be

twe

en

stu

dy g

rou

ps.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Ph

ysi

cia

ns

pe

rfo

rm a

gre

ate

r n

um

be

r o

f

pro

ce

du

res

at

ac

cid

en

t sc

en

es

with

ou

t

inc

rea

sin

g s

ce

ne

tim

e.

This

re

sults

in

sig

nific

an

tly lo

we

r m

ort

alit

y. C

ritic

al c

are

ph

ysi

cia

ns

sho

uld

be

ad

de

d t

o p

ara

me

dic

he

lico

pte

r se

rvic

es

for

sce

ne

re

spo

nse

to

blu

nt

tra

um

a.


27

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ca

me

ron

et

al.

20

05)

Au

stra

lia

Be

fore

an

d

aft

er

stu

dy.

Re

tro

spe

ctive

ch

art

re

vie

w.

Leve

l III-

3.

Stu

dy s

ett

ing

.

Ca

irn

s B

ase

ho

spita

l. U

ntil 2

00

1

he

lico

pte

rs in

clu

de

d e

me

rge

nc

y

ph

ysi

cia

ns.

Sin

ce

2001 t

he

y h

ave

be

en

sta

ffe

d b

y in

ten

sive

ca

re p

ara

me

dic

s.

Pa

rtic

ipa

nts

(n

=3

74

):

Inte

rve

ntio

n (

n=

16

3)

Sto

pp

ed

th

e p

rese

nc

e o

f a

n

em

erg

en

cy p

hysi

cia

n o

n a

n

em

erg

en

cy h

elic

op

ter

Co

mp

ara

tor

(n=

21

1)

Em

erg

en

cy p

hysi

cia

n o

n b

oa

rd

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e

Me

dia

n a

ge

(ye

ars

)

Ph

ysi

cia

n g

rou

p:

34

Pa

ram

ed

ic g

rou

p:

33

Ma

les

(%)

Ph

ysi

cia

n g

rou

p:

64.0

Pa

ram

ed

ic g

rou

p:

65.6

Me

an

RTS

Ph

ysi

cia

n g

rou

p:

7.7

22

Pa

ram

ed

ic g

rou

p:

7.7

30

Inc

lu/e

xcl c

rite

ria

.

An

y p

rim

ary

ta

skin

g o

f th

e

he

lico

pte

r. C

ase

s w

ere

ide

ntifie

d b

y

the

‘m

od

e o

f a

rriv

al’

fie

ld re

co

rde

d

on

th

e E

D in

form

atio

n s

yst

em

Da

ta c

olle

ctio

n

The

RTS

wa

s c

alc

ula

ted

usi

ng

th

e

initia

l clin

ica

l ob

serv

atio

ns

in t

he

me

dic

al r

ec

ord

s a

nd

am

bu

lan

ce

form

s.

Ou

tco

me

me

asu

res

30 d

ay m

ort

alit

y

Len

gth

of

in-h

osp

ita

l sta

y

Tra

nsf

er

rate

s

Ra

tes

of

dis

ch

arg

e d

ire

ctly f

rom

ho

spita

l

An

aly

sis

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

On

e w

ay a

na

lysi

s o

f va

ria

nc

e w

as

un

de

rta

ke

n a

nd

P v

alu

es

an

d a

χ2

test

with

Ya

tes

co

rre

ctio

n w

ere

ca

lcu

late

d w

he

re a

pp

rop

ria

te.

Ka

pp

a s

tatist

ic w

as

use

d t

o a

sse

ss

inte

r-ra

ter

relia

bili

ty in

th

e c

ha

rt

ext

rac

tio

n.

30 d

ay m

ort

alit

y p

rop

ort

ion

, b

y

stu

dy g

rou

p (

%)

Ph

ysi

cia

n g

rou

p:

2.8

%

Pa

ram

ed

ic g

rou

p:

2.5

%

P=

0.8

Me

an

ho

spita

l le

ng

th o

f st

ay, b

y

stu

dy g

rou

p (

da

ys)

Ph

ysi

cia

n g

rou

p:

2

Pa

ram

ed

ic g

rou

p:

1

P=

0.3

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

The

ac

cu

rac

y o

f c

od

ing

th

e m

od

e o

f

arr

iva

l (u

sed

to

ide

ntify

re

leva

nt

pa

tie

nts

) w

as

no

t d

oc

um

en

ted

.

Ho

we

ve

r, t

his

is n

ot

like

ly t

o b

e a

ma

jor

sou

rce

of

bia

s.

�

Un

cle

ar

if t

he

tim

ing

of

the

me

asu

rem

en

ts u

sed

to

ass

ess

RTS

wa

s

the

sa

me

in b

oth

gro

up

s.

�

Me

asu

res

of

inju

ry s

eve

rity

use

d w

ere

diffe

ren

t fr

om

th

e n

orm

al a

pp

roa

ch

.

No

use

of

the

ISS s

o it

wa

s n

ot

po

ssib

le

to a

sse

ss p

rob

ab

ility

of

surv

iva

l w

ith

refe

ren

ce

to

a s

uita

ble

po

pu

latio

n

suc

h a

s M

TOS.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

Ba

selin

e d

iffe

ren

ce

s in

pa

tie

nts

ad

mitte

d w

ith

hig

he

r p

rop

ort

ion

of

the

ph

ysi

cia

n g

rou

p b

ein

g a

dm

itte

d.

�

Ve

ry lo

w n

um

be

r o

f d

ea

ths

co

nsi

ste

nt

with

low

stu

dy p

ow

er

(10 d

ea

ths

in

tota

l).

�

Au

tho

rs s

ug

ge

ste

d R

TS w

as

a p

oo

r

pre

dic

tor

of

ne

ed

fo

r a

dm

issi

on

an

d

po

stu

late

d t

ha

t A

PA

CH

E m

ay h

ave

be

en

a b

ett

er

me

asu

re.

�

No

fo

llow

-up

of

pa

tie

nts

dis

ch

arg

ed

dire

ctly f

rom

ED


28

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ca

me

ron

et

al.

20

05)

Au

stra

lia

co

ntin

ue

d

P

rop

ort

ion

dis

ch

arg

ed

fro

m E

D

Ph

ysi

cia

n g

rou

p:

14.7

%

Pa

ram

ed

ic g

rou

p:

33.1

%

P=

0.0

00

1

Pro

po

rtio

n o

f tr

au

ma

pa

tie

nts

ad

mitte

d

Ph

ysi

cia

n g

rou

p:

86.7

%

Pa

ram

ed

ic g

rou

p:

68.9

%

P=

0.0

02

Pro

po

rtio

n o

f n

on

-tra

um

a p

atie

nts

ad

mitte

d

Ph

ysi

cia

n g

rou

p:

83.7

Pa

ram

ed

ic g

rou

p:

64.4

P=

0.0

04

Pro

po

rtio

n o

f tr

au

ma

pa

tie

nts

with

ma

xim

um

RTS

Ph

ysi

cia

n g

rou

p:

89.4

Pa

ram

ed

ic g

rou

p:

90.0

Co

mm

en

ts

�

Two

co

nse

cu

tive

ye

ars

fo

r e

ac

h g

rou

p

we

re e

xam

ine

d in

th

e c

ha

rt r

ev

iew

.

�

A s

ec

on

d r

ev

iew

er

va

lida

ted

th

e

ch

art

ext

rac

tio

n in

40 r

an

do

mly

sele

cte

d c

ha

rts

(10.7

% o

f th

e t

ota

l).

Exc

elle

nt

leve

l of

ag

ree

me

nt

ac

hie

ve

d (

Ka

pp

a 0

.937

).

�

All

pa

tie

nts

tra

nsp

ort

ed

to

th

e s

am

e

ba

se h

osp

ita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

sim

ilaritie

s in

ou

tco

me

s fo

r a

dm

itte

d

pa

tie

nts

su

pp

ort

th

e v

iew

th

at

bo

th g

rou

ps

ha

ve

sim

ilar

task

ing

crite

ria

fo

r h

igh

ac

uity

pa

tie

nts

an

d s

ug

ge

st t

ha

t p

ara

me

dic

s a

re

as

eff

ica

cio

us

as

ph

ysi

cia

ns

in d

eliv

erin

g

pre

-ho

spita

l ca

re in

th

is g

rou

p o

f p

atie

nts

.

Ho

we

ve

r, f

or

low

er

ac

uity p

atie

nts

, th

ere

is a

sta

tist

ica

lly s

ign

ific

an

t h

igh

er

rate

of

clin

ica

lly

un

ne

ce

ssa

ry t

ask

ing

s b

y t

he

am

bu

lan

ce

gro

up

. G

ive

n t

he

re

ce

nt

fata

l ae

rom

ed

ica

l

ac

cid

en

ts in

Qu

ee

nsl

an

d it

wo

uld

se

em

pru

de

nt

to r

ed

uc

e c

linic

ally

un

ne

ce

ssa

ry

retr

ieva

ls t

hro

ug

h c

linic

al c

oo

rdin

atio

n w

ith

ap

pro

pria

tely

qu

alif

ied

em

erg

en

cy

ph

ysi

cia

ns.


29

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bu

rne

y e

t a

l. 19

92)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Ae

rom

ed

ica

l tra

nsf

ers

be

twe

en

Se

pte

mb

er

1,

198

7 a

nd

Au

gu

st 3

1, 1

988.

Inc

lud

e t

ran

spo

rt f

rom

oth

er

ED

s

(n=

404

), t

he

sc

en

e (

n=

36)

an

d o

the

r

inp

atie

nt

un

its

(n=

21

8).

Pa

rtic

ipa

nts

:

Tota

l = 6

59

Inte

rve

ntio

n (

n=

41

8)

Ph

ysi

cia

n a

nd

nu

rse

te

am

Co

mp

ara

tor

(n=

24

1)

Two

nu

rse

te

am

.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

AP

AC

HE-I

I sc

ore

Ph

ysi

cia

n g

rou

p:

8.0

Nu

rse

gro

up

: 7

.8

RTS

1 s

co

re (

Ob

tain

ed

at

orig

in o

f

tra

nsp

ort

)

Ph

ysi

cia

n g

rou

p:

10.6

7

Nu

rse

gro

up

: 1

0.8

47

RTS

2 s

co

re (

ob

tain

ed

aft

er

arr

iva

l at

rec

eiv

ing

ho

spita

l)

Ph

ysi

cia

n g

rou

p:

10.8

77

Nu

rse

gro

up

: 1

0.9

72

TISS1 s

co

re

Ph

ysi

cia

n g

rou

p:

17.0

Nu

rse

gro

up

: 1

6.0

TISS2 s

co

re

Ph

ysi

cia

n g

rou

p:

14.9

Nu

rse

gro

up

: 1

2.6

P=

0.0

01

Inc

lu/e

xcl c

rite

ria

.

Inc

lud

ed

all

pa

tie

nts

tra

nsf

err

ed

du

rin

g t

he

stu

dy p

erio

d.

Exc

lud

ed

pa

tie

nts

un

de

r 1

6 y

ea

rs

an

d p

atie

nts

tra

nsf

err

ed

to

oth

er

ho

spita

ls

Da

ta c

olle

ctio

n

Se

ve

rity

of

illn

ess

me

asu

red

usi

ng

RTS

, A

PA

CH

E-I

I a

nd

th

e T

he

rap

eu

tic

Inte

rve

ntio

n S

co

rin

g S

yst

em

(TI

SS)

Two

TIS

S s

co

res

we

re o

bta

ine

d:

TISS1=

inte

rve

ntio

ns

ca

rrie

d o

ut

be

fore

th

e a

rriv

al o

f th

e f

ligh

t te

am

TISS2=

inte

rve

ntio

ns

co

ntin

ue

d o

r

initia

ted

by t

he

flig

ht

tea

m

Ou

tco

me

me

asu

res

Ho

spita

l mo

rta

lity

ICU

len

gth

of

sta

y

Ho

spita

l le

ng

th o

f st

ay

Follo

w-u

p in

terv

al

No

t st

ate

d

An

aly

sis

Pa

tie

nts

str

atifie

d in

to t

hre

e g

rou

ps:

ca

rdia

c, tr

au

ma

an

d o

the

r. O

rig

ins

of

the

tra

nsp

ort

an

d t

ran

sfe

r tim

es

we

re in

clu

de

d in

th

e a

na

lysi

s.

Da

ta a

na

lyse

d u

sin

g S

YSTA

T.

Pro

po

rtio

n d

isc

ha

rge

d a

live

(%

)

Ph

ysi

cia

n g

rou

p:

83

Nu

rse

gro

up

: 7

9

P=

0.2

Nu

mb

er

of

ho

spita

l da

ys

Ph

ysi

cia

n g

rou

p:

20.4

Nu

rse

gro

up

: 2

0.3

P=

0.9

45

Ou

tco

me

s w

ere

no

t si

gn

ific

an

tly

diffe

ren

t in

an

y o

f th

e c

linic

al

sub

gro

up

s o

r p

atie

nt

orig

in.

Lim

ita

tio

ns

�

The

orig

in o

f o

ne

pa

tie

nt

wa

s n

ot

ac

co

un

ted

fo

r.

�

On

ly 5

.5%

of

tra

nsp

ort

s o

rig

ina

ted

at

the

sc

en

e.

�

Sc

an

t d

eta

ils p

rovid

ed

ab

ou

t th

e

an

aly

sis

pro

ce

du

re in

th

e m

eth

od

s

sec

tio

n o

f th

e p

ap

er.

�

No

de

tails

pro

vid

ed

ab

ou

t th

e

pro

ce

du

res

co

nd

uc

ted

by e

ac

h

gro

up

.

�

Me

asu

res

of

inju

ry s

eve

rity

use

d w

ere

diffe

ren

t fr

om

th

e n

orm

al a

pp

roa

ch

.

No

use

of

the

ISS s

o it

wa

s n

ot

po

ssib

le

to a

sse

ss p

rob

ab

ility

of

surv

iva

l w

ith

refe

ren

ce

to

a s

uita

ble

po

pu

latio

n

suc

h a

s M

TOS.

�

Imp

erf

ec

t va

lidity o

f A

PA

CH

E,

TISS a

nd

ISS m

ay h

ave

imp

lica

tio

ns

reg

ard

ing

ba

lan

ce

of

gro

up

s a

t b

ase

line

.

�

Un

cle

ar

wh

eth

er

the

re w

as

an

y

va

riatio

n in

crite

ria

fo

r th

e d

isp

atc

h o

f

the

he

lico

pte

rs w

ith

diffe

ren

t c

rew

s.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Un

cle

ar

ho

w m

an

y r

ec

eiv

ing

ho

spita

ls

we

re in

clu

de

d –

if t

he

re w

as

mo

re

tha

n o

ne

it m

ay h

ave

re

sulte

d in

va

riatio

n in

qu

alit

y o

f c

are

at

diffe

ren

t

ho

spita

ls, re

sultin

g in

diffe

ren

t

ou

tco

me

s.


30

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bu

rne

y e

t a

l. 19

92)

USA

co

ntin

ue

d

M

ea

n a

ge

(ye

ars

)

Ph

ysi

cia

n g

rou

p:

45.7

Nu

rse

gro

up

: 5

0.9

P=

0.0

01

Ma

le (

%)

Ph

ysi

cia

n g

rou

p:

66

Nu

rse

gro

up

: 6

1

Pa

tie

nt

orig

in (

%)

ED

Ph

ysi

cia

n g

rou

p:

68

Nu

rse

gro

up

: 5

1

Inp

atie

nt

Ph

ysi

cia

n g

rou

p:

25

Nu

rse

gro

up

: 4

7

Sc

en

e

Ph

ysi

cia

n g

rou

p:

7

Nu

rse

gro

up

: 2

P=

0.0

01

Clin

ica

l ca

teg

ory

Ca

rdia

c

Ph

ysi

cia

n g

rou

p: 4

0

Nu

rse

gro

up

: 3

6

Tra

um

a

Ph

ysi

cia

n g

rou

p: 3

8

Nu

rse

gro

up

: 3

0

Oth

er

Ph

ysi

cia

n g

rou

p:

22

Nu

rse

gro

up

: 3

4

P=

0.0

03

�

Pa

tie

nt

orig

in m

ay b

e c

on

sist

en

t w

ith

the

nu

rse

gro

up

tra

nsp

ort

ing

a m

ore

sta

ble

gro

up

of

pa

tie

nts

.

�

Als

o v

aria

tio

n in

clin

ica

l ca

teg

ory

gro

up

tra

nsp

ort

ed

be

twe

en

th

e t

wo

gro

up

s.

�

Nu

rse

gro

up

tra

nsp

ort

ed

a

sig

nific

an

tly o

lde

r g

rou

p o

f p

atie

nts

–

oth

er

seve

rity

ind

ica

tors

be

ing

eq

ua

l

this

wo

uld

su

gg

est

a g

rou

p w

ith

a

po

ore

r p

rog

no

sis.

�

Str

atifie

d a

na

lysi

s b

y c

linic

al s

ub

gro

up

an

d p

atie

nt

orig

in s

ub

gro

up

ha

d lo

w

stu

dy p

ow

er.

�

No

su

bg

rou

p a

na

lyse

s p

rese

nte

d b

y

diffe

ren

t m

ark

ers

of

inju

ry s

eve

rity

.

Co

mm

en

ts

�

Ove

rall

sam

ple

siz

e o

f 659

is la

rge

en

ou

gh

to

ide

ntify

a 1

0%

diffe

ren

ce

in

surv

iva

l with

a p

ow

er

of

0.8

0.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

No

ob

jec

tive

diffe

ren

ce

s in

ou

tco

me

of

pa

tie

nts

we

re f

ou

nd

be

twe

en

ph

ysi

cia

n/n

urs

e a

nd

nu

rse

/nu

rse

te

am

s.

Alth

ou

gh

sm

all

diffe

ren

ce

s w

ere

fo

un

d in

typ

es

of

flig

hts

ta

ke

n b

y P

/N a

nd

N/N

te

am

s,

the

re w

ere

no

diffe

ren

ce

s in

ob

jec

tive

me

asu

res

of

seve

rity

be

twe

en

th

e t

wo

tea

ms.

We

fin

d n

o o

bje

ctive

ev

ide

nc

e t

o

pre

fer

on

e c

rew

co

mp

osi

tio

n o

ve

r a

no

the

r.


31

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bu

rne

y e

t a

l. 19

95)

USA

Pro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Follo

w o

n t

o p

rev

iou

s re

tro

spe

ctive

stu

dy b

y B

urn

ey e

t a

l. (1

992

). A

ll a

du

lt a

ir

me

dic

al p

atie

nts

be

twe

en

Ju

ly 1

, 19

90

an

d J

un

e 3

0, 19

92.

Pa

rtic

ipa

nts

: (n

=1,1

69)

Inte

rve

ntio

n (

n=

25

5)

Ph

ysi

cia

n a

nd

nu

rse

te

am

Co

mp

ara

tor

(n=

91

4)

Two

nu

rse

te

am

.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

Ag

e (

ye

ars

):

Ph

ysi

cia

n g

rou

p:

48.5

Nu

rse

gro

up

: 4

5.6

Ma

le (

%)

Ph

ysi

cia

n g

rou

p:

62

Nu

rse

gro

up

: 6

3

Pa

tie

nt

orig

in (

%)

ED

Ph

ysi

cia

n g

rou

p:

49

Nu

rse

gro

up

: 6

1

Ho

spita

l

Ph

ysi

cia

n g

rou

p:

38

Nu

rse

gro

up

: 3

2

Sc

en

e

Ph

ysi

cia

n g

rou

p:

12

Nu

rse

gro

up

: 6

AP

AC

HE1 s

co

re

Ph

ysi

cia

n g

rou

p:

10.5

Nu

rse

gro

up

: 1

0.4

Inc

lu/e

xcl c

rite

ria

.

Inc

lud

ed

all

pa

tie

nts

tra

nsf

err

ed

du

rin

g t

he

stu

dy p

erio

d.

Exc

lud

ed

pa

tie

nts

un

de

r 1

6 y

ea

rs

an

d p

atie

nts

tra

nsf

err

ed

to

oth

er

ho

spita

ls

Da

ta c

olle

ctio

n

Da

ta c

olle

cte

d p

rosp

ec

tive

ly b

y

flig

ht

nu

rse

s.

Se

ve

rity

of

illn

ess

me

asu

red

usi

ng

RTS

, A

PA

CH

E-I

I a

nd

th

e T

he

rap

eu

tic

.

Inte

rve

ntio

n S

co

rin

g S

yst

em

(TI

SS).

Two

AP

AC

HE s

co

res

we

re o

bta

ine

d:

the

first

(A

PA

CH

E1)

at

tim

e o

f

ho

spita

l arr

iva

l an

d t

he

se

co

nd

(AP

AC

HE2

) o

n t

he

ba

sis

of

the

first

24 h

ou

rs o

f h

osp

ita

lisa

tio

n a

fte

r

tra

nsf

er

Fou

r TI

SS s

co

res

we

re o

bta

ine

d:

TISS1=

inte

rve

ntio

ns

ca

rrie

d o

ut

in

the

24 h

ou

rs b

efo

re t

he

arr

iva

l of

the

flig

ht

tea

m

TISS2=

inte

rve

ntio

ns

co

ntin

ue

d o

r

initia

ted

by t

he

flig

ht

tea

m

TISS3=

inte

rve

ntio

ns

in t

he

first

24

ho

urs

aft

er

tra

nsf

er

TISS4=

co

mb

ine

d in

terv

en

tio

ns

du

rin

g f

ligh

t a

nd

th

e f

irst

24 h

ou

rs

aft

erw

ard

.

Thre

e R

TS s

co

res

we

re o

bta

ine

d:

RTS

1=

co

nd

itio

n b

efo

re t

ran

sfe

r

RTS

2=

co

nd

itio

n o

bta

ine

d d

urin

g

tra

nsf

er

RTS

3=

co

nd

itio

n a

fte

r tr

an

sfe

r.

Pro

po

rtio

n d

isc

ha

rge

d a

live

, a

ll

pa

tie

nts

(%

)

Ph

ysi

cia

n g

rou

p:

75

Nu

rse

gro

up

: 7

9

P=

0.1

2

Pro

po

rtio

n d

isc

ha

rge

d a

live

, o

rig

in

of

flig

ht

fro

m t

he

sc

en

e (

%)

Ph

ysi

cia

n g

rou

p:

84

Nu

rse

gro

up

: 8

4

Nu

mb

er

of

ho

spita

l da

ys

Pe

rio

d 1

990

-19

91

Ph

ysi

cia

n g

rou

p:

17.6

Nu

rse

gro

up

: 2

2.2

P=

0.0

9

Pe

rio

d 1

991

-19

92

Ph

ysi

cia

n g

rou

p:

33.4

Nu

rse

gro

up

: 2

2.5

P=

0.0

05

Lim

ita

tio

ns

�

On

ly 7

.6%

of

tra

nsp

ort

s o

rig

ina

ted

at

the

sc

en

e.

�

No

de

tails

pro

vid

ed

ab

ou

t th

e

pro

ce

du

res

co

nd

uc

ted

by e

ac

h

gro

up

.

�

Me

asu

res

of

inju

ry s

eve

rity

use

d w

ere

diffe

ren

t fr

om

th

e n

orm

al a

pp

roa

ch

.

No

use

of

the

ISS s

o it

wa

s n

ot

po

ssib

le

to a

sse

ss p

rob

ab

ility

of

surv

iva

l w

ith

refe

ren

ce

to

a s

uita

ble

po

pu

latio

n

suc

h a

s M

TOS.

�

Imp

erf

ec

t va

lidity o

f A

PA

CH

E,

TISS a

nd

ISS m

ay h

ave

imp

lica

tio

ns

reg

ard

ing

ba

lan

ce

of

gro

up

s a

t b

ase

line

.

�

Au

tho

rs d

oc

um

en

ted

, in

re

latio

n t

o

ye

ar

2 o

f th

e s

tud

y, th

at

ph

ysi

cia

ns

we

re in

vo

lve

d in

tra

nsf

err

ing

a s

ma

ller

nu

mb

er

of

pa

tie

nts

with

ve

ry

co

mp

lica

ted

inju

rie

s, t

he

refo

re t

he

y

exp

ec

ted

mo

rta

lity t

o b

e h

igh

er

in t

his

gro

up

. Th

ere

wa

s n

o a

na

lysi

s a

dju

stin

g

for

the

va

ria

tio

n in

se

ve

rity

.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Un

cle

ar

ho

w m

an

y r

ec

eiv

ing

ho

spita

ls

we

re in

clu

de

d –

if t

he

re w

as

mo

re

tha

n o

ne

it m

ay h

ave

re

sulte

d in

va

riatio

n in

qu

alit

y o

f c

are

at

diffe

ren

t

ho

spita

ls, re

sultin

g in

diffe

ren

t

ou

tco

me

s.

�

Str

atifie

d a

na

lysi

s b

y c

linic

al s

ub

gro

up

an

d p

atie

nt

orig

in s

ub

gro

up

like

ly h

ad

low

stu

dy p

ow

er.


32

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bu

rne

y e

t a

l. 19

95)

USA

co

ntin

ue

d

A

PA

CH

E2 s

co

re

Ph

ysi

cia

n g

rou

p:

12.0

Nu

rse

gro

up

: 1

1.9

Clin

ica

l ca

teg

ory

(%

)

Ca

rdia

c

Ph

ysi

cia

n g

rou

p:

20

Nu

rse

gro

up

: 1

6

Tra

um

a

Ph

ysi

cia

n g

rou

p:

41

Nu

rse

gro

up

: 4

5

Oth

er

Ph

ysi

cia

n g

rou

p:

39

Nu

rse

gro

up

: 3

9

RTS

1

Ph

ysi

cia

n g

rou

p:

10.9

Nu

rse

gro

up

: 1

0.5

RTS

2

Ph

ysi

cia

n g

rou

p:

10.8

Nu

rse

gro

up

: 1

0.5

RTS

3

Ph

ysi

cia

n g

rou

p:

11.5

Nu

rse

gro

up

: 1

1.0

TISS1

Ph

ysi

cia

n g

rou

p:

19.5

Nu

rse

gro

up

: 1

9.0

TISS2

Ph

ysi

cia

n g

rou

p:

17.4

Nu

rse

gro

up

: 1

6.5

Ou

tco

me

me

asu

res

Ho

spita

l mo

rta

lity

ICU

len

gth

of

sta

y

Ho

spita

l le

ng

th o

f st

ay

An

aly

sis

Pa

tie

nts

str

atifie

d in

to t

hre

e g

rou

ps:

ca

rdia

c, tr

au

ma

an

d o

the

r. O

rig

ins

of

the

tra

nsp

ort

an

d t

ran

sfe

r tim

es

we

re in

clu

de

d in

th

e a

na

lysi

s.

Da

ta a

na

lyse

d u

sin

g S

YSTA

T.

Co

ntin

uo

us

va

ria

ble

s w

ere

an

aly

sed

usi

ng

Stu

de

nt’

s t

test

an

d

ca

teg

oric

al d

ata

usi

ng

th

e χ

2 t

est

.

�

No

su

bg

rou

p a

na

lyse

s p

rese

nte

d b

y

diffe

ren

t m

ark

ers

of

inju

ry s

eve

rity

.

�

Sig

nific

an

t d

iffe

ren

ce

s in

ba

selin

e

me

asu

res

in t

he

tim

e p

erio

d 1

990

-

1991:

Ph

ysi

cia

n g

rou

p o

f p

atie

nts

we

re

old

er,

mo

re li

ke

ly t

o b

e f

rom

ho

spita

l

or

sce

ne

, a

nd

ha

d h

igh

er

RTS

3 s

co

re.

Du

rin

g s

ec

on

d p

erio

d (

1991

-19

92

),

ph

ysi

cia

n g

rou

p h

ad

hig

he

r TI

SS1 a

nd

TISS2 s

co

res.

�

No

ove

rall

resu

lts

pre

sen

ted

fo

r

du

ratio

n o

f st

ay (

all

resu

lts

stra

tifie

d b

y

the

tw

o s

tud

y y

ea

rs).

�

If t

he

stu

dy w

as

to h

ave

80%

po

we

r to

de

tec

t a

5%

diffe

ren

ce

in m

ort

alit

y a

t

the

5%

sig

nific

an

ce

leve

l,

ap

pro

xim

ate

ly 2

,00

0 p

atie

nts

wo

uld

ha

ve

be

en

re

qu

ired

.

Co

mm

en

ts

�

Co

mp

lete

ou

tco

me

da

ta o

bta

ine

d

on

1,1

69

of

1,1

70 e

ligib

le p

atie

nts

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Two

ye

ars

of

de

taile

d p

rosp

ec

tive

me

asu

rem

en

t o

f a

ir m

ed

ica

l pa

tie

nt

ch

ara

cte

ristic

s a

nd

ou

tco

me

s c

on

firm

ed

the

in

itia

l fin

din

g t

ha

t n

o s

ign

ific

an

t

diffe

ren

ce

s in

clin

ica

l ou

tco

me

s c

ou

ld b

e

ide

ntifie

d b

etw

ee

n p

atie

nts

ma

na

ge

d b

y

P/N

ve

rsu

s N

/N c

rew

s.


33

Ta

ble

7

Ev

iden

ce t

ab

les

of

stu

die

s co

mp

arin

g t

he

ou

tco

me

or

pa

tien

ts t

ran

spo

rted

by

hel

ico

pte

r w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

boa

rd (

con

tin

ued

)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bu

rne

y e

t a

l. 19

95)

USA

co

ntin

ue

d

TI

SS3

Ph

ysi

cia

n g

rou

p:

29.7

Nu

rse

gro

up

: 2

8.5

TISS4

Ph

ysi

cia

n g

rou

p:

32.7

Nu

rse

gro

up

: 3

2.0


34

Summary and Conclusions

Results were conflicting among the seven studies eligible for the review examining the effectiveness of including a doctor on board helicopter transportation. Some studies suggested there was some benefit to including doctors on board helicopters (Baxt and Moody 1987; Garner et al. 1999; Schmidt et al. 1992) while others found no difference in outcomes of interest between study groups (Burney et al. 1992; Cameron et al. 2005; Hamman et al. 1991). One study found no difference in mortality but found a significantly shorter hospital stay among the non-doctor group (Burney et al. 1995). It should be noted that the two papers by Burney et al are of marginal relevance to the use of air ambulances in New Zealand as the great preponderance of patients in these studies were inter-hospital transfers rather than transports directly from the scene. Each study had limitations that might help to explain the discrepant results.

There was variation in study design among the three studies that suggested some benefit from the inclusion of doctors on the helicopter. One made use of registry data (Schmidt et al. 1992), one used a pseudorandomised controlled trial design (Baxt and Moody 1987), and the other made use of a retrospective design (Garner et al. 1999). In relation to the pseudorandomised controlled trial, a particular consideration was the use of limiting the mortality outcome to deaths that were thought to be directly due to the trauma or complications of the trauma. This may have produced outcome misclassification with the potential for underestimating mortality in either group. The use of all-cause mortality as an outcome would have avoided this limitation. Nevertheless this was the strongest study in this section of the review. There were also limitations to the registry based study but the effect of these limitations on the study estimates was not clear. Firstly, a significant proportion of eligible patients were excluded from the doctor group due to missing charts. There were no deaths among these patients, thus the level of reduced mortality in the doctor group may have been underestimated. However, time to arrival on the scene was shorter in the doctor group. This may have resulted in an improved prognosis in the doctor group and may not be replicated in other settings where time to arrival of the doctor group may be delayed. Perhaps most significantly, the two groups (doctor and no doctor) were located in two different countries so there may have been other reasons that explain differences in outcomes other that the personnel supplying pre-hospital care. For example, level of hospital care may vary between the two settings. In the retrospective study (Garner et al. 1999), seven deaths in the non-doctor group were omitted due to missing case sheets, potentially underestimating the effectiveness of care provided by doctor crewed helicopters. Patients were also directed to different hospitals in this study depending on the helicopter crew mix. Based on the above limitations there is uncertainty about the robustness of the findings in these three studies.

There were similar limitations in the three studies that did not find any difference in outcome between the doctor and no doctor groups. Hamman et al. (1991) used a before and after design which is associated with low level evidence (level III-3). The study was underpowered to compare mortality across study groups, potentially explaining the lack of difference between groups. Burney et al. (1992) published a retrospective study that examined patient transfers to a base hospital. Unfortunately only 5.5% of these transfers were directly from the scene thus severely limiting the relevance of the study to this review. It should also be noted that the non-doctor group appeared to carry a more stable group of patients potentially underestimating any benefit from doctor involvement in helicopter transportation. Cameron et al. (2005) also published a retrospective chart review. This study was limited by the inclusion of limited injury severity data and low study power to detect a difference in mortality. Burney et al. also published a prospective study following on from their retrospective study and found no difference in mortality between doctor and no doctor groups but did find a reduced period of hospitalisation in the no doctor group. However, the authors noted that this wasn’t surprising given the less severe case mix in the no doctor group. Again this study was of limited applicability given the high proportion of inter-hospital transfers.

While overall there was more support for the inclusion of doctors on board helicopters there is a significant level of uncertainty across the literature examining this question. There certainly was not sufficient evidence to suggest that doctors should not be included on board helicopters. Other issues should also be noted:

1. The literature examined in this section does not answer the question whether wider training in procedures currently accessible to doctors would have an impact on patient outcome. For example, others have suggested that task specific crewing should be


35

adhered to and noted that the “utilisation of experienced critical care physicians, nurses or paramedics with enhanced skills, including rapid sequence intubation, must be entertained” (Rashford and Myers 2004). Others have documented the use of rapid sequence intubation by non-doctor personnel (Bernard et al. 2002; Bernard 2006; Sloane et al. 2000). However, such an approach does not consider the role of assessment in determining outcome. Hamman et al. (1991) commented in their conclusions in their study included in this review that “it appears that experienced nurses and paramedics, operating with well established protocols, can provide aggressive care that yields equal outcome results compared with those of a flight team that includes a physician. However, their study appeared to be underpowered to establish such equivalence.

2. Whether there may be variation in outcome across crew mixes for different clinical scenarios. For example, the inclusion of doctors on the helicopters tended to be associated with longer at scene times. This longer period of stabilization could be associated with improved outcome in some circumstances but not others. Cameron commented that advanced life support skills are more likely to be beneficial in the blunt trauma patient (Cameron 1999). Two of the three studies with reduced mortality among the doctor treated group were restricted to patients with blunt trauma.

3. The literature identified related to trauma only. There was no literature identified that was eligible for inclusion and examined the effect of different crew configurations in medical emergencies.

4. There was insufficient information to compare outcomes in paediatric and adult age groups or to stratify results by ISS score.

5. There was variation in the level of experience across both study teams in the selected studies.

Further research would be useful. Firstly, a well designed prospective study that follows clear and well documented dispatch criteria and provides for carriage to the same hospital (that is equipped to manage all patients). The study would need to have an adequate sample size to ensure meaningful results could be obtained. Appropriate control for injury severity and other factors associated with mortality would be required (a randomised controlled trial would be the ideal method of doing this). The study would need to encompass the study population for which any proposed helicopter services would provide coverage. Secondly, a study evaluating the role of training non-doctor crews to perform certain procedures such as rapid sequence intubation should be conducted. The specific comparison of interest would be doctors versus fully trained non-doctor personnel. The study should be adequately powered to detect equivalence in outcome.

Doctor versus no doctor on board road ambulances

From the search strategy for question two (outcome by time from ambulance callout to emergency department delivery) we identified, 516 potentially relevant articles/abstracts of which 34 were retrieved. Of these retrieved articles, 30 were excluded. These excluded papers are presented in Appendix 4. Reasons for exclusion of studies before retrieval in full text are outlined in Table 8. Reasons for exclusion of studies retrieved in full text are detailed in Table 9.


36

Table 8 Reasons for exclusion of studies before retrieval in full text: doctor versus no doctor

on board road ambulances











Neonatal study 2

Duplicate reference 1

Total 482

Table 9 Reasons for exclusion of studies retrieved in full text: doctor versus no doctor on

board road ambulances











Neonatal study 0

Total 30

Four retrieved articles were appraised and are listed in the references and Appendix 7. Included papers are presented in the evidence table below. Three studies were level III-3 (before and after studies) and the other was level III-2 (cohort study) according to NHMRC’s hierarchy of evidence.

Frandsen et al. 1991

This study (level III-3 evidence) compared three different emergency medical service configurations over three different time periods in the city of Odense, Denmark. The three configurations were:

1. Advanced EMS (n=85): included a tiered response service incorporating specifically trained doctors.

2. Intermediate EMS (n=160): included a tiered response service without doctors.

3. Basic EMS (n=148): a service without doctors.

The focus was on out of hospital cardiac arrest. The study measured mortality and cerebral status (amongst the survivors). Non-adjusted statistical analyses were conducted.

Survival between the three groups was:

• Advanced EMS (with doctors): 13% (95% CI 7-22)

• Intermediate EMS (without doctors): 1% (95% CI 0-6)

• Basic EMS (without doctors): 5% (95% CI 2-10)

• P<0.001.

However, there were significant limitations to this study:


37

� it was not possible to be certain that any difference in outcome was due to the different crew configurations

� the study was highly susceptible to confounding and the analyses have not adjusted for potential confounding factors

� the advanced EMS group treated a younger group of patients, which could explain the improved outcome in this group

� there was missing data.

These factors make it difficult to interpret the results of this study.

Koefoed-Nielsen et al. 2002

This before and after study (level III-3 evidence) compared 28 day mortality in acute myocardial infarction (MI) patients over two time periods:

1. A period preceding introduction of a mobile emergency care unit (MECU), (September to November 1996).

2. A period with a MECU (plus standard ambulances) that included anaesthetist staffing (September to November 1997).

There were 54 patients in each period.

The crude mortality rate was higher in the pre-MECU period (20.6% versus 11.1%). Multivariate regression (controlling for age, gender, pulse and systolic blood pressure) found a significantly lower odds of 28 day mortality in the time period that included a MECU (OR 0.3, P < 0.025). Forty-four percent of patients in the second time period were treated by the MECU. Another multivariate model also estimated lower odds of 28 day mortality in the group treated by MECU than the non-MECU group (OR 0.2, P < 0.05).

As with all before and after studies, there were limitations:

� it was not possible to be certain that any difference in outcome was due to the different crew configurations. The authors noted the increased proportion of patients undergoing angioplasty may explain the results.

� potential selection bias, with 25 people being excluded on the basis of insufficient information about the diagnosis.

� the multivariate models controlled for a limited range of potential confounders. Most importantly, the estimated odds ratio comparing MECU with non-MECU patients may underestimate the effectiveness of MECU due to the selection of more severely unwell patients for the MECU service.

� it was unclear if the differences in outcome would be maintained if non-anaesthetist groups were trained in further procedures.

Lee et al. 2003

A retrospective cohort study (level III-2 evidence) was conducted in Australia (Lee et al. 2003). This study compared the outcome across different levels of ambulance officer and physicians. There were three levels of ambulance officer (two classified as providing basic life support and the third providing advanced life support). Details of the procedures available to these groups are provided in Table 10. Logistic regression was used to control for confounding with the following predictor variables being included in initial models: level of pre-hospital care, time from injury to arrival in hospital, type of injury, mechanism of injury, age, sex, ISS, GCS and systolic blood pressure.

The effect of pre-hospital care on mortality was dependent on level of ICU care. Key results in the group that did not receive ICU care were (using basic life support as the reference group):

� level 5 ambulance: OR 2.18 (95% CI, 1.05-4.55)

� physician: OR 4.27 (95% CI 1.46-12.45).


38

The majority of these deaths occurred within 24 hours of admission, which the authors suggested was on the basis of not surviving initial resuscitation.

Key results in the group that did receive ICU care were (using basic life support as the reference group):

� level 5 ambulance: OR 0.70 (95% CI 0.53-1.18)

� physician: OR 0.63 (95% CI 0.28-1.39)


� the study used a retrospective design

� a selective dispatch strategy was used which probably explains the increased mortality rates in the level 5 ambulance group and the physician group in the non-ICU population

� paramedics may have involved the physician group when patient death was imminent

� the observational study design is susceptible to residual confounding.

Christenszen et al. 2003

Christenszen et al. (2003) examined the effect of introducing a mobile emergency care unit (MECU) in a before and after study (level III-3 evidence) set in Denmark. Two time periods were studied: in the first a consecutive sample of ambulance users was studied and in the second a consecutive sample of ambulance or MECU users was sampled. Twenty-eight percent of the second period sample used the MECU. There were 5,819 users overall. There was no significant difference in 180 day mortality between the two time periods although the mortality rate was significantly higher in the MECU group than the non-MECU group in the second study period. However, this was an unadjusted analysis and most notably did not control for injury severity.

There were significant unanswered questions in this study due to the nature of the study design. Most notably, due to the lack of control over injury severity it was not possible to form any conclusions about the effectiveness of MECU in reducing mortality. Other limitations included:

� poor control over confounding

� lack of statistical detail in some analyses

� a difference in outcome may have been noted if a higher proportion of users had been attended by a MECU in the second study period.

More detail is provided in Table 10.


39

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fra

nd

sen

et

al.

19

91

)

De

nm

ark

Be

fore

an

d

aft

er

Leve

l III-

3.

Stu

dy s

ett

ing

.

Se

t in

Od

en

se (

a c

ity w

ith

a p

op

ula

tio

n

of

23

8,0

00 in

De

nm

ark

). T

hre

e d

iffe

ren

t

tim

e p

erio

ds

we

re s

tud

ied

. O

ut

of

ho

spita

l ca

rdia

c a

rre

sts

we

re t

he

fo

cu

s

of

the

stu

dy.

Pa

rtic

ipa

nts

:

Inte

rve

ntio

n (

n=

85

).

Six

mo

nth

pe

rio

d o

f a

dva

nc

ed

EM

S w

ith

spe

cific

ally

tra

ine

d d

oc

tors

pro

vid

ing

ad

va

nc

ed

life

su

pp

ort

at

arr

iva

l in

a

tie

red

re

spo

nse

syst

em

.

Co

mp

ara

tor.

Two

tim

e p

erio

ds:

on

e p

rovid

ed

ba

sic

EM

S (

n=

16

0)

an

d a

n “

inte

rme

dia

te E

MS

pe

riod

” (n

=1

48)

wh

ich

inv

olv

ed

a t

iere

d

resp

on

se. N

eith

er

of

the

se p

erio

ds

use

d

do

cto

rs in

th

e p

re-h

osp

ita

l ca

re.

Ba

selin

e a

na

lyse

s

Me

an

ag

e: 65 y

ea

rs

Ma

les:

64%

Byst

an

de

r C

PR

: 1

4%

Co

llap

se t

ime

> 6

min

ute

s: 66

%

Inc

lu/e

xcl c

rite

ria

.

Ass

um

ed

to

be

all

tra

nsp

ort

s in

th

e

thre

e r

ele

va

nt

tim

e p

erio

ds

wh

o

suff

ere

d a

n o

ut

of

ho

spita

l ca

rdia

c

arr

est

.

Ou

tco

me

me

asu

res

Me

asu

red

mo

rta

lity a

nd

ce

reb

ral

sta

tus

An

aly

sis

χ2 t

est

, M

an

n-W

hitn

ey U

te

st a

nd

Kru

ska

l-W

alli

s a

na

lysi

s o

f va

rian

ce

.

Sig

nific

an

ce

leve

l se

t a

t 0

.05.

Su

rviv

al ra

te (

95%

CI)

Ba

sic

EM

S (

no

n-d

oc

tor)

: 5%

(2-1

0)

Inte

rme

dia

te E

MS (

no

n-d

oc

tor)

: 1%

(0-6

)

Ad

va

nc

ed

EM

S (

Do

cto

r): 1

3%

(7-2

2)

P<

0.0

01

Lim

ita

tio

ns

�

Low

stu

dy p

ow

er.

�

No

co

ntr

ol o

f c

on

fou

nd

ing

in t

he

an

aly

sis

wh

ich

is p

art

icu

larly

pro

ble

ma

tic

in t

his

be

fore

an

d a

fte

r

de

sig

n.

�

The

me

an

ag

e w

as

yo

un

ge

r in

th

e

ad

va

nc

ed

EM

S p

ha

se –

wh

ich

ma

y

exp

lain

th

e im

pro

ve

d s

urv

iva

l in

th

is

gro

up

.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

Da

ta m

issi

ng

fo

r so

me

ba

ckg

rou

nd

va

riab

les.

Fo

r e

xam

ple

, c

olla

pse

tim

e

do

cu

me

nte

d in

85

%. R

esp

on

se t

ime

an

d V

F ra

tio

at

arr

iva

l of

am

bu

lan

ce

sta

ff w

as

no

t d

oc

um

en

ted

in

th

e

ba

sic

EM

S g

rou

p (

41%

of

the

sa

mp

le).

�

Un

cle

ar

wh

at

the

P v

alu

e r

efe

rre

d t

o

in c

om

pa

rin

g s

urv

iva

l – w

as

it a

te

st f

or

tre

nd

or

a c

om

pa

riso

n o

f c

ert

ain

EM

S

ca

teg

orie

s.

Co

mm

en

ts

�

Aim

wa

s to

eva

lua

te (

1)

if m

ore

inte

nsi

ve

EM

S c

an

inc

rea

se s

urv

iva

l

aft

er

ou

t o

f h

osp

ita

l ca

rdia

c a

rre

st

an

d (

2)

if m

ore

inte

nsi

ve

EM

S c

an

dim

inis

h c

ere

bra

l da

ma

ge

aft

er

ou

t o

f

ho

spita

l ca

rdia

c a

rre

st.

�

Cle

ar

do

cu

me

nta

tio

n o

f d

efin

itio

ns

ap

plie

d in

th

e s

tud

y a

nd

th

e u

se o

f

the

se d

efin

itio

ns

me

an

s

mis

cla

ssific

atio

n o

f e

xpo

sure

is h

igh

ly

un

like

ly.


40

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fra

nd

sen

et

al.

19

91

)

De

nm

ark

co

ntin

ue

d

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

re

sults

of

the

in

ve

stig

atio

n d

em

on

stra

te

tha

t th

e m

ore

inte

nsi

ve

th

e p

re-h

osp

ita

l

tre

atm

en

t o

f o

ut-

of-

ho

spita

l ca

rdia

c a

rre

st,

the

mo

re p

atie

nts

su

rviv

e a

nd

th

e m

ore

pa

tie

nts

su

rviv

e w

ith

go

od

ce

reb

ral

fun

ctio

n.


41

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ko

efo

ed

-Nie

lse

n e

t a

l.

2002

)

De

nm

ark

Be

fore

an

d

aft

er

Leve

l III-

3.

Stu

dy s

ett

ing

.

A m

ob

ile e

me

rge

nc

y c

are

un

it (

MEC

U),

wh

ich

inc

lud

ed

an

an

ae

sth

etist

, w

as

set

up

in a

re

gio

n o

f D

en

ma

rk.

Co

nse

cu

tive

pa

tie

nts

with

ac

ute

MI w

ere

stu

die

d in

two

tim

e p

erio

ds

:

1. Se

pt

to N

ov 1

99

6 (

be

fore

MEC

U)

2. Se

pt

to N

ov 1

99

7 (

with

MEC

U).

The

are

a c

ove

red

ap

pro

xim

ate

d

800km

2. M

ed

ian

re

spo

nse

tim

e f

or

MEC

U

wa

s 8.9

min

ute

s a

nd

6.5

min

ute

s fo

r

sta

nd

ard

am

bu

lan

ce

s.

Pa

rtic

ipa

nts

:

Inte

rve

ntio

n (

n=

54

)

An

ae

sth

etist

pre

sen

t (w

ith

MEC

U)

MEC

U w

as

als

o c

rew

ed

with

a s

pe

cia

lly

tra

ine

d e

me

rge

nc

y t

ec

hn

icia

n.

Ad

ditio

na

l tre

atm

en

t p

rovid

ed

by

an

ae

sth

etist

: tr

ac

he

al i

ntu

ba

tio

n,

tra

nsc

uta

ne

ou

s p

ac

ing

, a

na

est

he

tic

s

(hyp

no

tic

s a

nd

mu

scle

re

laxa

nts

),

op

ioid

s, h

ea

rt s

tim

ula

tin

g d

rug

s a

nd

an

tia

rrh

yth

mic

s.

Pre

-ho

spita

l

thro

mb

oly

sis

wa

s n

ot

pro

vid

ed

.

Co

mp

ara

tor

(n=

54

)

No

n p

hysi

cia

n c

rew

(b

efo

re M

EC

U)

Inc

lu/e

xcl c

rite

ria

.

Use

rs o

f e

me

rge

nc

y s

erv

ice

s (b

ase

d

on

am

bu

lan

ce

co

mp

an

y’s

pa

tie

nt

rolls

) w

ho

we

re c

lass

ifie

d a

s a

dia

gn

osi

s o

f M

I (I

CD

-10 c

od

es

I21-

I22)

on

ad

mis

sio

n t

o a

n e

me

rge

nc

y

de

pa

rtm

en

t o

r c

oro

na

ry c

are

un

it in

on

e o

f th

ree

ho

spita

ls in

Aa

rhu

s,

De

nm

ark

.

The

pa

tie

nt

wa

s a

live

wh

en

rea

ch

ing

ho

spita

l.

The

ho

spita

l re

co

rd c

on

firm

ed

MI

dia

gn

osi

s, b

ase

d o

n e

nzy

me

te

sts

an

d E

CG

.

The

ho

spita

l re

co

rd p

rovid

ed

suff

icie

nt

info

rma

tio

n o

n d

iag

no

sis

an

d t

rea

tme

nt.

Ou

tco

me

me

asu

res

28 d

ay m

ort

alit

y

An

aly

sis

Pe

ars

on

’s χ

2, Fis

he

r’s

exa

ct

test

,

Sp

ea

rma

n’s

no

n-p

ara

me

tric

ra

nk

co

rre

latio

n, K

ap

lan

-Me

ier

surv

iva

l

an

aly

sis

an

d m

ultip

le lo

gis

tic

reg

ress

ion

we

re c

on

du

cte

d.

Cru

de

mo

rta

lity r

ate

at

28 d

ays,

by

stu

dy p

erio

d.

Pre

-MEC

U:

20.6

%

MEC

U: 11

.1%

Ad

just

ed

od

ds

ratio

by s

tud

y p

erio

d

(28 d

ay m

ort

alit

y),

(p

re-M

EC

U a

s

the

re

fere

nc

e)

OR

0.3

(P

<0.0

25

)

Ad

just

ed

fo

r a

ge

, g

en

de

r, p

uls

e

an

d s

yst

olic

blo

od

pre

ssu

re.

Ad

just

ed

od

ds

ratio

by u

se o

f M

EC

U

(28 d

ay m

ort

alit

y),

(p

re-M

EC

U a

s

the

re

fere

nc

e)

OR

0.2

(P

<0.0

5)

Ad

just

ed

fo

r a

ge

, g

en

de

r, p

uls

e

an

d s

yst

olic

blo

od

pre

ssu

re.

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

Exc

lusi

on

s: s

ix w

ere

eith

er

de

ad

on

arr

iva

l or

die

d im

me

dia

tely

aft

er

arr

iva

l (fo

ur

pre

MEC

U a

nd

tw

o w

ith

MEC

U),

sp

ec

ific

MI d

iag

no

sis

co

uld

no

t b

e c

on

firm

ed

in

25

. U

nc

lea

r if a

ny

we

re e

xclu

de

d d

ue

to

in

suff

icie

nt

info

rma

tio

n o

n t

rea

tme

nt.

�

Lim

ite

d d

ata

pro

vid

ed

at

tim

e o

f firs

t

att

en

da

nc

e o

f p

re-h

osp

ita

l ca

re.

�

Be

fore

an

d a

fte

r st

ud

y is

a lo

w q

ua

lity

de

sig

n –

fe

atu

res

oth

er

tha

n t

he

inte

rve

ntio

n o

f in

tere

st m

ay h

ave

ch

an

ge

d s

o it

is u

nc

lea

r to

wh

at

ext

en

t a

ny c

ha

ng

e in

ou

tco

me

is a

resu

lt o

f th

e in

tro

du

ctio

n o

f M

EC

U. In

this

stu

dy, th

e im

pro

ve

d p

rog

no

sis

ma

y h

ave

be

en

du

e t

o a

ng

iop

last

y

rath

er

tha

n M

EC

U.

�

Ba

sis

of

dia

gn

osi

s o

f M

I n

ot

pre

cis

ely

de

fin

ed

: p

ote

ntia

l fo

r in

ap

pro

pria

te

sele

ctio

n.

�

Po

ten

tia

l fo

r m

isc

lass

ific

atio

n o

f

ou

tco

me

, th

ou

gh

th

e s

ize

an

d

dire

ctio

n o

f su

ch

mis

cla

ssific

atio

n is

diffic

ult t

o d

ete

rmin

e.

�

Du

rin

g t

he

MEC

U p

erio

d,

som

e

pa

tie

nts

we

re n

ot

att

en

de

d b

y M

EC

U

(56%

). H

ow

eve

r, a

lth

ou

gh

MEC

U

pa

tie

nts

we

re m

ore

like

ly t

o b

e

co

nsi

de

red

as

seve

re,

MEC

U p

atie

nts

ha

d h

igh

er

ad

just

ed

od

ds

of

surv

iva

l

co

mp

are

d w

ith

no

n-M

EC

U p

atie

nts

.


42

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ko

efo

ed

-Nie

lse

n e

t a

l.

2002

)

De

nm

ark

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

Ag

e ≤

69 y

ea

rs (

%)

Pre

MEC

U: 4

1

MEC

U: 41

Ma

le s

ex

(%)

Pre

MEC

U: 6

3

MEC

U: 72

PTC

A p

erf

orm

ed

(%

)

Pre

MEC

U: 1

9

MEC

U: 26

Thro

mb

oly

sis

giv

en

(%

)

Pre

MEC

U: 2

8

MEC

U: 33

Co

mm

en

ts

�

Aim

ed

to

ass

ess

th

e im

pa

ct

of

a

MEC

U o

n s

urv

iva

l am

on

g p

atie

nts

with

ac

ute

MI.

�

MEC

U p

atie

nts

we

re m

ore

lik

ely

to

rec

eiv

e t

hro

mb

oly

sis.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

In t

he

pre

sen

t st

ud

y, M

I p

atie

nts

tre

ate

d in

a

MEC

U s

taff

ed

by a

n a

na

est

he

tist

an

d/o

r

ha

vin

g a

ng

iop

last

y w

as

fou

nd

to

be

ass

oc

iate

d w

ith

a r

ed

uc

ed

mo

rta

lity. Th

ese

ob

serv

atio

ns

ha

ve

be

en

ba

sed

on

qu

asi

-

exp

erim

en

tal r

ath

er

tha

n r

an

do

mis

ed

exp

erim

en

tal d

ata

, a

nd

ra

nd

om

ise

d d

ata

wo

uld

be

hig

hly

de

sira

ble

.


43

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

e e

t a

l. 2

003

)

Au

stra

lia

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Se

ve

re b

lun

t tr

au

ma

pa

tie

nts

ad

mitte

d

to W

est

me

ad

Ho

spita

l, Syd

ne

y b

etw

ee

n

July

1986 a

nd

De

ce

mb

er

20

00.

Am

bu

lan

ce

off

ice

rs t

rain

ed

to

th

ree

diffe

ren

t le

ve

ls: tw

o le

ve

ls o

f b

asi

c li

fe

sup

po

rt a

nd

on

e le

ve

l of

ad

va

nc

ed

life

sup

po

rt.

Als

o h

ad

tw

o p

hysi

cia

n-s

taff

ed

em

erg

en

cy m

ed

ica

l se

rvic

es

wh

ich

resp

on

d t

o a

cc

ide

nt

sce

ne

s e

ith

er

by

roa

d o

r h

elic

op

ter.

Leve

l 3 a

mb

ula

nc

e o

ffic

er:

ba

sic

life

sup

po

rt w

ith

ou

t e

xte

rna

l co

ntr

ol o

f

ha

em

orr

ha

ge

, sp

lintin

g,

no

n-in

va

siv

e

airw

ay m

an

oe

uvre

s a

nd

ba

g-v

alv

e-

ma

sk v

en

tila

tio

n.

Leve

l 4 a

mb

ula

nc

e o

ffic

er:

ba

sic

life

sup

po

rt p

lus

intr

ave

no

us

ca

nn

ula

tio

n

an

d a

dm

inis

tra

tio

n o

f IV

flu

ids,

in

ad

ditio

n t

o a

lim

ite

d r

an

ge

of

IV

me

dic

atio

ns

an

d n

ee

dle

th

ora

ce

nte

sis.

Leve

l 5 a

mb

ula

nc

e o

ffic

er:

Ad

va

nc

ed

life

su

pp

ort

. A

ble

to

pe

rfo

rm a

ll th

e

ab

ove

pro

ce

du

res

plu

s o

ral

en

do

tra

ch

ea

l in

tub

atio

n.

Ac

ce

ss t

o a

wid

er

ran

ge

of

IV m

ed

ica

tio

ns

(bu

t n

ot

ne

uro

mu

scu

lar

blo

cka

de

, a

na

est

he

tic

ag

en

ts o

r se

da

tive

ag

en

ts t

o f

ac

ilita

te

intu

ba

tio

n).

Ph

ysi

cia

ns

are

fre

e t

o e

xerc

ise

th

eir

clin

ica

l ju

dg

em

en

t in

ea

ch

ca

se.

Inc

lu/e

xcl c

rite

ria

.

Blu

nt

tra

um

a p

atie

nts

with

ISS >

15.

Ou

tco

me

me

asu

res

Mo

rta

lity d

urin

g h

osp

ita

l ad

mis

sio

n

An

aly

sis

Log

istic

re

gre

ssio

n u

sed

with

th

e

follo

win

g p

red

icto

r va

ria

ble

s: le

ve

l

of

pre

-ho

spita

l ca

re,

tim

e f

rom

inju

ry

to a

rriv

al i

n h

osp

ita

l, ty

pe

of

inju

ry,

me

ch

an

ism

of

inju

ry, a

ge

, se

x, ISS,

GC

S a

nd

syst

olic

blo

od

pre

ssu

re.

Mo

de

l ca

libra

tio

n a

sse

sse

d b

y t

he

Ho

sme

r-Le

me

sho

w g

oo

dn

ess

-of-

fit

χ2 t

est

an

d p

red

ictive

ac

cu

rac

y

ass

ess

ed

by t

he

are

a u

nd

er

the

rec

eiv

er

op

era

tin

g c

ha

rac

terist

ic

cu

rve

.

Ris

k in

od

ds

ratio

of

mo

rta

lity b

y

leve

l of

pre

-ho

spita

l ca

re a

nd

IC

U

tre

atm

en

t (b

asi

c li

fe s

up

po

rt a

s th

e

refe

ren

ce

), (

95

% C

I)

No

IC

U a

dm

issi

on

:

Leve

l 5 a

mb

ula

nc

e:

2.1

8 (

1.0

5-4

.55)

Ph

ysi

cia

n: 4.2

7 (

1.4

6-1

2.4

5)

ICU

ad

mis

sio

n:

Leve

l 5 a

mb

ula

nc

e:

0.7

0 (

0.5

3-1

.18)

Ph

ysi

cia

n: 0.6

3 (

0.2

8-1

.39

)

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

A s

ele

ctive

dis

pa

tch

str

ate

gy w

as

use

d w

hic

h p

rob

ab

ly e

xpla

ins

the

inc

rea

sed

mo

rta

lity r

ate

s in

th

e le

ve

l 5

am

bu

lan

ce

gro

up

an

d t

he

ph

ysi

cia

n

gro

up

in

th

e n

on

-IC

U p

op

ula

tio

n.

�

Pa

ram

ed

ics

ma

y h

ave

in

vo

lve

d t

he

ph

ysi

cia

n g

rou

p w

he

n p

atie

nt

de

ath

wa

s im

min

en

t.

�

Un

cle

ar

if s

om

e p

hysi

cia

n g

rou

p

pa

tie

nts

ma

y h

ave

be

en

tra

nsp

ort

ed

by h

elic

op

ter.

�

Diffe

ren

t le

ve

ls o

f a

cc

ess

to

pro

ce

du

res

in t

he

no

n-d

oc

tor

gro

up

s

ma

y n

ot

be

ge

ne

ralis

ab

le t

o o

the

r

sett

ing

s, m

ay g

uid

e d

isp

atc

h d

ec

isio

ns

an

d m

ay r

esu

lt in

bia

s in

co

mp

ara

tiv

e

est

ima

tes.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

Co

mm

en

ts

�

Aim

ed

to

de

term

ine

th

e a

sso

cia

tio

n

be

twe

en

mo

rta

lity a

nd

leve

l of

pre

-

ho

spita

l ca

re in

se

ve

rely

inju

red

blu

nt

tra

um

a p

atie

nts

with

or

with

ou

t se

ve

re

he

ad

tra

um

a.

�

No

on

line

me

dic

al c

on

tro

l pro

vid

ed

.

�

We

ll d

esc

ribe

d a

nd

co

nd

uc

ted

sta

tist

ica

l me

tho

do

log

y.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

leve

l of

pre

-ho

spita

l ca

re w

as

ass

oc

iate

d w

ith

th

e r

isk o

f m

ort

alit

y. Th

is w

as

mo

difie

d b

y w

he

the

r th

e p

atie

nt

surv

ive

d

lon

g e

no

ug

h t

o b

e a

dm

itte

d t

o t

he

IC

U.


44

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

e e

t a

l. 2

003

)

Au

stra

lia

co

ntin

ue

d

P

art

icip

an

ts (

n=

2010

):

Inte

rve

ntio

n.

Leve

l of

pre

-ho

spita

l ca

re w

as

cla

ssifie

d

into

th

ree

gro

up

s: b

asi

c li

fe s

up

po

rt,

ad

va

nc

ed

life

su

pp

ort

an

d p

hysi

cia

n

ca

re.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ISS b

y le

ve

l of

pre

-ho

spita

l ca

re:

No

n-E

MS t

ran

spo

rt:

20

Leve

l 3 a

mb

ula

nc

e:

24

Leve

l 4 a

mb

ula

nc

e:

25

Leve

l 5 a

mb

ula

nc

e:

31

Ph

ysi

cia

n: 3

1

Me

dia

n a

ge

30 y

ea

rs (

inte

rqu

art

ile

ran

ge

21-4

9 y

ea

rs)

Ma

les:

76%

Ca

use

s:

Ro

ad

tra

ffic

ac

cid

en

t 6

7%

Falls

: 13%

Ass

au

lt 5

%.

No

he

ad

inju

rie

s: 3

6%

Iso

late

d h

ea

d in

jury

: 52%

He

ad

inju

ry w

ith

ab

do

min

al/

ch

est

inju

rie

s: 1

2%

Pre

-ho

spita

l ca

re:

No

n-E

MS t

ran

spo

rt 5

%

Leve

l 3 a

mb

ula

nc

e 2

3%

Leve

l 4 a

mb

ula

nc

e 2

%

Leve

l 5 a

mb

ula

nc

e 5

9%

Ph

ysi

cia

n E

MS 1

1%


45

Ta

ble

10

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g t

he

ou

tco

me

of

pa

tien

ts t

ran

spo

rted

by

ro

ad

am

bu

lan

ce w

ith

an

d w

ith

ou

t a

med

ica

l d

oct

or

on

bo

ard

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ch

rist

en

sze

n e

t a

l.

2003

)

De

nm

ark

Be

fore

an

d

aft

er

stu

dy

Leve

l III-

3.

Stu

dy s

ett

ing

.

Co

ve

rs a

n a

rea

of

~8

00km

2 a

nd

a

po

pu

latio

n o

f ~

330,0

00.

Mo

st o

f th

e

po

pu

latio

n is

an

urb

an

ce

ntr

e

(~2

50,0

00

). A

mo

bile

em

erg

en

cy c

are

un

it (

sta

ffe

d w

ith

an

an

ae

sth

etist

) is

dis

pa

tch

ed

fo

r th

e m

ost

se

rio

us

ca

ses.

The

mo

bile

em

erg

en

cy c

are

un

it

(MEC

U)

wa

s se

t u

p in

1997

.

Am

bu

lan

ce

cre

w in

clu

de

s b

asi

c li

fe

sup

po

rt s

kill

s, s

em

iau

tom

atic

de

fib

rilla

tio

n,

ad

min

istr

atio

n o

f n

itro

-g

lyc

erin

e s

pra

ys,

inh

ala

tio

n o

f β

-2-

an

tag

on

ists

an

d t

he

ap

plic

atio

n o

f

dia

zep

am

re

cta

lly a

nd

nitro

us

oxi

de

/oxy

ge

n.

Pa

rtic

ipa

nts

(n

=5

819

):

Co

nse

cu

tive

am

bu

lan

ce

use

rs d

urin

g 1

Se

pte

me

br-

30

No

ve

mb

er

199

6 (

be

fore

MEC

U)

an

d 1

Se

pte

mb

er-

30 N

ove

mb

er

1997 (

aft

er

MEC

U in

tro

du

ctio

n).

Inte

rve

ntio

n.

MEC

U

Co

mp

ara

tor.

No

MEC

U

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Ma

les

52%

Me

an

ag

e: Fe

ma

le 5

0.7

ye

ars

, m

ale

45.0

Co

nse

cu

tive

use

rs in

pe

rio

d 1

: 295

0

Co

nse

cu

tive

use

rs in

pe

rio

d 2

: 286

9

(in

clu

de

d 2

7.6

% a

tte

nd

ed

by

am

bu

lan

ce

an

d M

EC

U)

Inc

lu/e

xcl c

rite

ria

.

Co

nse

cu

tive

am

bu

lan

ce

use

rs in

the

pe

rio

ds

of

inte

rest

.

Wh

en

th

ere

wa

s m

ore

th

an

on

e

am

bu

lan

ce

ca

ll d

urin

g t

he

stu

dy

pe

riod

th

e u

ser

wa

s in

clu

de

d o

nly

on

th

e d

ay o

f th

e in

itia

l ca

ll.

Ou

tco

me

me

asu

res

Ho

spita

lisa

tio

n

Dia

gn

ost

ic p

att

ern

Su

rviv

al a

t 1,

7, 2

8 a

nd

180

da

ys

An

aly

sis

Mu

ltip

le lo

gis

tic

re

gre

ssio

n w

as

use

d

for

da

ta a

na

lysi

s.

180 d

ay m

ort

alit

y: c

om

pa

riso

n

be

twe

en

pe

rio

d b

efo

re a

nd

aft

er

intr

od

uc

tio

n o

f M

EC

U (

pe

rio

d

be

fore

intr

od

uc

tio

n a

s th

e

refe

ren

ce

):

OR

1.0

6 (

no

t st

atist

ica

lly s

ign

ific

an

t)

Ad

just

ed

fo

r a

ge

an

d g

en

de

r

180 d

ay m

ort

alit

y: c

om

pa

riso

n

be

twe

en

MEC

U u

se a

nd

no

n-u

se in

the

pe

rio

d a

fte

r in

tro

du

ctio

n o

f

MEC

U.

MEC

U u

sers

: 14.7

% m

ort

alit

y

No

n-M

EC

U u

sers

8.9

% m

ort

alit

y

P<

0.0

01

Lim

ita

tio

ns

�

Be

fore

an

d a

fte

r st

ud

y is

a lo

w q

ua

lity

de

sig

n –

fe

atu

res

oth

er

tha

n t

he

inte

rve

ntio

n o

f in

tere

st m

ay h

ave

ch

an

ge

d s

o it

is u

nc

lea

r to

wh

at

ext

en

t a

ny c

ha

ng

e in

ou

tco

me

is a

resu

lt o

f th

e in

tro

du

ctio

n o

f M

EC

U.

�

The

se

co

nd

pe

rio

d in

clu

de

s a

mix

of

MEC

U a

nd

am

bu

lan

ce

re

trie

va

ls.

�

No

co

ntr

ol f

or

me

asu

res

of

inju

ry

seve

rity

in c

om

pa

rin

g t

he

pe

rio

d p

re

an

d p

ost

intr

od

uc

tio

n o

f M

EC

U o

r in

a

co

mp

ariso

n o

f M

EC

U w

ith

no

n-M

EC

U

use

rs in

th

e p

erio

d a

fte

r in

tro

du

ctio

n

of

MEC

U.

�

Co

nfid

en

ce

inte

rva

ls w

ere

no

t

pre

sen

ted

fo

r th

e a

dju

ste

d

co

mp

ariso

ns

of

mo

rta

lity.

Co

mm

en

ts

�

The

stu

dy a

im w

as

(1)

to d

esc

ribe

mo

rta

lity, h

osp

ita

lisa

tio

n, a

nd

th

e

dia

gn

ost

ic p

att

ern

am

on

g

em

erg

en

cy a

mb

ula

nc

e u

sers

an

d (

2)

to e

va

lua

te t

he

im

pa

ct

of

on

e M

EC

U

sta

ffe

d b

y a

n a

na

est

he

tist

.

�

Me

dia

n r

esp

on

se t

ime

fo

r M

EC

U 8

.9

min

ute

s a

nd

6.5

min

ute

s fo

r

am

bu

lan

ce

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Aft

er

the

MEC

U f

ew

er

we

re b

rou

gh

t to

ho

spita

l. Th

e o

ve

rall

mo

rta

lity f

or

all

am

bu

lan

ce

use

rs w

as

no

t in

flu

en

ce

d b

y t

he

MEC

U. Fo

r th

e s

ub

gro

up

s, e

spe

cia

lly A

MI,

mo

rta

lity w

as

low

er

aft

er

the

intr

od

uc

tio

n o

f

the

MEC

U.


46


Results were conflicting among the four studies eligible for the review examining the effectiveness of including a doctor on board ambulance transportation. Two studies estimated an increased level of effectiveness in services with a doctor on board (Frandsen et al. 1991; Koefoed-Nielsen et al. 2002), one had variable results depending on the comparator group and the use of ICU (Lee et al. 2003) and one found a significantly worse outcome associated with incorporation of a doctor (Christenszen et al. 2003). However, it is useful to consider the results more critically. Three of the four studies used a before and after design. This is a weak study design since it is not possible to be certain that the intervention of interest has resulted in any change in outcome or if some other factor has changed over time that has resulted in the change in outcome.

Three studies were set in Denmark. Two of these studies evaluated MECUs. Both these studies used before and after designs with the later period having access to MECUs. However, only 56% and 28% of patients in this later period were actually attended by a MECU. One of these studies reported on a population of patients with acute MI (Koefoed-Nielsen et al. 2002) and the other had no such restriction (Christenszen et al. 2003). In the study reporting on MI patients there were significantly lower odds of 28 day mortality both in the period with MECU available and in the direct comparison between MECU and non-MECU patients. The latter result is pertinent since the MECU was directed towards more severe patients. However, the results in the other evaluation of MECU were different. Specifically, there was no difference in 180 day mortality between the time periods with and without MECU and the 180 day mortality was significantly higher in MECU users compared with non-users. The latter may represent the casemix with more severe cases being attended by MECU. The other Danish study compared three different approaches for the pre-hospital care of out of hospital cardiac arrest (Frandsen et al. 1991). The survival rate was higher in the group that included a doctor on board compared with two non-doctor arrangements (survival rate 13% versus 5% and 1%). However, this study did not control for potential confounders. This lack of control of confounders was problematic because, amongst other potential issues, the mean age was lower in the doctor group compared with the other groups. This difference in age could explain the difference in outcome between groups. The fourth study was set in Australia (Lee et al. 2003). The study population consisted of patients with severe blunt trauma. Three types of pre-hospital care were studied. The most basic level of life support was used as the reference category and was compared with another non-medical configuration and a configuration with a doctor on board. The results varied by whether the patients were subsequently admitted to ICU. The odds of mortality were significantly higher in the two more advanced pre-hospital configurations when the patient was not admitted to ICU. However, when the patient was admitted to ICU there were no statistically significant differences between pre-hospital care groups. It should be noted that a selective dispatch strategy was used so the more severely injured patients were seen by the more advanced ambulance group and the doctor group. The authors also commented that paramedics may have involved the physician group when patient death was imminent. These factors may explain the poorer prognosis in the two more advance pre-hospital care groups among the patients who did not proceed on to ICU care.

Similar considerations applied in the comparison of effectiveness of doctors versus non-doctor configurations on road ambulances as they did on helicopters. Specifically, the literature identified had limitations as identified above. These limitations are such that no clear conclusions can be drawn on the question of benefit from having a doctor on board road ambulances. Further considerations also apply:

1. The literature examined in this section does not answer the question whether wider training in procedures accessible to doctors would have an impact on patient outcome.

2. Whether there may be variation in outcome across crew mixes for different clinical scenarios. For example, the inclusion of doctors on road ambulances may be associated with longer at scene times. This longer period of stabilization could be associated with improved outcome in some circumstances but not others.

3. There was insufficient information to compare outcomes in paediatric and adult age groups or to stratify results by ISS score. Note the one study limited to trauma patients was restricted to patients with an ISS>15.

Like the section above examining doctors versus no doctors on board helicopters further research is required to adequately answer the question about the effectiveness of doctors versus no doctors on


47

board road ambulances. A similar approach would be useful to that proposed in the helicopter section to further elucidate this issue.

Comparison of outcomes amongst crews that do and do not perform rapid sequence

intubation and/or thoracostomy

The identification of studies for inclusion in review question three (comparison of outcomes in patients transported by crews that do and do not have the ability to perform rapid sequence intubation and/or thoracostomy) were identified from three searches. The first two searches were the same as those used in questions 1 and 2. An additional search was conducted that identified 253 additional potentially relevant articles/abstracts. Therefore, there were 1837 potentially relevant articles/abstracts. One hundred of these were retrieved. Of these retrieved articles, 95 were excluded. These excluded papers are presented in Appendix 5. Two additional articles were identified from reference lists. Both these studies were excluded: one was not relevant to the review question and the other used an incorrect comparator. Reasons for exclusion of studies before retrieval in full text are outlined in Table 11. Reasons for exclusion of studies retrieved in full text are detailed in Table 12.

Table 11 Reasons for exclusion of studies before retrieval in full text (from additional search):

patients transported by crews that do and do not have the ability to perform rapid

sequence intubation and/or thoracostomy











Neonatal study 5

Total 1737

Table 12 Reasons for exclusion of studies retrieved in full text: patients transported by crews

that do and do not have the ability to perform rapid sequence intubation and/or

thoracostomy











Neonatal study 0

No documentation of procedures by different crews 25

Patient transfer 1

Total 95

Five retrieved articles were appraised and are listed in the references and Appendix 7. Included papers are presented in the evidence table below. Included studies were all level III-3 and above according to NHMRC’s hierarchy of evidence, including a pseudorandomised controlled trial, two cohort studies and two before and after studies.


48

Baxt et al. 1987

Baxt et al. (1987) conducted a pseudorandomised controlled trial (Level III-1 evidence) comparing mortality in a group of consecutive patients with blunt trauma transported by helicopter with physician and nurse on board with another group with paramedic and nurse on board. In relation to the procedures of interest, the physician group was able to perform both needle and tube thoracostomies whereas the paramedic group was not. There were also other differences in procedures performed and these are noted in Table 13. There were 574 participants (316 in the physician group and 258 in the comparator group). Actual mortality was compared with predicted mortality using TRISS methodology. The trauma score scale rather than the revised trauma score scale was included in this calculation. Actual mortality was not statistically significantly different from predicted mortality in the paramedic group while, in the physician group, actual mortality was significantly lower than predicted (P<0.05). There was a statistically significant difference in the Z statistic between the two groups, supporting reduced mortality in the physician group compared with the paramedic group. This reduction in mortality may have been related to the difference in procedures available to the physician group.

There were potential sources of confounding and bias that should be considered when interpreting this study. Key issues included:

� the study was not truly randomised so is susceptible to confounding. However, significant baseline differences were not observed for transport time, trauma score, ISS, GCS, predicted survival or patient age

� the key difference between the groups related to the category of patients who survived but were expected to die. There were only 22 patients who were expected to die in the physician group (five survived) and 16 who were expected to die in the paramedic group (none survived)

� the TRISS methodology used in this review did not assess the degree of match in injury severity between the two groups, or adjust for different casemix in the study groups

� the study did not use all-cause mortality, resulting in the potential for misclassification of outcome.

It should also be noted that although the results supported reduced mortality in the physician group and that this was the only group that was able to perform thoracostomies it does not necessarily mean that the difference in outcome is due to the ability to perform this procedure. It is possible that if the paramedic group had also been able to perform thoracostomies there may not have been support for reducing mortality in the physician group compared with the paramedic group but again this was not the focus of the study so no conclusion can be formed on this point. For such a conclusion to be made, a study with three arms would be required: the two arms included in the present study plus another arm for a paramedic group that was able to perform thoracostomy.

Garner et al. 1999

Garner et al. (1999) reported on a retrospective study set in Australia (level III-2 evidence). They compared the outcome (mortality) between groups transported via helicopter with a physician on board versus patients transported with a paramedic on board. The physician group performed a number of rapid sequence intubations (28 of the 34 intubations in this group were conducted after muscle relaxant drugs were provided). The use of muscle relaxant drugs was beyond the paramedic protocol. In general, the physician group was treated more aggressively. All patients were transported directly from the scene to the relevant hospital (hospitals varied by the crew mix on the helicopter). There were 67 patients in the physician group and 140 in the paramedic group. The patients were restricted to those with blunt trauma and an ISS score greater than 10. Actual mortality was compared with predicted mortality using TRISS methodology (MTOS as the reference population) and the adjusted W statistic was also used to directly compare the physician and paramedic groups. The degree of match on injury severity between the MTOS population and the physician and paramedic groups was poor, therefore the adjusted W statistic was appropriately presented. When comparing the physician group with the MTOS population it was estimated that 9.48 (95% CI 3.84-15.12) extra lives per 100 population were saved in the physician group. There was no significant difference in mortality between the MTOS population and the paramedic group. Direct comparison between the paramedic and physician groups suggested 13.44 (95% CI 7.80-19.08) extra lives per 100 population were saved in the physician group.


49

There were differences in the procedures performed between the two groups so these differences may have contributed to the difference in outcome. These are detailed in Table 13.

The study had limitations:

� it was a retrospective study

� there were sources of selection bias – with differences in baseline measures between the two study groups (the physician group appeared to manage a more severely injured group) and seven patients who died were excluded from the paramedic group due to missing case sheets

� confounding was a potential problem between the two groups although use of the W statistic should have partially controlled confounding (based on injury severity)

� the degree of match in injury severity with the MTOS study was poor and there were variations in methods adopted in the MTOS study and this study (most notably related to the timing of RTS measurement) which limits the usefulness of the MTOS cohort as a reference population.

It should also be noted that although the results supported reduced mortality in the physician group and that this was the only group that performed rapid sequence intubations it does not necessarily mean that the difference in outcome is due to the ability to perform this procedure. It is possible that if the paramedic group had performed rapid sequence intubations there may not have been support for reducing mortality in the physician group compared with the paramedic group but again this was not the focus of the study so no conclusion can be formed on this point. For such a conclusion to be made, a study with three arms would be required: the two arms included in the present study plus another arm for a paramedic group that performed rapid sequence intubation.

Koefoed-Nielsen et al. 2002

This before and after study (level III-3 evidence) compared 28 day mortality in acute MI patients over two time periods:

1. A period preceding introduction of a mobile emergency care unit (MECU), (September to November 1996).

2. A period with a MECU (plus standard ambulances) that included anaesthetist staffing (September to November 1997).

There were 54 patients in each period. The anaesthetist was able to perform the following additional procedures: tracheal intubation, transcutaneous pacing, anaesthetics (hypnotics and muscle relaxants), opioids, heart stimulating drugs and antiarrhythmics. On the basis of this description, it was interpreted that the anaesthetist was able to perform rapid sequence intubation in the field.

The crude mortality rate was higher in the pre-MECU period (20.6% versus 11.1%). Multivariate regression (controlling for age, gender, pulse and systolic blood pressure) found a significantly lower odds of 28 day mortality in the time period that included a MECU (OR 0.3, P < 0.025). Forty-four percent of patients in the second time period were treated by the MECU. Another multivariate model also estimated lower odds of 28 day mortality in the group treated by MECU than the non-MECU group (OR 0.2, P < 0.05).

As with all before and after studies, there were limitations:

� it is not possible to be certain that any difference in outcome is due to the different crew configurations. The authors noted the increased proportion of patients undergoing angioplasty may explain the results.

� potential selection bias, with 25 people being excluded on the basis of insufficient information about the diagnosis.

� the multivariate models controlled for a limited range of potential confounders. Most importantly, the estimated odds ratio comparing MECU with non-MECU patients may underestimate the effectiveness of MECU due to the selection of more severely unwell patients for this service.

� it was unclear if the differences in outcome would be maintained if non-anaesthetist groups were trained in further procedures.


50

It should also be noted that although the results supported reduced mortality in the anaesthetist group (MECU) and that this was the only group that performed rapid sequence intubations it does not necessarily mean that the difference in outcome is due to the ability to perform this procedure. It is possible that if the non-anaesthetist group had performed rapid sequence intubations there may not have been support for reducing mortality in the anaesthetist group compared with the non-anaesthetist group but again this was not the focus of the study so no conclusion can be formed on this point. For such a conclusion to be made, a study with three arms would be required: the two arms included in the present study plus another arm for a non-anaesthetist group that performed rapid sequence intubation.

Lee et al. 2003

A retrospective cohort study (level III-2 evidence) was conducted in Australia (Lee et al. 2003). This study compared the outcome across different levels of ambulance officer and physicians. There were three levels of ambulance officer (two classified as providing basic life support and the third providing advanced life support). Details of the procedures available to these groups are provided in Table 10. Most notably, physicians were able to perform any procedure they considered warranted, whereas other groups were not able to perform rapid sequence intubation or tube thoracocentesis. Logistic regression was used to control for confounding with the following predictor variables being included in initial models: level of pre-hospital care, time from injury to arrival in hospital, type of injury, mechanism of injury, age, sex, ISS, GCS and systolic blood pressure.

The effect of pre-hospital care on mortality was dependent on level of ICU care. Key results in the group that did not receive ICU care were (using basic life support as the reference group):

� level 5 ambulance: OR 2.18 (95% CI, 1.05-4.55)

� physician: OR 4.27 (95% CI 1.46-12.45).

The majority of these deaths occurred within 24 hours of admission, which the authors suggested was on the basis of not surviving initial resuscitation. Thus it is not clear if the logistic regression model adequately controlled for injury severity.

Key results in the group that did receive ICU care were (using basic life support as the reference group):

� level 5 ambulance: OR 0.70 (95% CI 0.53-1.18)

� physician: OR 0.63 (95% CI 0.28-1.39).


� the study used a retrospective design

� a selective dispatch strategy was used which probably explains the increased mortality rates in the level 5 ambulance group and the physician group in the non-ICU population

� paramedics may have involved the physician group when patient death was imminent

� the observational study is susceptible to residual confounding.

Cameron et al. 2005

A retrospective chart review was conducted in Australia (Cameron et al. 2005). This study used a before and after design (Level III-3 evidence) to compare outcome in a period where helicopters included emergency physicians with a subsequent period where the helicopter did not include an emergency physician (intensive care paramedics were used). The physicians were able to perform all the usual treatment and monitoring facilities that they could normally provide in the emergency department. In contrast, the non-physician team was unable to do rapid sequence intubations. Given the nature of the data recorded measures of injury severity were restricted to the RTS. Chart abstraction was primarily performed by one person but a 10% sample was validated by another abstractor. There was an excellent level of agreement between the two abstractors. There were 163 patients in the physician group and 211 in the paramedic group. Mortality was measured at 30 days. There were 10 deaths in total and no significant difference was detected between the study groups. There was also no


51

significant difference in the length of stay, although the length of stay was only two days and one day respectively in the physician and paramedic groups.


� the effects of a retrospective design were apparent. As the authors documented, they were unable to extract data that would have provided a better indication of injury/illness severity.

� there was no control over potential confounders. Given baseline differences suggestive of a more severe casemix in the physician group this may have led to bias in the comparison.

� the outcome of the group discharged from ED was not obtained.

� the study power was low, particularly for mortality, so the lack of a significant difference in outcome was not surprising.

Wirtz et al. 2002

There was one other study of interest identified, although it did not meet the eligibility criteria for the review. It is included here for completeness. Wirtz et al. (2002) noted that both paramedics and flight nurses performed rapid sequence intubation. There was therefore no comparison group in this study involving a crew that did not perform rapid sequence intubation. Therefore, the study could not be included. Mortality was similar in both groups. However, it leaves the question whether there truly would be a difference in outcome between:

1. Doctors and other crew configurations that are both able to conduct rapid sequence intubation.

2. Non-doctor crews able to perform rapid sequence intubation and non-doctor crews that are not able to perform rapid sequence intubation.


52

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ba

xt

an

d M

oo

dy 1

987

)

USA

Pse

ud

o-

ran

do

mis

ed

co

ntr

olle

d t

rial

Leve

l III-

1

Stu

dy s

ett

ing

.

Co

nse

cu

tive

pa

tie

nts

with

blu

nt

tra

um

a

ove

r a

24 m

on

th p

erio

d.

All

pa

tie

nts

we

re t

ran

spo

rte

d t

o t

he

Un

ive

rsity o

f

Ca

lifo

rnia

, Sa

n D

ieg

o,

Me

dic

al T

rau

ma

Ce

nte

r w

he

re t

he

y w

ere

ca

red

fo

r u

ntil

dis

ch

arg

e.

Pa

rtic

ipa

nts

:

Tota

l sa

mp

le 5

74

Inte

rve

ntio

n.

He

lico

pte

r st

aff

ed

by f

ligh

t n

urs

e a

nd

ph

ysi

cia

n (

n=

316

)

Pro

ce

du

res:

ora

l/n

asa

l* e

nd

otr

ac

he

al

intu

ba

tio

n

Ne

ed

le/t

ub

e t

ho

rac

ost

om

y p

lac

em

en

t*

Pe

rip

he

ral/

ce

ntr

al I

V li

ne

pla

ce

me

nt

Pe

ric

ard

ioc

en

tesi

s*

Cric

oth

yre

oto

my p

lac

em

en

t*

Exp

an

de

d m

ed

ica

tio

ns*

Co

mp

ara

tor.

He

lico

pte

r st

aff

ed

by f

ligh

t n

urs

e a

nd

pa

ram

ed

ic (

n=

25

8)

Pro

ce

du

res:

ora

l en

do

tra

ch

ea

l

intu

ba

tio

n

IV li

ne

pla

ce

me

nt

Lim

ite

d m

ed

ica

tio

ns

Pn

eu

ma

tic

an

tish

oc

k g

arm

en

t

pla

ce

me

nt*

Inc

lu/e

xcl c

rite

ria

.

Pa

tie

nts

with

blu

nt

tra

um

a

Exc

lud

ed

pa

tie

nts

wh

o d

id n

ot

ha

ve

an

y r

esu

scita

tive

pro

ce

du

res

in t

he

fie

ld

Da

ta c

olle

ctio

n

Tra

um

a s

co

re c

alc

ula

ted

on

pa

tie

nt

co

nta

ct

by t

he

cre

w.

Inju

ry s

eve

rity

sc

ore

ca

lcu

late

d f

rom

pa

tie

nt

rec

ord

s a

nd

au

top

sy

rep

ort

s.

Ou

tco

me

me

asu

res

Mo

rta

lity: d

efin

ed

as

de

ath

du

e t

o

the

in

itia

l in

jurie

s o

r c

om

plic

atio

ns

of

the

in

jurie

s

Follo

w-u

p in

terv

al

Min

imu

m o

f si

x m

on

ths

An

aly

sis

Pre

dic

ted

mo

rta

lity e

stim

ate

d u

sin

g

the

TR

ISS m

eth

od

olo

gy, u

tilis

ing

th

e

mo

st r

ec

en

t c

oe

ffic

ien

ts a

va

ilab

le

at

the

tim

e.

An

aly

tic

me

tho

ds

inc

lud

ed

usi

ng

χ2,

two

ta

iled

Stu

de

nt

t te

st,

Ma

nte

l-

He

an

sze

l te

st a

nd

th

e Z

sta

tist

ic o

f

co

mp

ariso

n b

etw

ee

n p

red

icte

d

an

d a

ctu

al s

urv

iva

l.

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Pa

ram

ed

ic g

rou

p:

Pre

dic

ted

= 1

9.5

Ac

tua

l=1

9

Z s

tatist

ic 0

.208 (

P>

0.0

5)

Ph

ysi

cia

n g

rou

p:

Pre

dic

ted

16.9

Ac

tua

l 11

Z s

tatist

ic 2

.284 (

P<

0.0

5)

Diffe

ren

ce

in Z

sta

tist

ic b

etw

ee

n t

he

two

gro

up

s: 2

.076 (

P<

0.0

5)

Ind

ica

tin

g a

sta

tist

ica

lly s

ign

ific

an

t

imp

rove

d o

utc

om

e in

th

e p

hysi

cia

n

gro

up

.

Dis

trib

utio

n o

f p

atie

nts

by

pro

ba

bili

ty o

f su

rviv

al

Exp

ec

ted

to

die

bu

t liv

ed

:

Pa

ram

ed

ic: 0

Ph

ysi

cia

n 5

Exp

ec

ted

to

live

bu

t d

ied

:

Pa

ram

ed

ic: 5

Ph

ysi

cia

n: 3

Exp

ec

ted

to

die

an

d d

ied

:

Pa

ram

ed

ic: 1

4

Ph

ysi

cia

n: 8

Lim

ita

tio

ns

�

No

t tr

uly

ra

nd

om

ise

d: d

isp

atc

h

de

pe

nd

ed

on

ro

tatio

n o

f c

alls

or

wh

ich

he

lico

pte

r w

as

clo

ser

to t

he

sce

ne

at

tim

e o

f d

isp

atc

h.

�

Ce

ntr

al c

om

mu

nic

atio

n d

id n

ot

ap

pe

ar

to b

e a

va

ilab

le f

or

eith

er

sta

ff

gro

up

.

�

The

leve

l of

exp

erie

nc

e o

f b

oth

sta

ff

gro

up

s w

as

un

cle

ar.

�

TRIS

S m

eth

od

olo

gy in

clu

de

d t

he

tra

um

a s

co

re (

TS)

rath

er

tha

n t

he

revis

ed

tra

um

a s

co

re (

RTS

). In

late

r

ye

ars

th

e R

TS s

co

re w

as

co

nsi

de

red

to

be

mo

re a

cc

ura

te t

ha

n t

he

TS s

co

re.

Su

bse

qu

en

t d

eve

lop

me

nts

in T

RIS

S

me

tho

do

log

y m

ad

e u

se o

f th

e M

sta

tist

ic w

hic

h a

llow

s a

n a

sse

ssm

en

t o

f

the

in

jury

se

ve

rity

mix

be

twe

en

stu

dy

gro

up

s a

nd

, if a

pp

rop

ria

te,

the

W

sta

tist

ic, w

hic

h d

ea

ls w

ith

diffe

ren

t

ca

se m

ixe

s a

cro

ss s

am

ple

s.

�

All-

ca

use

mo

rta

lity n

ot

use

d. M

ay

ha

ve

lea

d t

o o

utc

om

e

mis

cla

ssific

atio

n.

�

No

do

cu

me

nta

tio

n o

f b

lind

ing

in t

he

ass

ess

me

nt

of

ca

use

of

de

ath

.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps

(alth

ou

gh

all

rec

ord

ed

de

ath

s o

cc

urr

ed

with

in 4

8 h

ou

rs o

f

ad

mis

sio

n).


53

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ba

xt

an

d M

oo

dy 1

987

)

USA

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

.

Me

dia

n p

rob

ab

ility

of

surv

iva

l:

Pa

ram

ed

ic c

rew

: 0.9

5-1

.0

Ph

ysi

cia

n c

rew

: 0.9

5-1

.0

Me

dia

n G

lasg

ow

Co

ma

Sc

ore

of

seve

re

bra

in in

jure

d p

atie

nts

(G

CS≤8

)

Pa

ram

ed

ic c

rew

(n

=44,

17%

of

tota

l): 4

Ph

ysi

cia

n c

rew

(n

=5

4, 1

7%

of

tota

l):

5

Me

an

Tra

um

a S

co

re:

Pa

ram

ed

ic g

rou

p:

14.2

Ph

ysi

cia

n g

rou

p:

14.3

Me

an

In

jury

se

ve

rity

sc

ore

:

Pa

ram

ed

ic g

rou

p:

13.6

Ph

ysi

cia

n g

rou

p:1

3.9

Me

an

pre

dic

ted

su

rviv

al

Pa

ram

ed

ic g

rou

p:

0.9

24

Ph

ysi

cia

n g

rou

p:

0.9

47

Me

an

ag

e

Pa

ram

ed

ic g

rou

p:

27.6

ye

ars

Ph

ysi

cia

n g

rou

p:

27.8

ye

ars

Tra

nsp

ort

tim

es:

Me

an

re

spo

nse

tim

e:

Pa

ram

ed

ic g

rou

p 1

5.5

min

ute

s

Ph

ysi

cia

n g

rou

p:

14.9

min

ute

s

Me

an

sc

en

e t

ime

:

Pa

ram

ed

ic g

rou

p:

18.6

min

ute

s

Ph

ysi

cia

n g

rou

p:

19.1

min

ute

s

Me

an

de

live

ry t

ime

:

Pa

ram

ed

ic g

rou

p:

16.9

min

ute

s

Ph

ysi

cia

n g

rou

p:

16.1

min

ute

s.

B

lind

ed

ch

art

an

aly

sis

ass

ess

ing

ad

he

ren

ce

to

writt

en

me

dic

al

tre

atm

en

t p

roto

co

ls.

1.

pa

tie

nts

wh

o s

urv

ive

d

Pa

ram

ed

ic g

rou

p18 o

f 23

9 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

Ph

ysi

cia

n g

rou

p:

2 o

f 305

pa

tie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

2.

pa

tie

nts

wh

o d

ied

Pa

ram

ed

ic g

rou

p:

9 o

f 19 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

Ph

ysi

cia

n g

rou

p:

0 o

f 11 p

atie

nts

with

inc

on

sist

en

cie

s fr

om

th

e

rec

om

me

nd

ed

pro

toc

ols

.

No

te in

co

nsi

ste

nc

ies

inc

lud

ed

failu

re t

o c

on

du

ct

the

ind

ica

ted

pro

ce

du

re o

r m

ed

ica

l pro

ce

du

re

no

t fo

llow

ed

.

�

Po

ten

tia

l fo

r c

on

fou

nd

ing

, a

lth

ou

gh

ba

selin

e a

na

lyse

s in

dic

ate

litt

le

diffe

ren

ce

in t

ran

spo

rt t

ime

, tr

au

ma

sco

re, in

jury

se

ve

rity

sc

ore

, G

CS,

pre

dic

ted

su

rviv

al a

nd

pa

tie

nt

ag

e

be

twe

en

gro

up

s.

�

Ke

y r

esu

lts

are

ba

sed

on

a s

ma

ll

nu

mb

er

of

pa

tie

nts

wh

o s

urv

ive

d b

ut

we

re e

xpe

cte

d t

o d

ie (

5 o

f 22 in

th

e

ph

ysi

cia

n g

rou

p a

nd

0 o

f 1

6 in

th

e

pa

ram

ed

ic g

rou

p).

Co

mm

en

ts

�

All

pa

tie

nts

att

en

de

d a

sin

gle

tra

um

a

ce

ntr

e.

�

The

TR

ISS m

eth

od

olo

gy w

as

ap

plie

d t

o

the

tw

o s

tud

y g

rou

ps

usi

ng

da

ta t

ha

t

we

re c

olle

cte

d a

t th

e s

am

e t

ime

be

twe

en

th

e t

wo

gro

up

s.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

�

Thre

e p

atie

nts

we

re e

xclu

de

d in

ea

ch

gro

up

du

e t

o t

he

lac

k o

f re

susc

ita

tiv

e

me

asu

res

in t

he

fie

ld.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

A s

tatist

ica

lly s

ign

ific

an

t re

du

ctio

n in

th

e

mo

rta

lity o

f p

atie

nts

with

blu

nt

tra

um

a

tre

ate

d b

y a

me

dic

al h

elic

op

ter

em

erg

en

cy

ca

re s

erv

ice

sta

ffe

d b

y a

nu

rse

/ph

ysi

cia

n

co

mb

ina

tio

n c

ou

ld b

e d

em

on

stra

ted

co

mp

are

d w

ith

th

at

sta

ffe

d b

y a

nu

rse

/pa

ram

ed

ic c

om

bin

atio

n.

* D

iffer

ence

in p

roce

dure

s av

aila

ble

to th

e tw

o st

affin

g gr

oups


54

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ga

rne

r e

t a

l. 19

99

)

Au

stra

lia

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

He

lico

pte

r tr

an

spo

rta

tio

n o

f p

atie

nts

fro

m t

he

sc

en

e o

ve

r a

28 m

on

th p

erio

d.

Pa

rtic

ipa

nts

(n

=2

07

):

Inte

rve

ntio

n (

n=

67

).

Ph

ysi

cia

n s

taff

ed

NR

MA

Ca

reFl

igh

t

he

lico

pte

r

Co

mp

ara

tor

(n=

14

0).

Pa

ram

ed

ic s

taff

ed

We

stp

ac

Hu

nte

r

reg

ion

he

lico

pte

r.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

dia

n a

ge

(ye

ars

)

Ph

ysi

cia

n g

rou

p:

31 (

13

-70

)

Pa

ram

ed

ic g

rou

p:

33 (

2-8

9)

Me

ch

an

ism

of

inju

ry

Mo

tor

ve

hic

le o

cc

up

an

t

Ph

ysi

cia

n g

rou

p:

63%

Pa

ram

ed

ic g

rou

p:

63

%

Mo

tor

bik

e r

ide

r

Ph

ysi

cia

n g

rou

p:

9%

Pa

ram

ed

ic g

rou

p:

11

%

Pe

da

l cyc

list

or

pe

de

stria

n

Ph

ysi

cia

n g

rou

p:

3%

Pa

ram

ed

ic g

rou

p:

5%

Inc

lu/e

xcl c

rite

ria

.

Blu

nt

tra

um

a

ISS ≥

10

Tra

nsp

ort

ed

dire

ctly f

rom

th

e

inc

ide

nt

sce

ne

Inc

ide

nt

oc

cu

rre

d b

etw

ee

n

Jan

ua

ry 1

996 a

nd

Ap

ril 1

998.

Da

ta c

olle

ctio

n

Pa

ram

ed

ic g

rou

p r

etr

osp

ec

tive

ly

ide

ntifie

d f

rom

th

e t

rau

ma

re

gis

try

of

Joh

n H

un

ter

Ho

spita

l, N

ew

ca

stle

.

Ph

ysi

cia

n g

rou

p id

en

tifie

d f

rom

th

e

me

dic

al d

ata

ba

se o

f N

RM

A

Ca

reFl

igh

t/N

SW

Me

dic

al r

etr

ieva

l

Se

rvic

e w

ho

we

re t

ran

spo

rte

d t

o

We

stm

ea

d o

r N

ep

ea

n h

osp

ita

ls in

Syd

ne

y.

Pre

-ho

spita

l ca

se s

he

ets

we

re

exa

min

ed

to

allo

w t

he

ca

lcu

latio

n

of

the

RTS

(firs

t re

co

rde

d d

ata

use

d).

Oth

er

da

ta c

olle

cte

d

inc

lud

ed

de

mo

gra

ph

ics,

me

ch

an

ism

of

inju

ry, re

spo

nse

,

sce

ne

an

d t

ran

spo

rt t

ime

s,

en

tra

pm

en

t a

t th

e s

ce

ne

,

req

uire

me

nt

for

win

ch

extr

ac

tio

n,

flu

ids

ad

min

iste

red

an

d p

roc

ed

ure

s

pe

rfo

rme

d a

t th

e s

ce

ne

or

in t

ran

sit.

Ou

tco

me

me

asu

res

Mo

rta

lity. C

om

pa

red

with

TR

ISS

me

tho

do

log

y u

sin

g c

oe

ffic

ien

ts

de

rive

d f

rom

th

e M

TOS u

sin

g t

he

1990 a

bb

revia

ted

inju

ry s

ca

le.

Co

mp

aris

on

be

twe

en

ob

serv

ed

an

d p

red

icte

d m

ort

alit

y m

ad

e a

t

ho

spita

l dis

ch

arg

e.

Nu

mb

er

of

ac

tua

l de

ath

s b

y t

he

nu

mb

er

pre

dic

ted

to

die

(b

ase

d o

n

Ps

≤ 0

.50)

Ph

ysi

cia

n g

rou

p

Z s

tatist

ic +

2.7

2

P<

0.0

1

M s

tatist

ic 0

.62

Ad

just

ed

W s

tatist

ic 9

.48 (

95%

CI

3.8

4-1

5.1

2)

co

mp

are

d w

ith

th

e

MTO

S p

op

ula

tio

n.

Pa

ram

ed

ic g

rou

p

Z s

tatist

ic -

1.1

6

P=

0.2

5

M s

tatist

ic 0

.68

Ad

just

ed

W s

tatist

ic -

2.3

7 (

95%

CI -

6.8

1 t

o 2

.07-1

5.1

2)

co

mp

are

d w

ith

the

MTO

S p

op

ula

tio

n.

Dire

ct

co

mp

ariso

n b

etw

ee

n

ph

ysi

cia

n a

nd

pa

ram

ed

ic g

rou

p

Ad

just

ed

W s

tatist

ic 1

3.4

4 (

95%

CI

7.8

0-1

9.0

8)

sug

ge

stin

g a

n a

dd

itio

na

l

13 s

urv

ivo

rs p

er

10

0 p

atie

nts

tre

ate

d

in t

he

ph

ysi

cia

n g

rou

p c

om

pa

red

with

th

e p

ara

me

dic

gro

up

.

Ro

ad

tra

nsp

ort

ed

pa

tie

nts

: D

irec

t

co

mp

ariso

n b

etw

ee

n p

hysi

cia

n

an

d p

ara

me

dic

gro

up

Ad

just

ed

W s

tatist

ic 2

.11 (

95%

CI -

0.3

4 t

o 4

.56)

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

Gro

up

s tr

ea

ted

in d

iffe

ren

t h

osp

ita

ls –

pa

ram

ed

ic g

rou

p t

rea

ted

in a

leve

l 6

ho

spita

l, p

hysi

cia

n g

rou

p in

a le

ve

l 5

or

leve

l 6 h

osp

ita

l.

�

Sta

tist

ica

lly s

ign

ific

an

t d

iffe

ren

ce

in

ba

selin

e G

CS a

nd

ISS s

co

res

co

nsi

ste

nt

with

in

cre

ase

d s

eve

rity

in

the

ph

ysi

cia

n g

rou

p.

�

Un

cle

ar

if t

he

re w

ere

diffe

ren

ce

s in

the

du

ratio

n o

f fo

llow

-up

be

twe

en

stu

dy g

rou

ps.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g a

lth

ou

gh

use

of

the

W

sta

tist

ic h

elp

s a

dju

st f

or

TRIS

S v

aria

ble

s

be

twe

en

th

e s

tud

y p

op

ula

tio

ns.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

RTS

sc

ore

in t

he

MTO

S s

tud

y w

as

ca

lcu

late

d a

t a

dm

issi

on

ra

the

r th

an

at

the

sc

en

e a

s p

erf

orm

ed

in t

his

stu

dy. Th

e a

pp

roa

ch

use

d in

th

is s

tud

y

ha

s th

e a

dva

nta

ge

of

co

llec

tin

g R

TS

da

ta b

efo

re in

tub

atio

n a

nd

co

mp

arin

g R

TS in

bo

th g

rou

ps

at

ap

pro

xim

ate

ly t

he

sa

me

tim

e.

Ho

we

ve

r, c

om

pa

riso

n w

ith

MTO

S is

no

t so

va

lid g

ive

n t

he

diffe

ren

ce

in

tim

ing

.

�

Se

ve

n p

atie

nts

we

re e

xclu

de

d d

ue

to

mis

sin

g c

ase

sh

ee

ts.

All

7 w

ere

in t

he

pa

ram

ed

ic g

rou

p p

rod

uc

ing

a

sele

ctio

n b

ias.


55

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ga

rne

r e

t a

l. 19

99

)

Au

stra

lia

co

ntin

ue

d

Fa

lls

Ph

ysi

cia

n g

rou

p:

10%

Pa

ram

ed

ic g

rou

p:

6%

Me

dia

n t

ime

inte

rva

ls (

min

ute

s)

Ca

ll to

sc

en

e a

rriv

al

Ph

ysi

cia

n g

rou

p:

29

Pa

ram

ed

ic g

rou

p:

26

Sc

en

e t

ime

(e

xclu

din

g t

rap

pe

d a

nd

win

ch

ed

pa

tie

nts

)

Ph

ysi

cia

n g

rou

p:

33

Pa

ram

ed

ic g

rou

p:

34

Tra

nsp

ort

tim

e t

o h

osp

ita

l

Ph

ysi

cia

n g

rou

p:

15

Pa

ram

ed

ic g

rou

p:

12

Tota

l pre

-ho

spita

l tim

e

Ph

ysi

cia

n g

rou

p:

86

Pa

ram

ed

ic g

rou

p:

82

Me

dia

n R

TS

Ph

ysi

cia

n g

rou

p:

6.9

0

Pa

ram

ed

ic g

rou

p:

7.5

5

Me

dia

n G

CS

Ph

ysi

cia

n g

rou

p: 1

3

Pa

ram

ed

ic g

rou

p:

14

P=

0.0

5

Me

dia

n ISS

Ph

ysi

cia

n g

rou

p:

25

Pa

ram

ed

ic g

rou

p:

18

P=

0.0

5

Follo

w-u

p in

terv

al

No

t st

ate

d

An

aly

sis

Ca

teg

oric

al v

aria

ble

s: χ

2 o

r Fi

she

r’s

exa

ct

test

, a

s a

pp

rop

ria

te.

Co

ntin

uo

us

va

ria

ble

s: M

an

n-

Wh

itn

ey U

te

st.

Co

mp

aris

on

be

twe

en

pre

dic

ted

an

d o

bse

rve

d m

ort

alit

y u

sin

g Z

, W

an

d M

sta

tist

ics.

An

ad

just

ed

W

sta

tist

ic w

as

ca

lcu

late

d b

y t

he

me

tho

d o

f Y

ou

ng

e w

he

n t

he

M

sta

tist

ic in

dic

ate

d a

po

or

ma

tch

with

th

e M

TOS c

oh

ort

an

d t

o

dire

ctly c

om

pa

re t

he

pa

ram

ed

ic

an

d p

hysi

cia

n t

rea

ted

gro

up

s.

Pro

ce

du

res

at

sce

ne

Me

dia

n v

olu

me

of

flu

id in

fuse

d in

pa

tie

nts

wh

o r

ec

eiv

ed

> 5

0m

L

Ph

ysi

cia

n g

rou

p:

250

0

Pa

ram

ed

ic g

rou

p:

825

P<

0.0

01

Me

dia

n v

olu

me

of

flu

id (

mL)

infu

sed

in p

atie

nts

with

initia

l hyp

ote

nsi

on

(syst

olic

BP

<90

mm

Hg

)

Ph

ysi

cia

n g

rou

p:

503

5

Pa

ram

ed

ic g

rou

p:

147

5

P<

0.0

01

Nu

mb

er

of

pa

tie

nts

intu

ba

ted

Ph

ysi

cia

n g

rou

p:

34/6

7 (

1

cric

oth

yro

ido

tom

y, m

usc

le r

ela

xan

t

dru

gs

use

d in

28 o

f th

e 3

4

intu

ba

tio

ns)

Pa

ram

ed

ic g

rou

p:

14

/14

0

P<

0.0

01

Pro

po

rtio

n o

f p

atie

nts

with

GC

S<

9

intu

ba

ted

Ph

ysi

cia

n g

rou

p:

23/2

3

Pa

ram

ed

ic g

rou

p:

14

/36

P<

0.0

01

Tho

rac

ic d

ec

om

pre

ssio

ns

Ph

ysi

cia

n g

rou

p:

8/6

7 (

6 t

ub

e, 2

ne

ed

le)

Pa

ram

ed

ic g

rou

p:

2/1

40 (

bo

th

ne

ed

le)

P<

0.0

1

Co

mm

en

ts

�

All

pa

tie

nts

att

en

de

d a

sin

gle

tra

um

a

ce

ntr

e in

ea

ch

co

un

try.

�

TRIS

S m

eth

od

olo

gy a

pp

rop

ria

tely

use

d

RTS

sc

ore

.

�

TRIS

S a

na

lysi

s a

pp

rop

riate

ly in

clu

de

d

est

ima

tio

n o

f th

e M

an

d W

sta

tist

ics.

�

Use

d a

pp

rop

ria

te m

eth

od

olo

gy t

o

est

ima

te in

jury

se

ve

rity

.

�

Diffe

ren

ce

s in

pro

ce

du

res

pe

rfo

rme

d

be

twe

en

stu

dy g

rou

ps.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Ph

ysi

cia

ns

pe

rfo

rm a

gre

ate

r n

um

be

r o

f

pro

ce

du

res

at

ac

cid

en

t sc

en

es

with

ou

t

inc

rea

sin

g s

ce

ne

tim

e.

This

re

sults

in

sig

nific

an

tly lo

we

r m

ort

alit

y. C

ritic

al c

are

ph

ysi

cia

ns

sho

uld

be

ad

de

d t

o p

ara

me

dic

he

lico

pte

r se

rvic

es

for

sce

ne

re

spo

nse

to

blu

nt

tra

um

a.


56

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ko

efo

ed

-Nie

lse

n e

t a

l.

2002

)

De

nm

ark

Be

fore

an

d

aft

er

Leve

l III-

3.

Stu

dy s

ett

ing

.

A m

ob

ile e

me

rge

nc

y c

are

un

it (

MEC

U),

wh

ich

inc

lud

ed

an

an

ae

sth

etist

, w

as

set

up

in a

re

gio

n o

f D

en

ma

rk.

Co

nse

cu

tive

pa

tie

nts

with

ac

ute

MI w

ere

stu

die

d in

two

tim

e p

erio

ds

:

1. Se

pt

to N

ov 1

99

6 (

be

fore

MEC

U)

2. Se

pt

to N

ov 1

99

7 (

with

MEC

U).

The

are

a c

ove

red

ap

pro

xim

ate

ly

800km

2. M

ed

ian

re

spo

nse

tim

e f

or

MEC

U

wa

s 8.9

min

ute

s a

nd

6.5

min

ute

s fo

r

sta

nd

ard

am

bu

lan

ce

s.

Pa

rtic

ipa

nts

:

Inte

rve

ntio

n (

n=

54

)

An

ae

sth

etist

pre

sen

t (w

ith

MEC

U)

MEC

U w

as

als

o c

rew

ed

with

a s

pe

cia

lly

tra

ine

d e

me

rge

nc

y t

ec

hn

icia

n.

Ad

ditio

na

l tre

atm

en

t p

rovid

ed

by

an

ae

sth

etist

: tr

ac

he

al i

ntu

ba

tio

n,

tra

nsc

uta

ne

ou

s p

ac

ing

, a

na

est

he

tic

s

(hyp

no

tic

s a

nd

mu

scle

re

laxa

nts

),

op

ioid

s, h

ea

rt s

tim

ula

tin

g d

rug

s a

nd

an

tia

rrh

yth

mic

s.

Pre

-ho

spita

l

thro

mb

oly

sis

wa

s n

ot

pro

vid

ed

.

Co

mp

ara

tor

(n=

54

)

No

n p

hysi

cia

n c

rew

(b

efo

re M

EC

U)

Inc

lu/e

xcl c

rite

ria

.

Use

rs o

f e

me

rge

nc

y s

erv

ice

s (b

ase

d

on

am

bu

lan

ce

co

mp

an

y’s

pa

tie

nt

rolls

) w

ho

we

re c

lass

ifie

d a

s a

dia

gn

osi

s o

f M

I (I

CD

-10 c

od

es

I21-

I22)

on

ad

mis

sio

n t

o a

n e

me

rge

nc

y

de

pa

rtm

en

t o

r c

oro

na

ry c

are

un

it in

on

e o

f th

ree

ho

spita

ls in

Aa

rhu

s,

De

nm

ark

.

The

pa

tie

nt

wa

s a

live

wh

en

rea

ch

ing

ho

spita

l.

The

ho

spita

l re

co

rd c

on

firm

ed

MI

dia

gn

osi

s, b

ase

d o

n e

nzy

me

te

sts

an

d E

CG

.

The

ho

spita

l re

co

rd p

rovid

ed

suff

icie

nt

info

rma

tio

n o

n d

iag

no

sis

an

d t

rea

tme

nt.

Ou

tco

me

me

asu

res

28 d

ay m

ort

alit

y

An

aly

sis

Pe

ars

on

’s χ

2, Fis

he

r’s

exa

ct

test

,

Sp

ea

rma

n’s

no

n-p

ara

me

tric

ra

nk

co

rre

latio

n, K

ap

lan

-Me

ier

surv

iva

l

an

aly

sis

an

d m

ultip

le lo

gis

tic

reg

ress

ion

we

re c

on

du

cte

d.

Cru

de

mo

rta

lity r

ate

at

28 d

ays,

by

stu

dy p

erio

d.

Pre

-MEC

U:

20.6

%

MEC

U: 11

.1%

Ad

just

ed

od

ds

ratio

by s

tud

y p

erio

d

(28 d

ay m

ort

alit

y),

(p

re-M

EC

U a

s

the

re

fere

nc

e)

OR

0.3

(P

<0.0

25

)

Ad

just

ed

fo

r a

ge

, g

en

de

r, p

uls

e

an

d s

yst

olic

blo

od

pre

ssu

re.

Ad

just

ed

od

ds

ratio

by u

se o

f M

EC

U

(28 d

ay m

ort

alit

y),

(p

re-M

EC

U a

s

the

re

fere

nc

e)

OR

0.2

(P

<0.0

5)

Ad

just

ed

fo

r a

ge

, g

en

de

r, p

uls

e

an

d s

yst

olic

blo

od

pre

ssu

re.

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

Exc

lusi

on

s: s

ix w

ere

eith

er

de

ad

on

arr

iva

l or

die

d im

me

dia

tely

aft

er

arr

iva

l (fo

ur

pre

MEC

U a

nd

tw

o w

ith

MEC

U),

sp

ec

ific

MI d

iag

no

sis

co

uld

no

t b

e c

on

firm

ed

in

25

. U

nc

lea

r if a

ny

we

re e

xclu

de

d d

ue

to

in

suff

icie

nt

info

rma

tio

n o

n t

rea

tme

nt.

�

Lim

ite

d d

ata

pro

vid

ed

at

tim

e o

f firs

t

att

en

da

nc

e o

f p

re-h

osp

ita

l ca

re.

�

Be

fore

an

d a

fte

r st

ud

y is

a lo

w q

ua

lity

de

sig

n –

fe

atu

res

oth

er

tha

n t

he

inte

rve

ntio

n o

f in

tere

st m

ay h

ave

ch

an

ge

d s

o it

is u

nc

lea

r to

wh

at

ext

en

t a

ny c

ha

ng

e in

ou

tco

me

is a

resu

lt o

f th

e in

tro

du

ctio

n o

f M

EC

U. In

this

stu

dy, th

e im

pro

ve

d p

rog

no

sis

ma

y h

ave

be

en

du

e t

o a

ng

iop

last

y

rath

er

tha

n M

EC

U.

�

Ba

sis

of

dia

gn

osi

s o

f M

I n

ot

pre

cis

ely

de

fin

ed

: p

ote

ntia

l fo

r in

ap

pro

pria

te

sele

ctio

n.

�

Po

ten

tia

l fo

r m

isc

lass

ific

atio

n o

f

ou

tco

me

, th

ou

gh

th

e s

ize

an

d

dire

ctio

n o

f su

ch

mis

cla

ssific

atio

n is

diffic

ult t

o d

ete

rmin

e.

�

Du

rin

g t

he

MEC

U p

erio

d,

som

e

pa

tie

nts

we

re n

ot

att

en

de

d b

y M

EC

U

(56%

). H

ow

eve

r, a

lth

ou

gh

MEC

U

pa

tie

nts

we

re m

ore

like

ly t

o b

e

co

nsi

de

red

as

seve

re,

MEC

U p

atie

nts

ha

d h

igh

er

ad

just

ed

od

ds

of

surv

iva

l

co

mp

are

d w

ith

no

n-M

EC

U p

atie

nts

.

�

Un

cle

ar

if r

esu

lts

co

uld

be

re

plic

ate

d if

oth

er

cre

ws

we

re s

taff

ed

with

pe

op

le

ca

pa

ble

of

pe

rfo

rmin

g r

ap

id

seq

ue

nc

e in

tub

atio

n.


57

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ko

efo

ed

-Nie

lse

n e

t a

l.

2002

)

De

nm

ark

co

ntin

ue

d

A

na

lyse

s c

om

pa

ring

gro

up

s a

t b

ase

line

Ag

e ≤

69 y

ea

rs (

%)

Pre

MEC

U: 4

1

MEC

U: 41

Ma

le s

ex

(%)

Pre

MEC

U: 6

3

MEC

U: 72

PTC

A p

erf

orm

ed

(%

)

Pre

MEC

U: 1

9

MEC

U: 26

Thro

mb

oly

sis

giv

en

(%

)

Pre

MEC

U: 2

8

MEC

U: 33

Co

mm

en

ts

�

aim

ed

to

ass

ess

th

e im

pa

ct

of

a

MEC

U o

n s

urv

iva

l am

on

g p

atie

nts

with

ac

ute

MI.

�

MEC

U p

atie

nts

we

re m

ore

lik

ely

to

rec

eiv

e t

hro

mb

oly

sis.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

In t

he

pre

sen

t st

ud

y, M

I p

atie

nts

tre

ate

d in

a

MEC

U s

taff

ed

by a

n a

na

est

he

tist

an

d/o

r

ha

vin

g a

ng

iop

last

y w

as

fou

nd

to

be

ass

oc

iate

d w

ith

a r

ed

uc

ed

mo

rta

lity. Th

ese

ob

serv

atio

ns

ha

ve

be

en

ba

sed

on

qu

asi

-

exp

erim

en

tal r

ath

er

tha

n r

an

do

mis

ed

exp

erim

en

tal d

ata

, a

nd

ra

nd

om

ise

da

ta

wo

uld

be

hig

hly

de

sira

ble

.


58

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

e e

t a

l. 2

003

)

Au

stra

lia

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Se

ve

re b

lun

t tr

au

ma

pa

tie

nts

ad

mitte

d

to W

est

me

ad

Ho

spita

l, Syd

ne

y b

etw

ee

n

July

1986 a

nd

De

ce

mb

er

20

00.

Am

bu

lan

ce

off

ice

rs t

rain

ed

to

th

ree

diffe

ren

t le

ve

ls: tw

o le

ve

ls o

f b

asi

c li

fe

sup

po

rt a

nd

on

e le

ve

l of

ad

va

nc

ed

life

sup

po

rt.

Als

o h

ad

tw

o p

hysi

cia

n-s

taff

ed

em

erg

en

cy m

ed

ica

l se

rvic

es

wh

ich

resp

on

d t

o a

cc

ide

nt

sce

ne

s e

ith

er

by

roa

d o

r h

elic

op

ter.

Leve

l 3 a

mb

ula

nc

e o

ffic

er:

ba

sic

life

sup

po

rt w

ith

ou

t e

xte

rna

l co

ntr

ol o

f

ha

em

orr

ha

ge

, sp

lintin

g,

no

n-in

va

siv

e

airw

ay m

an

oe

uvre

s a

nd

ba

g-v

alv

e-

ma

sk v

en

tila

tio

n.

Leve

l 4 a

mb

ula

nc

e o

ffic

er:

ba

sic

life

sup

po

rt p

lus

intr

ave

no

us

ca

nn

ula

tio

n

an

d a

dm

inis

tra

tio

n o

f IV

flu

ids,

in

ad

ditio

n t

o a

lim

ite

d r

an

ge

of

IV

me

dic

atio

ns

an

d n

ee

dle

th

ora

ce

nte

sis.

Leve

l 5 a

mb

ula

nc

e o

ffic

er:

Ad

va

nc

ed

life

su

pp

ort

. A

ble

to

pe

rfo

rm a

ll th

e

ab

ove

pro

ce

du

res

plu

s o

ral

en

do

tra

ch

ea

l in

tub

atio

n.

Ac

ce

ss t

o a

wid

er

ran

ge

of

IV m

ed

ica

tio

ns

(bu

t n

ot

ne

uro

mu

scu

lar

blo

cka

de

, a

na

est

he

tic

ag

en

ts o

r se

da

tive

ag

en

ts t

o f

ac

ilita

te

intu

ba

tio

n).

Ph

ysi

cia

ns

are

fre

e t

o e

xerc

ise

th

eir

clin

ica

l ju

dg

em

en

t in

ea

ch

ca

se.

Inc

lu/e

xcl c

rite

ria

.

Blu

nt

tra

um

a p

atie

nts

with

ISS >

15.

Ou

tco

me

me

asu

res

Mo

rta

lity d

urin

g h

osp

ita

l ad

mis

sio

n

An

aly

sis

Log

istic

re

gre

ssio

n u

sed

with

th

e

follo

win

g p

red

icto

r va

ria

ble

s: le

ve

l

of

pre

-ho

spita

l ca

re,

tim

e f

rom

inju

ry

to a

rriv

al i

n h

osp

ita

l, ty

pe

of

inju

ry,

me

ch

an

ism

of

inju

ry, a

ge

, se

x, ISS,

GC

S a

nd

syst

olic

blo

od

pre

ssu

re.

Mo

de

l ca

libra

tio

n a

sse

sse

d b

y t

he

Ho

sme

r-Le

me

sho

w g

oo

dn

ess

-of-

fit

χ2 t

est

an

d p

red

ictive

ac

cu

rac

y

ass

ess

ed

by t

he

are

a u

nd

er

the

rec

eiv

er

op

era

tin

g c

ha

rac

terist

ic

cu

rve

.

Ris

k in

od

ds

ratio

of

mo

rta

lity b

y

leve

l of

pre

-ho

spita

l ca

re a

nd

IC

U

tre

atm

en

t (b

asi

c li

fe s

up

po

rt a

s th

e

refe

ren

ce

), (

95

% C

I)

No

IC

U a

dm

issi

on

:

Leve

l 5 a

mb

ula

nc

e:

2.1

8 (

1.0

5-4

.55)

Ph

ysi

cia

n: 4.2

7 (

1.4

6-1

2.4

5)

ICU

ad

mis

sio

n:

Leve

l 5 a

mb

ula

nc

e:

0.7

0 (

0.5

3-1

.18)

Ph

ysi

cia

n: 0.6

3 (

0.2

8-1

.39

)

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

A s

ele

ctive

dis

pa

tch

str

ate

gy w

as

use

d w

hic

h p

rob

ab

ly e

xpla

ins

the

inc

rea

sed

mo

rta

lity r

ate

s in

th

e le

ve

l 5

am

bu

lan

ce

gro

up

an

d t

he

ph

ysi

cia

n

gro

up

in

th

e n

on

-IC

U p

op

ula

tio

n.

�

Pa

ram

ed

ics

ma

y h

ave

in

vo

lve

d t

he

ph

ysi

cia

n g

rou

p w

he

n p

atie

nt

de

ath

wa

s im

min

en

t.

�

Un

cle

ar

if s

om

e p

hysi

cia

n g

rou

p

pa

tie

nts

ma

y h

ave

be

en

tra

nsp

ort

ed

by h

elic

op

ter.

�

Diffe

ren

t le

ve

ls o

f a

cc

ess

to

pro

ce

du

res

in t

he

no

n-d

oc

tor

gro

up

s

ma

y n

ot

be

ge

ne

ralis

ab

le t

o o

the

r

sett

ing

s, m

ay g

uid

e d

isp

atc

h d

ec

isio

ns

an

d m

ay r

esu

lt in

bia

s in

co

mp

ara

tiv

e

est

ima

tes.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

Co

mm

en

ts

�

Aim

ed

to

de

term

ine

th

e a

sso

cia

tio

n

be

twe

en

mo

rta

lity a

nd

leve

l of

pre

-

ho

spita

l ca

re in

se

ve

rely

inju

red

blu

nt

tra

um

a p

atie

nts

with

or

with

ou

t se

ve

re

he

ad

tra

um

a.

�

No

on

line

me

dic

al c

on

tro

l pro

vid

ed

.

�

We

ll d

esc

ribe

d a

nd

co

nd

uc

ted

sta

tist

ica

l me

tho

do

log

y.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

leve

l of

pre

-ho

spita

l ca

re w

as

ass

oc

iate

d w

ith

th

e r

isk o

f m

ort

alit

y. Th

is w

as

mo

difie

d b

y w

he

the

r th

e p

atie

nt

surv

ive

d

lon

g e

no

ug

h t

o b

e a

dm

itte

d t

o t

he

IC

U.


59

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

e e

t a

l. 2

003

)

Au

stra

lia

co

ntin

ue

d

P

art

icip

an

ts (

n=

2010

):

Inte

rve

ntio

n.

Leve

l of

pre

-ho

spita

l ca

re w

as

cla

ssifie

d

into

th

ree

gro

up

s: b

asi

c li

fe s

up

po

rt,

ad

va

nc

ed

life

su

pp

ort

an

d p

hysi

cia

n

ca

re.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ISS b

y le

ve

l of

pre

-ho

spita

l ca

re:

No

n-E

MS t

ran

spo

rt:

20

Leve

l 3 a

mb

ula

nc

e:

24

Leve

l 4 a

mb

ula

nc

e:

25

Leve

l 5 a

mb

ula

nc

e:

31

Ph

ysi

cia

n: 3

1

Me

dia

n a

ge

30 y

ea

rs (

inte

rqu

art

ile

ran

ge

21-4

9 y

ea

rs)

Ma

les:

76%

Ca

use

s:

Ro

ad

tra

ffic

ac

cid

en

t 6

7%

Falls

: 13%

Ass

au

lt 5

%.

No

he

ad

inju

rie

s: 3

6%

Iso

late

d h

ea

d in

jury

: 52%

He

ad

inju

ry w

ith

ab

do

min

al/

ch

est

inju

rie

s: 1

2%

Pre

-ho

spita

l ca

re:

No

n-E

MS t

ran

spo

rt 5

%

Leve

l 3 a

mb

ula

nc

e 2

3%

Leve

l 4 a

mb

ula

nc

e 2

%

Leve

l 5 a

mb

ula

nc

e 5

9%

Ph

ysi

cia

n E

MS 1

1%


60

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ca

me

ron

et

al.

20

05)

Au

stra

lia

Be

fore

an

d

aft

er

stu

dy.

Re

tro

spe

ctive

ch

art

re

vie

w.

Leve

l III-

3.

Stu

dy s

ett

ing

.

Ca

irn

s B

ase

ho

spita

l. U

ntil 2

00

1

he

lico

pte

rs in

clu

de

d e

me

rge

nc

y

ph

ysi

cia

ns.

Sin

ce

2001 t

he

y h

ave

be

en

sta

ffe

d b

y in

ten

sive

ca

re p

ara

me

dic

s.

Pa

rtic

ipa

nts

(n

=3

74

):

Inte

rve

ntio

n (

n=

16

3)

Sto

pp

ed

th

e p

rese

nc

e o

f a

n

em

erg

en

cy p

hysi

cia

n o

n a

n

em

erg

en

cy h

elic

op

ter.

C

rew

mix

wa

s

un

ab

le t

o p

erf

orm

ra

pid

se

qu

en

ce

intu

ba

tio

n o

r tu

be

th

ora

co

sto

my.

Co

mp

ara

tor

(n=

21

1)

Em

erg

en

cy p

hysi

cia

n (

EP

) o

n b

oa

rd.

The

EP

wa

s a

ble

to

pe

rfo

rm a

ny

pro

ce

du

re t

ha

t th

ey c

an

pe

rfo

rm in

th

e

em

erg

en

cy d

ep

art

me

nt.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e

Me

dia

n a

ge

(ye

ars

)

Ph

ysi

cia

n g

rou

p:

34

Pa

ram

ed

ic g

rou

p:

33

Ma

les

(%)

Ph

ysi

cia

n g

rou

p:

64.0

Pa

ram

ed

ic g

rou

p:

65.6

Me

an

RTS

Ph

ysi

cia

n g

rou

p:

7.7

22

Pa

ram

ed

ic g

rou

p:

7.7

30

Inc

lu/e

xcl c

rite

ria

.

An

y p

rim

ary

ta

skin

g o

f th

e

he

lico

pte

r. C

ase

s w

ere

ide

ntifie

d b

y

the

‘m

od

e o

f a

rriv

al’

fie

ld re

co

rde

d

on

th

e E

D in

form

atio

n s

yst

em

Da

ta c

olle

ctio

n

The

RTS

wa

s c

alc

ula

ted

usi

ng

th

e

initia

l clin

ica

l ob

serv

atio

ns

in t

he

me

dic

al r

ec

ord

s a

nd

am

bu

lan

ce

form

s.

Ou

tco

me

me

asu

res

30 d

ay m

ort

alit

y

Len

gth

of

in-h

osp

ita

l sta

y

Tra

nsf

er

rate

s

Ra

tes

of

dis

ch

arg

e d

ire

ctly f

rom

ho

spita

l

An

aly

sis

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

On

e w

ay a

na

lysi

s o

f va

ria

nc

e w

as

un

de

rta

ke

n a

nd

P v

alu

es

an

d a

χ2

test

with

Ya

tes

co

rre

ctio

n w

ere

ca

lcu

late

d w

he

re a

pp

rop

ria

te

Ka

pp

a s

tatist

ic w

as

use

d t

o a

sse

ss

inte

r-ra

ter

relia

bili

ty in

th

e c

ha

rt

ext

rac

tio

n.

30 d

ay m

ort

alit

y p

rop

ort

ion

, b

y

stu

dy g

rou

p (

%)

Ph

ysi

cia

n g

rou

p:

2.8

%

Pa

ram

ed

ic g

rou

p:

2.5

%

P=

0.8

Me

an

ho

spita

l le

ng

th o

f st

ay, b

y

stu

dy g

rou

p (

da

ys)

Ph

ysi

cia

n g

rou

p:

2

Pa

ram

ed

ic g

rou

p:

1

P=

0.3

Lim

ita

tio

ns

�

Re

tro

spe

ctive

stu

dy.

�

The

ac

cu

rac

y o

f c

od

ing

th

e m

od

e o

f

arr

iva

l (u

sed

to

ide

ntify

re

leva

nt

pa

tie

nts

) w

as

no

t d

oc

um

en

ted

.

Ho

we

ve

r, t

his

is n

ot

like

ly t

o b

e a

ma

jor

sou

rce

of

bia

s.

�

Un

cle

ar

if t

he

tim

ing

of

the

me

asu

rem

en

ts u

sed

to

ass

ess

RTS

wa

s

the

sa

me

in b

oth

gro

up

s.

�

Me

asu

res

of

inju

ry s

eve

rity

use

d w

ere

diffe

ren

t fr

om

th

e n

orm

al a

pp

roa

ch

.

No

use

of

the

ISS s

o it

wa

s n

ot

po

ssib

le

to a

sse

ss p

rob

ab

ility

of

surv

iva

l w

ith

refe

ren

ce

to

a s

uita

ble

po

pu

latio

n

suc

h a

s M

TOS.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Ce

ntr

al c

om

mu

nic

atio

n w

as

no

t

do

cu

me

nte

d f

or

eith

er

gro

up

.

�

Ba

selin

e d

iffe

ren

ce

s in

pa

tie

nts

ad

mitte

d w

ith

hig

he

r p

rop

ort

ion

of

the

ph

ysi

cia

n g

rou

p b

ein

g a

dm

itte

d.

�

Ve

ry lo

w n

um

be

r o

f d

ea

ths

co

nsi

ste

nt

with

low

stu

dy p

ow

er

(10 d

ea

ths

in

tota

l).

�

Au

tho

rs s

ug

ge

ste

d R

TS w

as

a p

oo

r

pre

dic

tor

of

ne

ed

fo

r a

dm

issi

on

an

d

po

stu

late

d t

ha

t A

PA

CH

E m

ay h

ave

be

en

a b

ett

er

me

asu

re.

�

No

fo

llow

-up

of

pa

tie

nts

dis

ch

arg

ed

dire

ctly f

rom

ED

.


61

Ta

ble

13

E

vid

ence

ta

ble

s o

f st

ud

ies

com

pa

rin

g c

rew

s th

at

do

an

d d

o n

ot

per

form

ra

pid

seq

uen

ce i

ntu

ba

tio

n a

nd

/or

tho

raco

sto

my

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ca

me

ron

et

al.

20

05)

Au

stra

lia

co

ntin

ue

d

P

rop

ort

ion

dis

ch

arg

ed

fro

m E

D

Ph

ysi

cia

n g

rou

p:

14.7

%

Pa

ram

ed

ic g

rou

p:

33.1

%

P=

0.0

00

1

Pro

po

rtio

n o

f tr

au

ma

pa

tie

nts

ad

mitte

d

Ph

ysi

cia

n g

rou

p:

86.7

%

Pa

ram

ed

ic g

rou

p:

68.9

%

P=

0.0

02

Pro

po

rtio

n o

f n

on

-tra

um

a p

atie

nts

ad

mitte

d

Ph

ysi

cia

n g

rou

p:

83.7

Pa

ram

ed

ic g

rou

p:

64.4

P=

0.0

04

Pro

po

rtio

n o

f tr

au

ma

pa

tie

nts

with

ma

xim

um

RTS

Ph

ysi

cia

n g

rou

p:

89.4

Pa

ram

ed

ic g

rou

p:

90.0

Co

mm

en

ts

�

Two

co

nse

cu

tive

ye

ars

fo

r e

ac

h g

rou

p

we

re e

xam

ine

d in

th

e c

ha

rt r

ev

iew

.

�

A s

ec

on

d r

ev

iew

er

va

lida

ted

th

e

ch

art

ext

rac

tio

n in

40 r

an

do

mly

sele

cte

d c

ha

rts

(10.7

% o

f th

e t

ota

l).

Exc

elle

nt

leve

l of

ag

ree

me

nt

ac

hie

ve

d (

Ka

pp

a 0

.937

).

�

All

pa

tie

nts

tra

nsp

ort

ed

to

th

e s

am

e

ba

se h

osp

ita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

sim

ilaritie

s in

ou

tco

me

s fo

r a

dm

itte

d

pa

tie

nts

su

pp

ort

th

e v

iew

th

at

bo

th g

rou

ps

ha

ve

sim

ilar

task

ing

crite

ria

fo

r h

igh

ac

uity

pa

tie

nts

an

d s

ug

ge

st t

ha

t p

ara

me

dic

s a

re

as

eff

ica

cio

us

as

ph

ysi

cia

ns

in d

eliv

erin

g

pre

-ho

spita

l ca

re in

th

is g

rou

p o

f p

atie

nts

.

Ho

we

ve

r, f

or

low

er

ac

uity p

atie

nts

, th

ere

is a

sta

tist

ica

lly s

ign

ific

an

t h

igh

er

rate

of

clin

ica

lly

un

ne

ce

ssa

ry t

ask

ing

s b

y t

he

am

bu

lan

ce

gro

up

. G

ive

n t

he

re

ce

nt

fata

l ae

rom

ed

ica

l

ac

cid

en

ts in

Qu

ee

nsl

an

d it

wo

uld

se

em

pru

de

nt

to r

ed

uc

e c

linic

ally

un

ne

ce

ssa

ry

retr

ieva

ls t

hro

ug

h c

linic

al c

oo

rdin

atio

n w

ith

ap

pro

pria

tely

qu

alif

ied

em

erg

en

cy

ph

ysi

cia

ns.


62


Five studies were identified that were eligible for examining the question “In adults and children with a medical or trauma related emergency, does the presence of a medical crew able to perform rapid sequence intubation and/or thoracostomy improve health outcome when compared with a medical crew unable to perform rapid sequence intubation and/or tube thoracostomy and/or thoracotomy?” In three studies the mortality rate was lower in the group receiving care from crews able to provide at least one of the stipulated procedures (Baxt and Moody 1987; Garner et al. 1999; Koefoed-Nielsen et al. 2002). One other study demonstrated mixed results when stratified by admission to ICU (Lee et al. 2003) and the final study found no significant difference in mortality or hospital length of stay between the two groups (Cameron et al. 2005).

The included studies were only partially helpful in answering this review question as the additional procedures (rapid sequence intubation, tube thoracostomy or thoracotomy) were performed by crews with a doctor present. That is, there were no studies that compared one non doctor crew that was able to perform any of these procedures with another non doctor crew that were not able to perform those procedures. Therefore, although the overall results suggest an improved outcome when being attended by crews that were able to perform at least one of rapid sequence intubation, tube thoracostomy or thoracotomy, there could be other reasons for any such improved outcome. For example, if the assessment process differed between crews and subsequent management decisions differed based on the assessments made by the different groups then this may explain any difference in outcome (rather than a difference in procedures).

There were other inherent limitations in the five studies selected. There was variation in study design among the three studies that suggested some benefit from being attended by crews with the additional procedural capabilities. One used a pseudorandomised controlled trial design (Baxt and Moody 1987), one used a retrospective design (Garner et al. 1999) and the other used a before and after design (Koefoed-Nielsen et al. 2002). In relation to the pseudorandomised controlled trial, a particular consideration was the use of limiting the mortality outcome to deaths that were thought to be directly due to the trauma or complications of the trauma. This may have produced outcome misclassification with the potential for underestimating mortality in either group. The use of all-cause mortality as an outcome would have avoided this limitation. There were also limitations to the registry based study but the effect of these limitations on the study estimates was not clear. Firstly, a significant proportion of eligible patients were excluded from the doctor group due to missing charts. There were no deaths among these patients, thus the level of reduced mortality in the doctor group may have been underestimated. However, time to arrival on the scene was shorter in the doctor group. This may have resulted in an improved prognosis in the doctor group and may not be replicated in other settings where time to arrival of the doctor group may be delayed. Perhaps most significantly, the two groups (doctor and no doctor) were located in two different countries so there may have been other reasons that explain differences in outcomes other than the personnel supplying pre-hospital care. For example, level of hospital care may vary between the two settings. In the retrospective study (Garner et al. 1999), seven deaths in the non-doctor group were omitted due to missing case sheets, potentially underestimating the effectiveness of care provided by doctor crewed helicopters. Patients were also directed to different hospitals in this study depending on the helicopter crew mix. The before and after study was restricted to patients with acute MI. The before and after design (Koefoed-Nielsen et al. 2002) is the biggest limitation since it is not possible to be certain the improved outcome was a result of the different crewing mix or other factors that had changed over time. Based on the above limitations there is uncertainty about the robustness of the findings in these three studies.

Cameron et al. (2005) published a retrospective chart review. This study found no significant difference in outcome between study groups. However, it was limited by the inclusion of limited injury severity data and low study power to detect a difference in mortality.

The other study was set in Australia (Lee et al. 2003). The study population consisted of patients with severe blunt trauma. Three types of pre-hospital care were studied. The most basic level of life support was used as the reference category and was compared with another non-medical configuration and a configuration with a doctor on board. It was only the latter group that was able to perform rapid sequence intubation or tube thoracentesis. The results varied by whether the patients were subsequently admitted to ICU. The odds of mortality were significantly higher in the two more advanced pre-hospital configurations when the patient was not admitted to ICU. However, when the patient was admitted to


63

ICU there were no statistically significant differences between pre-hospital care groups. It should be noted that a selective dispatch strategy was used so the more severely injured patients were seen by the more advanced ambulance group and the doctor group. The authors also commented that paramedics may have involved the physician group when patient death was imminent. These factors may explain the poorer prognosis in the two more advance pre-hospital care groups among the patients who did not proceed on to ICU care.

As in previous sections, more research is required to answer this question. A useful study would be to compare doctors versus fully trained non-doctor personnel. The study should be adequately powered to detect equivalence in outcome. Other considerations include

1. Whether there may be variation in outcome across crew mixes for different clinical scenarios. For example, the inclusion of doctors on road ambulances may be associated with longer at scene times. This longer period of stabilisation could be associated with improved outcome in some circumstances but not others.

2. Comparing outcomes in paediatric and adult age groups and stratifying results by ISS score.

Outcomes by time from ambulance call out to emergency department delivery

From the search strategy for question four (outcome by time from ambulance callout to emergency department delivery) we identified, 1863 potentially relevant articles/abstracts of which 152 were retrieved. Of these retrieved articles, 132 were excluded. These papers, annotated with the reason for exclusion, are presented in Appendix 6. One study was identified from reference lists and was included in the papers for appraisal. Reasons for exclusion of studies before retrieval in full text are outlined in Table 14. Reasons for exclusion of studies retrieved in full text are detailed in Table 15.

Table 14 Reasons for exclusion of studies before retrieval in full text: time from callout to

emergency department delivery











Neonatal study 1

Duplicate abstract 1

Wrong or uncertain time component presented 10

Total 1711

Table 15 Reasons for exclusion of studies retrieved in full text: time from callout to emergency

department delivery











Neonatal study 0

Wrong or uncertain time component presented 55

Duplicate study 1

Total 132


64

Twenty-one retrieved articles were appraised and are listed in the references and Appendix 7. Included papers are presented in the evidence table below. Included studies were all level III-2 according to NHMRC’s hierarchy of evidence, including 21 cohort studies and one case control study.

Pepe et al. 1987

Pepe et al. (1987) investigated the outcome of patients following penetrating trauma. The 498 consecutive patients were stratified into four groups based on trauma scores and survival was assessed by total pre-hospital time across these four trauma group categories. Chi square analysis was used to compare survival by four groups of total pre-hospital time. The sample size varied across the four categories of trauma score (35-317). The study was set in Texas, average age was 31 years and the mean total pre-hospital time was 32.6 minutes. There was no significant difference in survival across the four categories of total pre-hospital time within each trauma score stratum.


� stratifying by trauma score reduced study power. The trauma score category with the largest number of participants consisted of patients with >90% probability of surviving (based on TRISS methodology).

� the observational design is susceptible to confounding. Although the authors have stratified by trauma score, this measure provides a physiological assessment (rather than anatomical assessment) and can change rapidly over time. Therefore, injury severity may vary within the trauma score strata. Although some reassurance is provided by the relative uniformity of predicted survival (based on TRISS methodology) within trauma score strata, small sample sizes in some categories mean residual confounding could remain as an issue.

� the results of this study can only be applied to patients with penetrating trauma given the selection criteria for the study population.

Schiller et al. 1988

Schiller et al. (1988) compared helicopter with ground transportation in a group of patients with blunt trauma and ISS of 20-39. This was a retrospective chart review and total pre-hospital time had to be estimated in 15% of patients due to lack of recording. There were 259 patients transported by ground ambulance and 347 by helicopter. Statistical analysis was restricted to univariate analyses by Student’s t test and chi square test. The mean mission time for ambulances was 39 minutes and 50 minutes for the helicopters. Mortality was significantly higher in the helicopter group (18% versus 13%, P<0.05). There was no difference in length of stay between the two groups (26 days in both groups). There was no comparison between pre-hospital time and the outcome of interest.

Potential sources of bias in this study included:

� the arbitrary estimation of pre-hospital time in 15% of participants.

� the ecological nature of comparing two different transportation groups and basing any association with pre-hospital time on the mean pre-hospital times of the two groups.

� the lack of control over potential confounders.

� uncertainty whether consecutive patients were used in the study.

� uncertainty whether any difference in mortality between the two groups is due to a difference in pre-hospital time or other factors such as the dispatch strategy. The skill mix may also result in bias in the comparative outcome estimates if the crews in the two different modes of transport have different levels of skill.

Sloan et al. 1989

A study set in Chicago investigated the effects of taking trauma patients directly to a level 1 trauma centre whether or not they bypassed other hospitals (Sloan et al. 1989). Thus two groups were formed: a direct group and a bypass group with the latter group bypassing other hospitals en route to the level 1 trauma centre. On that basis the total run time (time from dispatch to trauma centre arrival) varied by patient. Data comparing the mortality and survival groups were presented including total run time in the two groups. Eligibility criteria included one of the following three categories:


65

1. Life threatening injury, including traumatic arrest, penetrating neck trauma, and/or blunt or penetrating chest or abdominal trauma with systolic BP < 100mmHg

2. Field trauma score ≤ 12

3. Limb threatening injury.

Patients were excluded if they arrived at the trauma centre without vital signs and were unable to be resuscitated or if outcome data were missing.

There was no significant difference in total run time between the mortality and survival groups (32 minutes versus 35 minutes respectively) although the study had 90% power to detect a 6% difference in survival.

However, there were limitations:

� the observational design was susceptible to confounding and the method of analysis did not help to control confounding (multivariate or stratified analyses were not presented)

� the method of deriving time data was unclear therefore the accuracy of the total run time was uncertain

� fifty five (21%) patients were excluded, including 48 who presented to the trauma centre without vital signs.

Schwartz et al. 1990

Schwartz et al. (1990) compared air and ground ambulance programmes to determine whether pre-hospital time or pre-hospital care was the major contributor towards survival in a group of patients following blunt trauma. All patients were transported directly from the scene to definitive care. There were 126 patients, 93 transported by air and 33 by ground. The authors extracted data from three registries and chart review.

The average pre-hospital time was longer in the air group (65 minutes versus 34 minutes). However, Z scores (based on TRISS methodology via comparison with the MTOS cohort) found a significantly improved survival in the air group (Z=2.23) and significantly poorer outcome in the ground group (Z=-2.69).

There were a number of limitations to this study:

� confounding was a potential problem in this observational study. In this study it is likely that the crew mix was a more important determinant of outcome than the pre-hospital time. On this basis the study results are difficult to interpret.

� the basis for dispatching a helicopter rather than a ground crew was not clear. There may have been a potential selection bias as a result, although it is noted that if there is a bias other studies have tended to dispatch helicopters to the more severely unwell patients.

� the degree of match between the study data and the MTOS cohort data was unclear as the M statistic was not documented.

� the ecological nature of comparing two different transportation groups and basing any association with pre-hospital time on the mean pre-hospital times of the two groups.

Sampalis et al. 1992

Sampalis et al. (1992) conducted a study that aimed to compare observed and expected mortality (using TRISS methodology and based on the MTOS cohort) overall and by subgroups (including pre-hospital time, level of pre-hospital care and level of in hospital care). The investigators used a complicated three stage sampling process. The aim of this sampling process was to select patients with severe but survivable injuries. It was unclear if the approach was set prior to selecting the sample. The final sample consisted of 355 patients with a mean ISS of 13.7. However, there was a wide range of ISS scores (1-59) so some patients did not fulfil the severe injury aim of the sampling strategy. There were three broad data analyses:


66

1. Comparison of the overall data (using the Z score and the SMR)

2. A subanalysis of different strata defined on pre-hospital time, pre-hospital care and in-hospital care (using the Z score and the SMR)

3. An adjusted comparison of the above strata using logistic regression.

Overall there were more deaths than expected as reflected in the Z score (6.77, P<0.0001) and SMR (1.81, 95% CI 1.42-2.21). Within strata defined by pre-hospital time (>60 minutes compared with reference of 0-60 minutes) there were greater excess deaths in the > 60 minute group with an SMR ratio (> 60 minutes versus 0-60 minutes) of 6.41 (95% CI 1.69-17.37). Logistic regression also found an association between prolonged pre-hospital time and poor survival (OR 29.9, 95% CI 2.7-33.3).

There were strong aspects to this study, not the least of which was to incorporate a multivariate analysis within the methodology. However, limitations also existed:

� it was unclear from the selection process whether the criteria were preset and just how representative the selected sample was of the group the investigators were aiming to explore. The reviewers are unclear why an approach such as stratified random selection based on ISS score were not used.

� there were further sources of selection bias due to missing data. In particular, 30 of 385 potential patients from the final sample were excluded due to missing charts and pre-hospital time data was only available on 270 of 355 eligible participants (76%).

� there were only 13 patients with a pre-hospital time > 60 minutes.

� the accuracy of the pre-hospital time was not clear.

� although a multivariate analysis was conducted this study is still susceptible to confounding by other unknown and uncontrolled factors. However, it is important to recognize that these findings suggest in this study population that prolonged pre-hospital time (> 60 minutes) was associated with poorer outcome even after controlling for different mixes of pre-hospital crews (including the presence of a physician). Furthermore, the level of care provided by the hospital (based on the American College of Surgeons criteria for trauma centre categorisation) was also controlled.


Sampalis et al. (1993) also published a case control study where cases died within 6 days of injury and controls survived more than 6 days after injury. There were similarities to their 1992 study and it is assumed the study populations overlapped. Specifically, the first two stages of the three stage sampling process were the same with the same number of patients selected at the end of sampling stage 2. The difference was at stage 3 of the sampling process where the selection was based on specific criteria to fulfil either a case or a control. More detail is provided in Table 16. There were 72 cases and 288 controls. Again the study focused on three factors, total pre-hospital time, the crew mix at the pre-hospital phase and the level of care provided in-hospital. Both univariate and multivariate analyses were conducted. A wide range of variables were considered in the stepwise selection of the logistic regression model before finishing with a relatively simple model.

There were increased odds of survival of more than six days in association with a pre-hospital time under 60 minutes (OR 3.01, 95% CI 1.27-5.06).

Similar quality considerations apply to this study as they did with the 1992 study. There were strong aspects to this study, not the least of which was to incorporate a multivariate analysis within the methodology. However, limitations also existed:

� it was unclear from the selection process whether the criteria were preset and just how representative the selected sample was of the group the investigators were aiming to explore. The reviewers are unclear why an approach such as stratified random selection based on ISS score was not used.

� there were further sources of selection bias due to missing data. In particular, 34 of 337 potential patients with a PHI>3 were excluded from the final sample due to missing charts.

� the accuracy of the pre-hospital time was not clear.


67

� although a multivariate analysis was conducted this study is still susceptible to confounding by other unknown factors. However, it is important to recognize that these findings suggest in this study population that prolonged pre-hospital time (> 60 minutes) was associated with poorer outcome even after controlling for different mixes of pre-hospital crews (including the presence of a physician). Furthermore, the level of care provided by the hospital (based on the American College of Surgeons criteria for trauma centre categorisation) was also controlled. It is interesting to note the odds ratio in this study was much smaller than in the previous study, although still indicating that pre-hospital time more than 60 minutes appears to have an adverse outcome on survival.

Bonatti et al. 1995

Bonatti et al. (1995) aimed to identify predictors of short-term survival in their Austrian setting. The study was based on a HEMS unit that included physician staffing. There were no specific eligibility criteria listed but it is presumed that all missions between 1989 and 1991 were included. There were 2139 participants with a preponderance of sporting injuries (53.7%). Data were extracted from medical records, flight logs and discharge summaries. Univariate and multivariate analyses were conducted.

The univariate analyses showed decreasing survival with increasing total mission time (P=0.0001):

� 0-20 minutes: 95.5% survival




� > 80 minutes: 78.8% survival.

However, on multivariate analysis, there was no association between total mission time and survival after adjusting for cause of injury/emergency, flight time to scene, scene time, patient age, patient gender, NACA score, state of consciousness, respiratory status, circulatory status, emergency physician.

There was relatively little methodological data given. It is difficult to assess the probability of selection bias given the lack of details about the process although it should be recognised that there were a high proportion of sporting injuries in this population. The accuracy of data recording in the sources used for data extraction was not documented. Adjusting for components of the total mission time in the multivariate model may have contributed to the lack of association observed. Finally, despite the use of a multivariate model, the possibility of residual confounding cannot be discounted.

Feero et al. 1995

This group used a different study design where they compared unexpected survivors with unexpected deaths (based on TRISS methodology), (Feero et al. 1995). While there were 848 eligible trauma victims in the study time period, there were only 13 unexpected survivors and 20 unexpected deaths. The study relied on identifying patients from a local registry. Entry on this registry consisted of a set of mandatory criteria and a set of optional criteria. The mandatory criteria were divided into physiological, anatomical and mechanism of injury characteristics. The key result was a prolonged total EMS time interval in the unexpected death group when compared with the unexpected survivor group (29.3 minutes compared with 20.8 minutes, P=0.02). However, the unexpected death group was also significantly older than the unexpected survivor group (50.8 years versus 29.5 years, P=0.01).

Study limitations included:

� while the TRISS methodology was useful for identifying the two unexpected outcome groups (and there was a high degree of concordance between the reference population and the actual population), the lack of control for potential confounders in this study was problematic

� there were small numbers in the two study groups of interest (total of 33 patients)

� probability of selection bias due to the lack of consistent use of objective criteria for study inclusion (a reflection of the retrospective design that made use of registry data).


68

Young et al. 1998

Young et al. (1998) presented a study comparing mortality and length of hospital stay across two groups: trauma patients transported directly to a level 1 trauma centre versus trauma patients transferred via another hospital en route to the level 1 trauma centre. The study was restricted to patients over 18 years with an ISS greater than 15. The study used a retrospective design and relied on registry plus hospital note data. The time from injury to arrival at the trauma centre averaged 480 minutes in the transfer group and 92 minutes in the direct group. There was no significant difference in length of hospital stay or overall mortality between the two groups. The authors studied two mortality subgroups (deaths within 24 hours and deaths more than 24 hours after injury). Unexpected deaths (based on TRISS methods) were analysed further in the group of deaths in the first 24 hours and found a higher proportion of unexpected deaths in the transfer patients (75% versus 28%, P<0.05).

The study had significant limitations

� there were discrepancies in the number of deaths within the paper (three deaths were not accounted for when categorizing to the first 24 hours and more than 24 hours after injury)

� M statistic indicated a poor match between the reference data and the actual data casting doubt on the validity of the probability of survival data

� no direct comparison between pre-hospital time and outcome

� observational study design that is susceptible to confounding.

Frezza et al. 1999

A study set in the USA examined the effect of pre-hospital time on outcome in patients undergoing emergency room thoracotomy (ERT), (Frezza and Mezghebe 1999). There were 58 adult patients with penetrating chest trauma, although the analysis was restricted to 33 of these patients who actually received ERT. Twenty-four hour survival was higher in the group with a pre-hospital time <30 minutes compared with the group with a pre-hospital time >30 minutes (20/27 (63%) versus 0/6 (0%)).

This study had significant limitations making it difficult to interpret the above results. In particular:

� scant details were presented in the methods

� there were apparent deaths > 24 hours post injury thus incorporation of those deaths in the analysis of pre-hospital time would have reduced the difference in survival between the two groups

� the components of the pre-hospital time recorded were not documented so it is unclear if this time refers to time from dispatch of emergency services to time of arrival at hospital

� data were missing on nine patients and a further 16 were excluded due to lack of vital signs being recorded from the field.

Phillips et al. 1999

Phillips et al. (1999) compared the outcome in patients transported by road and air ambulance. The aim of this study was to assess outcome against national standards. Patients retrieved by air ambulance had, on average, a longer period of pre-hospital time than the road ambulance group (77 minutes versus 54 minutes). The air ambulance crew was permitted to conduct more advanced procedures than the road ambulance crew so therefore were preferentially given the more severely injured patients to transport.

As background, based on TRISS methodology, the mean predicted survival rate was 93.9% in the road ambulance group and 83.1% in the air ambulance group. There was a significantly longer hospital stay in the air ambulance group than the road ambulance group (4.21 days versus 8.97 days, P<0.001). Estimation of Z scores (based on comparison with the MTOS cohort using TRISS methodology) showed no significant difference in actual versus predicted mortality either within the road ambulance group or the air ambulance group. The M statistic was not estimated so the degree of fit between the MTOS cohort and the study cohort was not clear.

Other limitations also existed. The most significant was that this study was not designed to compare pre-hospital time against outcome thus the differences in outcome (especially hospital length of stay) could be fully explained by the increased severity of injury in the air transport group. It is not possible


69

to determine whether the prolonged hospital stay in the air ambulance group was partially due to the longer pre-hospital period in this group. There were other limitations that are outlined in Table 16. These limitations are such that it would be inadvisable to form any conclusions regarding the effect of different pre-hospital times on patient outcome based on this study.


Following the two earlier studies by Sampalis et al. (1992; 1993) a programme of regionalization of the trauma services occurred. This cohort study aimed to assess the impact of regionalization on mortality. It also investigated the association between pre-hospital time and mortality. There were 12,208 patients included in the study. Selection criteria were

� treated for injuries at acute care hospitals in Montreal and Quebec

� one of:

• Death as a result of injury

• ISS>12

• At least two injuries with AIS ≥3

• Hospital stays > 3 days.

Patients who died at the scene were excluded.

The analytic method varied by hypothesis tested. In relation to the association between pre-hospital time and mortality, logistic regression was used. The mean age of the sample was 48 years. This decreased over the six study years. Mean ISS was 26.1. The majority were discharged alive (72%).

During the six years of the study 3,453 (28%) of the patients died. The adjusted odds ratio for the association between each additional minute of pre-hospital time and mortality was 1.046 (95% CI 1.044-1.050). This was adjusted for time to admission, trauma centre designation, transfer versus direct transport, patient age and ISS. This result indicates a 5% increase in risk of mortality for every minute’s pre-hospital delay.

This was a well conducted study and it had the advantage of a large sample size. Any biases present are likely to be small and could arise for inaccurate pre-hospital times, confounding (given the observational design) and potential selection biases resulting from the selection methods.

Grzybowski et al. 2000

This study compared patients surviving at least seven days with deaths within that time period after presenting with chest pain or shortness of breath (Grzybowski et al. 2000). Ninety-six percent of patients had an acute MI. The study population was restricted to patients 18 years and over and nine patients were excluded due to missing outcome data. There were 244 patients in the final study population. Bivariate comparison found a significantly longer total pre-hospital time in the deaths than

the survivors (50.6 minutes versus 42.8 minutes, P≤0.01).

Limitations of this study included:

� while multivariate regression was used, indices of pre-hospital time were not included in this model, so there was no control over potential confounders in the pre-hospital time to outcome relationship

� it is unclear if the prolonged time in the non-survivors may have been due to more severe disease requiring immediate management on the scene.

Berns et al. 2001

Berns et al. (2001) compared outcome amongst a group of cardiac patients transported by helicopter with another group transported by road ambulance. There were 266 helicopter patients and 28 road ambulance patients. Most patients were transferred from referring emergency departments rather than the site of onset of cardiac symptoms and the pre-hospital time reflected time from the request for


70

transfer to arrival at the definitive treatment hospital. Data on hospital length of stay and mortality were presented.

The pre-hospital times were a mean of 104 minutes in the helicopter group and 142 minutes in the ground ambulance group. Hospital length of stay was significantly shorter in the helicopter group (6.4 versus 8 days, P=0.04) but there was no significant difference in mortality between the two groups (7% in helicopter group and 4% in ground group).

In regards to the review question of interest, this study was difficult to interpret as the difference in hospital length of stay may have reflected differences in staffing mixes by the two different forms of transport. The study was underpowered to detect a difference in mortality. There were no comparisons with individual pre-hospital time data further limiting this study.

Clarke et al. 2002

This study focused on patients with severe abdominal trauma who had systolic BP < 90mmHg on ED arrival and either died in ED or were transferred to the operating room for laparotomy. There were 243 patients included in the study. Interval risk ratios for death were estimated based on pre-hospital time. All risk ratios included one indicating no increased risk of death for any pre-hospital time interval. However, the time interval 31-60 minutes neared significance (RR 1.268, 95% CI 0.980-1.641). There were 117 patients in this time category (including 63 who survived and 54 who died).

There were potential sources of selection bias (omission of patients who did not have appropriate time intervals recorded in the registry and exclusion of patients who had extreme time intervals), misclassification (due to reliance on registry data that the authors may have been frequently rounded to the nearest five minutes) and confounding based on the observational design and lack of analyses that could control for known confounding factors. In general, the study was well conducted but the main focus was on ED time rather than pre-hospital time.

Lim et al. 2002

A study set in Singapore investigated survival following out of hospital cardiac arrest (OHCA), (Lim and Seow 2002). This study divided the 93 patients into two groups: patients who survived post ED resuscitation and patients who did not survive beyond this time. There were 15 survivors using this classification although only one survived to hospital discharge. All patients with non-traumatic OHCA presenting during a three month period from November 2001 were included. Data were extracted from various records, including ambulance case records, ED resuscitation charts and in-patient records. Univariate analyses were conducted using the t test and the chi square test. There was no significant difference in total pre-hospital time between the survivors and non-survivors (38.3 minutes versus 35.4 minutes respectively).


� the sample size was low so the power to detect a significant difference in pre-hospital time across the two groups was low

� there is likely to be misclassification of the total pre-hospital time since the recordings were only documented once the ambulance arrived at the hospital

� the combination of the observational design and the lack of multivariate analysis mean the results are highly susceptible to confounding.

Osterwalder et al. 2002

Osterwalder et al. (2002), in their Swiss based study, aimed to test the hypothesis that exceeding the 60 minute limit for the entire pre-hospital time increases mortality of blunt trauma patients. There were 254 participants including 107 with a rescue time up to 60 minutes and 147 with a rescue time over 60 minutes. All the blunt trauma patients were treated at a single hospital (St Gallen Cantonal Hospital) and had an AIS of at least two for at least two of six defined body regions. The actual 30 day mortality was compared with predicted mortality (based on ASCOT score). Flora’s z statistic was used to compare actual with expected mortality. Univariate and multivariate analyses (logistic regression) were also conducted. The adjusted odds ratio found increased odds of dying in the group with a transport


71

time up to 60 minutes compared with the group with a transport time more than 60 minutes (OR 8, 95% CI 1.7-38.5).


� the observational design is susceptible to confounding and while a multivariate analysis was conducted, the variables included in the model were not documented. There were differences between the two groups at baseline

� while consecutive patients were included in the study, 9% were omitted due to missing time data.

Lerner et al. 2003

Lerner et al. (2003) conducted a chart review examining the association between total pre-hospital time and mortality. Charts were selected for patients who were transported directly to the study hospital and were either admitted to hospital or died in ED. Patients were excluded if there was incomplete data, more than one day between the time of injury and time of admission, CPR was initiated in the field, or the patient was transported from a correctional facility. Most of the data were extracted from the trauma registry. Univariate and multivariate analyses (logistic regression) were conducted. There were 1877 participants.

On univariate analysis the total pre-hospital time was longer in the survivors than the non-survivors (35.26 minutes versus 31.58 minutes, difference 3.69 minutes, 95% CI -.52-6.85 minutes). There was no association between total out of hospital time and mortality on multivariate analysis (OR 0.987, 0.97-1.00).

Limitations of the study included:

� the limitations associated with chart reviews including missing data and inconsistent recording of data

� the authors commented they expected the pre-hospital time to be randomly misclassified, thus diluting any association

� approximately 20% of the study population had to be omitted due to missing data, resulting in selection bias

� the design is susceptible to confounding although the use of logistic regression helps to control this.

Biewener et al. 2004

Biewener et al. (2004) studied four pathways for the transportation of polytrauma patients:

1. Helicopter transportation to a level 1 trauma centre (HEMS-UNI group).

2. Ambulance transportation to a level 1 trauma centre (AMB-UNI group).

3. Ambulance transportation to a level 2 or 3 trauma centre (AMB-REG group).

4. Ambulance transportation to a level 2 or 3 trauma centre followed by transfer to a level 1 trauma centre (INTER group).

In the context of this review, the first two categories were of interest. The AMB-REG group did not fulfil the criteria for definitive care and the pre-hospital times were not stated for the INTER group.

There were 403 patients in total, including 140 in the HEMS-UNI group and 70 in the AMB-UNI group. Inclusion was limited to patients with an ISS at least 16, alive at time of hospital arrival and had complete documentation of all patent data. Patients over 75 years were excluded as were patients with an ISS score over 67. There were no differences in age, gender or ISS across the groups.

The mean pre-hospital time in the HEMS-UNI and AMB-UNI groups were 90 minutes and 68 minutes respectively. There was no significant difference in mortality between the two groups on logistic regression (controlling for study group, age and ISS).


72


� in relation to the review question of interest our primary interest is with pre-hospital time whereas the authors’ primary interest was with comparing different transport pathways. As a consequence there was no comparison between different pre-hospital times at an individual level. This also had the consequence of lack of control over other differences between study groups, such as aggressiveness of management during the pre-hospital phase.

� the study made use of registry collected data but there was no indication of the accuracy of data recording.

� it is noted that there were no cases of penetrating trauma within the study population.

Gao et al. 2006

A retrospective review of polytrauma patients was conducted in China (Gao et al. 2006). There were 15,340 eligible patients during the ten year study period (1993-2003). To be included there needed to be injuries to more than two ISS regions and at least one region needed to include an AIS of at least three. Data were extracted on sex, age, causes of injury, duration of preadmission and injured regions, shock state on admission, amount of blood transfusion, severity of injuries, method of diagnosis, therapeutic procedures. The chi square test was used in data analysis. Mortality was significantly higher in the group with a preadmission time of at least one hour (7.7% versus 3.9%, P<0.01).


� potential limitations of this type of retrospective review, including inconsistent recording of data (e.g. some may round pre-hospital time and others may not) and uncertainty about the accuracy of data

� no control of potential confounders (absence of multivariate analysis in this observational design)

� potential lack of applicability to other populations.

Hartl et al. 2006

Hartl et al. (2006) explored the effect of pre-hospital management decisions on early mortality following severe traumatic brain injury (TBI). A major focus was on the decision to indirectly transport patients to definitive care via an intermediate hospital. Nevertheless the investigators also examined the pre-hospital time and its influence on two week mortality. Patients were restricted to those with head injury with GCS<9 for at least six hours after injury and arrival at a level 1 or level 2 trauma centre within 24 hours of injury. More detail is provided in Table 16. Data were extracted from the TBI-trac registry. There was no significant association between pre-hospital time and two week mortality (OR for each minute increase in transport time 1.00, 95% CI 1.00, 1.00).

There were potential sources of bias:

� there were patient exclusions, some of which fulfilled the a priori eligibility criteria and some did not. Specifically from 1449 patients entered in the TBI-trac data base, 1123 were included in the study. From the list of exclusion criteria documented in the study methods, 210 of the 326 exclusions appeared to fulfil those pre-set criteria. There was also incomplete participation of the trauma centres throughout New York State with 54% of all trauma centres participating in the New York State quality improvement program (which was used as the source of trauma centres) being included in the study. These limitations indicate potential selection biases.

� while a multivariate model was developed that included some potential confounders other potential confounders may have been present in this observational study that were overlooked.

� there was no information presented on the accuracy of the registry data. Misclassification may have occurred due to for example coding errors. Misclassification of pre-hospital time may also have occurred if, for example there was a tendency to round times.


73

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Pe

pe

et

al.

198

7)

USA

Co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Su

bje

cts

with

pe

ne

tra

tin

g in

jurie

s

de

live

red

to

a s

ing

le r

eg

ion

al tr

au

ma

ce

ntr

e in

Ho

ust

on

, Te

xas.

Re

spo

nse

are

a

co

ve

rs c

lose

to

700 s

qu

are

mile

s a

nd

tra

nsp

ort

tim

es

to t

he

tra

um

a c

en

tre

ca

n e

xce

ed

30 m

inu

tes.

Pa

rtic

ipa

nts

:

498 c

on

sec

utive

pa

rtic

ipa

nts

Ba

selin

e m

ea

sure

me

nts

Ave

rag

e a

ge

31.3

ye

ars

Me

an

tra

um

a s

co

re 1

1.1

Me

an

ISS 2

0.2

Me

an

re

spo

nse

tim

e 5

.3 m

inu

tes

Me

an

sc

en

e t

ime

15.8

min

ute

s

Me

an

tra

nsp

ort

tim

e 1

1.7

min

ute

s

Me

an

to

tal p

re-h

osp

ita

l tim

e 3

2.6

min

ute

s

Inc

lu/e

xcl c

rite

ria

.

Pe

ne

tra

tin

g t

rau

ma

Syst

olic

BP

≤ 9

0m

mH

g

Tra

nsp

ort

ed

dire

ctly t

o B

en

Ta

ub

Ge

ne

ral H

osp

ita

l

Da

ta c

olle

ctio

n

Fire

De

pa

rtm

en

t re

co

rds

tim

es

of

rec

eip

t o

f c

all,

dis

pa

tch

of

pa

ram

ed

ics,

pa

ram

ed

ics’

arr

iva

l at

inc

ide

nt

loc

atio

n, le

av

ing

in

cid

en

t

loc

atio

n, a

nd

ho

spita

l arr

iva

l. Th

e

tota

l pre

-ho

spita

l tim

e (

TPT)

wa

s

de

fin

ed

as

the

tim

e f

rom

re

ce

ipt

of

the

ca

ll u

ntil th

e t

ime

of

arr

iva

l at

the

tra

um

a c

en

tre

.

Ou

tco

me

me

asu

res

Dis

ch

arg

ed

aliv

e f

orm

ho

spita

l

An

aly

sis

Pa

tie

nts

str

atifie

d in

to f

ou

r g

rou

ps

ba

sed

on

th

eir in

itia

l pre

-ho

spita

l

tra

um

a s

co

re (

1, 2-6

, 7

-11

, 12

-15

).

Ea

ch

of

the

se g

rou

ps

wa

s st

ud

ied

ind

ep

en

de

ntly in

re

latio

n t

o T

PT.

Fou

r g

rou

ps

of

TPT

we

re f

orm

ed

: 0

-

20 m

inu

tes,

21-3

0 m

inu

tes,

31

-40

min

ute

s a

nd

> 4

0 m

inu

tes.

All

pa

tie

nts

we

re a

lso

ass

ess

ed

in

term

s o

f th

eir p

rob

ab

ility

of

surv

iva

l

usi

ng

TR

ISS m

eth

od

olo

gy.

Su

rviv

al ra

tes

we

re c

om

pa

red

usi

ng

ch

i-sq

ua

re a

na

lysi

s (o

r Fi

she

r’s

exa

ct

test

wh

en

n<

75).

Co

ntin

uo

us

va

riab

les

we

re a

sse

sse

d u

sin

g

an

aly

sis

of

va

rian

ce

.

With

in t

he

fo

ur

tra

um

a s

co

re

gro

up

ing

s, t

he

re w

as

no

sig

nific

an

t

diffe

ren

ce

in s

urv

iva

l by T

PT

Tra

um

a s

co

re=

1

TPT

(Min

s)

Ob

serv

ed

(%)

Pre

dic

te

d (

%)

≤20

0

2

21-3

0

0

2

31-4

0

0

2

>40

0

2

Tra

um

a s

co

re=

2-6

TPT

(Min

s)

Ob

serv

ed

(%)

Pre

dic

te

d (

%)

≤20

14

14

21-3

0

31

24

31-4

0

24

28

>40

0

3

Tra

um

a s

co

re=

7-1

1

TPT

(Min

s)

Ob

serv

ed

(%)

Pre

dic

te

d (

%)

≤20

50

71

21-3

0

52

64

31-4

0

52

69

>40

69

70

Tra

um

a s

co

re=

12

-15

TPT

(Min

s)

Ob

serv

ed

(%)

Pre

dic

te

d (

%)

≤20

90

96

21-3

0

91

90

31-4

0

91

92

>40

91

91

Lim

ita

tio

ns

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. Li

mite

d c

on

tro

l of

po

ten

tia

l co

nfo

un

din

g (

stra

tific

atio

n

by p

re-h

osp

ita

l tra

um

a s

co

re).

�

Tim

ing

of

BP

me

asu

rem

en

t u

nc

lea

r

an

d m

ay h

ave

re

sulte

d in

va

riab

le

ap

plic

atio

n o

f se

lec

tio

n c

rite

ria

.

�

Tra

um

a s

co

re is

a p

hysi

olo

gic

al

me

asu

re t

ha

t c

an

ch

an

ge

ove

r tim

e,

thu

s tim

e o

f m

ea

sure

me

nt

is c

ritic

al t

o

en

surin

g c

on

sist

en

cy in

gro

up

ing

�

Sm

all

sam

ple

s in

so

me

ce

lls li

mitin

g

po

we

r to

de

tec

t si

gn

ific

an

t

diffe

ren

ce

s a

cro

ss t

he

tim

e

ca

teg

orie

s.

�

Stu

dy li

mite

d t

o p

en

etr

atin

g in

jurie

s so

lac

ks

ge

ne

ralis

ab

ility

to

blu

nt

tra

um

a

an

d m

ed

ica

l em

erg

en

cie

s.


74

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Pe

pe

et

al.

198

7)

USA

co

ntin

ue

d

Co

mm

en

ts

�

Aim

ed

to

exa

min

e t

he

re

latio

nsh

ip

be

twe

en

th

e s

urv

iva

l ra

tes

of

pa

tie

nts

with

pre

sum

ed

ha

em

orr

ha

gic

sh

oc

k

du

e t

o p

en

etr

atin

g in

jurie

s a

nd

th

e

tota

l pre

-ho

spita

l tim

e r

eq

uire

d t

o

ma

na

ge

an

d d

eliv

er

tho

se p

atie

nts

to

a s

ing

le r

eg

ion

al tr

au

ma

ce

ntr

e in

a

larg

e u

rba

n a

rea

.

�

Pa

ram

ed

ics

op

era

te u

nd

er

a

ma

na

ge

me

nt

pla

n.

Ke

y p

oin

ts f

or

pe

ne

tra

tin

g in

jurie

s in

clu

de

co

ntr

ol o

f

an

y s

ign

ific

an

t e

xte

rna

l ha

em

orr

ha

ge

,

ag

gre

ssiv

e a

irw

ay m

an

ag

em

en

t

(pre

fera

bly

by in

tub

atio

n),

larg

e b

ore

IV c

an

nu

latio

n e

n r

ou

te, p

atie

nts

pla

ce

d o

n b

ac

kb

oa

rds

an

d c

erv

ica

l

spin

e is

im

mo

bili

sed

. P

rim

ary

go

al i

s

rap

id e

va

cu

atio

n. P

atie

nt’

s c

on

ditio

n

is r

ad

ioe

d e

n r

ou

te.

�

Co

nse

cu

tive

pa

rtic

ipa

nts

se

lec

ted

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

tim

e f

ac

tor

invo

lve

d in

ma

na

gin

g a

nd

tra

nsp

ort

ing

hyp

ote

nsi

ve

pe

ne

tra

tin

g in

jury

vic

tim

s d

ire

ctly t

o a

re

gio

na

l tra

um

a c

en

tre

do

es

no

t a

pp

ea

r to

be

re

late

d t

o a

n

ad

ve

rse

ou

tco

me

, a

t le

ast

du

rin

g t

he

first

ho

ur

of

inju

ry.


75

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hill

er

et

al.

198

8)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Pa

tie

nts

ad

mitte

d t

o t

he

tra

um

a c

en

tre

at

St

Jose

ph

’s H

osp

ita

l, P

ho

en

ix,

Arizo

na

du

rin

g 1

983-1

98

6.

Ap

pro

xim

ate

ly 8

00

Ca

teg

ory

1 t

rau

ma

pa

tie

nts

ad

mitte

d

an

nu

ally

.

Pa

rtic

ipa

nts

:

606 p

art

icip

an

ts w

ith

25

9 t

ran

spo

rte

d b

y

gro

un

d a

mb

ula

nc

e a

nd

347

by

he

lico

pte

r.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ag

e (

ye

ars

):

Am

bu

lan

ce

31

He

lico

pte

r 3

0

Ma

le (

%)

Am

bu

lan

ce

74

He

lico

pte

r 7

8

Me

an

tra

um

a s

co

re

Am

bu

lan

ce

12.7

He

lico

pte

r 1

2.1

Me

an

GC

S

Am

bu

lan

ce

10.4

He

lico

pte

r 9

.6

Inc

lu/e

xcl c

rite

ria

.

ISS 2

0-3

9

Blu

nt

tra

um

a

Da

ta c

olle

ctio

n

Mo

de

of

tra

nsp

ort

atio

n,

site

of

orig

in, e

lap

sed

tim

e o

f th

e r

esc

ue

mis

sio

n, G

CS, a

ge

, g

en

de

r a

nd

inju

rie

s.

Ou

tco

me

me

asu

res

Ho

spita

l da

ys

Mo

rta

lity

An

aly

sis

Stu

de

nt’

s u

np

aire

d t

te

st, c

hi s

qu

are

test

Me

an

mis

sio

n t

ime

s (m

inu

tes)

Am

bu

lan

ce

39

He

lico

pte

r 5

0

Mo

rta

lity (

%)

Am

bu

lan

ce

13

He

lico

pte

r 1

8

P<

0.0

5

Ho

spita

l le

ng

th o

f st

ay (

da

ys)

Am

bu

lan

ce

26

He

lico

pte

r 2

6

Lim

ita

tio

ns

�

Tota

l pre

-ho

spita

l tim

e f

req

ue

ntly

mis

sin

g.

Wh

en

ne

ce

ssa

ry (

~1

5%

), t

his

wa

s e

stim

ate

d b

y d

ou

blin

g t

he

tim

e

fro

m t

he

sc

en

e a

nd

ad

din

g 1

0

min

ute

s. T

hu

s p

re-h

osp

ita

l tim

e is

sub

jec

t to

mis

cla

ssific

atio

n a

nd

th

e

dire

ctio

n o

f a

ny b

ias

is u

nc

lea

r

alth

ou

gh

th

e a

uth

ors

fe

lt t

his

wo

uld

un

de

rest

ima

te m

issi

on

tim

e in

th

e

he

lico

pte

r g

rou

p.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

.

Sp

ec

ific

ally

, th

ere

wa

s n

o c

om

pa

riso

n

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d

ou

tco

me

, ra

the

r th

e c

om

pa

riso

n w

as

be

twe

en

he

lico

pte

r a

nd

gro

un

d

tra

nsp

ort

atio

n.

�

Re

tro

spe

ctive

an

aly

sis.

�

Un

cle

ar

if c

on

sec

utive

pa

tie

nts

we

re

use

d.

�

No

do

cu

me

nta

tio

n a

bo

ut

the

exp

erie

nc

e a

nd

skill

s o

f th

e c

rew

s in

the

air v

ers

us

gro

un

d c

om

pa

riso

n.

�

No

do

cu

me

nta

tio

n o

f th

e b

asi

s fo

r

de

cid

ing

wh

eth

er

a h

elic

op

ter

or

gro

un

d a

mb

ula

nc

e s

ho

uld

be

dis

pa

tch

ed

.


76

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hill

er

et

al.

198

8)

USA

co

ntin

ue

d

Co

mm

en

ts

�

Prim

ary

aim

wa

s to

ass

ess

wh

eth

er

tra

nsp

ort

atio

n o

f p

atie

nts

with

ISS 2

0-

39 b

y h

elic

op

ter

resu

lte

d in

im

pro

ve

d

surv

iva

l wh

en

co

mp

are

d w

ith

gro

un

d

tra

nsp

ort

atio

n.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

re is

no

su

rviv

al a

dva

nta

ge

in t

he

he

lico

pte

r tr

an

spo

rte

d g

rou

p in

an

urb

an

are

a w

ith

a s

op

his

tic

ate

d p

re-h

osp

ita

l ca

re

syst

em

. P

atie

nts

of

rura

l orig

in d

ese

rve

furt

he

r st

ud

y.


77

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Slo

an

et

al.

19

89

)

USA

Co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Stu

dy in

Ch

ica

go

th

at

co

mp

are

d

tra

um

a p

atie

nts

tra

nsp

ort

ed

dire

ctly t

o

the

ne

are

st le

ve

l 1 t

rau

ma

ce

ntr

e w

ith

pa

tie

nts

wh

o h

ad

to

byp

ass

oth

er

em

erg

en

cy d

ep

art

me

nts

to

ge

t to

th

e

ne

are

st le

ve

l 1 t

rau

ma

ce

ntr

e S

tud

y

pe

riod

: M

arc

h t

o N

ove

mb

er

19

87.

Pa

rtic

ipa

nts

(n

=2

03

):

Inte

rve

ntio

n (

n=

66

).

Dire

ct

tra

nsp

ort

gro

up

. Ta

ke

n d

irec

tly t

o

Co

ok C

ou

nty

Ho

spita

l with

ou

t

byp

ass

ing

an

y o

the

r h

osp

ita

ls.

Co

mp

ara

tor

(n=

13

7).

Re

qu

ire

d h

osp

ita

l byp

ass

. P

atie

nts

byp

ass

ed

oth

er

ho

spita

ls o

n r

ou

te t

o

Co

ok C

ou

nty

Ho

spita

l.

Ba

selin

e a

na

lysi

s

Ave

rag

e a

ge

26 y

ea

rs

Ma

le 8

3%

Blu

nt

tra

um

a 5

7%

Op

era

tive

tra

um

a 6

3%

Ave

rag

e ISS 1

7

Ho

spita

l tra

um

a s

co

re 1

3

Ave

rag

e t

ota

l ru

n t

ime

35 m

inu

tes

Ave

rag

e h

osp

ita

l sta

y 1

2 d

ays

Inc

lu/e

xcl c

rite

ria

.

Fell

into

on

e o

f th

e f

ollo

win

g t

hre

e

ca

teg

orie

s:

1.

Life

th

rea

ten

ing

inju

ry,

inc

lud

ing

tra

um

atic

arr

est

,

pe

ne

tra

tin

g n

ec

k t

rau

ma

,

an

d/o

r b

lun

t o

r p

en

etr

atin

g

ch

est

or

ab

do

min

al tr

au

ma

with

syst

olic

BP

< 1

00

mm

Hg

.

2.

Fie

ld t

rau

ma

sc

ore

≤ 1

2.

3.

Lim

b t

hre

ate

nin

g in

jury

.

Exc

lud

ed

: 1. p

atie

nts

wh

o a

rriv

ed

at

Co

ok C

ou

nty

Ho

spita

l in

tra

um

atic

arr

est

wh

o w

ere

un

ab

le t

o b

e

resu

scita

ted

. 2. p

atie

nts

with

mis

sin

g

ou

tco

me

da

ta.

Da

ta c

olle

ctio

n

Tota

l ru

n t

ime

wa

s th

e t

ime

fro

m

Ch

ica

go

Fire

De

pa

rtm

en

t (C

FD

)

dis

pa

tch

to

arr

iva

l at

tra

um

a

ce

ntr

e. O

ve

rall

pre

-ho

spita

l tim

e

als

o in

clu

de

d d

ela

y t

ime

fro

m t

ime

of

inju

ry t

o C

FD d

isp

atc

h.

Re

gis

try d

ata

use

d t

o e

xtr

ac

t

me

ch

an

ism

of

inju

ry,

initia

l h

osp

ita

l

tra

um

a s

co

re, e

stim

ate

d

ab

bre

via

ted

inju

ry s

ca

le a

nd

inju

ry

seve

rity

sc

ore

.

Ou

tco

me

me

asu

res

Su

rviv

al

An

aly

sis

Pro

po

rtio

ns

co

mp

are

d b

y c

hi

squ

are

te

st. Stu

de

nt’

s t

test

use

d t

o

co

mp

are

diffe

ren

ce

be

twe

en

me

an

s.

Co

mp

aris

on

of

mo

rta

lity a

nd

surv

iva

l gro

up

s

Tota

l ru

n t

ime

(m

inu

tes)

:

Mo

rta

lity g

rou

p 3

2

Su

rviv

al g

rou

p 3

5

no

sig

nific

an

t d

iffe

ren

ce

Lim

ita

tio

ns

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. M

ultiv

aria

te a

na

lysi

s

wa

s n

ot

co

nd

uc

ted

th

us

furt

he

r

limitin

g c

on

tro

l ove

r p

ote

ntia

l

co

nfo

un

de

rs (

alth

ou

gh

th

ere

we

re n

o

sig

nific

an

t d

iffe

ren

ce

s in

ba

selin

e

da

ta b

etw

ee

n d

irec

t a

nd

byp

ass

pa

tie

nts

).

�

Sm

all

nu

mb

er

of

de

ath

s (n

=29

)

red

uc

ing

po

we

r o

f th

e s

tud

y.

�

Ac

cu

rac

y o

f tim

e d

ata

un

ce

rta

in.

�

Inju

ry s

eve

rity

sc

ore

on

ly k

no

wn

fo

r 1

90

of

the

203 p

atie

nts

.

�

Me

asu

rem

en

t o

f tr

au

ma

sc

ore

co

nd

uc

ted

on

ho

spita

l arr

iva

l ra

the

r

tha

n o

n s

ce

ne

.

�

Me

ch

an

ism

of

tra

um

a d

oc

um

en

ted

in

196 o

f 20

3 p

atie

nts

.

�

Un

suc

ce

ssfu

lly r

esu

scita

ted

tra

um

atic

arr

est

wa

s th

e r

ea

son

fo

r e

xclu

sio

n in

48 (

19%

) w

ith

hig

he

r p

rop

ort

ion

exc

lud

ed

fo

r th

is r

ea

son

in d

irec

t

tra

nsp

ort

gro

up

(28

% v

ers

us

14%

,

P<

0.0

5).

Se

ve

n p

atie

nts

exc

lud

ed

du

e

to in

ad

eq

ua

te o

utc

om

e d

ata

.

Co

mm

en

ts

�

Stu

die

d t

he

in

flu

en

ce

of

ho

spita

l

byp

ass

on

pre

-ho

spita

l tim

es

an

d le

ve

l

1 t

rau

ma

pa

tie

nt

surv

iva

l.

�

Stu

dy h

ad

90%

po

we

r to

de

tec

t a

diffe

ren

ce

in s

urv

iva

l of

6%

or

mo

re.


78

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Slo

an

et

al.

19

89

)

USA

co

ntin

ue

d

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

urb

an

use

of

ho

spita

l byp

ass

do

es

no

t

de

cre

ase

tra

um

a p

atie

nt

surv

iva

l in

th

ose

wh

o a

rriv

e a

t th

e t

rau

ma

ce

ntr

e w

ith

vita

l

sig

ns.

We

als

o c

on

clu

de

th

at

att

em

pts

sho

uld

be

ma

de

to

sh

ort

en

de

lay in

CFD

co

nta

ct

to r

ed

uc

e o

ve

rall

pre

-ho

spita

l tim

e

an

d m

axim

ise

pa

tie

nt

surv

iva

l. Fu

rth

er

stu

dy

in b

oth

urb

an

an

d r

ura

l se

ttin

gs

sho

uld

de

term

ine

wh

eth

er

byp

ass

allo

ws

de

ath

to

oc

cu

r d

urin

g t

ran

spo

rt a

nd

wh

eth

er

lon

ge

r

byp

ass

tim

es

influ

en

ce

ove

rall

pre

-ho

spita

l

tim

e a

nd

mo

rta

lity.


79

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hw

art

z e

t a

l. 1

990

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

EM

S s

yst

em

in C

on

ne

ctic

ut:

60

% o

f

po

pu

latio

n c

ove

red

by g

rou

nd

am

bu

lan

ce

an

d e

ntire

po

pu

latio

n c

an

be

re

ac

he

d b

y a

ir a

mb

ula

nc

e.

Pa

rtic

ipa

nts

:

126 p

atie

nts

, 93 t

ran

spo

rte

d b

y a

ir a

nd

33 b

y g

rou

nd

.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Ave

rag

e r

esp

on

se t

ime

(tim

e t

o s

ce

ne

)

Air 3

4 m

inu

tes

Gro

un

d 6

min

ute

s

Ave

rag

e s

ce

ne

tim

e

Air 2

2 m

inu

tes

Gro

un

d 1

8 m

inu

tes

Ave

rag

e t

ime

to

ho

spita

l

Air 10 m

inu

tes

Gro

un

d 1

1 m

inu

tes

Ave

rag

e t

ota

l pre

-ho

spita

l tim

e

Air 6

5 m

inu

tes

Gro

un

d 3

4 m

inu

tes

Inc

lu/e

xcl c

rite

ria

.

Mu

ltis

yst

em

inju

red

pa

tie

nts

Blu

nt

tra

um

a

Tra

nsp

ort

ed

dire

ctly f

rom

sc

en

e t

o

Ha

rtfo

rd H

osp

ita

l

Tre

ate

d b

y g

rou

nd

pa

ram

ed

ics

or

LIFE

STA

R.

Da

ta c

olle

ctio

n

Da

ta o

bta

ine

d f

rom

th

ree

se

pa

rate

co

mp

ute

rise

d d

ata

ma

na

ge

me

nt

reg

istr

ies

an

d m

ed

ica

l re

co

rds.

Pa

ram

ed

ic a

nd

LIF

E S

TAR

re

gis

trie

s

co

nta

in d

ata

on

pre

-ho

spita

l tim

es,

inte

rve

ntio

ns

an

d in

div

idu

als

wh

o

pe

rfo

rme

d t

he

pro

ce

du

res.

Tra

um

a

reg

istr

y c

on

tain

s in

pa

tie

nt

da

ta

inc

lud

ing

len

gth

of

sta

y. P

atie

nt

ch

art

s w

ere

exa

min

ed

fo

r o

utc

om

e

an

d c

on

firm

atio

n o

f re

gis

try d

ata

.

Da

ta c

olle

cte

d in

clu

de

d a

ge

, se

x,

mo

de

of

tra

nsp

ort

, m

ec

ha

nis

m o

f

inju

ry, p

re-h

osp

ita

l tim

es,

me

dic

al

inte

rve

ntio

ns,

tra

um

a s

co

re, IS

S a

nd

ou

tco

me

. Ti

me

of

dis

pa

tch

wa

s

use

d t

o a

pp

roxim

ate

th

e t

ime

of

inju

ry.

An

aly

sis

An

aly

sis

of

pre

-ho

spita

l tim

es

wa

s

ma

de

by t

he

t t

est

. TR

ISS m

eth

od

s

we

re u

sed

to

co

mp

are

su

rviv

al

be

twe

en

th

e t

wo

gro

up

s.

Su

rviv

al (Z

sc

ore

co

mp

are

d w

ith

MTO

S d

ata

set)

Air a

mb

ula

nc

e

Z=2.2

3

Sig

nific

an

tly im

pro

ve

d c

om

pa

red

with

MTO

S c

oh

ort

Gro

un

d a

mb

ula

nc

e

Z=

-2.6

9

Sig

nific

an

tly w

ors

e t

ha

n M

TOS

co

ho

rt

Lim

ita

tio

ns

�

Diffe

ren

ce

s in

cre

win

g b

etw

ee

n t

he

air

an

d g

rou

nd

se

rvic

es.

Th

ere

fore

,

diffic

ult t

o e

sta

blis

h if

diffe

ren

ce

in

ou

tco

me

wa

s d

ue

to

th

e d

iffe

ren

ce

in

pre

-ho

spita

l tim

e,

the

cre

w m

ix o

r

som

eth

ing

els

e.

The

re w

ere

diffe

ren

ce

s in

pro

ce

du

res

pe

rfo

rme

d

be

twe

en

th

e t

wo

gro

up

s (t

he

air

gro

up

wa

s m

ore

inte

rve

ntio

na

l).

On

the

ba

sis

tha

t th

e p

oo

rer

ou

tco

me

wa

s a

sso

cia

ted

with

th

e g

rou

p w

ith

sho

rte

r p

re-h

osp

ita

l tim

es

it s

ee

ms

like

ly t

ha

t sk

ill m

ix w

as

a b

igg

er

co

ntr

ibu

tor

tha

n t

he

pre

-ho

spita

l tim

e.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. M

ultiv

aria

te a

na

lysi

s

wa

s n

ot

co

nd

uc

ted

th

us

furt

he

r

limitin

g c

on

tro

l ove

r p

ote

ntia

l

co

nfo

un

de

rs

�

No

do

cu

me

nta

tio

n o

f th

e b

asi

s fo

r

de

cid

ing

wh

eth

er

a h

elic

op

ter

or

gro

un

d a

mb

ula

nc

e s

ho

uld

be

dis

pa

tch

ed

, th

ere

fore

po

ten

tia

l

sele

ctio

n b

ias

in b

etw

ee

n g

rou

p

co

mp

ariso

ns.

�

M s

tatist

ic c

om

pa

rin

g m

atc

h b

etw

ee

n

the

ac

tua

l da

ta a

nd

th

e r

efe

ren

ce

da

ta w

as

no

t p

rese

nte

d s

o t

he

re m

ay

ha

ve

be

en

a p

oo

r m

atc

h b

etw

ee

n

da

ta.

The

refo

re z

sta

tist

ic d

ata

ma

y

be

un

relia

ble

. W

sta

tist

ic a

lso

no

t

pre

sen

ted

.

�

Ac

cu

rac

y o

f d

ata

extr

ac

ted

fro

m

reg

istr

ies

wa

s n

ot

do

cu

me

nte

d

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

. Sp

ec

ific

ally

,

the

re w

as

no

co

mp

aris

on

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d o

utc

om

e,

rath

er

the

co

mp

ariso

n w

as

be

twe

en

air

an

d

gro

un

d t

ran

spo

rta

tio

n


80

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sc

hw

art

z e

t a

l. 1

990

)

USA

co

ntin

ue

d

Co

mm

en

ts

�

Co

mp

are

d a

ho

spita

l ba

sed

ae

rom

ed

ica

l pro

gra

mm

e t

o a

gro

un

d

pa

ram

ed

ic s

erv

ice

in o

rde

r to

de

term

ine

wh

eth

er

pre

-ho

spita

l tim

e

or

pre

-ho

spita

l ca

re is

th

e m

ajo

r

co

ntr

ibu

tor

tow

ard

s su

rviv

al.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Sin

ce

th

e s

ce

ne

tim

e o

f b

oth

ae

rom

ed

ica

l

an

d g

rou

nd

se

rvic

es

we

re s

imila

r, t

he

imp

rove

d s

urv

iva

l of

the

air

pa

tie

nts

ma

y b

e

du

e t

o t

he

te

ch

nic

al i

nte

rve

ntio

n

pro

ce

du

res

pe

rfo

rme

d.


81

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

92)

Ca

na

da

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

All

EM

S in

Mo

ntr

ea

l is

co

ntr

olle

d b

y

Urg

en

ce

s-sa

nte

. U

rge

nc

es-

san

te

co

ord

ina

tes

pre

-ho

spita

l em

erg

en

cy

serv

ice

s, c

oo

rdin

ate

s a

mb

ula

nc

e

tra

nsp

ort

, p

lan

s e

me

rge

nc

y r

oo

m u

se in

co

llab

ora

tio

n w

ith

ho

spita

ls a

nd

co

ntr

ols

ad

mittin

g p

olic

ies

an

d d

ata

ac

cu

mu

latio

n o

n t

he

re

gio

na

l

ava

ilab

ility

of

ho

spita

l be

ds.

Cre

ws

att

en

din

g a

re d

ep

en

de

nt

on

th

e

seve

rity

of

inju

ry. Fo

r c

ritic

ally

ill,

a

ph

ysi

cia

n is

inc

lud

ed

wh

ere

as

in le

ss

seve

re c

ase

s a

n a

mb

ula

nc

e a

nd

em

erg

en

cy m

ed

ica

l te

ch

nic

ian

will

be

dis

pa

tch

ed

.

Pa

rtic

ipa

nts

:

Sa

mp

le 1

: 32

93 (

29

56 w

ith

min

or

tra

um

a

an

d 3

37 w

ith

ma

jor

tra

um

a)

Sa

mp

le 2

: 92

8

Sa

mp

le 3

: 35

5.

Ba

selin

e a

na

lyse

s

Me

an

ag

e 3

3 y

ea

rs (

ran

ge

0-8

4)

Inju

ry s

ust

ain

ed

at

ho

me

26%

Mo

tor

ve

hic

le c

rash

37%

He

ad

inju

ry 3

7%

Ch

est

in

jury

29%

Ab

do

min

al i

nju

ry 2

5%

Pe

ne

tra

tin

g in

jury

22%

Me

an

ISS 1

3.7

(ra

ng

e 1

-59

)

Me

an

to

tal p

re-h

osp

ita

l tim

e 3

6 m

inu

tes

Me

an

tim

e o

n s

ce

ne

20 m

inu

tes

De

ath

s 70/3

55

Inc

lu/e

xcl c

rite

ria

.

Thre

e s

tud

y s

am

ple

s.

Sa

mp

le 1

. R

etr

ieve

d r

ec

ord

s o

f 47

22

of

55

53 p

atie

nts

tre

ate

d b

y a

ph

ysi

cia

n a

t th

e s

ce

ne

, 147

7

pa

tie

nts

fo

r w

hic

h a

nu

rse

req

ue

ste

d a

ph

ysi

cia

n b

ut

no

ne

we

re a

va

ilab

le a

nd

of

977 p

atie

nts

for

wh

ich

on

ly a

n E

MT

wa

s

req

ue

ste

d a

nd

dis

pa

tch

ed

. Th

e

latt

er

gro

up

wa

s se

lec

ted

by

ran

do

mly

sa

mp

ling

on

e o

f e

igh

t

da

ys

for

the

last

se

ve

n m

on

ths

of

the

stu

dy. Exc

lusi

on

s: d

ec

lare

d

de

ad

at

sce

ne

or

no

t ta

ke

n t

o

ho

spita

l.

Sa

mp

le 2

. A

ra

nd

om

10%

sub

sam

ple

of

pa

tie

nts

with

min

or

tra

um

a (

Pre

-ho

spita

l in

de

x ≤3

) a

nd

tre

ate

d b

y a

ph

ysi

cia

n, a

ll p

atie

nts

with

ma

jor

tra

um

a a

nd

on

e 1

3%

ran

do

m s

am

ple

of

pa

tie

nts

tre

ate

d

by E

MT

on

ly.

Sa

mp

le 3

(fin

al s

am

ple

). D

erive

d

fro

m s

am

ple

2. Se

lec

ted

pa

tie

nts

aliv

e a

t th

e t

ime

th

e a

mb

ula

nc

e

arr

ive

d a

t th

e s

ce

ne

, tr

an

spo

rte

d t

o

a h

osp

ita

l by a

n U

rge

nc

es-

san

te

am

bu

lan

ce

an

d o

ne

of:

ad

mitte

d

to h

osp

ita

l, h

ad

su

rge

ry, tr

ea

ted

in

ICU

or

on

site

pre

-ho

spita

l in

de

x>3.

Ou

tco

me

me

asu

res

Mo

rta

lity

Ove

rall

resu

lts

(ob

serv

ed

co

mp

are

d

with

exp

ec

ted

ba

sed

on

MTO

S

po

pu

latio

n a

s re

fere

nc

e)

Z=

6.7

7

P=

0.0

00

1

SM

R 1

.81 (

95%

CI 1.4

2-2

.21

)

Bo

th in

dic

ate

hig

he

r o

bse

rve

d t

ha

n

exp

ec

ted

de

ath

s.

Re

sults

by t

ota

l pre

-ho

spita

l tim

e

Pre

-ho

spita

l tim

e 0

-60 m

inu

tes

SM

R (

ob

serv

ed

co

mp

are

d w

ith

exp

ec

ted

): 1

.56 (

95

% C

I 1.1

3-1

.97

)

Z=

3.9

2

Pre

-ho

spita

l tim

e >

60

min

ute

s

SM

R (

ob

serv

ed

co

mp

are

d w

ith

exp

ec

ted

): 1

0.0

(95

% C

I 2.9

6-1

9.9

6)

Z=

5.0

0

SM

R r

atio

(p

re-h

osp

ita

l tim

e 0

-60

min

ute

s ve

rsu

s >

60 m

inu

tes)

(95%

CI 1.6

9-1

7.3

7)

Ind

ica

tin

g in

cre

ase

d e

xce

ss

mo

rta

lity in

th

e lo

ng

er

pre

-ho

spita

l

tim

e g

rou

p.

Ad

just

ed

od

ds

ratio

(95%

CI)

fo

r p

re-

ho

spita

l tim

e >

60 m

inu

tes

co

mp

are

d w

ith

0-6

0 m

inu

tes:

OR

29.9

(2

.7-3

3.3

)

Co

ntr

olle

d f

or

pre

-ho

spita

l cre

w

mix

, in

-ho

spita

l ca

re le

ve

l an

d ISS.

Lim

ita

tio

ns

�

Co

nfu

sin

g s

am

plin

g a

pp

roa

ch

. It

wa

s

un

cle

ar

if c

om

po

ne

nts

of

sam

ple

1

we

re r

an

do

mly

se

lec

ted

– s

pe

cific

ally

the

re w

as

no

do

cu

me

nta

tio

n a

bo

ut

ho

w t

he

pa

tie

nts

tre

ate

d a

t th

e s

ce

ne

by a

ph

ysi

cia

n w

ere

se

lec

ted

o

r

wh

eth

er

the

147

7 c

ase

s w

he

re a

nu

rse

req

ue

ste

d a

ph

ysi

cia

n t

o a

tte

nd

th

e

sce

ne

bu

t n

on

e w

ere

ava

ilab

le

rep

rese

nte

d t

he

en

tire

co

ho

rt o

r

wh

eth

er

som

e u

nd

esc

ribe

d s

ele

ctio

n

pro

ce

ss a

pp

lied

. O

n t

his

ba

sis

sele

ctio

n b

ias

ca

nn

ot

be

exc

lud

ed

. It

wa

s u

nc

lea

r if s

am

plin

g r

ule

s w

ere

se

t

be

fore

se

lec

tin

g t

he

sa

mp

le.

�

Furt

he

r so

urc

es

of

sele

ctio

n b

ias

am

on

gst

so

me

exc

lusi

on

s. F

or

exa

mp

le, 30 o

f 38

5 p

atie

nts

we

re

exc

lud

ed

fro

m s

am

ple

3 d

ue

to

ho

spita

l ch

art

s b

ein

g n

ot

ava

ilab

le.

�

M s

tatist

ic c

om

pa

rin

g m

atc

h b

etw

ee

n

the

ac

tua

l da

ta a

nd

th

e r

efe

ren

ce

da

ta w

as

no

t p

rese

nte

d (

alth

ou

gh

in

the

qu

est

ion

an

d a

nsw

er

sec

tio

n

Sa

mp

alis

sta

ted

it

wa

s n

ot

sig

nific

an

tly

low

er

tha

n 0

.9)

so t

he

re m

ay h

ave

be

en

a p

oo

r m

atc

h b

etw

ee

n d

ata

.

The

refo

re z

sta

tist

ic d

ata

ma

y b

e

un

relia

ble

.


82

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

92)

Ca

na

da

co

ntin

ue

d

An

aly

sis

Two

me

tho

ds

use

d t

o c

om

pa

re t

he

ob

serv

ed

mo

rta

lity w

ith

th

e

exp

ec

ted

mo

rta

lity (

ba

sed

on

th

e

MTO

S p

op

ula

tio

n).

First

wa

s

est

ima

tin

g t

he

Z s

co

re a

nd

th

e

sec

on

d b

y e

stim

atin

g t

he

sta

nd

ard

ise

d m

ort

alit

y r

atio

(SM

R).

Thre

e s

tep

s to

th

e d

ata

an

aly

sis:

1.

The

Z s

co

re a

nd

th

e S

MR

we

re

est

ima

ted

in

sa

mp

le 3

.

2.

Aim

ed

to

eva

lua

te t

he

diffe

ren

ce

be

twe

en

exp

ec

ted

an

d o

bse

rve

d d

ea

ths

in

diffe

ren

t st

rata

de

fin

ed

to

rep

rese

nt

pre

-ho

spita

l ca

re,

in-

ho

spita

l ca

re a

nd

to

tal p

re-

ho

spita

l tim

e. Z s

co

res

an

d

SM

Rs

we

re e

stim

ate

d f

or

ea

ch

of

the

se s

tra

ta.

3.

Aim

ed

to

pe

rfo

rm a

dju

ste

d

co

mp

ariso

ns

of

the

SM

Rs

in

diffe

ren

t le

ve

ls o

f th

e v

aria

ble

s

de

scrib

ed

ab

ove

. Lo

gis

tic

reg

ress

ion

wa

s u

sed

in t

his

an

aly

sis.

Th

e m

od

el c

om

pa

red

the

ou

tco

me

ra

te in

on

e le

ve

l

of

an

in

de

pe

nd

en

t va

riab

le

with

th

e r

ate

exp

ec

ted

ac

co

rdin

g t

o in

dire

ct

sta

nd

ard

isa

tio

n w

hile

co

ntr

olli

ng

fo

r th

e e

ffe

ct

of

oth

er

co

va

ria

tes.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g t

ho

ug

h m

ultiv

aria

te

an

aly

sis

wa

s c

on

du

cte

d w

hic

h is

an

imp

rove

me

nt

on

oth

er

stu

die

s.

�

Pre

-ho

spita

l tim

e d

ata

re

stric

ted

to

270 o

f 35

5 (

76%

) w

ith

co

mp

lete

da

ta.

�

On

ly 1

3 p

atie

nts

with

a p

re-h

osp

ita

l

tim

e >

60

min

ute

s so

de

alin

g w

ith

sma

ll p

atie

nt

nu

mb

ers

in t

his

gro

up

.

Co

mm

en

ts

�

Prim

ary

aim

wa

s to

ap

ply

Flo

ra’s

Z

sta

tist

ic a

nd

in

dire

ct

sta

nd

ard

isa

tio

n

to t

he

MTO

S in

a s

am

ple

of

seve

rely

inju

red

pa

tie

nts

. A

se

co

nd

aim

wa

s to

ass

ess

th

e a

sso

cia

tio

n b

etw

ee

n p

re-

ho

spita

l an

d in

-ho

spita

l co

mp

on

en

ts

of

the

Mo

ntr

ea

l EM

S w

ith

th

e S

MR

in

this

sa

mp

le o

f tr

au

ma

vic

tim

s.

�

Sa

mp

ling

ha

d t

he

aim

of

sele

ctin

g

pa

tie

nts

with

se

ve

re b

ut

surv

iva

ble

inju

rie

s.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Sta

nd

ard

isa

tio

n t

o t

he

MTO

S p

op

ula

tio

n

ind

ica

ted

a s

ign

ific

an

tly h

igh

ove

rall

exc

ess

mo

rta

lity in

th

e M

on

tre

al s

am

ple

. B

ein

g

tre

ate

d in

a le

ve

l I o

r le

ve

l II c

om

pa

tib

le

ho

spita

l wa

s a

sso

cia

ted

with

low

er

exc

ess

mo

rta

lity.

Tota

l pre

-ho

spita

l tim

e o

ve

r 60

min

ute

s w

as

ass

oc

iate

d w

ith

a s

ign

ific

an

t

inc

rea

se in

exc

ess

mo

rta

lity.


83

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

93)

Ca

na

da

Ca

se c

on

tro

l

stu

dy

Leve

l III-

2

Stu

dy s

ett

ing

.

All

EM

S in

Mo

ntr

ea

l is

co

ntr

olle

d b

y

Urg

en

ce

s-sa

nte

. U

rge

nc

es-

san

te

co

ord

ina

tes

pre

-ho

spita

l em

erg

en

cy

serv

ice

s, c

oo

rdin

ate

s a

mb

ula

nc

e

tra

nsp

ort

, p

lan

s e

me

rge

nc

y r

oo

m u

se in

co

llab

ora

tio

n w

ith

ho

spita

ls a

nd

co

ntr

ols

ad

mittin

g p

olic

ies

an

d d

ata

ac

cu

mu

latio

n o

n t

he

re

gio

na

l

ava

ilab

ility

of

ho

spita

l be

ds.

Cre

ws

att

en

din

g a

re d

ep

en

de

nt

on

th

e

seve

rity

of

inju

ry. Fo

r c

ritic

ally

ill,

a

ph

ysi

cia

n is

inc

lud

ed

wh

ere

as

in le

ss

seve

re c

ase

s a

n a

mb

ula

nc

e a

nd

em

erg

en

cy m

ed

ica

l te

ch

nic

ian

will

be

dis

pa

tch

ed

.

Pa

rtic

ipa

nts

:

Ca

ses

72

Co

ntr

ols

288

Ba

selin

e a

na

lyse

s.

Me

an

ag

e 3

3.9

ye

ars

Ma

les

71%

At

lea

st o

ne

co

mo

rbid

co

nd

itio

n 1

2%

Mo

tor

ve

hic

le c

rash

32%

Tota

l pre

-ho

spita

l tim

e 3

5.6

min

ute

s

Inc

lu/e

xcl c

rite

ria

.

Thre

e s

tud

y s

am

ple

s.

Sa

mp

le 1

. R

etr

ieve

d r

ec

ord

s o

f 47

22

of

55

53 p

atie

nts

tre

ate

d b

y a

ph

ysi

cia

n a

t th

e s

ce

ne

, 147

7

pa

tie

nts

fo

r w

hic

h a

nu

rse

req

ue

ste

d a

ph

ysi

cia

n b

ut

no

ne

we

re a

va

ilab

le a

nd

of

977 p

atie

nts

for

wh

ich

on

ly a

n E

MT

wa

s

req

ue

ste

d a

nd

dis

pa

tch

ed

. Th

e

latt

er

gro

up

wa

s se

lec

ted

by

ran

do

mly

sa

mp

ling

on

e o

f e

igh

t

da

ys

for

the

last

se

ve

n m

on

ths

of

the

stu

dy. Exc

lusi

on

s: d

ec

lare

d

de

ad

at

sce

ne

or

no

t ta

ke

n t

o

ho

spita

l.

Sa

mp

le 2

. A

ra

nd

om

10%

sub

sam

ple

of

pa

tie

nts

with

min

or

tra

um

a (

Pre

-ho

spita

l in

de

x ≤3

) a

nd

tre

ate

d b

y a

ph

ysi

cia

n, a

ll p

atie

nts

with

ma

jor

tra

um

a a

nd

on

e 1

3%

ran

do

m s

am

ple

of

pa

tie

nts

tre

ate

d

by E

MT

on

ly.

Od

ds

of

6 d

ay s

urv

iva

l (p

re-h

osp

ita

l

tim

e ≤

60 m

inu

tes

ve

rsu

s >

60

min

ute

s)

OR

2.0

9 (

95%

CI 0.6

7-6

.29)

Ad

just

ed

od

ds

of

6 d

ay s

urv

iva

l (p

re-h

osp

ita

l tim

e ≤

60

min

ute

s

ve

rsu

s >

60 m

inu

tes)

OR

3.0

1 (

95%

CI 1.2

7-5

.06)

Ad

just

ed

fo

r a

ge

, IS

S, M

VC

, fire

arm

,

leve

l of

pre

-ho

spita

l an

d in

ho

spita

l

ca

re.

Re

sults

ind

ica

te in

cre

ase

d o

dd

s o

f

surv

iva

l be

yo

nd

6 d

ays

with

a p

re-

ho

spita

l tim

e <

60 m

inu

tes.

Lim

ita

tio

ns

�

Co

nfu

sin

g s

am

plin

g a

pp

roa

ch

. It

wa

s

un

cle

ar

if c

om

po

ne

nts

of

sam

ple

1

we

re r

an

do

mly

se

lec

ted

– s

pe

cific

ally

the

re w

as

no

do

cu

me

nta

tio

n a

bo

ut

ho

w t

he

pa

tie

nts

tre

ate

d a

t th

e s

ce

ne

by a

ph

ysi

cia

n w

ere

se

lec

ted

or

wh

eth

er

the

147

7 c

ase

s w

he

re “

a

nu

rse

re

qu

est

ed

a p

hysi

cia

n t

o a

tte

nd

the

sc

en

e b

ut

no

ne

we

re a

va

ilab

le”

rep

rese

nte

d t

he

en

tire

co

ho

rt o

r

wh

eth

er

som

e u

nd

esc

ribe

d s

ele

ctio

n

pro

ce

ss a

pp

lied

. O

n t

his

ba

sis

sele

ctio

n b

ias

ca

nn

ot

be

exc

lud

ed

. It

wa

s u

nc

lea

r if s

am

plin

g r

ule

s w

ere

se

t

be

fore

se

lec

tin

g t

he

sa

mp

le.

�

Furt

he

r so

urc

es

of

sele

ctio

n b

ias

am

on

gst

so

me

exc

lusi

on

s. F

or

exa

mp

le, 34 o

f 33

7 w

ith

a P

HI>

3 w

ere

exc

lud

ed

fro

m s

am

ple

3 d

ue

to

ho

spita

l ch

art

s b

ein

g n

ot

ava

ilab

le.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g t

ho

ug

h m

ultiv

aria

te

an

aly

sis

wa

s c

on

du

cte

d w

hic

h is

an

imp

rove

me

nt

on

oth

er

stu

die

s.

.


84

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

93)

Ca

na

da

co

ntin

ue

d

Sa

mp

le 3

(fin

al s

am

ple

). C

ase

s fr

om

sam

ple

2 w

ho

fu

lfill

ed

th

e c

rite

ria

fo

r

be

ing

a c

ase

or

a c

on

tro

l we

re

inc

lud

ed

. C

ase

s: a

live

at

tim

e o

f

am

bu

lan

ce

arr

iva

l at

sce

ne

,

tra

nsp

ort

ed

to

a h

osp

ita

l by a

n

Urg

en

ce

s-sa

nte

am

bu

lan

ce

, d

ied

≤6 d

ays

of

tim

e o

f in

jury

. C

on

tro

ls:

aliv

e a

t tim

e o

f a

mb

ula

nc

e a

rriv

al

at

sce

ne

, tr

an

spo

rte

d t

o a

ho

spita

l

by a

n U

rge

nc

es-

san

te a

mb

ula

nc

e,

surv

ive

d >

6 d

ays

of

tim

e o

f in

jury

,

fulfill

ed

an

y o

f: a

dm

itte

d t

o

ho

spita

l, h

ad

su

rge

ry,

tre

ate

d in

IC

U

or

on

site

pre

-ho

spita

l in

de

x>3.

Ou

tco

me

me

asu

res

Six

da

y s

urv

iva

l

An

aly

sis

Un

iva

ria

te m

eth

od

s u

sed

to

co

mp

are

ca

ses

an

d c

on

tro

ls.

Mu

ltip

le lo

gis

tic

re

gre

ssio

n u

sed

to

co

ntr

ol f

or

po

ten

tia

lly c

on

fou

nd

ing

fac

tors

. P

re-s

et

va

ria

ble

s a

dd

ed

to

the

mo

de

l an

d f

ina

l m

od

el w

as

de

rive

d f

rom

th

e s

tep

wis

e s

ele

ctio

n

me

tho

d.

C

om

me

nts

�

Aim

ed

to

ass

ess

th

e a

sso

cia

tio

n

be

twe

en

use

of

on

-site

ad

va

nc

ed

life

sup

po

rt,

tota

l pre

-ho

spita

l tim

e a

nd

leve

l of

in-h

osp

ita

l ca

re w

ith

six

da

y

surv

iva

l in

se

ve

rely

inju

red

pa

tie

nts

.

�

Sim

ilar

sam

plin

g s

tra

teg

y t

o t

ha

t u

sed

in S

am

pa

lis e

t a

l. 200

2 w

ith

th

e f

irst

two

sta

ge

s a

pp

ea

rin

g t

o id

en

tify

th

e

sam

e p

atie

nts

. U

nd

ou

bte

dly

co

nsi

de

rab

le o

ve

rla

p in

pa

tie

nts

.

�

Hig

he

r p

rop

ort

ion

of

ca

ses

ha

d a

t

lea

st o

ne

co

mo

rbid

ity (

22%

v 1

0%

).

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

No

be

ne

fit

is a

sso

cia

ted

with

th

e u

se o

f

ph

ysi

cia

n-p

rov

ide

d o

n-s

ite

ad

va

nc

e li

fe

sup

po

rt in

re

du

cin

g t

he

ris

k o

f d

ea

th in

seve

rely

inju

red

pa

tie

nts

. Th

e d

ata

str

on

gly

sup

po

rts

the

sig

nific

an

ce

of

red

uc

ed

pre

-

ho

spita

l tim

e a

nd

hig

h le

ve

l in

-ho

spita

l ca

re

for

the

co

ntr

ol o

f tr

au

ma

re

late

d m

ort

alit

y


85

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bo

na

tti e

t a

l. 19

95

)

Au

stria

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Co

nd

uc

ted

at

the

HEM

S u

nit

Ch

risto

ph

oru

s I b

ase

d a

t In

nsb

ruc

k,

Au

stria

du

rin

g a

th

ree

ye

ar

pe

rio

d f

rom

1989 t

o 1

991 in

clu

sive

.

Pa

rtic

ipa

nts

:

2139 p

art

icip

an

ts

Ba

selin

e a

na

lyse

s.

Sp

ort

ing

ac

cid

en

ts 5

3.7

%

Mo

tor

ve

hic

le a

cc

ide

nts

11.1

%

Oc

cu

pa

tio

na

l in

jurie

s 5.6

%

Me

dic

al e

me

rge

nc

ies

16.1

%

Ne

uro

log

ica

l em

erg

en

cie

s 2.8

%

Oth

er

mis

sio

ns

10.7

%

30 d

ay s

urv

iva

l 87

.9%

Inc

lu/e

xcl c

rite

ria

.

No

t st

ate

d

Da

ta c

olle

ctio

n

Ob

tain

ed

fro

m H

EM

S o

pe

ratio

n

pro

toc

ols

an

d b

y w

ritt

en

, p

ers

on

al

or

tele

ph

on

e r

eq

ue

st f

rom

me

dic

al

rec

ord

s o

f th

e a

dm

ittin

g h

osp

ita

ls.

Flig

ht

log

s w

ere

re

co

rde

d in

ele

ctr

on

ic f

orm

an

d f

ollo

w-u

p d

ata

ad

de

d a

s h

osp

ita

l dis

ch

arg

e

sum

ma

rie

s w

ere

re

ce

ive

d.

Initia

l vita

l sig

ns

(sta

te o

f

co

nsc

iou

sne

ss,

resp

irato

ry s

tatu

s,

circ

ula

tory

sta

tus)

we

re a

sse

sse

d o

n

a f

ou

r p

oin

t sc

ale

An

aly

sis

Un

iva

ria

te s

urv

iva

l an

aly

sis

by t

he

life

ta

ble

me

tho

d a

nd

Wilc

oxo

n

test

. C

ox

pro

po

rtio

na

l h

aza

rds

mo

de

l fo

r m

ultiv

aria

te a

na

lysi

s.

30 d

ay s

urv

iva

l by t

ota

l mis

sio

n t

ime

0-2

0 m

inu

tes:

95.5

% s

urv

iva

l

21-4

0 m

inu

tes:

91.7

% s

urv

iva

l

41-6

0 m

inu

tes:

87.6

% s

urv

iva

l

61-8

0 m

inu

tes:

86.8

% s

urv

iva

l

> 8

0 m

inu

tes:

78.8

% s

urv

iva

l

Un

iva

ria

te a

na

lysi

s, P

=0.0

00

1

Mu

ltiv

aria

te a

na

lysi

s, n

o s

ign

ific

an

t

ass

oc

iatio

n, a

dju

ste

d f

or

ca

use

of

inju

ry/e

me

rge

nc

y, flig

ht

tim

e t

o

sce

ne

, sc

en

e t

ime

, p

atie

nt

ag

e,

pa

tie

nt

ge

nd

er,

NA

CA

sc

ore

, st

ate

of

co

nsc

iou

sne

ss, re

spira

tory

sta

tus,

circ

ula

tory

sta

tus,

em

erg

en

cy

ph

ysi

cia

n.

Lim

ita

tio

ns

�

No

do

cu

me

nta

tio

n o

f se

lec

tio

n

crite

ria

. It

is a

ssu

me

d a

ll p

atie

nts

we

re

inc

lud

ed

.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g t

ho

ug

h m

ultiv

aria

te

an

aly

sis

wa

s c

on

du

cte

d w

hic

h is

an

imp

rove

me

nt

on

oth

er

stu

die

s.

�

Ac

cu

rac

y o

f d

ata

re

co

rdin

g in

th

e

da

ta b

ase

s u

sed

wa

s n

ot

rec

ord

ed

.

�

Pa

tie

nts

ap

pe

are

d t

o b

e t

ake

n t

o

mu

ltip

le h

osp

ita

ls –

crite

ria

fo

r ta

kin

g

pa

tie

nts

to

sp

ec

ific

ho

spita

ls w

ere

no

t

sta

ted

an

d m

ay h

ave

in

flu

en

ce

d

ou

tco

me

.

�

Ad

just

ing

fo

r flig

ht

tim

e t

o t

he

sc

en

e

an

d s

ce

ne

tim

e m

ay h

ave

dim

inis

he

d

the

ass

oc

iatio

n b

etw

ee

n t

ota

l mis

sio

n

tim

e a

nd

su

rviv

al s

inc

e t

he

se

me

asu

res

are

su

bse

ts o

f to

tal m

issi

on

tim

e.

Co

mm

en

ts

�

Aim

ed

to

ide

ntify

ea

sily

ob

tain

ab

le

pre

dic

tors

of

sho

rt-t

erm

ou

tco

me

fo

r

em

erg

en

cy v

ictim

s tr

ea

ted

by a

ph

ysi

cia

n s

taff

ed

he

lico

pte

r

em

erg

en

cy m

ed

ica

l syst

em

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

fo

llow

ing

pa

ram

ete

rs c

an

be

use

d in

an

initia

l pre

dic

tive

ass

ess

me

nt

by t

he

flig

ht

ph

ysi

cia

n a

nd

th

e a

dm

ittin

g in

stitu

tio

n:

seve

rity

of

em

erg

en

cy, in

itia

l re

spira

tory

sta

tus,

tim

e a

t sc

en

e,

pa

tie

nt

ag

e a

nd

pa

tie

nt

ge

nd

er.


86

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fe

ero

et

al.

199

5)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2.

Stu

dy s

ett

ing

.

Stu

dy s

et

in P

ort

lan

d O

reg

on

(po

pu

latio

n 5

00,0

00 2

00 s

qu

are

mile

s).

Em

erg

en

cy s

erv

ice

s w

ork

un

de

r a

tw

o

tie

r sy

ste

m: b

asi

c a

nd

ad

va

nc

ed

life

sup

po

rt)

Pa

rtic

ipa

nts

:

The

re w

ere

848 m

ajo

r tr

au

ma

ca

ses

bu

t

the

stu

dy c

on

ce

ntr

ate

d o

n t

he

un

exp

ec

ted

su

rviv

ors

an

d t

he

un

exp

ec

ted

de

ath

s (b

ase

d o

n T

RIS

S

me

tho

do

log

y)

Un

exp

ec

ted

su

rviv

ors

(n

=1

3)

Un

exp

ec

ted

de

ath

s (n

=20).

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

ch

an

ism

of

inju

ry:

Mo

tor

ve

hic

le c

olli

sio

n 3

7%

Sta

bb

ing

17%

Fall

12%

Au

to v

pe

de

stria

n 1

2%

Gu

nsh

ot

wo

un

d 9

%

Au

to v

bic

yc

le 2

%

Bic

yc

le c

olli

sio

n 1

%

Oth

er

8%

Un

kn

ow

n 2

%

Me

an

ag

e (

ye

ars

)

Un

exp

ec

ted

su

rviv

or

29

.5

Un

exp

ec

ted

de

ath

50.8

P=

0.0

1

Ma

le

Un

exp

ec

ted

su

rviv

or

67

%

Un

exp

ec

ted

de

ath

61%

NS

Inc

lu/e

xcl c

rite

ria

.

All

ma

jor

tra

um

a c

ase

s fo

r 199

0

we

re o

bta

ine

d f

rom

th

e S

tate

of

Ore

go

n In

jury

Re

gis

try a

nd

th

e

Me

dic

al R

eso

urc

e H

osp

ita

l. M

ajo

r

tra

um

a d

efin

ed

as

tho

se c

ase

s

en

tere

d b

y E

MS p

rovid

ers

into

th

e

loc

al t

rau

ma

syst

em

. M

an

da

tory

crite

ria

inc

lud

ed

:

Syst

olic

BP

< 9

0 m

mH

g

Re

spira

tory

ra

te <

10 o

r >

29

bre

ath

s/m

in

GC

S<

13

Pe

ne

tra

tin

g in

jury

of

he

ad

, n

ec

k,

tors

o, g

roin

<20%

to

tal s

urf

ac

e a

rea

bu

rns

Am

pu

tatio

n a

bo

ve

th

e w

rist

or

an

kle

Sp

ina

l co

rd in

jury

with

lim

b p

ara

lysi

s

Fla

il c

he

st

Two

or

mo

re o

bv

iou

s p

roxim

al l

on

g

bo

ne

fra

ctu

res

De

ath

of

sam

e c

ar

oc

cu

pa

nt

Eje

ctio

n f

rom

en

clo

sed

ve

hic

le

Ext

rica

tio

n t

ime

lon

ge

r th

an

20

min

ute

s.

Pa

tie

nts

co

uld

be

en

tere

d if

:

Hig

h e

ne

rgy t

ran

sfe

r

Ba

sed

on

co

mo

rbid

co

nd

itio

ns

Da

ta c

olle

ctio

n

Ag

e,

sex,

me

ch

an

ism

of

inju

ry, EM

S

resp

on

se t

ime

inte

rva

ls, e

me

rge

nc

y

de

pa

rtm

en

t a

nd

inp

atie

nt

dis

po

sitio

n,

rev

ise

d t

rau

ma

sc

ore

, IS

S

Ou

tco

me

me

asu

res

Su

rviv

al to

lea

ve

ho

spita

l

Tota

l EM

S t

ime

inte

rva

l (m

inu

tes)

un

exp

ec

ted

su

rviv

ors

an

d

un

exp

ec

ted

de

ath

s

Un

exp

ec

ted

su

rviv

ors

: 20.8

min

ute

s

Un

exp

ec

ted

de

ath

s: 2

9.3

min

ute

s

P=

0.0

2

Lim

ita

tio

ns

�

Sm

all

nu

mb

ers

of

pa

tie

nts

in t

he

un

exp

ec

ted

su

rviv

or

an

d u

ne

xpe

cte

d

de

ath

gro

up

s.

�

Po

ssib

le m

isc

lass

ific

atio

n o

f th

e

un

exp

ec

ted

su

rviv

or

an

d u

ne

xpe

cte

d

de

ath

gro

up

s (a

lth

ou

gh

M s

tatist

ic

wa

s 0.9

8 s

ug

ge

stin

g a

go

od

fit

be

twe

en

th

e r

efe

ren

ce

da

ta a

nd

th

e

ac

tua

l da

ta).

�

Po

ten

tia

l fo

r in

co

nsi

ste

nc

y in

th

e

sele

ctio

n p

roc

ess

giv

en

a c

ate

go

ry o

f

no

n-m

an

da

tory

re

po

rtin

g.

�

Po

ten

tia

l fo

r c

on

fou

nd

ing

in t

his

ob

serv

atio

na

l stu

dy.

�

Mix

ture

of

RTS

da

ta u

sed

– s

om

e f

rom

sce

ne

an

d s

om

e f

rom

ho

spita

l,

alth

ou

gh

sim

ilar

pro

po

rtio

ns

use

d

ho

spita

l da

ta in

th

e u

ne

xpe

cte

d

surv

ivo

rs a

nd

th

e u

ne

xpe

cte

d d

ea

ths

(42%

v 3

9%

).

Co

mm

en

ts

�

Aim

ed

to

de

term

ine

if o

ut

of

ho

spita

l

em

erg

en

cy m

ed

ica

l se

rvic

es

tim

e

inte

rva

ls a

re a

sso

cia

ted

with

un

exp

ec

ted

de

ath

an

d s

urv

iva

l in

urb

an

ma

jor

tra

um

a.

�

All

tra

um

a c

ase

s ta

ke

n t

o a

sin

gle

ho

spita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Sh

ort

ou

t o

f h

osp

ita

l tim

e in

terv

al m

ay

po

sitive

ly a

ffe

ct

pa

tie

nt

su

rviv

al i

n s

ele

cte

d

urb

an

ma

jor

tra

um

a p

atie

nts


87

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fe

ero

et

al.

199

5)

USA

co

ntin

ue

d

An

aly

sis

TRIS

S m

eth

od

olo

gy u

sed

to

ca

lcu

late

pro

ba

bili

ty o

f su

rviv

al,

Flo

ra’s

Z s

tatist

ic u

sed

to

co

mp

are

exp

ec

ted

an

d o

bse

rve

d d

ea

ths.

Un

pa

ire

d t

te

st u

sed

to

co

mp

are

me

an

tim

e in

terv

als

fo

r th

e

un

exp

ec

ted

su

rviv

or

an

d d

ea

th

gro

up

s.


88

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Yo

un

g e

t a

l. 199

8)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Se

t in

th

e U

niv

ers

ity o

f V

irg

inia

He

alth

Sc

ien

ce

s C

en

ter

wh

ich

is a

leve

l 1

tra

um

a c

en

tre

se

rvin

g c

en

tra

l an

d

we

ste

rn V

irgin

ia.

An

ae

rom

ed

ica

l

pro

gra

mm

e e

xist

s th

at

tra

nsf

ers

pa

tie

nts

fro

m t

he

sc

en

e a

nd

fro

m o

utlyin

g

ho

spita

ls. Stu

dy c

on

du

cte

d d

urin

g 1

994-

1995.

Pa

rtic

ipa

nts

:

316 p

art

icip

an

ts d

ivid

ed

into

dire

ct

pa

tie

nts

(n

=16

5)

an

d t

ran

sfe

r p

atie

nts

(n=

151

)

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Ag

e (

ye

ars

)

Tra

nsf

er

gro

up

46

Dire

ct

gro

up

44

ISS

Tra

nsf

er

gro

up

23

.1

Dire

ct

gro

up

24.8

GC

S in

ED

Tra

nsf

er

gro

up

11

.4

Dire

ct

gro

up

11.4

Exp

ec

ted

nu

mb

er

of

de

ath

s

Tra

nsf

er

gro

up

23

Dire

ct

gro

up

34

Inc

lu/e

xcl c

rite

ria

.

Ag

e >

18 y

ea

rs

ISS >

15

Da

ta c

olle

ctio

n

Da

ta e

xtr

ac

ted

fro

m t

rau

ma

re

gis

try

an

d p

atie

nt

rec

ord

s.

ISS, d

ela

y a

t o

uts

ide

ho

spita

l,

pa

tie

nt

de

mo

gra

ph

ics

co

llec

ted

.

Ou

tco

me

me

asu

res

Ho

spita

l le

ng

th o

f st

ay a

nd

mo

rta

lity.

An

aly

sis

TRIS

S c

alc

ula

ted

with

re

fere

nc

e t

o

MTO

S c

oe

ffic

ien

ts.

Ch

i sq

ua

re t

est

use

d f

or

ca

teg

oric

al d

ata

an

d

Stu

de

nt’

s t

test

fo

r c

on

tin

uo

us

da

ta.

Tim

e f

rom

inju

ry t

o a

rriv

al a

t tr

au

ma

ce

ntr

e

Tra

nsf

er

gro

up

48

0 m

inu

tes

Dire

ct

gro

up

92 m

inu

tes

Len

gth

of

ho

spita

l sta

y

Tra

nsf

er

gro

up

19

.1 d

ays

Dire

ct

gro

up

15.4

da

ys

No

sig

nific

an

t d

iffe

ren

ce

Mo

rta

lity >

24 h

ou

rs a

fte

r in

jury

Tra

nsf

er

gro

up

12

de

ath

s

Dire

ct

gro

up

10 d

ea

ths

No

sig

nific

an

t d

iffe

ren

ce

Mo

rta

lity <

24 h

ou

rs a

fte

r in

jury

Tra

nsf

er

gro

up

16

de

ath

s

Dire

ct

gro

up

25 d

ea

ths

No

sig

nific

an

t d

iffe

ren

ce

De

ath

s w

ith

pro

ba

bili

ty o

f su

rviv

al >

50%

in f

irst

24 h

ou

rs

Tra

nsf

er

gro

up

12

of

16

Dire

ct

gro

up

7 o

f 25

P<

0.0

5

Lim

ita

tio

ns

�

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

.

Sp

ec

ific

ally

, th

ere

wa

s n

o c

om

pa

riso

n

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d

ou

tco

me

, ra

the

r th

e c

om

pa

riso

n w

as

be

twe

en

ou

tco

me

tra

nsf

er

an

d d

ire

ct

da

ta (

an

d t

hu

s u

sed

a s

a p

roxy

fo

r

tim

e,

giv

en

th

e lo

ng

er

tra

nsp

ort

atio

n

tim

e in

th

e t

ran

sfe

r g

rou

p t

ha

n t

he

dire

ct

gro

up

, 480 m

inu

tes

ve

rsu

s 9

2

min

ute

s).

�

Ac

cu

rac

y o

f tr

au

ma

re

gis

try n

ot

sta

ted

.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. M

ultiv

aria

te a

na

lysi

s

wa

s n

ot

co

nd

uc

ted

th

us

furt

he

r

limitin

g c

on

tro

l ove

r p

ote

ntia

l

co

nfo

un

de

rs (

alth

ou

gh

th

ere

we

re n

o

sig

nific

an

t d

iffe

ren

ce

s in

ba

selin

e

da

ta b

etw

ee

n d

irec

t a

nd

byp

ass

pa

tie

nts

).

�

Dis

cre

pa

nt

resu

lts

pre

sen

ted

in t

ha

t

the

to

tal d

ea

ths

pre

sen

ted

in t

he

dire

ct

gro

up

(n

=38

) is

no

t c

on

sist

en

t

with

to

tal d

ea

ths

in f

irst

24 h

ou

rs in

dire

ct

gro

up

(n

=10

) p

lus

tota

l de

ath

s

aft

er

24 h

ou

rs in

th

e d

ire

ct

gro

up

(n=

25

).


89

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Yo

un

g e

t a

l. 199

8)

USA

co

ntin

ue

d

�

Did

no

t p

rese

nt

un

exp

ec

ted

de

ath

s

mo

re t

ha

n 2

4 h

ou

rs a

fte

r in

jury

an

d

did

no

t p

rese

nt

un

exp

ec

ted

su

rviv

ors

.

�

M s

co

re w

as

< 0

.88 in

dic

atin

g a

po

or

ma

tch

with

th

e M

TOS d

ata

set.

Co

mm

en

ts

�

Exa

min

ed

th

e h

yp

oth

esi

s th

at

de

lay a

t

the

re

ferr

ing

ho

spita

l is

de

trim

en

tal t

o

pa

tie

nt

ou

tco

me

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Pa

tie

nts

with

ma

jor

tra

um

a t

ake

n d

ire

ctly t

o

the

tra

um

a c

en

tre

ha

d s

ho

rte

r h

osp

ita

l sta

y

an

d lo

we

r m

ort

alit

y. Th

e s

tud

y s

up

po

rts

tra

nsf

err

ing

ma

jor

tra

um

a p

atie

nts

dire

ctly t

o

tra

um

a c

en

tre

s fr

om

th

e in

jury

sc

en

e.


90

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fre

zza

an

d M

ezg

he

be

1999

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Tra

um

a p

atie

nts

att

en

din

g t

he

Ho

wa

rd

Un

ive

rsity H

osp

ita

l Em

erg

en

cy

De

pa

rtm

en

t b

etw

ee

n 1

992 a

nd

199

5.

Pa

rtic

ipa

nts

:

58 a

du

lt p

atie

nts

with

pe

ne

tra

tin

g c

he

st

tra

um

a

Ba

selin

e a

na

lyse

s

Gu

nsh

ot

wo

un

ds

70

%

Sta

b w

ou

nd

s 1

2%

Syst

olic

BP

< 7

0m

mH

g 2

4%

Tra

nsf

er

to IC

U 6

3%

Ave

rag

e p

re-h

osp

ita

l tim

e:

At

the

sc

en

e 1

1 m

inu

tes

Tra

nsi

t 8

min

ute

s

ED

10 m

inu

tes

Inc

lu/e

xcl c

rite

ria

.

Pa

tie

nts

wh

o u

nd

erw

en

t

em

erg

en

cy r

oo

m t

ho

rac

oto

my

(ER

T) a

nd

ha

d v

ita

l sig

ns

in t

he

fie

ld

Pe

ne

tra

tin

g c

he

st t

rau

ma

.

Da

ta c

olle

ctio

n

Pre

-ad

mis

sio

n d

ata

extr

ac

ted

fro

m

EM

S r

ep

ort

s.

Ou

tco

me

me

asu

res

Mo

rta

lity

An

aly

sis

Fish

er

co

rre

cte

d c

hi s

qu

are

te

st

Su

rviv

al w

ith

in 2

4 h

ou

rs b

y p

re-

ho

spita

l tim

e

Pre

-ho

spita

l tim

e <

30 m

inu

tes:

63%

surv

iva

l (20/2

7)

Pre

-ho

spita

l tim

e >

30 m

inu

tes:

0%

surv

iva

l (0/6

)

Lim

ita

tio

ns

�

Sc

an

t d

eta

ils p

rese

nte

d in

me

tho

ds.

�

Re

vie

we

r is

no

t a

wa

re o

f Fis

he

r

co

rre

cte

d c

hi s

qu

are

te

st –

oth

er

co

rre

ctio

ns

are

ava

ilab

le –

th

e

me

tho

d is

no

t re

fere

nc

ed

.

�

Ac

cu

rac

y o

f p

re-h

osp

ita

l tim

e

un

cle

ar.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. M

ultiv

aria

te a

na

lysi

s

wa

s n

ot

co

nd

uc

ted

th

us

furt

he

r

limitin

g c

on

tro

l ove

r p

ote

ntia

l

co

nfo

un

de

rs.

�

Da

ta o

nly

pre

sen

ted

on

33 o

f th

e 5

8

pa

tie

nts

(d

ata

we

re m

issi

ng

on

nin

e

pa

tie

nts

an

d 1

6 w

ere

exc

lud

ed

fro

m

furt

he

r a

na

lysi

s d

ue

to

lac

k o

f v

ita

l

sig

ns

in t

he

fie

ld.

�

Pre

-ho

spita

l tim

e n

ot

cle

arly d

efin

ed

.

In p

art

icu

lar

it w

as

un

cle

ar

if it

inc

lud

ed

tim

e f

rom

dis

pa

tch

to

tim

e o

f

arr

iva

l at

sce

ne

.

�

Ap

pe

are

d t

o b

e d

ea

ths

aft

er

24 h

ou

rs

bu

t it w

as

no

t p

oss

ible

to

est

ab

lish

ho

w m

an

y d

ea

ths

the

re w

ere

in t

his

tim

e p

erio

d.

The

refo

re, if t

he

nu

mb

er

of

de

ath

s p

rese

nte

d b

y p

re-h

osp

ita

l

tim

e h

ad

be

en

exte

nd

ed

be

yo

nd

24

ho

urs

, th

e d

iffe

ren

ce

in s

urv

iva

l

be

twe

en

th

e t

wo

tim

e g

rou

ps

wo

uld

ha

ve

be

en

less

.


91

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Fre

zza

an

d M

ezg

he

be

1999

)

USA

co

ntin

ue

d

Co

mm

en

ts

�

Aim

ed

to

ass

ess

if p

re-h

osp

ita

l tim

e

co

uld

be

use

d a

s th

e p

rin

cip

le

pa

ram

ete

r to

pre

dic

t w

he

the

r

em

erg

en

cy r

oo

m t

ho

rac

oto

my in

pe

ne

tra

tin

g c

he

st t

rau

ma

is u

sefu

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

on

ly r

ole

of

ER

T in

ou

r o

pin

ion

is in

pa

tie

nts

wh

o a

rriv

e w

ith

in 3

0 m

inu

tes

of

pre

-

ho

spita

l tim

e.


92

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ph

illip

s e

t a

l. 19

99)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Pa

tie

nts

tra

nsp

ort

ed

to

Bro

oke

Arm

y

Me

dic

al C

en

tre

leve

l 1 t

rau

ma

ce

ntr

e

by e

ith

er

air o

r ro

ad

am

bu

lan

ce

du

rin

g

1995

-6.

Pa

rtic

ipa

nts

:

792 c

on

sec

utive

pa

tie

nts

(6

87 d

eliv

ere

d

by r

oa

d a

mb

ula

nc

e a

nd

105 b

y a

ir

am

bu

lan

ce

).

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ag

e (

ye

ars

)

Gro

un

d: 3

2.2

3

Air 3

2.1

4

Ma

le

Gro

un

d 7

1.8

%

Air 6

7.6

%

Blu

nt

tra

um

a

Gro

un

d 7

1.3

%

Air 8

4.7

%

Tim

e f

rom

inju

ry t

o h

osp

ita

l arr

iva

l

Gro

un

d 5

4 m

inu

tes

Air 7

7 m

inu

tes

Inc

lu/e

xcl c

rite

ria

.

Co

nse

cu

tive

am

bu

lan

ce

tra

nsp

ort

ed

tra

um

a p

atie

nts

.

Exc

lud

ed

pa

tie

nts

wh

o d

id n

ot

ha

ve

TR

ISS v

alu

es.

Da

ta c

olle

ctio

n

Ext

rac

ted

da

ta r

eq

uire

d f

or

TRIS

S

ca

lcu

latio

ns

Ou

tco

me

me

asu

res

Su

rviv

al

An

aly

sis

Z st

atist

ic c

alc

ula

ted

ba

sed

on

co

mp

ariso

n w

ith

th

e M

TOS

po

pu

latio

n.

Co

mp

aris

on

of

ac

tua

l an

d

exp

ec

ted

mo

rta

lity b

y t

ran

spo

rt

gro

up

Gro

un

d t

ran

spo

rt (

sho

rte

r tim

e)

Exp

ec

ted

de

ath

s 39.1

Ac

tua

l de

ath

s 41

Z=

0.0

4

Air t

ran

spo

rt (

lon

ge

r a

ve

rag

e

tra

nsp

ort

tim

e)

Exp

ec

ted

de

ath

s 16.3

Ac

tua

l de

ath

s 15

Z=

-0.1

51

Me

an

len

gth

of

sta

y

Gro

un

d t

ran

spo

rt 4

.21 d

ays

Air a

mb

ula

nc

e 8

.97

da

ys

P<

0.0

01

Lim

ita

tio

ns

�

Mis

sin

g in

form

atio

n o

n 3

8 p

atie

nts

led

to t

he

ir e

xclu

sio

n (

4.8

%).

�

Diffe

ren

t st

aff

ing

on

th

e t

wo

mo

de

s o

f

de

live

ry: ro

ad

am

bu

lan

ce

ha

d t

wo

pa

ram

ed

ics,

air

am

bu

lan

ce

ha

d a

pa

ram

ed

ic a

nd

a f

ligh

t n

urs

e.

�

Mo

re s

eve

rely

inju

red

pa

tie

nts

we

re

pre

fere

ntia

lly t

rea

ted

by a

ir

am

bu

lan

ce

- t

his

co

uld

exp

lain

th

e

pro

lon

ge

d s

tay in

th

is g

rou

p.

TRIS

S

va

lue

s in

dic

ate

d a

diffe

ren

ce

in

seve

rity

be

twe

en

th

e t

wo

gro

up

s w

ith

pro

ba

bili

ty o

f su

rviv

al b

ein

g lo

we

r in

the

air g

rou

p (

93.9

% v

ers

us

83

.1%

).

�

Air a

mb

ula

nc

e p

atie

nts

re

ce

ive

d a

hig

he

r le

ve

l of

ca

re e

n r

ou

te.

�

M s

tatist

ic n

ot

pre

sen

ted

so

de

gre

e o

f

fit

with

MTO

S d

ata

wa

s u

nc

lea

r.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g. M

ultiv

aria

te a

na

lysi

s

wa

s n

ot

co

nd

uc

ted

th

us

furt

he

r

limitin

g c

on

tro

l ove

r p

ote

ntia

l

co

nfo

un

de

rs.

�

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

.

Sp

ec

ific

ally

, th

ere

wa

s n

o c

om

pa

riso

n

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d

ou

tco

me

, ra

the

r th

e c

om

pa

riso

n w

as

be

twe

en

air a

nd

ro

ad

am

bu

lan

ce

(an

d t

hu

s u

sed

as

a p

roxy

fo

r tim

e,

giv

en

th

e lo

ng

er

tra

nsp

ort

atio

n t

ime

in

the

air g

rou

p.


93

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ph

illip

s e

t a

l. 19

99)

USA

co

ntin

ue

d

Co

mm

en

ts

�

Aim

ed

to

re

vie

w w

he

the

r a

ir

am

bu

lan

ce

tra

nsp

ort

atio

n o

f tr

au

ma

pa

tie

nts

to

a le

ve

l 1 t

rau

ma

ce

ntr

e

co

ntr

ibu

ted

to

ma

inta

inin

g n

atio

na

l

mo

rta

lity s

tan

da

rds

in t

he

tra

um

a c

are

of

the

se p

atie

nts

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Re

sults

sug

ge

st t

ha

t a

ero

me

dic

al

eva

cu

atio

n o

f th

e m

ore

se

ve

rely

inju

red

pa

tie

nts

fa

rth

est

fro

m t

he

tra

um

a c

en

tre

resu

lte

d in

mo

rta

lity r

ate

s th

at

me

t n

atio

na

l

sta

nd

ard

s.


94

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

99)

Ca

na

da

Pro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Stu

dy s

et

in M

on

tre

al a

nd

Qu

eb

ec

du

rin

g a

nd

aft

er

a t

ime

of

reg

ion

alis

atio

n o

f tr

au

ma

ca

re s

erv

ice

s

Pa

rtic

ipa

nts

(n

=1

2,2

08):

Ba

selin

e a

na

lyse

s

Me

an

ag

e 4

8 y

ea

rs

Ma

le 6

7.6

%

Me

an

ISS 2

6.1

Dis

ch

arg

ed

aliv

e 7

1.6

%

Inc

lu/e

xcl c

rite

ria

.

Tre

ate

d f

or

inju

rie

s a

t a

cu

te c

are

ho

spita

ls in

Mo

ntr

ea

l an

d Q

ue

be

c.

On

e o

f:

De

ath

as

a r

esu

lt o

f in

jury

ISS>

12

PH

I>3

≥2 in

jurie

s w

ith

AIS

≥3

Ho

spita

l sta

y >

3 d

ays

Exc

lusi

on

s: d

ied

at

sce

ne

Da

ta c

olle

ctio

n

Da

ta e

xtr

ac

ted

fro

m r

ec

ord

s u

sin

g

a s

tan

da

rdis

ed

da

ta e

xtr

ac

tio

n

form

an

d t

he

n e

nte

red

into

a

cu

sto

mis

ed

da

ta m

an

ag

em

en

t

soft

wa

re p

rog

ram

. Fin

al a

na

lysi

s

wa

s c

on

du

cte

d in

SP

SS.

Ou

tco

me

me

asu

res

De

ath

du

rin

g h

osp

ita

l ad

mis

sio

n

An

aly

sis

The

an

aly

tic

al m

eth

od

s w

ere

div

ide

d b

y h

yp

oth

esi

s.

Hyp

oth

esi

s 1 t

est

ed

th

at

tra

um

a

ca

re r

eg

ion

alis

atio

n is

ass

oc

iate

d

with

a r

ed

uc

tio

n in

tra

um

a r

ela

ted

mo

rta

lity.

An

aly

sis

co

nsi

ste

d o

f

co

mp

ariso

n o

f m

ort

alit

y r

ate

s d

urin

g

ea

ch

fis

ca

l ye

ar

be

fore

an

d a

fte

r

reg

ion

alis

atio

n, c

om

pa

riso

n o

f

mo

rta

lity r

ate

s d

urin

g e

ac

h p

ha

se

of

imp

lem

en

tatio

n o

f th

e

reg

ion

alis

ed

Qu

eb

ec

syst

em

.

Log

istic

re

gre

ssio

n w

as

use

d.

Ove

rall

mo

rta

lity r

ate

28

% t

hro

ug

h

the

six

ye

ars

.

Ad

just

ed

od

ds

of

de

ath

by p

re-

ho

spita

l tim

e (

OR

fo

r e

ac

h

ad

ditio

na

l min

ute

of

pre

-ho

spita

l

tim

e)

OR

1.0

46 (

95%

CI 1.0

44-1

.05

0)

Ad

just

ed

fo

r tim

e t

o a

dm

issi

on

,

tra

um

a c

en

tre

de

sig

na

tio

n,

tra

nsf

er

ve

rsu

s d

ire

ct

tra

nsp

ort

, p

atie

nt

ag

e

an

d ISS.

Lim

ita

tio

ns

�

Ac

cu

rac

y o

f p

re-h

osp

ita

l tim

e

un

cle

ar.

�

Ob

serv

atio

n s

tud

y is

su

sce

ptib

le t

o

co

nfo

un

din

g a

lth

ou

gh

with

th

e

mu

ltiv

aria

te m

od

elli

ng

th

e r

isk o

f

co

nfo

un

din

g b

y k

no

wn

co

nfo

un

de

rs is

red

uc

ed

.

�

Po

ten

tia

l fo

r in

ap

pro

pria

te s

ele

ctio

n

as

the

fin

al p

roc

ess

re

lied

on

ch

art

revie

w.

Co

mm

en

ts

�

This

stu

dy a

ime

d t

o a

sse

ss t

he

imp

ac

t

of

reg

ion

alis

atio

n o

f tr

au

ma

ca

re

serv

ice

s o

n m

ort

alit

y. R

eg

ion

alis

atio

n

wa

s in

itia

ted

in 1

993. N

ote

th

e t

wo

ea

rlie

r Sa

mp

alis

stu

die

s (1

99

2 a

nd

1993

) su

pp

ort

ed

su

ch

re

gio

na

lisa

tio

n.

�

No

ris

k o

f lo

ss t

o f

ollo

w-u

p d

ue

to

th

e

na

ture

of

the

stu

dy d

esi

gn

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

This

stu

dy p

rod

uc

ed

em

piric

al e

vid

en

ce

tha

t th

e in

teg

ratio

n o

f tr

au

ma

ca

re s

erv

ice

s

into

a r

eg

ion

alis

ed

syst

em

re

du

ce

s m

ort

alit

y.

The

re

sults

sho

w t

ha

t te

rtia

ry t

rau

ma

ce

ntr

es

an

d r

ed

uc

ed

pre

-ho

spita

l tim

es

are

th

e

ess

en

tia

l co

mp

on

en

ts o

f a

n e

ffic

ien

t tr

au

ma

ca

re s

yst

em

.


95

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Sa

mp

alis

et

al.

19

99)

Ca

na

da

co

ntin

ue

d

Hyp

oth

esi

s 2a

. te

ste

d m

ort

alit

y r

ate

s

of

pa

tie

nts

tre

ate

d a

t te

rtia

ry c

are

ce

ntr

es

co

mp

are

d w

ith

pa

tie

nts

at

less

sp

ec

ialis

ed

ho

spita

ls.

Als

o

test

ed

mo

rta

lity r

ate

s in

th

ose

wh

o

we

re t

ran

spo

rte

d d

ire

ctly f

rom

th

e

sce

ne

to

th

e t

ert

iary

ce

ntr

e a

nd

tho

se w

ho

we

re t

ran

spo

rte

d

ind

irec

tly.

Hyp

oth

esi

s 2b

(te

ste

d t

he

ass

oc

iatio

n b

etw

ee

n p

roc

ess

of

tra

um

a c

are

re

gio

na

lisa

tio

n a

nd

mo

rta

lity a

s it r

ela

ted

to

th

e

Qu

eb

ec

tra

um

a s

yst

em

- w

he

re t

he

rate

of

ap

pro

pria

te p

atie

nt

tria

ge

will

inc

rea

se a

nd

pre

-ho

spita

l tim

e

will

de

cre

ase

ove

r tim

e).

Th

is

co

mp

on

en

t w

as

no

t re

leva

nt

to t

his

revie

w.

Hyp

oth

esi

s 2c

(re

du

ce

d p

re-h

osp

ita

l

tim

e is

ass

oc

iate

d w

ith

re

du

ce

d

mo

rta

lity).

Lo

gis

tic

re

gre

ssio

n u

sed

.


96

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Grz

yb

ow

ski e

t a

l. 2

000

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l II-

2

Stu

dy s

ett

ing

.

Stu

dy c

on

du

cte

d in

19

96 a

nd

19

97

co

nsi

stin

g o

f p

atie

nts

ha

vin

g a

n A

MI

wh

o w

ere

tra

nsp

ort

ed

by a

mb

ula

nc

e t

o

on

e o

f th

ree

ho

spita

ls in

th

e s

ub

urb

s o

f

De

tro

it.

Pa

rtic

ipa

nts

:

253 e

ligib

le b

ut

244 s

ele

cte

d d

ue

to

mis

sin

g o

utc

om

e d

ata

in t

he

oth

er

nin

e

pa

tie

nts

Ba

selin

e a

na

lysi

s

Me

an

ag

e 6

6.6

ye

ars

Ma

le 6

0.2

%

Ca

rdia

c d

ea

th w

ith

in s

eve

n d

ays

14.8

%

Ac

ute

MI 96%

Ca

rdia

c a

rre

st 7

%

Me

an

to

tal E

MS t

ime

44 m

inu

tes

EM

S v

ita

l sig

ns:

He

art

ra

te 7

8.9

Re

spira

tory

ra

te 2

3.1

Syst

olic

BP

128

.1

Dia

sto

lic B

P 8

0

Inc

lu/e

xcl c

rite

ria

.

≥18 y

ea

rs

Ch

ief

co

mp

lain

t o

f c

he

st p

ain

or

sho

rtn

ess

of

bre

ath

Pa

tie

nts

with

su

ita

ble

ou

tco

me

da

ta

Da

ta c

olle

ctio

n

Am

bu

lan

ce

ru

n s

he

ets

: a

ge

,

ge

nd

er,

ra

ce

, EM

S v

ita

l sig

ns,

am

bu

lan

ce

ru

n t

ime

s, t

yp

e o

f

ho

spita

l.

Ou

tco

me

me

asu

res

De

ath

with

in s

eve

n d

ays

of

ED

arr

iva

l

An

aly

sis

Pre

dic

tor

va

riab

les

we

re c

om

pa

red

by s

urv

iva

l sta

tus.

Od

ds

ratio

s a

nd

95%

co

nfid

en

ce

inte

rva

ls w

ere

est

ima

ted

fo

r e

ac

h p

red

icto

r

va

riab

le.

A f

orw

ard

ste

pw

ise

log

istic

reg

ress

ion

mo

de

l wa

s fitt

ed

.

Me

an

to

tal E

MS t

ime

, su

rviv

ors

v

de

ath

s (m

inu

tes)

Su

rviv

ors

42.8

De

ath

s 50.6

P

≤0.0

1

Lim

ita

tio

ns

�

Initia

l po

pu

latio

n o

f 291

se

lec

ted

fo

r

stu

dy. Exc

lusi

on

s: in

elig

ible

ch

ief

co

mp

lain

t (n

=37

), m

issi

ng

ED

ch

ief

co

mp

lain

t (n

=1),

mis

sin

g o

utc

om

e

da

ta (

n=

9).

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g.

Wh

ile m

ultiv

aria

te

mo

de

llin

g w

as

co

nd

uc

ted

, p

re-

ho

spita

l tim

e w

as

no

t in

clu

de

d in

th

e

mo

de

l, th

us

co

mp

ariso

n o

f p

re-

ho

spita

l tim

es

wa

s b

ase

d o

n

un

iva

ria

te a

na

lysi

s o

nly

.

Co

mm

en

ts

�

Ass

ess

ed

wh

ich

ind

ep

en

de

nt

va

riab

les

pre

dic

t d

ea

th w

ith

in s

eve

n

da

ys

in –

pa

tie

nts

with

su

spe

cte

d A

MI

tra

nsp

ort

ed

by E

MS.

�

Ro

bu

stn

ess

of

the

da

ta f

or

est

ima

tin

g

tota

l pre

-ho

spita

l tim

e w

as

no

t c

lea

r

bu

t a

ny e

rro

r se

em

s lik

ely

to

be

sm

all.

�

Low

ris

k o

f o

utc

om

e m

isc

lass

ific

atio

n.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

A t

riag

e r

ule

ba

sed

on

a m

ultiv

aria

te m

od

el

ca

n id

en

tify

th

e g

rou

p a

t h

igh

ris

k o

f e

arly

ca

rdia

c d

ea

th.

This

de

cis

ion

ru

le n

ee

ds

to

be

pro

spe

ctive

ly v

alid

ate

d.


97

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Be

rns

et

al.

20

01)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

A h

osp

ita

l ba

sed

he

lico

pte

r p

rog

ram

me

in M

inn

eso

ta.

Co

nd

uc

ted

ove

r Ja

nu

ary

1998 t

o J

un

e 1

999.

Pa

rtic

ipa

nts

:

266 h

elic

op

ter

pa

tie

nts

an

d 2

8 r

oa

d

am

bu

lan

ce

pa

tie

nts

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ag

e (

ye

ars

)

He

lico

pte

r 65

Gro

un

d 6

7

Ma

le:

He

lico

pte

r 67%

Gro

un

d 6

1%

Tim

e f

rom

ca

ll to

ho

spita

l arr

iva

l

(min

ute

s)

He

lico

pte

r 104

Gro

un

d 1

42

Inc

lu/e

xcl c

rite

ria

.

All

ca

rdia

c p

atie

nts

tra

nsp

ort

ed

by

the

Ma

yo

On

e R

oc

he

ste

r

he

lico

pte

r, in

clu

din

g p

atie

nts

with

ch

est

pa

in, a

ng

ina

, M

I a

nd

arr

hyth

mia

s. E

xclu

sio

n c

rite

ria

inc

lud

ed

pa

tie

nts

tra

nsp

ort

ed

to

a

diffe

ren

t h

osp

ita

l, b

y t

he

Ma

yo

fix

ed

win

g s

erv

ice

, b

y g

rou

nd

with

a f

ligh

t

nu

rse

on

bo

ard

, b

y t

he

Ma

yo

On

e

Ea

u C

laire

he

lico

pte

r a

nd

by

an

oth

er

he

lico

pte

r se

rvic

e.

Da

ta c

olle

ctio

n

Ch

art

re

vie

w

Ou

tco

me

me

asu

re

Ho

spita

l le

ng

th o

f st

ay (

LOS)

Mo

rta

lity

An

aly

sis

t te

st p

erf

orm

ed

on

tra

nsp

ort

tim

e,

tim

e f

rom

ca

ll u

ntil h

osp

ita

l arr

iva

l,

CC

U L

OS a

nd

ho

spita

l LO

S

Ho

spita

l le

ng

th o

f st

ay (

da

ys)

He

lico

pte

r 6.4

Gro

un

d 8

P=

0.0

4

Mo

rta

lity d

ata

He

lico

pte

r 7%

Gro

un

d 4

%

Au

tho

rs d

id n

ot

pre

sen

t a

sta

tist

ica

l

an

aly

sis

of

this

re

sult.

Re

vie

we

rs s

elf

an

aly

sis,

usi

ng

Fis

he

r’s

exa

ct

test

fou

nd

no

sig

nific

an

t d

iffe

ren

ce

be

twe

en

th

e t

wo

gro

up

s

Lim

ita

tio

ns

�

Wh

ile t

he

re w

ere

50 e

ligib

le g

rou

nd

am

bu

lan

ce

pa

tie

nts

re

co

rds

we

re

on

ly r

ec

eiv

ed

in 2

8 o

f th

ese

re

sult

ing

in

a s

ign

ific

an

t se

lec

tio

n b

ias.

�

Diffe

ren

t st

aff

ing

mix

es

be

twe

en

he

lico

pte

r a

nd

ro

ad

am

bu

lan

ce

s

me

an

s it is

no

t p

oss

ible

to

asc

rib

e a

ny

diffe

ren

ce

in o

utc

om

e t

o d

iffe

ren

ce

in

pre

-ho

spita

l tim

e.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

.

Sp

ec

ific

ally

, th

ere

wa

s n

o c

om

pa

riso

n

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d

ou

tco

me

, ra

the

r th

e c

om

pa

riso

n w

as

be

twe

en

air a

nd

ro

ad

am

bu

lan

ce

(an

d t

hu

s u

sed

as

a p

roxy

fo

r tim

e,

giv

en

th

e lo

ng

er

tra

nsp

ort

atio

n t

ime

in

the

gro

un

d g

rou

p).

Co

mm

en

ts

�

Aim

ed

to

inve

stig

ate

th

e o

utc

om

e in

ca

rdia

c p

atie

nts

tra

nsp

ort

ed

by

he

lico

pte

r ve

rsu

s g

rou

nd

am

bu

lan

ce

.

�

Gro

un

d t

ran

spo

rta

tio

n w

as

use

d

un

de

r c

on

ditio

ns

tha

t d

id n

ot

allo

w f

or

he

lico

pte

r flig

ht

(eg

we

ath

er,

ma

inte

na

nc

e, a

nd

airc

raft

in

use

or

dis

tan

ce

in

vo

lve

d).

�

Mo

st p

atie

nts

tra

nsf

err

ed

fro

m t

he

refe

rrin

g f

ac

ility

’s E

D.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

He

lico

pte

r tr

an

spo

rt b

en

efits

th

e c

ard

iac

pa

tie

nt

with

de

cre

ase

d c

he

st p

ain

as

a

resu

lt o

f m

ore

tre

atm

en

ts e

n r

ou

te;

de

cre

ase

d t

ime

fro

m c

all

to a

rriv

al,

resu

ltin

g

in d

ec

rea

sed

tim

e t

o in

terv

en

tio

n; a

nd

sho

rte

r p

re-h

osp

ita

l tim

e a

nd

ho

spita

l sta

ys.


98

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Cla

rke

et

al.

2002

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Inju

red

pa

tie

nts

wh

o w

ere

ca

red

fo

r a

t

a P

en

nsy

lva

nia

tra

um

a c

en

tre

Pa

rtic

ipa

nts

:

250 p

atie

nts

me

t a

prio

ri e

ligib

ility

crite

ria

bu

t se

ve

n w

ere

exc

lud

ed

du

e t

o

ext

rem

e p

re-h

osp

ita

l or

ED

tim

es.

Ba

selin

e a

na

lyse

s

Syst

olic

BP

ra

ng

e 3

0-9

0m

mH

g

Ela

pse

d t

ime

to

ED

ra

ng

e 7

-185

min

ute

s

Tim

e in

ED

ra

ng

e 7

-91

5 m

inu

tes

Inc

lu/e

xcl c

rite

ria

.

Ma

de

use

of

tra

um

a r

eg

istr

y d

ata

wh

ich

om

its

info

rma

tio

n o

n p

atie

nts

wh

o d

ied

, w

ere

tra

nsf

err

ed

to

oth

er

tra

um

a c

en

tre

s o

r h

ad

a h

osp

ita

l

sta

y o

f m

ore

th

an

tw

o d

ays.

Pa

tie

nts

with

iso

late

d h

ip f

rac

ture

s

we

re e

xclu

de

d.

Pa

tie

nts

we

re r

est

ric

ted

to

th

ose

bro

ug

ht

dire

ctly t

o t

he

tra

um

a

ce

ntr

e f

rom

th

e s

ce

ne

, w

ere

no

t

tra

nsf

err

ed

fro

m t

he

ED

to

an

oth

er

ho

spita

l, d

id n

ot

ha

ve

co

nfo

un

din

g

bu

rns

or

pre

-exi

stin

g c

on

ditio

ns.

Pa

tie

nts

we

re t

he

n s

ele

cte

d b

ase

d

on

:

Syst

olic

BP

< 9

0m

mH

g o

n a

rriv

al a

t

ED

Pa

tie

nt

eith

er

die

d in

ED

or

wa

s

tra

nsf

err

ed

to

th

e o

pe

ratin

g r

oo

m

for

lap

aro

tom

y

Ab

do

min

al v

asc

ula

r, s

olid

org

an

or

wa

ll in

jury

with

an

ab

bre

via

ted

inju

ry s

ca

le s

co

re (

AIS

) o

f 3

-6

No

oth

er

inju

ries

with

an

AIS

>2

exc

ep

t fo

r a

lac

era

ted

dia

ph

rag

m

or

op

en

, d

isp

lac

ed

, c

om

min

ute

d

pe

lvic

fra

ctu

re

Eith

er

the

tim

e o

f in

jury

or

am

bu

lan

ce

dis

pa

tch

an

d t

he

tim

e

of

arr

iva

l in

ED

an

d t

ime

of

de

pa

rtu

re f

rom

ED

or

de

ath

in E

D o

r

arr

iva

l at

OR

.

Ris

k r

atio

s fo

r d

ea

th b

y m

inu

tes

to

ED

(95%

CI)

1-3

0 m

inu

tes

RR

0.7

73 (

0.5

01-1

.194)

31-6

0 m

inu

tes

RR

1.2

68 (

0.9

80

-1.6

41

)

61-9

0 m

inu

tes

RR

0.8

32 (

0.4

96

-1.3

96

)

91-1

85 m

ins

RR

0.7

40

(0

.18

9-2

.888

)

Lim

ita

tio

ns

�

Re

lied

on

ac

cu

rac

y o

f re

gis

try d

ata

.

No

da

ta p

rese

nte

d o

n t

he

ac

cu

rac

y

of

tha

t so

urc

e.

�

Ge

ne

ralis

ab

ility

re

stric

ted

to

a n

arr

ow

ran

ge

of

inju

rie

s re

sultin

g f

rom

tra

um

a

du

e t

o t

he

na

rro

w s

ele

ctio

n c

rite

ria

.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Foc

us

of

the

stu

dy w

as

mo

re o

n E

D

tim

e t

ha

n t

he

pre

-ho

spita

l pe

rio

d.

�

Se

ve

n o

f th

e 2

50 e

ligib

le p

atie

nts

we

re e

xclu

de

d d

ue

to

extr

em

e p

re-

ho

spita

l tim

ing

s (8

ho

urs

fo

r o

ne

an

d 7

da

ys

20h

ou

rs f

or

an

oth

er)

an

d

pro

lon

ge

d E

D t

ime

(>

24 h

ou

rs).

�

Po

ten

tia

l bia

ses

ide

ntifie

d b

y t

he

au

tho

rs in

clu

de

d m

isc

lass

ific

atio

n o

f

tim

ing

inte

rva

ls d

ue

to

a t

en

de

nc

y t

o

rou

nd

to

th

e n

ea

rest

5 m

inu

tes,

fa

ilure

to id

en

tify

pre

-exis

tin

g c

on

ditio

ns

in

ind

ivid

ua

ls w

ho

die

d s

ho

rtly

aft

er

arr

iva

l, se

lec

tio

n b

ias

du

e t

o m

issi

ng

tim

e d

ata

.


99

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Cla

rke

et

al.

2002

)

USA

co

ntin

ue

d

Da

ta c

olle

ctio

n

Use

d d

ata

fro

m t

he

Pe

nn

sylv

an

ia

Tra

um

a S

yst

em

s Fo

un

da

tio

n t

rau

ma

reg

istr

y. D

ata

extr

ac

ted

inc

lud

ed

:

Tim

e o

f in

jury

, tim

e o

f a

mb

ula

nc

e

dis

pa

tch

, tim

e p

atie

nt

arr

ive

d a

t ED

,

syst

olic

BP

, d

iag

no

ses,

pre

-exi

stin

g

co

nd

itio

ns,

tim

e p

atie

nt

left

ED

, tim

e

pa

tie

nt

arr

ive

d in

op

era

tin

g r

oo

m,

op

era

tive

pro

ce

du

res,

pa

tie

nts

ou

tco

me

.

Ou

tco

me

me

asu

res

Mo

rta

lity

An

aly

sis

Ris

k r

atio

s fo

r d

ea

th w

ere

ca

lcu

late

d f

or

the

tim

e t

o t

he

ED

,

tim

e in

th

e E

D a

nd

to

tal tim

e (

tim

e

to E

D a

nd

tim

e in

ED

) a

nd

th

e S

BP

on

arr

iva

l in

th

e E

D. Lo

gis

tic

reg

ress

ion

wa

s u

sed

to

mo

de

l

pre

dic

tio

ns

of

ou

tco

me

usi

ng

co

ntin

uo

us

va

ria

ble

s o

f tim

e a

nd

SB

P w

ith

in t

he

tim

e in

terv

als

th

at

we

re f

ou

nd

to

ha

ve

sig

nific

an

t risk

ratio

s.

C

om

me

nts

�

Exa

min

ed

th

e r

ela

tio

nsh

ip b

etw

ee

n

surv

iva

l an

d t

ime

in

th

e e

me

rge

nc

y

de

pa

rtm

en

t b

efo

re la

pa

roto

my f

or

hyp

ote

nsi

ve

pa

tie

nts

ble

ed

ing

fro

m

ab

do

min

al i

nju

ries.

�

Pre

-ho

spita

l tim

e p

refe

ren

tia

lly u

sed

tim

e f

rom

dis

pa

tch

ra

the

r th

an

tim

e o

f

inju

ry a

s th

e s

tart

of

the

pre

-ho

spita

l

pe

riod

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Am

on

g p

atie

nts

in a

tra

um

a r

eg

istr

y w

ho

we

re h

yp

ote

nsi

ve

on

arr

iva

l in

ED

an

d h

ad

ma

jor

inju

rie

s is

ola

ted

to

th

e a

bd

om

en

req

uirin

g e

me

rge

nc

y la

pa

roto

my,

the

pro

ba

bili

ty o

f d

ea

th s

ho

we

d a

re

latio

nsh

ip

to b

oth

th

e e

xte

nt

of

hyp

ote

nsi

on

an

d t

he

len

gth

of

tim

e in

th

e E

D f

or

pa

tie

nts

wh

o

we

re in

th

e E

D f

or

90 m

inu

tes

or

less

.


10

0

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Lim

an

d S

eo

w 2

00

2)

Sin

ga

po

re

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Am

bu

lan

ce

se

rvic

e in

Sin

ga

po

re u

ses

a

sin

gle

tie

r sy

ste

m. Stu

dy s

et

in E

D o

f Y

an

Toc

k S

en

g H

osp

ita

l, w

hic

h s

ee

s a

bo

ut

350 p

atie

nts

pe

r d

ay.

Pa

rtic

ipa

nts

:

n=

93 (

15 s

urv

ivo

rs, 78

no

n-s

urv

ivo

rs)

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ag

e (

ye

ars

)

Su

rviv

ors

63.1

No

n-s

urv

ivo

rs 6

5.5

Ma

le (

%)

Su

rviv

ors

53.3

No

n-s

urv

ivo

rs 6

2.8

Initia

l ca

rdia

c r

hyth

m: a

syst

ole

(%

)

Su

rviv

ors

53.3

No

n-s

urv

ivo

rs 7

1.3

Byst

an

de

r C

PR

(%

)

Su

rviv

ors

6.6

No

n-s

urv

ivo

rs 1

8.0

Pre

-ho

spita

l de

fib

rilla

tio

n (

%)

Su

rviv

ors

6.6

No

n-s

urv

ivo

rs 1

8.0

RO

SC

(%

)

Su

rviv

ors

46.7

No

n-s

urv

ivo

rs 0

.0

P<

0.0

01

Inc

lu/e

xcl c

rite

ria

.

All

ou

t-o

f-h

osp

ita

l ca

rdia

c a

rre

st

(OH

CA

) p

atie

nts

pre

sen

tin

g f

rom

No

v 2

001 t

hro

ug

h J

an

20

02 w

ith

no

n-t

rau

ma

tic

OH

CA

.

Da

ta c

olle

ctio

n

Da

ta c

olle

cte

d f

rom

am

bu

lan

ce

ca

se r

ec

ord

s, E

D r

esu

scita

tio

n

ch

art

s a

nd

ED

VH

F C

ase

Lo

g S

he

ets

,

in-p

atie

nt

ho

spita

l re

co

rds.

Co

llec

ted

de

mo

gra

ph

ic

info

rma

tio

n,

tim

e r

ela

ted

da

ta,

initia

l ca

rdia

c r

hyth

m,

use

of

au

tom

atic

exte

rna

l de

fib

rilla

tor,

resu

lt o

f re

susc

ita

tio

n o

n s

ce

ne

or

en

-ro

ute

to

ho

spita

l, p

atie

nt’

s

pre

mo

rbid

co

nd

itio

n.

Ou

tco

me

me

asu

res

Su

rviv

al p

ost

ED

re

susc

ita

tio

n

An

aly

sis

Da

ta a

na

lyse

d u

sin

g t

wo

ta

iled

t

test

fo

r a

ll c

on

tin

uo

us

va

ria

ble

s a

nd

ch

i-sq

ua

re t

est

fo

r a

ll d

isc

rete

va

riab

les.

Co

mp

aris

on

of

surv

ivo

rs (

po

st E

D

resu

scita

tio

n)

an

d n

on

-su

rviv

ors

,

tota

l pre

-ho

spita

l tim

e (

min

ute

s)

Su

rviv

ors

: 3

8.3

No

n-s

urv

ivo

rs: 3

5.4

P=

0.9

2

No

te o

nly

on

e p

atie

nt

surv

ive

d t

o

ho

spita

l dis

ch

arg

e.

Lim

ita

tio

ns

�

Re

lied

on

ac

cu

rac

y o

f re

co

rde

d d

ata

.

The

pa

ram

ed

ics

are

on

ly a

ble

to

co

mp

lete

th

e r

ele

va

nt

form

co

nta

inin

g t

he

tim

e d

ata

on

arr

iva

l a

t

ED

so

th

e t

ime

re

co

rde

d is

like

ly t

o b

e

susc

ep

tib

le t

o m

isc

lass

ific

atio

n.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g a

nd

me

tho

d o

f a

na

lysi

s

wa

s n

ot

use

ful f

or

co

ntr

olli

ng

po

ten

tia

l

co

nfo

un

de

rs.

�

Sm

all

pa

tie

nt

nu

mb

ers

, p

art

icu

larly

am

on

gst

th

e s

urv

ivo

rs (

n=

15

), r

ed

uc

ed

stu

dy p

ow

er.

�

No

t th

e m

ost

use

ful o

utc

om

e m

ea

sure

(su

rviv

al p

ost

ED

re

susc

ita

tio

n).

It

is

no

tew

ort

hy t

ha

t o

nly

1 o

f th

e 1

5

pa

tie

nts

su

rviv

ing

ED

re

susc

ita

tio

n

ac

tua

lly s

urv

ive

d t

o h

osp

ita

l

dis

ch

arg

e.

�

So

me

inc

on

sist

en

cie

s in

th

e p

ap

er

eg

.

On

e s

tate

me

nt

sug

ge

ste

d 1

5 s

urv

ive

d

ED

re

susc

ita

tio

n a

nd

an

oth

er

sug

ge

ste

d s

eve

n s

urv

ive

d E

D

resu

scita

tio

n.


10

1

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Lim

an

d S

eo

w 2

00

2)

Sin

ga

po

re

co

ntin

ue

d

Co

mm

en

ts

�

Aim

ed

to

eva

lua

te c

ha

rac

terist

ics

an

d o

utc

om

e o

f o

ut-

of-

ho

spita

l

ca

rdia

c a

rre

st p

atie

nts

pre

sen

tin

g t

o

the

ED

, a

nd

to

exa

min

e f

ac

tors

th

at

co

uld

be

use

d t

o d

ete

rmin

e w

he

the

r

to p

rolo

ng

or

ab

ort

re

susc

ita

tio

n f

or

the

se p

atie

nts

.

�

Inc

lud

ed

co

nse

cu

tive

pa

tie

nts

.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

su

rviv

al ra

te f

or

pa

tie

nts

with

OH

CA

aft

er

ED

re

susc

ita

tio

n is

sim

ilar

to t

he

re

sults

fro

m o

the

r st

ud

ies.

Pro

lon

ge

d r

esu

scita

tio

n

eff

ort

s a

pp

ea

r to

be

fu

tile

fo

r O

HC

A p

atie

nts

if t

he

tim

e f

rom

ca

rdia

c a

rre

st u

ntil a

rriv

al i

n

the

ED

is a

t le

ast

30 m

inu

tes

co

up

led

with

RO

SC

, a

nd

if c

on

tin

uo

us

asy

sto

le h

as

be

en

do

cu

me

nte

d f

or

mo

re t

ha

n 1

0 m

inu

tes.


10

2

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ost

erw

ald

er

20

02

)

Sw

itze

rla

nd

Pro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Stu

dy h

osp

ita

l (S

t G

alle

n C

an

ton

al

Ho

spita

l) s

erv

es

Ea

ste

rn S

witze

rla

nd

, h

as

an

imm

ed

iate

ca

tch

me

nt

po

pu

latio

n o

f

ab

ou

t 100,0

00

an

d h

as

ab

ou

t 80

0 b

ed

s

(in

clu

din

g t

wo

IC

Us)

.

EM

S in

clu

de

d h

elic

op

ters

(o

fte

n w

ith

a

ph

ysi

cia

n o

n b

oa

rd)

an

d g

rou

nd

am

bu

lan

ce

s

Pa

rtic

ipa

nts

:

N=

254

inc

lud

ing

107

with

a r

esc

ue

p

erio

d ≤

60 m

inu

tes

an

d 1

47 w

ith

a

resc

ue

pe

rio

d >

60 m

inu

tes.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

Me

an

ag

e (

ye

ars

)

≤60 m

inu

tes

gro

up

: 30

>60 m

inu

tes

gro

up

: 29

Ma

le (

%)

≤60 m

inu

tes

gro

up

: 66

>60 m

inu

tes

gro

up

: 78

Me

dia

n ISS

≤60 m

inu

tes

gro

up

: 24

>60 m

inu

tes

gro

up

: 24

Inc

lu/e

xcl c

rite

ria

.

Blu

nt

tra

um

a

Tre

atm

en

t in

th

e s

ho

ck r

oo

m,

ED

, St

Ga

llen

Ca

nto

na

l Ho

spita

l

Pre

sen

ce

of

inju

ries

with

a m

inim

um

AIS

of

≥2 in

at

lea

st t

wo

of

six

de

fin

ed

bo

dy r

eg

ion

s

Eith

er

tra

nsf

er

to IC

U o

r a

sta

y o

f a

t

lea

st t

hre

e d

ays

in h

osp

ita

l or

de

ath

follo

win

g a

dm

issi

on

.

Ou

tco

me

me

asu

res

30 d

ay m

ort

alit

y

Pre

dic

ted

mo

rta

lity b

ase

d o

n

ASC

OT

sco

re

Re

sultin

g e

xce

ss m

ort

alit

y r

ate

(ac

tua

l – e

xpe

cte

d d

ea

ths)

An

aly

sis

Flo

ra’s

Z s

tatist

ic u

sed

to

co

mp

are

ac

tua

l with

exp

ec

ted

mo

rta

lity.

Po

ssib

le c

on

fou

nd

ing

va

riab

les

an

d

furt

he

r c

om

pa

riso

ns

we

re t

est

ed

usi

ng

th

e in

de

pe

nd

en

t Stu

de

nt’

s t

test

, M

an

n W

hitn

ey U

te

st, c

hi

squ

are

te

st a

nd

log

istic

re

gre

ssio

n

Ac

tua

l ve

rsu

s p

red

icte

d m

ort

alit

y

(30 d

ays)

Tr

an

spo

rt t

ime

≤60 m

inu

tes

Ac

tua

l: 1

4%

Pre

dic

ted

: 9.5

%

P=

0.0

6

Tra

nsp

ort

tim

e >

60 m

inu

tes

Ac

tua

l: 1

0.2

%

Pre

dic

ted

: 13.1

%

P=

0.1

9

Ad

just

ed

co

mp

aris

on

be

twe

en

tra

nsp

ort

tim

e ≤

60 m

inu

tes

an

d >

60

min

ute

s, a

nd

mo

rta

lity

OR

(>

60 m

inu

tes

as

refe

ren

ce

) 8

(95%

CI 1.7

-38.5

)

Lim

ita

tio

ns

�

M s

tatist

ic n

ot

pre

sen

ted

so

de

gre

e o

f

fit

with

MTO

S d

ata

wa

s u

nc

lea

r.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g a

lth

ou

gh

mu

ltiv

aria

te

an

aly

sis

is a

use

ful m

eth

od

to

co

ntr

ol

for

kn

ow

n c

on

fou

nd

ers

. H

ow

eve

r, t

he

va

riab

les

inc

lud

ed

in t

he

mu

ltiv

aria

te

mo

de

l we

re n

ot

do

cu

me

nte

d.

�

Stu

dy r

est

ric

ted

to

blu

nt

tra

um

a

pa

tie

nts

wh

ich

ne

ed

s to

be

rec

og

nis

ed

wh

en

co

nsi

de

rin

g

ge

ne

ralis

ab

ility

.

�

Re

lied

on

ac

cu

rac

y o

f re

gis

try d

ata

.

No

da

ta p

rese

nte

d o

n t

he

ac

cu

rac

y

of

tha

t so

urc

e.


10

3

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ost

erw

ald

er

20

02

)

Sw

itze

rla

nd

co

ntin

ue

d

C

au

se o

f tr

au

ma

:

roa

d t

raff

ic a

cc

ide

nt

(%)

≤60 m

inu

tes

gro

up

: 77

>60 m

inu

tes

gro

up

: 69

Wo

rk (

%)

≤60 m

inu

tes

gro

up

: 11

>60 m

inu

tes

gro

up

: 12

Sp

ort

(%

):

≤60 m

inu

tes

gro

up

: 1

>60 m

inu

tes

gro

up

: 13

P=

0.0

05

Su

icid

e/v

iole

nc

e (

%)

≤60 m

inu

tes

gro

up

: 9

>60 m

inu

tes

gro

up

: 1

P=

0.0

04

�

Sig

nific

an

t d

iffe

ren

ce

s in

ba

selin

e

me

asu

res

in r

ela

tio

n t

o c

au

se o

f

tra

um

a:

spo

rt w

as

ass

oc

iate

d w

ith

lon

ge

r tr

an

spo

rt t

ime

an

d

suic

ide

/vio

len

ce

with

sh

ort

er

tra

nsp

ort

tim

e.

Als

o f

ew

er

pa

tie

nts

we

re t

rea

ted

w

ith

a p

hysi

cia

n in

th

e ≤

60 m

inu

tes

gro

up

.

�

26 p

atie

nts

we

re o

mitte

d d

ue

to

mis

sin

g t

ime

da

ta (

ove

rall

the

re w

as

mis

sin

g d

ata

in 9

% o

f th

e s

tud

y

po

pu

latio

n).

Co

mm

en

ts

�

Aim

ed

to

eva

lua

te t

he

hyp

oth

esi

s th

at

exc

ee

din

g t

he

60 m

inu

te li

mit f

or

the

en

tire

pre

-ho

spita

l tim

e in

cre

ase

s

mo

rta

lity o

f b

lun

t p

oly

tra

um

a p

atie

nts

.

�

All

pa

tie

nts

tre

ate

d a

t th

e s

am

e L

ev

el

1 t

rau

ma

ho

spita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

It a

pp

ea

rs in

th

is t

rau

ma

syst

em

, in

wh

ich

em

erg

en

cy p

hysi

cia

ns

oft

en

are

de

plo

ye

d,

tha

t th

e g

old

en

ho

ur

of

sho

ck c

an

be

ext

en

de

d s

afe

ly in

ma

ny b

lun

t p

oly

tra

um

a

pa

tie

nts

, si

nc

e t

his

wa

s a

sso

cia

ted

with

be

tte

r su

rviv

al f

igu

res

tha

n in

th

ose

pa

tie

nts

for

wh

om

th

e t

ime

wa

s <

1 h

ou

r.


10

4

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

rne

r e

t a

l. 20

03

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Da

ta o

bta

ine

d f

rom

re

co

rds

ma

inta

ine

d

by t

he

on

ly a

du

lt r

eg

ion

al t

rau

ma

ce

ntr

e in

we

ste

rn N

ew

Yo

rk.

The

ce

ntr

e

wa

s a

389 b

ed

te

rtia

ry c

are

te

ac

hin

g

fac

ility

with

ab

ou

t 1

3,5

00 a

nn

ua

l

ad

mis

sio

ns

(ap

pro

xim

ate

ly 1

,600 o

f

wh

ich

re

sulte

d f

rom

tra

um

a).

Pa

rtic

ipa

nts

:

N=

187

7

Ba

selin

e a

na

lyse

s (p

atie

nts

with

co

mp

lete

da

ta)

Tra

nsp

ort

ed

by h

elic

op

ter

8%

Pe

ne

tra

tin

g in

jury

23%

Ma

le 7

2%

Me

an

ag

e 3

8 y

ea

rs

Me

an

re

vis

ed

tra

um

a s

co

re 7

.5

Me

an

ISS 1

0

Me

an

to

tal o

ut-

of-

ho

spita

l tim

e 3

5

min

ute

s

Inc

lu/e

xcl c

rite

ria

.

All

pa

tie

nts

fro

m J

an

199

3 t

o O

ct

1996 if

th

e p

atie

nt

ha

d b

ee

n

tra

nsp

ort

ed

dire

ctly f

rom

th

e s

ce

ne

by a

mb

ula

nc

e o

r h

elic

op

ter.

Pa

tie

nts

we

re a

dm

itte

d f

rom

ED

or

die

d in

th

e E

D.

Exc

lud

ed

pa

tie

nts

with

inc

om

ple

te

da

ta, w

ith

mo

re t

ha

n o

ne

da

y

diffe

ren

ce

be

twe

en

th

e d

ate

of

inju

ry a

nd

th

e d

ate

of

ad

mis

sio

n,

CP

R in

itia

ted

in t

he

fie

ld o

r

tra

nsp

ort

ed

fro

m a

co

rre

ctio

na

l

fac

ility

.

Da

ta c

olle

ctio

n

Mo

st d

ata

we

re e

xtr

ac

ted

fro

m t

he

tra

um

a r

eg

istr

y. O

ut-

of-

ho

spita

l

pa

tie

nt

ca

re r

ep

ort

an

d d

isp

atc

h

ag

en

cy r

ec

ord

s u

sed

to

sup

ple

me

nt

the

re

gis

try t

ime

da

ta

as

it w

as

oft

en

in

co

mp

lete

.

Ou

t o

f h

osp

ita

l va

ria

ble

s: t

ran

spo

rt

mo

de

, to

tal o

ut-

of-

ho

spita

l tim

e,

pa

tie

nt’

s C

UP

S s

tatu

s.

Ho

spita

l va

ria

ble

s in

clu

de

d r

ev

ise

d

tra

um

a s

ca

le, IS

S, E c

od

e, a

dm

issi

on

da

te, a

ge

, se

x, t

yp

e o

f in

jury

.

Ou

tco

me

me

asu

res

Mo

rta

lity

Diffe

ren

ce

in m

ea

n t

ota

l ou

t-o

f-

ho

spita

l tim

e b

etw

ee

n s

urv

ivo

rs a

nd

no

n-s

urv

ivo

rs

Su

rviv

ors

: 3

5.2

6 m

inu

tes

No

n-s

urv

ivo

rs: 3

1.5

8 m

inu

tes

Diffe

ren

ce

3.6

9 m

inu

tes

(95%

CI

0.5

2-6

.85 m

inu

tes)

Ad

just

ed

OR

, to

tal o

ut

of

ho

spita

l

tim

e,

inc

rea

sin

g p

re-h

osp

ita

l tim

e

an

d o

dd

s o

f m

ort

alit

y

OR

0.9

87 (

95%

CI 0.9

7-1

.00).

Lim

ita

tio

ns

�

Me

dic

al r

ec

ord

s re

vie

w w

ith

inh

ere

nt

limita

tio

ns

of

this

so

urc

e. P

oss

ible

limita

tio

ns

inc

lud

e m

issi

ng

da

ta a

nd

inc

on

sist

en

t m

eth

od

s o

f re

co

rdin

g.

�

Po

ten

tia

l fo

r m

isc

lass

ific

atio

n o

f to

tal

pre

-ho

spita

l tim

e o

r p

ote

ntia

l

co

nfo

un

din

g v

aria

ble

s. M

ost

like

ly t

o

be

ra

nd

om

, re

sultin

g in

dilu

tio

n o

f th

e

eff

ec

t.

�

Like

ly s

ele

ctio

n b

ias

du

e t

o t

he

om

issi

on

of

pa

tie

nts

with

mis

sin

g d

ata

(48

2 o

f 235

9 w

ere

om

itte

d d

ue

to

mis

sin

g d

ata

, 2

0%

).

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g a

lth

ou

gh

mu

ltiv

aria

te

an

aly

sis

is a

use

ful m

eth

od

to

co

ntr

ol

for

kn

ow

n c

on

fou

nd

ers

.

�

It is

po

ssib

le t

ha

t p

rov

ide

rs t

ran

spo

rt

the

pa

tie

nts

th

ey b

elie

ve

d t

o b

e m

ost

seve

rely

inju

red

qu

icke

r, t

hu

s b

iasi

ng

the

re

sults.


10

5

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Le

rne

r e

t a

l. 20

03

)

USA

co

ntin

ue

d

An

aly

sis

Biv

aria

te a

na

lyse

s c

on

du

cte

d t

o

de

term

ine

wh

ich

va

ria

ble

s w

ere

ass

oc

iate

d w

ith

mo

rta

lity. Stu

de

nt’

s

t te

st u

sed

fo

r c

on

tin

uo

us

va

riab

les

an

d c

hi s

qu

are

or

Fis

he

r’s

exa

ct

test

use

d f

or

ca

teg

oric

al v

aria

ble

s.

Str

atific

atio

n b

y in

jury

se

ve

rity

an

d

typ

e a

lso

co

nd

uc

ted

. M

ultip

le

pre

dic

tors

log

istic

re

gre

ssio

n u

sed

to

de

term

ine

if t

ota

l ou

t o

f h

osp

ita

l

tim

e w

as

a s

ign

ific

an

t p

red

icto

r o

f

tra

um

a m

ort

alit

y. V

aria

ble

s

ass

oc

iate

d w

ith

mo

rta

lity o

n

un

iva

ria

te a

na

lysi

s w

ere

inc

lud

ed

in

the

mo

de

l.

C

om

me

nts

�

Aim

wa

s to

de

term

ine

if t

he

re is

an

ass

oc

iatio

n b

etw

ee

n t

ota

l ou

t o

f

ho

spita

l tim

e a

nd

tra

um

a m

ort

alit

y.

�

We

ll c

on

du

cte

d s

tatist

ica

l an

aly

sis.

�

All

pa

tie

nts

tra

nsp

ort

ed

to

a s

ing

le

ho

spita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Pro

vid

er

ass

ign

ed

CU

PS s

tatu

s, p

atie

nt

ag

e,

Inju

ry S

eve

rity

Sc

ore

, a

nd

Re

vis

ed

Tra

um

a

Sc

ore

all

we

re s

ign

ific

an

t p

red

icto

rs o

f

tra

um

a p

atie

nt

mo

rta

lity. To

tal o

ut

of

ho

spita

l tim

e w

as

no

t a

sso

cia

ted

with

mo

rta

lity.


10

6

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bie

we

ne

r e

t a

l. 2

004

)

Ge

rma

ny

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Stu

die

d f

ou

r p

oss

ible

pa

thw

ays

of

po

lytr

au

ma

pa

tie

nts

in D

resd

en

,

Ge

rma

ny: 1

. H

elic

op

ter

tra

nsp

ort

atio

n

to L

eve

l 1 t

rau

ma

ce

ntr

e (

HEM

S-U

NI)

, 2.

Am

bu

lan

ce

tra

nsp

ort

atio

n t

o L

eve

l 1

tra

um

a c

en

tre

(A

MB

-UN

I),

3.

Am

bu

lan

ce

tra

nsp

ort

atio

n t

o L

eve

l 2 o

r

3 t

rau

ma

ce

ntr

e (

AM

B-R

EG

), 4

.

Am

bu

lan

ce

tra

nsp

ort

atio

n t

o L

eve

l 2 o

r

3 tr

au

ma

ce

ntr

e w

ith

su

bse

qu

en

t

tra

nsf

er

to le

ve

l 1 t

rau

ma

ce

ntr

e (

INTE

R).

Pa

rtic

ipa

nts

:

403 p

art

icip

an

ts, H

EM

S-U

NI 140,

AM

B-

REG

10

2,

AM

B-U

NI 70, IN

TER

92.

An

aly

ses

co

mp

arin

g g

rou

ps

at

ba

selin

e.

No

diffe

ren

ce

s b

etw

ee

n t

he

fo

ur

gro

up

s

by a

ge

, g

en

de

r o

r IS

S.

Me

an

ag

e 3

6.8

ye

ars

Ma

le 7

3.3

%

Inc

lu/e

xcl c

rite

ria

.

ISS ≥

16

Arr

iva

l of

pa

tie

nt

aliv

e a

t th

e

ho

spita

l

Co

mp

lete

do

cu

me

nta

tio

n o

f a

ll

pa

tie

nt

da

ta.

Exc

lusi

on

crite

ria: a

ge

> 7

5 y

ea

rs,

ISS>

67.

Da

ta c

olle

ctio

n

Da

ta e

xtr

ac

tio

n f

or

all

bu

t th

e A

MB

-

REG

gro

up

fro

m p

oly

tra

um

a

da

tab

ase

of

a s

ing

le L

eve

l 1 t

rau

ma

ce

ntr

e (

co

mp

iled

pro

spe

ctive

ly).

Da

ta c

olle

ctio

n p

erf

orm

ed

retr

osp

ec

tive

ly in

th

e A

MB

-REG

gro

up

.

Do

cu

me

nte

d ISS, a

ge

, g

en

de

r,

resc

ue

tim

e.

Ou

tco

me

me

asu

res

Mo

rta

lity a

t 30 d

ays

An

aly

sis

Sta

tist

ica

l an

aly

sis

of

the

diffe

ren

ce

s

be

twe

en

gro

up

s w

as

co

nd

uc

ted

usi

ng

ch

i-sq

ua

re a

nd

Fis

he

r’s

exa

ct

test

s.

Log

istic

re

gre

ssio

n u

sed

to

ad

just

mo

rta

lity r

isk d

iffe

ren

ce

fo

r a

ge

, IS

S

an

d g

rou

p.

Pre

-ho

spita

l tim

es

(min

ute

s)

HEM

S-U

NI g

rou

p:

90

AM

B-U

NI 68

Un

iva

ria

te c

om

pa

riso

n o

f 30 d

ay

mo

rta

lity:

HEM

S-U

NI 22.1

%

AM

B-U

NI: 1

5.7

%

Mu

ltiv

aria

te c

om

pa

riso

n o

f m

ort

alit

y

(HEM

S-U

NI a

s re

fere

nc

e)

AM

B-U

NI: O

R 1

.06 (

95%

CI 0.4

27-

2.6

35

)

Inte

rpre

tatio

n:

no

sig

nific

an

t

diffe

ren

ce

in m

ort

alit

y b

etw

ee

n t

he

two

tra

nsp

ort

atio

n m

eth

od

s

invo

lvin

g d

ire

ct

tra

nsp

ort

atio

n t

o

leve

l 1 t

rau

ma

ce

ntr

e d

esp

ite

pro

lon

ge

d t

ran

spo

rta

tio

n t

ime

in

the

he

lico

pte

r g

rou

p.

Lim

ita

tio

ns

�

Re

gis

try b

ase

d s

tud

y w

ith

so

me

retr

osp

ec

tive

da

ta c

olle

ctio

n a

nd

inc

om

ple

te T

RIS

S d

ata

ava

ilab

le

(la

ckin

g in

th

e A

MB

-REG

gro

up

).

�

Ac

cu

rac

y o

f d

ata

un

cle

ar.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g.

�

Ele

me

nts

of

ec

olo

gic

al a

na

lysi

s fo

r th

e

resu

lts

of

inte

rest

to

th

is r

ev

iew

.

Sp

ec

ific

ally

, th

ere

wa

s n

o c

om

pa

riso

n

be

twe

en

ind

ivid

ua

l tim

e d

ata

an

d

ou

tco

me

, ra

the

r th

e c

om

pa

riso

n w

as

be

twe

en

diffe

ren

t g

rou

ps

(an

d t

hu

s

use

d a

s a

pro

xy f

or

tim

e).

�

Diffe

ren

t st

aff

ing

mix

es

be

twe

en

tra

nsp

ort

atio

n m

eth

od

s (m

ore

pro

ce

du

res

ten

d t

o b

e p

erf

orm

ed

in

the

he

lico

pte

r p

atie

nts

) a

nd

diffe

ren

ce

s in

ma

na

ge

me

nt

ac

ross

the

diffe

ren

t h

osp

ita

ls m

ea

ns

it is

no

t

po

ssib

le t

o a

scrib

e a

ny d

iffe

ren

ce

in

ou

tco

me

to

diffe

ren

ce

in p

re-h

osp

ita

l

tim

e.

�

Pre

-ho

spita

l tim

e t

o le

ve

l 1 t

rau

ma

ca

re n

ot

giv

en

in IN

TER

gro

up

.


10

7

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Bie

we

ne

r e

t a

l. 2

004

)

Ge

rma

ny

co

ntin

ue

d

Co

mm

en

ts

�

Aim

ed

to

co

mp

are

th

e m

ort

alit

y o

f

fou

r ty

pic

al a

nd

co

mp

lete

pa

thw

ays

of

po

lytr

au

ma

pa

tie

nts

: a

ir o

r g

rou

nd

tra

nsp

ort

to

a le

ve

l 1 t

rau

ma

ce

ntr

e,

gro

un

d t

ran

spo

rt in

to le

ve

l II o

r III

co

mm

un

ity h

osp

ita

ls o

r in

terh

osp

ita

l

tra

nsf

er.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

Prim

ary

tra

nsf

er

into

a L

eve

l 1 t

rau

ma

ce

ntr

e

red

uc

es

mo

rta

lity m

ark

ed

ly. In

prin

cip

le, th

is

be

ne

fit

ca

n b

e a

ttrib

ute

d t

o s

up

erio

r

pre

clin

ica

l th

era

py, p

rim

ary

ad

mis

sio

n t

o a

Leve

l 1 t

rau

ma

ce

ntr

e o

r b

oth

. H

ow

eve

r, t

he

ide

ntic

al p

rob

ab

ility

of

surv

iva

l of

the

AM

B-

UN

I a

nd

HEM

S-U

NI g

rou

ps

in t

his

an

d

co

mp

ara

ble

stu

die

s d

oe

s n

ot

co

nfirm

ge

ne

rally

be

tte

r su

rviv

al r

ate

s o

n a

cc

ou

nt

of

a m

ore

ag

gre

ssiv

e o

n-s

ite

ap

pro

ac

h.


10

8

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ga

o e

t a

l. 200

6)

Ch

ina

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Po

lytr

au

ma

pa

tie

nts

with

th

ora

cic

an

d/o

r a

bd

om

ina

l in

jurie

s tr

ea

ted

at

Ch

on

gq

ing

Em

erg

en

cy M

ed

ica

l Ce

nte

r,

Ch

ina

fro

m O

ct

19

93 t

o S

ep

t 2

003.

Pa

rtic

ipa

nts

:

n=

1540

Ba

selin

e a

na

lyse

s.

Me

an

ag

e 2

8.8

ye

ars

Ma

les

79%

Du

ratio

n o

f p

rea

dm

issi

on

(%

)

< 1

ho

ur:

38.4

%

1-6

ho

urs

: 4

0.6

%

>6 h

ou

rs: 21.0

%

Blu

nt

tra

um

a:

61

.7%

Inc

lu/e

xcl c

rite

ria

.

Inju

ries

to m

ore

th

an

tw

o ISS b

od

y

reg

ion

s a

nd

at

lea

st o

ne

re

gio

n h

ad

AIS

≥3.

Da

ta c

olle

ctio

n

Ext

rac

ted

da

ta o

n s

ex,

ag

e, c

au

ses

of

inju

ry, d

ura

tio

n o

f p

rea

dm

issi

on

an

d in

jure

d r

eg

ion

s, s

ho

ck s

tate

on

ad

mis

sio

n, a

mo

un

t o

f b

loo

d

tra

nsf

usi

on

, se

ve

rity

of

inju

ries,

me

tho

d o

f d

iag

no

sis,

th

era

pe

utic

pro

ce

du

res.

Ou

tco

me

me

asu

res

Mo

rta

lity

An

aly

sis

Ch

i sq

ua

re t

est

.

Re

latio

nsh

ip b

etw

ee

n p

rea

dm

issi

on

tim

e a

nd

mo

rta

lity

Pre

ad

mis

sio

n <

1 h

ou

r: 3

.9%

mo

rta

lity

Pre

ad

mis

sio

n ≥

1 h

ou

r: 7

.7%

mo

rta

lity

P<

0.0

1

Lim

ita

tio

ns

�

Re

tro

spe

ctive

re

vie

w. P

ote

ntia

l

limita

tio

ns

of

this

ap

pro

ac

h in

clu

de

mis

sin

g d

ata

, a

nd

inc

on

sist

en

t

rec

ord

ing

of

da

ta.

�

Ac

cu

rac

y o

f d

ata

un

cle

ar.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g. N

o m

ultiv

aria

te a

na

lysi

s

to c

on

tro

l fo

r c

on

fou

nd

ing

.

Co

mm

en

ts

�

Aim

ed

to

inve

stig

ate

th

e e

arly

dia

gn

osi

s a

nd

tre

atm

en

t o

f

po

lytr

au

ma

pa

tie

nts

with

th

ora

cic

an

d/o

r a

bd

om

ina

l in

jurie

s.

�

All

pa

tie

nts

ta

ke

n t

o t

he

sa

me

ho

spita

l.

Re

po

rte

d c

on

clu

sio

ns

(by a

uth

ors

).

The

first

go

lde

n h

ou

r a

fte

r tr

au

ma

sh

ou

ld b

e

gra

spe

d, si

nc

e t

he

tre

atm

en

t in

th

is h

ou

r

ca

n d

ete

rmin

e g

rea

tly w

he

the

r th

e c

ritic

ally

inju

red

vic

tim

co

uld

su

rviv

e.

Pro

mp

t

dia

gn

osi

s a

nd

pro

pe

r tr

ea

tme

nt

co

ntr

ibu

te

mo

re g

rea

tly t

o t

he

su

rviv

al o

f th

e v

ictim

tha

n t

he

se

ve

rity

of

inju

ry.


10

9

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ha

rtl e

t a

l. 2

006

)

USA

Re

tro

spe

ctive

co

ho

rt s

tud

y

Leve

l III-

2

Stu

dy s

ett

ing

.

Pa

tie

nts

tre

ate

d a

t 22 t

rau

ma

ce

ntr

es

in

a N

ew

Yo

rk S

tate

qu

alit

y im

pro

ve

me

nt

pro

gra

m b

etw

ee

n 2

000 a

nd

20

04.

Pa

rtic

ipa

nts

:

1123 p

atie

nts

Ba

selin

e a

na

lyse

s

Me

an

ag

e 3

6 y

ea

rs

Ma

le 7

5%

GC

S: 2-5

54%

6-8

33%

≥9 1

3%

Pu

pill

ary

ab

no

rma

litie

s 21

%

Tra

um

a lo

ca

tio

n u

rba

n 2

8%

Air t

ran

spo

rt m

od

e 3

6%

Dire

ct

tra

nsp

ort

tim

e m

ea

n 1

.1 h

ou

rs

Ind

irec

t tr

an

spo

rt t

ime

me

an

4.5

ho

urs

Am

bu

lan

ce

intu

ba

tio

n 4

2%

Inc

lu/e

xcl c

rite

ria

.

Da

ta e

xtr

ac

ted

fro

m T

BI-

tra

c.

TBI-

tra

c in

clu

de

s:

Arr

iva

l at

leve

l 1 o

r le

ve

l 2 t

rau

ma

ce

ntr

e w

ith

in 2

4 h

ou

rs o

f in

jury

GC

S<

9 f

or

at

lea

st 6

ho

urs

aft

er

inju

ry a

nd

aft

er

resu

scita

tio

n e

ffo

rts

inc

lud

ing

airw

ay m

an

ag

em

en

t,

ve

ntila

tory

su

pp

ort

, a

nd

circ

ula

tory

sup

po

rt.

Me

ch

an

ism

of

inju

ry m

ust

be

co

nsi

ste

nt

with

tra

um

a. Exc

lud

es

sub

ara

ch

no

id h

ae

mo

rrh

ag

e

sec

on

da

ry t

o a

ne

ury

sm o

r st

roke

,

pa

tie

nts

wh

o e

xpire

d in

th

e E

D,

or

ad

mitte

d w

ith

a d

iag

no

sis

of

bra

in

de

ath

, o

r w

ere

tra

nsf

err

ed

to

th

e

stu

dy h

osp

ita

l > 2

4 h

ou

rs a

fte

r in

jury

.

No

np

ara

lyse

d p

atie

nts

with

a G

CS

of

3 o

r 4

an

d f

ixe

d a

nd

dila

ted

pu

pils

we

re e

xclu

de

d f

rom

an

aly

sis.

Da

ta c

olle

ctio

n

Da

ta e

xtr

ac

ted

fro

m T

BI-

tra

c:

Tim

e o

f in

jury

, tim

e o

f a

rriv

al a

t

tra

um

a c

en

tre

, m

od

e o

f tr

an

spo

rt,

typ

e o

f EM

S p

rovid

er,

dire

ct

or

ind

irec

t tr

an

spo

rt, b

loo

d p

ress

ure

an

d p

uls

e o

xim

etr

y v

alu

es,

GC

S,

pu

pill

ary

ass

ess

me

nt

an

d a

irw

ay

ma

na

ge

me

nt

pro

ce

du

res.

Ou

tco

me

me

asu

res

Two

we

ek m

ort

alit

y

Ass

oc

iatio

n b

etw

ee

n t

ime

to

tra

um

a c

en

tre

a

nd

tw

o w

ee

k

mo

rta

lity, a

dju

ste

d o

dd

s ra

tio

(9

5%

CI)

OR

1.0

0 (

1.0

0,

1.0

0)

for

ea

ch

on

e

min

ute

in

cre

ase

in t

ran

spo

rt t

ime

.

Ad

just

ed

fo

r h

yp

ote

nsi

on

sta

tus

on

da

y 1

, <

or

> 6

0 y

ea

rs o

f a

ge

,

pu

pill

ary

sta

tus

on

da

y 1

an

d in

itia

l

GC

S.

Lim

ita

tio

ns

�

Stu

dy c

on

du

cte

d in

54%

of

the

to

tal

tra

um

a c

en

tre

s in

th

e s

tate

.

�

Un

cle

ar

if c

on

sec

utive

pa

tie

nts

we

re

sele

cte

d in

th

e p

art

icip

atin

g t

rau

ma

ce

ntr

es.

�

Re

gis

try b

ase

d d

ata

with

inh

ere

nt

limita

tio

ns

ab

ou

t re

lian

ce

on

da

ta

co

llec

ted

.

�

Ac

cu

rac

y o

f d

ata

co

llec

ted

no

t

sta

ted

.

�

Ob

serv

atio

na

l stu

dy is

su

sce

ptib

le t

o

co

nfo

un

din

g.

A lim

ite

d r

an

ge

of

po

ten

tia

l co

nfo

un

de

rs w

ere

ad

just

ed

for

in t

he

mu

ltiv

aria

te m

od

el.

�

Da

ta f

or

144

9 p

atie

nts

en

tere

d in

th

e

da

tab

ase

bu

t fu

rth

er

exc

lusi

on

s

red

uc

ed

sa

mp

le s

ize

to

11

23.

Exc

lusi

on

s w

ere

: G

CS>

9 (

n=

71),

GC

S

mo

tor

sco

re >

6 (

n=

14),

fix

ed

an

d

dila

ted

pu

pils

an

d n

ot

pa

raly

sed

(n=

126

), G

CS>

3 a

nd

bila

tera

lly f

ixe

d

an

d d

ilate

d o

r m

issi

ng

pu

pil

info

rma

tio

n (

n=

79

), t

ime

to

ho

spita

l >

24 h

ou

rs (

n=

13),

tra

nsp

ort

tim

e <

10

min

ute

s (n

=17

), m

issi

ng

ou

tco

me

ass

ess

me

nt

(n=

6).

Co

mm

en

ts

�

Aim

ed

to

exp

lore

th

e e

ffe

ct

of

pre

-

ho

spita

l ma

na

ge

me

nt

de

cis

ion

s o

n

ea

rly m

ort

alit

y a

fte

r se

ve

re T

BI.

�

Tim

e o

f in

jury

wa

s b

ase

d o

n t

ime

of

dis

pa

tch

of

EM

S.

�

20 o

f 22 t

rau

ma

ce

ntr

es

we

re le

ve

l 1

tra

um

a c

en

tre

s.


11

0

Ta

ble

16

E

vid

ence

ta

ble

s o

f st

ud

ies

exa

min

ing

tim

e fr

om

am

bu

lan

ce c

all

ou

t to

em

erg

ency

dep

art

men

t a

rriv

al

(co

nti

nu

ed)

Au

tho

rs

Co

un

try

Stu

dy D

esig

n

Sam

ple

an

d In

terv

en

tio

ns

Meth

od

s

Resu

lts

Lim

itati

on

s a

nd

Co

nclu

sio

ns

(Ha

rtl e

t a

l. 2

006

)

USA

Co

ntin

ue

d

An

aly

sis

Ch

i sq

ua

re t

est

use

d t

o e

va

lua

te

pre

-ho

spita

l ch

ara

cte

ristic

s ve

rsu

s

dire

ct/

ind

ire

ct

tra

nsp

ort

atio

n.

Stu

de

nt’

s t

test

use

d f

or

co

mp

arin

g

me

an

s o

f c

on

tin

uo

us

va

riab

les.

Ma

nn

-Wh

itn

ey t

est

wa

s u

sed

wh

en

sam

ple

siz

es

we

re t

oo

sm

all

for

the

t

test

. Lo

gis

tic

re

gre

ssio

n a

na

lyse

s

pre

dic

tin

g t

wo

we

ek m

ort

alit

y w

ere

use

d t

o e

stim

ate

od

ds

ratio

s.

R

ep

ort

ed

co

nc

lusi

on

s (b

y a

uth

ors

).

The

pre

sen

t st

ud

y p

rov

ide

s c

lass

II e

vid

en

ce

tha

t d

em

on

stra

tes

a 5

0%

inc

rea

se in

mo

rta

lity a

sso

cia

ted

with

ind

irec

t tr

an

sfe

r o

f

TBI p

atie

nts


111


There were 21 articles selected for the time component of the review. The selection criteria required that the time from receipt of alarm to the time of arrival at hospital should be recorded. This criterion effectively resulted in a focus on trauma. Medical emergencies were frequently ruled out because the time interval started with the time symptoms started (rather than the time the alarm was received by emergency services) or, in the case of out of hospital cardiac arrest, the time interval recorded was usually from the time of alarm until either arrival at the scene or time of first defibrillation. There was a large body of literature on cardiac arrest that was excluded for that reason. As a consequence there was only one study included that examined out of hospital cardiac arrest.

In general, most studies were retrospective and thus relied on the accuracy of recording. The original purpose of recording was for reasons other than the studies of this nature. There is also likely to be variation in recording practices by different staff within the settings studied. For example, in some studies there may have been rounding of the time component by some staff and not by others. There is certainly the potential for misclassification of time components and also potentially in other variables that may have resulted in residual confounding.

There was wide variation in sample sizes and eligibility criteria as shown through Table 16. The results have been summarised into four sections:

1. Studies that found an association between prolonged pre-hospital time and poor prognosis.

2. Studies where there was no direct comparison between pre-hospital time and outcome. These studies assessed various groupings that happened to have different mean pre-hospital times.

3. Studies that did not identify a statistically significant association between pre-hospital time and outcome.

4. Studies that found an association between shorter pre-hospital time and increased mortality.

Three of the seven studies that found a statistically significant association between prolonged pre-hospital time and poor outcome included multivariate analyses. These three studies were all conducted by Sampalis et al and two had overlapping populations. The largest study (Sampalis et al. 1999) had over 12,000 participants and the study population was distinct from two earlier studies (Sampalis et al. 1992; Sampalis et al. 1993) by the same group. In this study there was a linear association between pre-hospital duration and odds of death such that the odds of death increased by 1.046 for an additional minute of pre-hospital time. On that basis, the odds of death would increase by 1.252 for a five minute increase in pre-hospital time. This study was well conducted. The focus was on patients with severe trauma. This large study was conducted during and after regionalisation of emergency services. The two other studies were conducted before regionalisation. One used a case control design and the other a cohort design. Although there was overlap in the study populations, the estimated odds ratios were quite different (varying between 3 and 30) although both were statistically significant.

The other four studies in this section were limited by their lack of control over potential confounders. Two of these studies focussed on set pre-hospital times. One dichotomised pre-hospital time at 1 hour (Gao et al. 2006) and one at 30 minutes (Frezza and Mezghebe 1999). In both cases mortality was significantly higher in the group with longer pre-hospital times. The other two studies focussed on survivors or unexpected survivors versus deaths and unexpected deaths. The mean times for each respective group across the two studies were quite different although there was also a difference in the mean pre-hospital time between survivors and deaths within each study. For example the mean pre-hospital time in the unexpected survivors in one study was 20.8 minutes (Feero et al. 1995) whereas the mean pre-hospital time in the survivors from the other study was 42.8 minutes (Grzybowski et al. 2000). Based on the information presented from these studies it is not clear that there is a threshold in the pre-hospital time that should be aimed at. The findings from the largest study would support the hypothesis that the shorter the pre-hospital time the better. These results are summarised in Table 17.


112

Table 17 Key results for studies examining time from ambulance callout to emergency

department arrival that found an association between prolonged pre-hospital time

and poor outcome

Reference Sample size Multivariate analysis

Results

(Sampalis et al. 1999) 12,208 Yes Odds of death for each additional minute

of pre-hospital time: OR 1.046 (1.044-1.050)

(Sampalis et al. 1993) 360 Yes OR (mortality by 6 days, pre-hospital time >

60 minutes compared with up to 60

minutes): 3.01 (1.27-5.06)

(Sampalis et al. 1992) 355 Yes OR (mortality, pre-hospital time > 60 minutes

compared with up to 60 minutes): 29.9 (2.7-

33.3)

(Gao et al. 2006) 1,540 No Preadmission < 1 hour: 3.9% mortality

Preadmission ≥ 1 hour: 7.7% mortality

P<0.01

(Feero et al. 1995) 848 No Mean pre-hospital time

Unexpected survivors: 20.8 minutes

Unexpected deaths: 29.3 minutes

P=0.02

(Grzybowski et al.

2000)

244 No Mean total EMS time

Survivors: 42.8 minutes

Deaths: 50.6 minutes

P≤0.01

(Frezza and

Mezghebe 1999)

58 No Pre-hospital time < 30 minutes: 63% survival

Pre-hospital time > 30 minutes: 0% survival

P=0.002

The second group of studies that compared groups that had different mean pre-hospital times have the limitation that it could not be established if there was a relationship between pre-hospital time and outcome. For example, the studies comparing ground ambulance with helicopters, although having different pre-hospital times may also have other factors that explain any difference in outcome. Such differences could include a difference in crew mix that also contributes to a difference in outcome. In general, crews with a doctor on board may tend to spend a longer time at the scene than other crews. It is therefore possible that the longer pre-hospital time may be balanced by the presence of a doctor. These studies have elements of an ecological analysis weakening the study design. There were six studies in this group. A statistically significant difference in mortality between groups was not observed in these studies. Two studies found a significantly longer length of hospital stay in groups transported by the mode of transport with the longer pre-hospital time (Berns et al. 2001; Phillips et al. 1999). However, neither of these studies conducted multivariate analyses so confounding is likely to be a problem. These results are summarised in Table 18.


113


department arrival in studies that did not directly compare pre-hospital time with

outcome


Results

(Biewener et al.

2004)

403 Yes Mortality comparing ambulance use with

HEMS use): OR 1.06 (0.427-2.635).

Pre-hospital times: helicopter 90 minutes,

ambulance 68 minutes

(Phillips et al. 1999) 792 No Ground transport: 54 minutes, length of stay

4.21 days

Air transport, 77 minutes, length of stay 8.97

days

P<0.001

Nil significant in mortality findings

(Schiller et al. 1988) 606 No Ambulance: mission time 39 minutes,

mortality 13%

Helicopter: mission time 50 minutes,

mortality 18%

(Young et al. 1998) 316 No Transfer group: 480 minutes pre-hospital

time

Direct group: 92 minutes pre-hospital time

No significant difference in mortality within

24 hours of injury or beyond 24 hours of

injury

(Berns et al. 2001) 294 No Ground, 142 minutes, length of stay 8 days

Helicopter, 104 minutes, length of stay 6.4

days

P=0.04

Nil significant in mortality

(Schwartz et al. 1990) 126 No Air ambulance, 65 minutes, Z=2.23

(improved compared with MTOS)

Ground ambulance, 34 minutes, Z=-2.69

(worse than MTOS)

There were also seven studies that found no association between pre-hospital time and outcome. It should be noted that two of these studies found an association between improved survival and shorter pre-hospital times on univariate analysis but both disappeared on multivariate analysis (Lerner et al. 2003). It should be observed that Lerner et al. (2003) estimated the odds of mortality were higher in patients with shorter pre-hospital time, but this finding was of borderline significance (OR 0.987, 95% CI 0.97-1.00). There were two other multivariate analyses conducted. One examined risk ratios for death by minutes to ED and found increased risk of death in the pre-hospital period 31-60 minutes but this association was not sustained in the 61-90 minutes and 91-185 minutes groups (Clarke et al. 2002). There were also three other studies that did not find a significant association on univariate analysis.


114


department arrival in studies that did not find an association between pre-hospital

time and outcome


Results

(Bonatti et al. 1995) 2139 Yes No significant difference between total

mission time and survival on multivariate

analysis (there was a significant decrease in

survival with longer mission times on

univariate analysis)

(Lerner et al. 2003) 1877 Yes Total out of hospital time and mortality

OR 0.987 (0.97-1.00).

Note univariate analysis significantly longer

pre-hospital time in the survivors (difference

3.69 minutes, 95% CI 0.52-6.85 minutes)

(Hartl et al. 2006) 1123 Yes Mortality for each extra minute in transport

time

OR 1.00 (1.00-1.00)

(Clarke et al. 2002) 250 Yes Risk ratios for death by minutes to ED (95%

CI)

1-30 minutes RR 0.773 (0.501-1.194)

31-60 minutes RR 1.268 (0.980-1.641)

61-90 minutes RR 0.832 (0.496-1.396)

91-185 minutes RR 0.740 (0.189-2.888)

(Pepe et al. 1987) 498 No Results stratified across four different trauma

score categories. No association between

pre-hospital time and mortality within each

trauma score stratum

(Sloan et al. 1989) 203 No Total run time (minutes):

Mortality group 32

Survival group 35

(Lim and Seow 2002) 93 No Pre-hospital time

Survivors: 38.3 minutes

Non-survivors: 35.4 minutes

One study estimated that a pre-hospital time up to 60 minutes was associated with an increased odds of mortality when compared with a pre-hospital time more than 60 minutes (OR 8, 95% CI 1.7-38.5), (Osterwalder 2002). This finding may have been due to a tendency to shorten the on scene time in patients who appeared critical, recognising the need for urgent definitive care. This study had a sample size of 254 and included a multivariate analysis.

There was some information provided that was helpful in considering whether pre-hospital time had an effect on outcome after controlling for crew mix. The two smaller studies by Sampalis et al. (1992; 1993) both controlled for crew mix and in-hospital parameters. In these studies, physicians were available to attend the scene although their attendance was reserved for severe cases. In both studies (which had overlapping populations), there was a significant association between shorter pre-hospital time and improved survival after controlling for crew mix but there was no association between crew mix and survival after controlling for pre-hospital time. However, Bonatti et al. (1995) did not identify any association between pre-hospital time and outcome after controlling for crew mix and they also did not find any association between the attendance of a physician and outcome after controlling for pre-hospital time.

In conclusion, there was some inconsistent support for shorter pre-hospital times being associated with improved survival. However, it is not clear if there is a threshold time to aim for or if any reduction in pre-hospital time is associated with improved outcome. There is therefore no clear pre-hospital time to aim for based on the literature reviewed. It seems biologically plausible that severity of injury may operate as an effect modifier in the relationship between pre-hospital time and outcome. In other words, the relationship between pre-hospital time and outcome may be influenced by measures of severity. Unfortunately, there was insufficient information to investigate this further within the studies eligible for this review.


115

OVERVIEW

Main findings

This report has four main areas of assessment:

1. In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on emergency helicopter services improve health outcome when compared with transportation by emergency helicopter without a medical doctor?

2. In adults and children with a medical or trauma related emergency, does the presence of a medical doctor on a road ambulance service improve health outcome when compared with transportation by a road ambulance service without a medical doctor?

3. In adults and children with a medical or trauma related emergency, does the presence of a medical crew able to perform rapid sequence intubation and/or thoracostomy improve health outcome when compared with a medical crew unable to perform rapid sequence intubation and/or tube thoracostomy and/or thoracotomy?

4. In adults and children with a medical or trauma related emergency how does variation in the time from callout to arrival at a medical facility with definitive care influence health outcome?

Summaries and conclusions have been included for each question earlier in this report. This overview provides links across the sections and identifies unanswered issues.

The first three questions were intended to examine the usefulness of including medical doctors on emergency transportation. The inclusion of medical doctors on emergency transportation is more consistent with the “stay and treat” strategy described earlier. That is, a period of stabilisation is implied before transportation to a definitive care hospital. In contrast, question four was particularly designed to assess whether there was a set time to aim for in relation to total pre-hospital transportation time. A short time would be consistent with the “scoop and run” strategy. Key results of these issues have been summarised earlier but, to reiterate the major points:

1. There was generally more support for the inclusion of doctors on helicopters in the seven studies appraised in this section. However, there were uncertainties due to study design issues (levels of evidence ranged between III-1 and III-3), lack of consideration about whether non-doctor groups can be trained to perform certain procedures that would improve patient outcome and whether there may be different clinical scenarios that would favour one crew mix type over another.

2. Similar considerations applied in the studies examining the use of doctors on board road ambulances. There were four studies in this section with levels of evidence ranging between III-2 and III-3.

3. When considering the outcome in patients who were treated by crews able to perform rapid sequence intubation and/or thoracostomy with other crews who were not able to perform these procedures, the only studies identified that met the study eligibility criteria included doctors amongst those able to perform the procedures of interest. It was therefore not possible to examine this issue in relation to non doctor groups. There were five studies in this section with levels of evidence ranging between III-1 and III-3.

4. There was inconsistent evidence on the association between pre-hospital time and patient outcome. There were 21 studies in this section with levels of evidence all being III-2. However, the general direction was to support improved outcome in association with shorter pre-hospital times. There was no clear time threshold to aim for. Two studies provided information to consider whether crew mix or rapid transport had a more significant bearing on outcome. The results were conflicting across these two studies.

5. Most of the studies included related to trauma rather than medical emergencies.

6. There was insufficient information to consider subgroups based on injury severity or age group.


116

Limitations

The studies included ranged from evidence level III-1 to III-3. Study limitations are described in earlier sections of this report. In addition, it is important to note that this Technical Brief is not a full systematic review, although a systematic approach to search for and retrieve relevant studies was used. This report constitutes a rapidly produced assessment and summary of the best available evidence. Wider searches of the Internet, hand searching of journals and contacting of authors for unpublished research were not undertaken.

Research gaps

Specific study designs that would be useful to further consider the review questions have been detailed throughout this report. Some general areas of future research that would be helpful include:

1. Is there some form of interaction between pre-hospital time and pre-hospital crew that has impact on patient outcome? Linked to this is whether the same pre-hospital approach (time and crew) results in improved outcome in all emergency patients or whether the best approach is dependent on the clinical situation.

2. Given differences in procedures performed and clinical assessment processes adopted by doctors compared with non-doctor pre-hospital personnel, to what extent would enhanced procedure training for non-doctor groups be helpful?

3. There are cost differences between the “scoop and run” and “stay and treat” approaches, along with the crew mixes used that ideally should be examined in relation to cost effectiveness of different approaches. However, given current uncertainties in effectiveness of the different strategies, incremental cost effectiveness can not be robustly examined at this time.

Conclusions

While the balance of studies support improved outcome associated with doctors on board emergency transportation, the robustness of these studies and the areas of uncertainty that remain (see under research gaps) provide uncertainty about the best approach. The best study supported the use of doctors on board helicopters. The balance of studies supported improved outcome associated with more rapid pre-hospital times. The studies identifying such improved outcome frequently assessed the linear relationship between pre-hospital outcome and time, meaning that the focus was on any improvement in outcome rather than a set threshold of pre-hospital time to meet in order to achieve improved outcome.


117

REFERENCES

Baker, S. P., O'Neill, B., Haddon, W. Jr., & Long, W. B. (1974). The injury severity score: a method for describing patients with multiple injuries and evaluating emergency care. Journal of

Trauma-Injury Infection & Critical Care, 14, 187-196.

Bartolacci, R. A., Munford, B. J., Lee, A., & McDougall, P. A. (1998). Air medical scene response to blunt trauma: effect on early survival. Medical Journal of Australia, 169, 612-616.

Baxt, W. G., & Moody, P. (1987). The impact of a physician as part of the aeromedical prehospital team in patients with blunt trauma. JAMA, 257, 3246-3250.

Bernard, S., Smith, K., Foster, S., Hogan, P., & Patrick, I. (2002). The use of rapid sequence intubation by ambulance paramedics for patients with severe head injury. Emergency Medicine

(Fremantle, W A ), 14, 406-411.

Bernard, S. A. (2006). Paramedic intubation of patients with severe head injury: a review of current Australian practice and recommendations for change. Emergency Medicine Australasia, 18, 221-228.

Berns, K. S., Hankins, D. G., & Zietlow, S. P. (2001). Comparison of air and ground transport of cardiac patients. Air Medical Journal, 20, 33-36.

Biewener, A., Aschenbrenner, U., Rammelt, S., Grass, R., & Zwipp, H. (2004). Impact of helicopter transport and hospital level on mortality of polytrauma patients. Journal of Trauma-Injury

Infection and Critical Care, 56, 94-98.

Bonatti, J., Goschl, O., Larcher, P., Wodlinger, R., & Flora, G. (1995). Predictors of short-term survival after helicopter rescue. Resuscitation, 30, 133-140.

Boyd, C. R., Tolson, M. A., & Copes, W. S. (1987). Evaluating trauma care: the TRISS method. Trauma Score and the Injury Severity Score. Journal of Trauma-Injury Infection & Critical

Care, 27, 370-378.

Burney, R. E., Hubert, D., Passini, L., & Maio, R. (1995). Variation in air medical outcomes by crew composition: a two-year follow-up. Annals of Emergency Medicine, 25, 187-192.

Burney, R. E., Passini, L., Hubert, D., & Maio, R. (1992). Comparison of aeromedical crew performance by patient severity and outcome. Annals of Emergency Medicine, 21, 375-378.

Cameron, P. A. (1999). Helicopter transport: can physicians save lives? Australian & New Zealand

Journal of Surgery, 69, 690-691.

Cameron, P. A., & Zalstein, S. (1998). Transport of the critically ill. Medical Journal of Australia, 169, 610-611.

Cameron, S., Pereira, P., Mulcahy, R., & Seymour, J. (2005). Helicopter primary retrieval: tasking who should do it? Emergency Medicine Australasia, 17, 387-391.

Champion, H. R., Copes, W. S., Sacco, W. J., Lawnick, M. M., Bain, L. W., Gann, D. S., Gennarelli, T., et al. (1990a). A new characterization of injury severity. Journal of Trauma-Injury

Infection & Critical Care, 30, 539-545; discussion 545-536.

Champion, H. R., Copes, W. S., Sacco, W. J., Lawnick, M. M., Keast, S. L., Bain, L. W., Jr., Flanagan, M. E., et al. (1990b). The Major Trauma Outcome Study: establishing national norms for trauma care. Journal of Trauma-Injury Infection & Critical Care, 30, 1356-1365.


118

Christenszen, E. F., Melchiorsen, H., Kilsmark, J., Foldspang, A., & Sogaard, J. (2003). Anesthesiologists in prehospital care make a difference to certain groups of patients. Acta

Anaesthesiologica Scandinavica, 47, 146-152.

Clarke, J. R., Trooskin, S. Z., Doshi, P. J., Greenwald, L., & Mode, C. J. (2002). Time to laparotomy for intra-abdominal bleeding from trauma does affect survival for delays up to 90 minutes. Journal of Trauma-Injury Infection & Critical Care, 52, 420-425.

Corfield, A. R., Thomas, L., Inglis, A., & Hearns, S. (2006). A rural emergency medical retrieval service: the first year. Emergency Medicine Journal, 23, 679-683.

Dalton, A. M., Botha, A., Coats, T., Spalding, T., Hodkinson, S., Warren, C., & Hodgson, R. (1992). Helicopter doctors? Injury, 23, 249-250.

Dretzke, J., Sandercock, J., Bayliss, S., & Burls, A. (2004). Clinical effectiveness and cost-effectiveness of prehospital intravenous fluids in trauma patients. Health Technology

Assessment (Winchester, England), 8, 1-103.

Feero, S., Hedges, J. R., Simmons, E., & Irwin, L. (1995). Does out-of-hospital EMS time affect trauma survival? American Journal of Emergency Medicine, 13, 133-135.

Flora, J. D. (1978). Method for comparing survival of burn patients to a standard survival curve. Journal of Trauma-Injury Infection and Critical Care, 18, 701-705.

Frandsen, F., Nielsen, J. R., Gram, L., Larsen, C. F., Jorgensen, H. R., Hole, P., & Haghfelt, T. (1991). Evaluation of intensified prehospital treatment in out-of-hospital cardiac arrest: survival and cerebral prognosis. The Odense ambulance study. Cardiology, 79, 256-264.

Frankema, S. P., Ringburg, A. N., Steyerberg, E. W., Edwards, M. J., Schipper, I. B., & van Vugt, A. B. (2004). Beneficial effect of helicopter emergency medical services on survival of severely injured patients. British Journal of Surgery, 91, 1520-1526.

Frankema, S. P., Steyerberg, E. W., Edwards, M. J., & van Vugt, A. B. (2005). Comparison of current injury scales for survival chance estimation: an evaluation comparing the predictive performance of the ISS, NISS, and AP scores in a Dutch local trauma registration. Journal of


Frezza, E. E., & Mezghebe, H. (1999). Is 30 minutes the golden period to perform emergency room thoratomy (ERT) in penetrating chest injuries? Journal of Cardiovascular Surgery, 40, 147-151.

Gabbe, B. J., Cameron, P. A., Wolfe, R., Simpson, P., Smith, K. L., & McNeil, J. J. (2005). Predictors of mortality, length of stay and discharge destination in blunt trauma. ANZ Journal of Surgery, 75, 650-656.

Gao, J. M. Gao, Y. H., Zeng, J. B., Wang, J. B., He, P., Wei, G. B., & Xiang, Z. (2006). Polytrauma with thoracic and/or abdominal injuries: experience in 1 540 cases. Chinese Journal of

Traumatology, 9, 108-114.

Garner, A., Rashford, S., Lee, A., & Bartolacci, R. (1999). Addition of physicians to paramedic helicopter services decreases blunt trauma mortality. Australian & New Zealand Journal of

Surgery, 69, 697-701.

Garner, A. A. (2004). The role of physician staffing of helicopter emergency medical services in prehospital trauma response. Emergency Medicine Australasia, 16, 318-323.

Gisvold, S. E. (2002). Helicopter emergency medical service with specially trained physicians - does it make a difference? Acta Anaesthesiologica Scandinavica, 46, 757-758.


119

Gries, A., Zink, W., Bernhard, M., Messelken, M., & Schlechtriemen, T. (2006). Realistic assessment of the physician-staffed emergency services in Germany. Anaesthesist, 55(10), 1080-1086

Grzybowski, M., Zalenski, R. J., Ross, M. A., & Bock, B. (2000). A prediction model for prehospital triage of patients with suspected cardiac ischemia. Journal of Electrocardiology, 33 Suppl, 253-258.

Hamman, B. L., Cue, J. I., Miller, F. B., O'Brien, D. A., House, T., Polk, H. C., Jr., & Richardson, J. D. (1991). Helicopter transport of trauma victims: does a physician make a difference? Journal of


Hartl, R., Gerber, L. M., Iacono, L., Ni, Q. H., Lyons, K., & Ghajar, J. (2006). Direct transport within an organized state trauma system reduces mortality in patients with severe traumatic brain injury. Journal of Trauma-Injury Infection and Critical Care, 60, 1250-1256.

Helm, M., Hossfeld, B., Schafer, S., Hoitz, J., & Lampl, L. (2006). Factors influencing emergency intubation in the pre-hospital setting - a multicentre study in the German Helicopter Emergency Medical Service. British Journal of Anaesthesia, 96, 67-71.

Kerr, W. A., Kerns, T. J., & Bissell, R. A. (1999). Differences in mortality rates among trauma patients transported by helicopter and ambulance in Maryland. Prehospital & Disaster Medicine, 14, 159-164.

Koefoed-Nielsen, J., Christensen, E. F., Melchiorsen, H., & Foldspang, A. (2002). Acute myocardial infarction: does pre-hospital treatment increase survival? European Journal of Emergency

Medicine, 9, 210-216.

Lee, A., Garner, A., Fearnside, M., & Harrison, K. (2003). Level of prehospital care and risk of mortality in patients with and without severe blunt head injury. Injury, 34, 815-819.

Lerner, E. B., Billittier, A. J., Dorn, J. M., & Wu, Y. W. (2003). Is total out-of-hospital time a significant predictor of trauma patient mortality? Academic Emergency Medicine, 10, 949-954.

Lerner, E. B., & Moscati, R. M. (2001). The golden hour: scientific fact or medical "urban legend"? Academic Emergency Medicine, 8, 758-760.

Lim, G. H., & Seow, E. (2002). Resuscitation for patients with out-of-hospital cardiac arrest: Singapore. Prehospital & Disaster Medicine, 17, 96-101.

Matsumoto, H., Mashiko, K., Hara, Y., Sakamoto, Y., Kutsukata, N., Takei, K., Tomita, Y., et al. (2006). Effectiveness of a "doctor-helicopter" system in Japan. Israel Medical Association

Journal: IMAJ, 8, 8-11.

Ministerial Taskforce on Trauma and Emergency Services (1999). Review of trauma and emergency services Victoria 1999. Final report of the Ministerial Taskforce on Trauma and Emergency Services and the Department Working Party on Emergency and Trauma Services. Melbourne: Human Services Victoria.

Moreno, R., & Morais, P. (1997). Outcome prediction in intensive care: results of a prospective, multicentre, Portuguese study. Intensive Care Medicine, 23, 177-186.

Muckart, D. J., Bhagwanjee, S., & Gouws, E. (1997). Validation of an outcome prediction model for critically ill trauma patients without head injury. Journal of Trauma-Injury Infection &

Critical Care, 43, 934-938; discussion 938-939.

Murphy-Macabobby, M., Marshall, W. J., Schneider, C., & Dries, D. (1992). Neuromuscular blockade in aeromedical airway management. Annals of Emergency Medicine, 21, 664-668.


120

National Health and Medical Research Council (2000). How to use the evidence: assessment and

application of scientific evidence. Canberra: NHMRC.

Nicholl, J. P., Brazier, J. E., & Snooks, H. A. (1995). Effects of London helicopter emergency medical service on survival after trauma. BMJ, 311, 217-222.

Nielsen, E. W., Ulvik, A., Carlsen, A. W., & Rannestad, B. (2002). When is an anesthesiologist needed in a helicopter emergency medical service in northern Norway? Acta Anaesthesiologica

Scandinavica, 46, 785-788.

Osterwalder, J. J. (2002). Can the "golden hour of shock" safely be extended in blunt polytrauma patients? Prospective cohort study at a level I hospital in eastern Switzerland. Prehospital &

Disaster Medicine, 17, 75-80.

Osterwalder, J. J. (2003). Mortality of blunt polytrauma: A comparison between emergency physicians and emergency medical technicians - prospective cohort study at a level I hospital in eastern Switzerland. Journal of Trauma-Injury Infection and Critical Care, 55, 355-361.

Pepe, P. E., Wyatt, C. H., Bickell, W. H., Bailey, M. L., & Mattox, K. L. (1987). The relationship between total prehospital time and outcome in hypotensive victims of penetrating injuries. Annals of Emergency Medicine, 16, 293-297.

Phillips, R. T., Conaway, C., Mullarkey, D., & Owen, J. L. (1999). One year's trauma mortality experience at Brooke Army Medical Center: is aeromedical transportation of trauma patients necessary? Military Medicine, 164, 361-365.

Rashford, S., & Myers, C. (2004). Optimal staffing of helicopter emergency medical services is controversial. Emergency Medicine Australasia, 16, 269-270.

Rhee, K. J., Strozeski, M., Burney, R. E., Mackenzie, J. R., & LaGreca-Reibling, K. (1986). Is the flight physician needed for helicopter emergency medical services? Annals of Emergency

Medicine, 15, 174-177.

Sampalis, J. S., Denis, R., Lavoie, A., Frechette, P., Boukas, S., Nikolis, A., Benoit, D., et al. (1999). Trauma care regionalization: a process-outcome evaluation. Journal of Trauma-Injury


Sampalis, J. S., Lavoie, A., Williams, J. I., Mulder, D. S., & Kalina, M. (1992). Standardized mortality ratio analysis on a sample of severely injured patients from a large Canadian city without regionalized trauma care. Journal of Trauma-Injury Infection & Critical Care, 33, 205-211; discussion 211-202.

Sampalis, J. S., Lavoie, A., Williams, J. I., Mulder, D. S., & Kalina, M. (1993). Impact of on-site care, prehospital time, and level of in-hospital care on survival in severely injured patients. Journal

of Trauma-Injury Infection & Critical Care, 34, 252-261.

Sampalis, J. S., Tamim, H., Denis, R., Boukas, S., Ruest, S. A., Nikolis, A., Lavoie, A., et al. (1997). Ineffectiveness of on-site intravenous lines: is prehospital time the culprit? Journal of

Trauma-Injury Infection & Critical Care, 43, 608-615; discussion 615-607.

Sanson, G., Di Bartolomeo, S., Nardi, G., Albanese, P., Diani, A., Raffin, L., Filippetto, C., et al. (1999). Road traffic accidents with vehicular entrapment: incidence of major injuries and need for advanced life support. European Journal of Emergency Medicine, 6, 285-291.

Schiller, W. R., Knox, R., Zinnecker, H., Jeevanandam, M., Sayre, M., Burke, J., & Young, D. H. (1988). Effect of helicopter transport of trauma victims on survival in an urban trauma center. Journal of Trauma-Injury Infection & Critical Care, 28, 1127-1134.


121

Schmidt, U., Frame, S. B., Nerlich, M. L., Rowe, D. W., Enderson, B. L., Maull, K. I., & Tscherne, H. (1992). On-scene helicopter transport of patients with multiple injuries--comparison of a German and an American system. Journal of Trauma-Injury Infection & Critical Care, 33, 548-553; discussion 553-545.

Schwartz, R. J., Jacobs, L. M., & Juda, R. J. (1990). A comparison of ground paramedics and aeromedical treatment of severe blunt trauma patients. Connecticut Medicine, 54, 660-662.

Sloan, E. P., Callahan, E. P., Duda, J., Sheaff, C. M., Robin, A. P., & Barrett, J. A. (1989). The effect of urban trauma system hospital bypass on prehospital transport times and Level 1 trauma patient survival. Annals of Emergency Medicine, 18, 1146-1150.

Sloane, C., Vilke, G. M., Chan, T. C., Hayden, S. R., Hoyt, D. B., & Rosen, P. (2000). Rapid sequence intubation in the field versus hospital in trauma patients. Journal of Emergency Medicine, 19, 259-264.

Snow, N., Hull, C., & Severns, J. (1986). Physician presence on a helicopter emergency medical service: necessary or desirable? Aviation Space & Environmental Medicine, 57, 1176-1178.

Spanjersberg, W., Ringburg, A., Bergs, B., Krijen, P., Schipper, I., Ringburg, A. N., Steyerberg, E. W., et al. (2005). Prehospital chest tube thoracostomy: effective treatment or additional trauma? Journal of Trauma-Injury Infection and Critical Care, 59, 96-101.

Suistomaa, M., Niskanen, M., Kari, A., Hynynen, M., & Takala, J. (2002). Customized prediction models based on APACHE II and SAPS II scores in patients with prolonged length of stay in the ICU. Intensive Care Medicine, 28, 479-485.

Thomas, S. H., Harrison, T. H., Buras, W. R., Ahmed, W., Cheema, F., & Wedel, S. K. (2002). Helicopter transport and blunt trauma mortality: a multicenter trial. Journal of Trauma-Injury


Trunkey, D. D. (1983). Trauma. Accidental and intentional injuries account for more years of life lost in the U.S. than cancer and heart disease. Among the prescribed remedies are improved preventive efforts, speedier surgery and further research. Scientific American, 249, 2, 28-35.

Vassar, M. J., Lewis, F. R., Jr., Chambers, J. A., Mullins, R. J., O'Brien, P. E., Weigelt, J. A., Hoang, M. T., et al. (1999). Prediction of outcome in intensive care unit trauma patients: a multicenter study of Acute Physiology and Chronic Health Evaluation (APACHE), Trauma and Injury Severity Score (TRISS), and a 24-hour intensive care unit (ICU) point system. Journal of


Wang, H. E., Kupas, D. F., Hostler, D., Cooney, R., Yealy, D. M., & Lave, J. R. (2005). Procedural experience with out-of-hospital endotracheal intubation. Critical Care Medicine, 33, 1718-1721.

Wirtz, M. H., Cayten, C. G., Kohrs, D. A., Atwater, R., & Larsen, E. A. (2002). Paramedic versus nurse crews in the helicopter transport of trauma patients. Air Medical Journal, 21, 17-21.

Young, J. S., Bassam, D., Cephas, G. A., Brady, W. J., Butler, K., & Pomphrey, M. (1998). Interhospital versus direct scene transfer of major trauma patients in a rural trauma system. American Surgeon, 64, 88-91.

Younge, P. A., Coats, T. J., Gurney, D., & Kirk, C. J. (1997). Interpretation of the Ws statistic: application to an integrated trauma system. Journal of Trauma-Injury Infection & Critical

Care, 43, 511-515.


122

APPENDIX 1: SEARCH STRATEGY

Medline 1

1 air ambulances/ (1024) 2 aircraft/ (5261) 3 helicopter$.mp. (1541) 4 (medevac or medivac or casivac or casevac).mp. (31) 5 aeromedic$.mp. (673) 6 (air ambulanc$ or flight ambulan$).mp. (1125) 7 or/1-6 (7198) 8 "personnel staffing and scheduling"/ (9920) 9 personnel selection/ (7837) 10 patient care team/ (36954) 11 physician's role/ (19199) 12 (doctor$ or staff$ or personnel$ or physician$).tw. (291307) 13 exp physicians/ (56725) 14 medical practitioner$.tw. (2394) 15 (medical$ adj qualif$).tw. (187) 16 ma.fs. (43025) 17 or/8-16 (398761) 18 7 and 17 (1113) 19 limit 18 to english (918) 20 limit 19 to yr=1980-2006 (853) 21 (letter or news or historical article).pt. (916855) 22 20 not 21 (802) 23 (commercial adj (airline$ or aircrew)).tw. (151) 24 (spaceflight or space flight).tw. (3422) 25 22 not (23 or 24) (788) 26 allied health personnel/ (8563) 27 (paramedic$ or medic or medics).tw. (4114) 28 nurse's role/ (15545) 29 nurses/ (21869) 30 flight nurse$.mp. (138) 31 or/26-30 (48520) 32 7 and 31 (267) 33 limit 32 to english (249) 34 limit 33 to yr=1980-2006 (241) 35 34 not (21 or 23 or 24) (226) 36 35 not 25 (118) 37 emergency medical technicians/ (3409) 38 7 and 37 (160) 39 limit 38 to english (155) 40 limit 39 to yr=1980-2006 (153) 41 40 not (21 or 23 or 24) (150) 42 41 not (25 or 36) (52)

Medline 2

1 ambulances/ (3697) 2 ambulance$.tw. (4057) 3 1 or 2 (5887) 4 aircraft/ (5261) 5 air ambulances/ (1024) 6 helicopter$.mp. (1541) 7 or/4-6 (6776) 8 3 not 7 (5131)


123

9 (doctor$ or staff$ or personnel).mp. (303172) 10 personnel selection/ (7837) 11 "personnel staffing and scheduling"/ (9920) 12 patient care team/ (36954) 13 physician's role/ (19199) 14 exp physicians/ (56725) 15 (medical practitioner$ or physician$).tw. (167891) 16 (medic$ adj qualif$).tw. (214) 17 emergency medical technicians/ (3409) 18 allied health personnel/ (8563) 19 (paramedic$ or medic or medics).tw. (4114) 20 ma.fs. (43025) 21 nurse's role/ (15545) 22 nurses/ (21869) 23 or/9-22 (547961) 24 8 and 23 (2153) 25 mortality/ (25412) 26 survival analysis/ (65471) 27 survival rate/ (83848) 28 length of stay/ (35751) 29 (mortality or survival).tw. (544897) 30 exp treatment outcome/ (300617) 31 "outcome assessment (health care)"/ (25927) 32 or/25-31 (873880) 33 24 and 32 (392) 34 limit 33 to english (319) 35 (letter or news).pt. (689225) 36 34 not 35 (317) 37 neonat$.ti. (61473) 38 36 not 37 (315)

Medline 3

1 *time factors/ (904) 2 (time adj3 delay$).tw. (6849) 3 (time or delay).ti. (88449) 4 ((prehospital or pre-hospital) adj (time or care or treatment)).tw. (1157) 5 ((call-out or callout) and (arrival or admission or admit$ or hospital or emergency department or

ED)).tw. (20) 6 (delay$ adj3 (arrival or admission or admit$ or hospital or medical facility or definitive care or

emergency department or ED)).tw. (1246) 7 (time adj3 (arrival or admission or admit$ or hospital or medical facility or definitive care or

emergency department or ED)).tw. (8465) 8 ((call-out or callout) and (arrival or admission or admit$ or hospital or medical facility or

definitive care or emergency department or ED)).tw. (20) 9 (scene time or "out of hospital time").tw. (105) 10 (transport adj time$).tw. (506) 11 (transfer$ adj time$).tw. (281) 12 (prehospital index or pre-hospital index).tw. (21) 13 or/1-12 (105221) 14 survival analysis/ (65345) 15 exp treatment outcome/ (299801) 16 length of stay/ (35699) 17 patient discharge/ (12043) 18 morbidity/ (17555) 19 mortality/ (25394) 20 "Outcome Assessment (Health Care)"/ (25869) 21 (survival or outcome).tw. (599960) 22 or/14-21 (900262) 23 *emergencies/ (7692) 24 exp emergency medical services/ (58193)


124

25 emergency service, hospital/ (24370) 26 transportation of patients/ (6348) 27 ambulances/ (3694) 28 air ambulances/ (1024) 29 helicopters/ (5256) 30 or/23-29 (68719) 31 13 and 22 and 30 (908) 32 (news or letter).pt. (688417) 33 infant, newborn/ or infant, premature/ or neonat$.ti. (406994) 34 case reports.pt. (1291342) 35 or/32-34 (2194216) 36 31 not 35 (855) 37 (golden hour or golden minute$).mp. (76) 38 36 or 37 (927) 39 limit 38 to yr=1980-2006 (918) 40 limit 39 to english (832)

Embase 1

1 air medical transport/ (70) 2 AIRCRAFT/ (1958) 3 HELICOPTER/ (786) 4 aeromedic$.tw. (384) 5 flight ambulance$.tw. (0) 6 air ambulance$.tw. (91) 7 (medevac or medivac or casivac or casevac).tw. (21) 8 or/1-7 (3043) 9 medical personnel/ (2810) 10 Airplane Crew/ (1080) 11 health care personnel/ (23462) 12 manpower/ (1093) 13 Health Care Manpower/ (1025) 14 (doctor$ or staff$ or personnel$ or physician$).tw. (164842) 15 medical practitioner$.tw. (1560) 16 (medical$ adj qualif$).tw. (135) 17 patient care team.tw. (37) 18 physician/ or emergency physician/ (33974) 19 or/9-18 (199007) 20 8 and 19 (687) 21 limit 20 to english (621) 22 letter.pt. (330316) 23 21 not 22 (600) 24 (commercial adj (airline$ or aircrew or attendant$)).tw. (113) 25 (spaceflight or space flight).tw. (1262) 26 24 or 25 (1375) 27 23 not 26 (580) 28 Paramedical Personnel/ (1706) 29 paramedical personnel/ (1706) 30 rescue personnel/ (1043) 31 (medic or medics or paramedic$).tw. (2030) 32 Nursing Role/ (25) 33 nurse/ (11528) 34 nursing staff/ (2233) 35 flight nurse$.tw. (34) 36 or/28-35 (17487) 37 8 and 36 (224) 38 limit 37 to english (196) 39 38 not (22 or 26) (181) 40 39 not 27 (68)


125

Embase 2

1 ambulance/ (1942) 2 ambulance$.tw. (2198) 3 1 or 2 (2915) 4 air medical transport/ (70) 5 aircraft/ (1958) 6 helicopter/ (786) 7 (aeromedic$ or flight ambulance$ or air ambulance$).tw. (460) 8 or/4-7 (3031) 9 3 not 8 (2606) 10 medical personnel/ (2810) 11 health care personnel/ (23462) 12 manpower/ (1093) 13 health care manpower/ (1025) 14 (doctor$ or staff$ or personnel$ or physician$).tw. (164842) 15 medical practitioner$.tw. (1560) 16 (medic$ adj qualif$).tw. (154) 17 patient care team.tw. (37) 18 physician/ or emergency physician/ (33974) 19 paramedical personnel/ (1706) 20 rescue personnel/ (1043) 21 (medic or medics or paramedic$).tw. (2030) 22 nursing role/ (25) 23 nurse/ (11528) 24 nursing staff/ (2233) 25 or/10-24 (207253) 26 9 and 25 (1149) 27 MORTALITY/ (124790) 28 Survival/ (46091) 29 Survival Rate/ (42525) 30 (mortality or survival).tw. (372312) 31 "Length of Stay"/ (17989) 32 exp Treatment Outcome/ (348062) 33 Outcome Assessment/ (10452) 34 or/27-33 (730660) 35 26 and 34 (298) 36 limit 35 to english (256) 37 letter.pt. (330316) 38 36 not 37 (250)

Embase 3

1 exp *time/ (2667) 2 time factor$.tw. (455) 3 (time or delay).ti. (51255) 4 (time adj3 delay).tw. (3194) 5 ((prehospital or pre-hospital) adj (time or care or treatment or delay)).tw. (859) 6 (delay$ adj3 (arrival or admission or admit$ or hospital or medical facility or definitive care or

emergency department or ED)).tw. (922) 7 (time adj3 (arrival or admission or admit$ or hospital or medical facility or definitive care or

emergency department or ED)).tw. (6337) 8 ((callout or call-out) and (arrival or admission or admit$ or hospital or medical facility or

definitive care or emergency department or ED)).tw. (15) 9 (scene time or "out of hospital time").tw. (68) 10 (transport adj time$).tw. (360) 11 (transfer$ adj time$).tw. (211) 12 (prehospital index or pre-hospital index).tw. (17) 13 or/1-12 (63376) 14 Survival/ (45956) 15 exp treatment outcome/ (345878)


126

16 length of stay/ (17884) 17 hospital discharge/ (18041) 18 morbidity/ (63032) 19 mortality/ (124143) 20 (survival or outcome).tw. (452728) 21 survival rate/ (42297) 22 Outcome Assessment/ (9819) 23 or/14-22 (830528) 24 Emergency Care/ (1145) 25 Emergency Health Service/ (9413) 26 emergency/ (3558) 27 Emergency Treatment/ (7782) 28 ambulance/ (1930) 29 HELICOPTER/ (777) 30 Patient Transport/ (5227) 31 air ambulan$.tw. (91) 32 or/24-31 (26070) 33 13 and 23 and 32 (597) 34 letter.pt. (329150) 35 Case Report/ (680614) 36 33 not (34 or 35) (575) 37 Newborn/ (127209) 38 Prematurity/ (20771) 39 neonat$.ti. (33939) 40 or/37-39 (146969) 41 36 not 40 (565) 42 (golden hour or golden minutes).mp. (58) 43 41 or 42 (620) 44 limit 43 to english (539)

Cinahl 1

1 Aeromedical Transport/ (1390) 2 aircraft/ (528) 3 helicopter$.mp. (352) 4 (medevac or casevac or medivac or casivac or evac).tw. (19) 5 (aeromedic$ or aero medic$).tw. (94) 6 (air ambulance$ or flight ambulance$).tw. (77) 7 or/1-6 (1918) 8 "Personnel Staffing and Scheduling"/ (7243) 9 Multidisciplinary Care Team/ (9627) 10 personnel selection/ (1476) 11 exp PHYSICIANS/ (20310) 12 Physician's Role/ (1736) 13 (medical practitioner$ or (medic$ adj qualif$)).tw. (427) 14 (doctor$ or physician$ or staff$ or personnel$).tw. (69968) 15 (medical$ adj2 (staff$ or personnel)).tw. (1303) 16 or/8-15 (96867) 17 7 and 16 (249) 18 limit 17 to english (249) 19 letter.pt. (37113) 20 (book or book chapter).pt. (18610) 21 pamphlet.pt. (2446) 22 18 not (19 or 20 or 21) (247) 23 (commercial adj2 (airline or aircraft)).tw. (22) 24 (spaceflight or space flight).tw. (43) 25 22 not (23 or 24) (243) 26 biography.pt. (2712) 27 25 not 26 (242)


127

Cinahl 2

1 AMBULANCES/ (1030) 2 ambulance$.tw. (1408) 3 1 or 2 (1916) 4 AIRCRAFT/ (528) 5 helicopter$.mp. (352) 6 Aeromedical Transport/ (1391) 7 air ambulance$.tw. (77) 8 or/4-7 (1884) 9 3 not 8 (1747) 10 (doctor$ or staff$ or personnel).mp. (95811) 11 "personnel staffing and scheduling"/ (7270) 12 Personnel Selection/ (1485) 13 Multidisciplinary Care Team/ (9658) 14 Physician's Role/ (1742) 15 exp physicians/ (20428) 16 (medical practitioner$ or physician$).tw. (26538) 17 (medic$ adj qualif$).tw. (17) 18 Emergency Medical Technicians/ (3793) 19 allied health personnel/ (824) 20 (paramedic$ or medic or medics).tw. (1367) 21 ma.fs. (4504) 22 Nursing Role/ (18909) 23 nurses/ (23300) 24 or/10-23 (176466) 25 9 and 24 (687) 26 mortality/ (5012) 27 survival/ (4141) 28 Survival Analysis/ (3532) 29 (survival or mortality).tw. (25643) 30 "Length of Stay"/ (6142) 31 exp Treatment Outcomes/ (32058) 32 Outcome Assessment/ (4204) 33 or/26-32 (67621) 34 25 and 33 (67) 35 limit 34 to english (66) 36 neonat$.ti. (5674) 37 35 not 36 (66) 38 letter.pt. (37444) 39 37 not 38 (66)

Cinahl 3

1 *Time Factors/ (1377) 2 (time adj3 delay$).tw. (283) 3 (time or delay).ti. (12147) 4 ((prehospital or pre-hospital) adj (time or care or treatment or index)).tw. (427) 5 ((call-out or callout) and (arrival or admission or admit$ or hospital or emergency department or

ED)).tw. (1) 6 (delay adj3 (arrival or admission or admit$ or hospital or definitive care or medical facility or

emergency department or ED)).tw. (78) 7 (time adj3 (arrival or admission or admit$ or hospital or definitive care or medical facility or

emergency department or ED)).tw. (1124) 8 ((call-out or callout) and (definitive care or medical facility)).tw. (0) 9 (scene time or "out of hospital time").tw. (48) 10 (transport$ adj time$).tw. (71) 11 (transfer$ adj time$).tw. (12) 12 or/1-11 (14842) 13 Survival Analysis/ (3508) 14 exp Treatment Outcomes/ (31790)


128

15 length of stay/ (6118) 16 Patient Discharge/ (3187) 17 Outcome Assessment/ (4172) 18 (survival or outcome).tw. (42380) 19 morbidity/ (1503) 20 mortality/ (4997) 21 or/13-20 (81805) 22 *EMERGENCIES/ (1205) 23 Emergency Service/ (8794) 24 exp Emergency Medical Services/ (23036) 25 "Transportation of Patients"/ (1527) 26 AMBULANCES/ (1025) 27 Aircraft/ (528) 28 Aeromedical Transport/ (1390) 29 (ambulan$ or helicopter$).tw. (1791) 30 or/22-29 (24871) 31 12 and 21 and 30 (162) 32 (biography or book or book chapter or interview or case study).pt. (93767) 33 31 not 32 (162) 34 letter.pt. (37113) 35 33 not 34 (160) 36 (golden hour or golden minutes).mp. (35) 37 35 or 36 (195) 38 limit 37 to english (195) 39 Infant, Newborn/ (29078) 40 Infant, Premature/ (4504) 41 neonat$.ti. (5660) 42 or/39-41 (31082) 43 38 not 42 (189)

Current Contents/ Citation Indexes 1

1. Helicopter* OR air ambulance* OR flight ambulance* 2. Aircraft 3. Medevac OR medivac OR casevac OR casivac OR evac 4. Aeromedic* 5. Commercial SAME (airline OR aircraft OR aircrew) 6. Spaceflight OR space flight 7. Personnel OR medical practitioner* 8. (medic* SAME qualif*) 9. Doctor * OR physician* 10. Staffing OR staffed OR staff 11. #1 OR #2 OR #3 OR #4 12. #7 OR #8 OR #9 OR #10 13. #11 AND #12 14. Allied health OR paramedic* OR medic OR medics 15. Flight nurse* 16. Nurse OR nurses OR nursing 17. #14 OR #15 OR #16 18. #12 AND #17 19. #18 NOT #13 20. #19 NOT (#5 OR #6)


1. Ambulance* 2. Aircraft OR helicopter* OR aeromedical or air ambulance* 3. #1 NOT #2 4. personnel OR staff OR staffing OR staffed 5. patient SAME care SAME team 6. doctor* OR physician* 7. medical practitioner* OR (medic* SAME qualif*) OR paramedic* OR medic OR medics


129

8. emergency technician* OR allied health OR manpower 9. nurse’s role OR physician’s role OR nursing role 10. nurse OR nurses 11. #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10 12. #3 AND #11 13. #12 AND (survival OR mortality) 14. #12 AND outcome 15. #12 AND (length SAME stay) 16. #13 OR #14 OR #15


1. (Prehospital OR pre-hospital) SAME (index OR care OR time OR treatment OR delay) 2. (transport* OR transfer*) SAME time 3. Scene time OR “out of hospital time” 4. (callout OR call out) SAME (arrival OR admission OR admit* OR hospital OR emergency

department OR emergency room OR definitive care OR medical facilit* OR ED OR ER) 5. (time OR delay) SAME (arrival OR admission OR admit* OR hospital OR emergency

department OR emergency room OR definitive care OR medical facilit* OR ED OR ER) 6. Emergency OR emergencies 7. Patient SAME (transport* OR transfer*) 8. Ambulan* OR helicopter* 9. #6 OR #7 OR #8 10. (newborn OR neonat*) 11. (infant* OR baby OR babies) SAME premature 12. Case study 13. Case report 14. Golden hour 15. #1 OR #2 OR #3 OR #4 OR #5 16. #9 AND #15 17. #16 NOT (#10 OR #11 OR #12 OR #13) 18. #17 AND (survival OR outcome) 19. #14 OR #18

PubMed (last 90 days)

PubMed searches were substantially the same as the strategies for Current Contents and the Citation Indexes.

Additional searching

Several small additional searches were carried out as required during the course of the project to obtain information on the merits of the respective scales for predicting mortality and morbidity after trauma.


130

APPENDIX 2: LEVELS OF EVIDENCE∗∗∗∗

Level I Evidence obtained from a systematic review (or meta-analysis) of relevant randomised controlled trials.

Level II Evidence obtained from at least one randomised controlled trial.

Level III. 1 Evidence obtained from pseudorandomised controlled trials (alternate allocation or some other method).

2 Evidence obtained from comparative studies (including systematic reviews of such studies) with concurrent controls and allocation not randomised, cohort studies, case control studies or interrupted time series with a control group).

3 Evidence obtained from comparative studies with historical control, two or more single-arm studies or interrupted time series without a parallel control group.

Level IV Evidence obtained from case series, either post-test or pre-test/post-test.

∗ From National Health and Medical Research Council (2000)


131

APPENDIX 3: EXCLUDED RETRIEVED PAPERS: DOCTOR VERSUS

NO DOCTOR ON HELICOPTERS


Baxt, W. G., & Moody, P. (1983). The impact of a rotorcraft aeromedical emergency care service on trauma mortality. JAMA, 249, 3047-3051.

Baxt, W. G., & Moody, P. (1987). The impact of advanced prehospital emergency care on the mortality of severely brain-injured patients. Journal of Trauma-Injury Infection & Critical Care, 27, 365-369.

Blumen, I. J., & Gordon, R. S. (1989). Taking to the skies. Emergency, 21, 32-38.

Brismar, B., Alveryd, A., Johnsson, O., & Ohrvall, U. (1986). The ambulance helicopter is a prerequisite for centralised emergency care. Acta Chirurgica Scandinavica - Supplementum, 530, 89-93.

Burillo-Putze, G., Duarte, I. H., & Alvarez Fernandez, J. A. (2001). Helicopter emergency medical service in Spain. Air Medical Journal, 20, 21-23.

Cameron, P. A., Flett, K., Kaan, E., Atkin, C., & Dziukas, L. (1993). Helicopter retrieval of primary trauma patients by a paramedic helicopter service. Australian and New Zealand Journal of

Surgery, 63, 790-797.

Cannell, H., Silvester, K. C., & O'Regan, M. B. (1993). Early management of multiply injured patients with maxillofacial injuries transferred to hospital by helicopter. British Journal of Oral &

Maxillofacial Surgery, 31, 207-212.

Celli, P., Fruin, A., & Cervoni, L. (1997). Severe head trauma. Review of the factors influencing the prognosis. Minerva Chirurgica, 52, 1467-1480.

Cline Jr, C. T., Smith, S., & Davenport, P. (2006). Carilion Life-Guard: A quarter century of air medical service. Air Medical Journal, 25, 118-121.

Cocanour, C. S., Fischer, R. P., & Ursic, C. M. (1997). Are scene flights for penetrating trauma justified? Journal of Trauma-Injury Infection & Critical Care, 43, 83-86; discussion 86-88.

Collier, J. (2006). Air Evac Services: improving Arizona's health over 36 years. Air Medical Journal, 25, 196-199.

Conroy, M. B., Rodriguez, S. U., Kimmel, S. E., & Kasner, S. E. (1999). Helicopter transfer offers a potential benefit to patients with acute stroke. Stroke, 30, 2580-2584.


Davis, D. P., Ochs, M., Hoyt, D. B., Bailey, D., Marshall, L. K., & Rosen, P. (2003). Paramedic-administered neuromuscular blockade improves prehospital intubation success in severely head-injured patients. Journal of Trauma-Injury Infection & Critical Care, 55, 713-719.

Davis, D. P., Pettit, K., Rom, C. D., Poste, J. C., Sise, M. J., Hoyt, D. B., & Vilke, G. M. (2005a). The safety and efficacy of prehospital needle and tube thoracostomy by aeromedical personnel. Prehospital Emergency Care, 9, 191-197.


132

Davis, D. P., Vadeboncoeur, T. F., Ochs, M., Poste, J. C., Vilke, G. M., & Hoyt, D. B. (2005b). The association between field Glasgow Coma Scale score and outcome in patients undergoing paramedic rapid sequence intubation. Journal of Emergency Medicine, 29, 391-397.

Garner, A., Crooks, J., Lee, A., & Bishop, R. (2001). Efficacy of prehospital critical care teams for severe blunt head injury in the Australian setting. Injury, 32, 455-460.


Gomes, E., Araujo, R., Soares-Oliveira, M., & Pereira, N. (2004). International EMS systems: Portugal. Resuscitation, 62, 257-260.

Hachinski, V. (2001). Sky doc. CMAJ Canadian Medical Association Journal, 165, 1621-1622.

Holcomb, J. B., Niles, S. E., Miller, C. C., Hinds, D., Duke, J. H., & Moore, F. A. (2005). Prehospital physiologic data and lifesaving interventions in trauma patients. Military Medicine, 170, 7-13.

Isakov, A. P. (2006). Souls on board: helicopter emergency medical services and safety. Annals of

Emergency Medicine, 47, 357-360.

Isenberg, D. L., & Bissell, R. (2005). Does advanced life support provide benefits to patients?: a literature review. Prehospital & Disaster Medicine, 20, 265-270.

Kuper, P. (2005). Austin-Travis County STAR flight. Air Medical Journal, 24, 192-194.

Lechleuthner, A., Bouillon, B., Neugebauer, E., Mennigen, R., & Tiling, T. (1994). Prehospital chest tubes incidence and analysis of iatrogenic injuries in the Emergency Medical-Service Cologne. Theoretical Surgery, 9, 220-226.

Little, M. (1994). Another kind of flying doctor. Medical Journal of Australia, 160, 214-216.

Lyons, T. J., & Connor, S. B. (1995). Increased flight surgeon role in military aeromedical evacuation. Aviation Space & Environmental Medicine, 66, 927-929.

Macintyre, I. (1994). Another kind of flying doctor. BMJ, 309, 1745-1746.

Mackenzie, C. F., Shin, B., & Matjasko, M. J. (1987). Physicians on aeromedical teams. JAMA-Journal

of the American Medical Association, 258, 2377-2378.

Macrae, D. J. (1994). Paediatric intensive care transport. Archives of Disease in Childhood, 71, 175-178.



Munford, B., & Manning, R. (1994). Paramedic helicopter retrieval of trauma patients. Australian and

New Zealand Journal of Surgery, 64, 640-641.

Nocera, A., & Dalton, A. M. (1994). Disaster alert! The role of physician-staffed helicopter emergency medical services. Medical Journal of Australia, 161, 689-692.

Oppe, S., & De Charro, F. T. (2001). The effect of medical care by a helicopter trauma team on the probability of survival and the quality of life of hospitalised victims. Accident Analysis &

Prevention, 33, 129-138.

Orlando, R., Schwartz, R., Lee, M., & Jacobs, L. (1987). The role of the flight physician in helicopter critical care transport. Critical Care Medicine, 15, 367-367.


133



Osterwalder, J. J. (2003). Mortality of blunt polytrauma: a comparison between emergency physicians and emergency medical technicians - prospective cohort study at a level I hospital in eastern Switzerland. Journal of Trauma-Injury Infection and Critical Care, 55, 355-361.

Rodenberg, H. (1992a). Effect of aeromedical aircraft on care of trauma patients - evaluation using the revised trauma score. Southern Medical Journal, 85, 1065-1071.

Rodenberg, H. (1992b). The revised trauma score: a means to evaluate aeromedical staffing patterns. Aviation Space & Environmental Medicine, 63, 308-313.


Shufflebarger, C., & Townsend, R. (1987). Physicians on aeromedical teams. JAMA-Journal of the

American Medical Association, 258, 2378-2378.

Slagt, C., Zondervan, A., Patka, P., & de Lange, J. J. (2004). A retrospective analysis of the intubations performed during 5 years of helicopter emergency medical service in Amsterdam. Air Medical

Journal, 23, 36-37.


Spangler, D. E., Jr., Rogers, W. J., Gore, J. M., Griffith, M., Maske, L. E., Morgan, T. E., & Corrao, J. (1991). Early tPA treatment and aeromedical transport of patients with acute myocardial infarction. Journal of Interventional Cardiology, 4, 81-89.

Stansbury, D. (1996). Flying high: EMS in the air. Journal of Emergency Medical Services, 21, 59-61.

Stauffer, U. G. (1995). Surgical and critical care management of children with life-threatening injuries: the Swiss experience. Journal of Pediatric Surgery, 30, 903-910.

Suominen, P., Baillie, C., Kivioja, A., Korpela, R., Rintala, R., Silfvast, T., & Olkkola, K. T. (1998). Prehospital care and survival of pediatric patients with blunt trauma. Journal of Pediatric

Surgery, 33, 1388-1392.

Thomas, S. H., Harrison, T. H., Buras, W. R., Ahmed, W., Cheema, F., & Wedel, S. K. (2002). Helicopter transport and blunt trauma mortality: A multicenter trial. Journal of Trauma-Injury


van Wijngaarden, M., Kortbeek, J., Lafreniere, R., Cunningham, R., Joughin, E., & Yim, R. (1996). Air ambulance trauma transport: a quality review. Journal of Trauma-Injury Infection & Critical

Care, 41, 26-31.

Vilke, G. M., Hoyt, D. B., Epperson, M., Fortlage, D., Hutton, K. C., & Rosen, P. (1994). Intubation techniques in the helicopter. Journal of Emergency Medicine, 12, 217-224.


Zalstein, S., & Cameron, P. A. (1997). Helicopter emergency medical services: their role in integrated trauma care. Australian & New Zealand Journal of Surgery, 67, 593-598.


134

APPENDIX 4: EXCLUDED RETRIEVED PAPERS: DOCTOR VERSUS

NO DOCTOR ON ROAD AMBULANCES

Adams, J., Aldag, G., & Wolford, R. (1996). Does the level of prehospital care influence the outcome of patients with altered levels of consciousness? Prehospital & Disaster Medicine, 11, 101-104.

Arntz, H. R., Oeff, M., Willich, S. N., Storch, W. H., & Schroder, R. (1993). Establishment and results of an EMT-D program in a two-tiered physician-escorted rescue system. The experience in Berlin, Germany. Resuscitation, 26, 39-46.

Bjerre, S. K., Hansen, T. M., Melchiorsen, H., & Christensen, E. F. (2002). Prehospital treatment of patients with acute exacerbation of chronic pulmonary disease - before and after introduction of a Mobile Emergency Care Unit. Ugeskrift for Laeger, 164, 1349-1352.

Bjorklund, P., & O'Rourke, M. F. (1984). Pre-hospital emergency care: evaluation of an Australian system. Australian & New Zealand Journal of Medicine, 14, 419-423.

Einav, S., Donchin, Y., Weissman, C., & Drenger, B. (2003). Anesthesiologists on ambulances: where do we stand? Current Opinion in Anaesthesiology, 16, 585-591.

Eisen, J. S., & Dubinsky, I. (1998). Advanced life support vs. basic life support field care: an outcome study. Academic Emergency Medicine, 5, 592-598.

Eisenburger, P., Czappek, G., Sterz, F., Vergeiner, G., Losert, H., Holzer, M., & Laggner, A. N. (2001). Cardiac arrest patients in an alpine area during a six year period. Resuscitation, 51, 39-46.

Erich, J. (2003). Road trauma. Emergency Medical Services, 32, 55-56, 58, 60 passim.

Giraud, F., Rascle, C., & Guignand, M. (1996). Out-of-hospital cardiac arrest. Evaluation of one year of activity in Saint-Etienne's emergency medical system using the Utstein style. Resuscitation, 33, 19-27.

Guly, U. M., Mitchell, R. G., Cook, R., Steedman, D. J., & Robertson, C. E. (1995). Paramedics and technicians are equally successful at managing cardiac arrest outside hospital. BMJ, 310, 1091-1094.

Hillis, M., Sinclair, D., Butler, G., & Cain, E. (1993). Prehospital cardiac arrest survival and neurologic recovery. Journal of Emergency Medicine, 11, 245-252.

Ladwig, K. H., Schoefinius, A., Danner, R., Gurtler, R., Herman, R., Koeppel, A., & Hauber, P. (1997). Effects of early defibrillation by ambulance personnel on short- and long-term outcome of cardiac arrest survival: the Munich experiment. Chest, 112, 1584-1591.

Liberman, M., Mulder, D., & Sampalis, J. (2000). Advanced or basic life support for trauma: meta-analysis and critical review of the literature. Journal of Trauma-Injury Infection and Critical

Care, 49, 584-599.

Maio, R. F., Green, P. E., Becker, M. P., Burney, R. E., & Compton, C. (1992). Rural motor vehicle crash mortality: the role of crash severity and medical resources. Accident Analysis &


Martin, S. K., Shatney, C. H., Sherck, J. P., Ho, C. C., Homan, S. J., & Neff, J. (2002). Blunt trauma patients with prehospital pulseless electrical activity (PEA): poor ending assured. Journal of



135

Nguyen-Van-Tam, J. S., Dove, A. F., Bradley, M. P., Pearson, J. C., Durston, P., & Madeley, R. J. (1997). Effectiveness of ambulance paramedics versus ambulance technicians in managing out of hospital cardiac arrest. Journal of Accident & Emergency Medicine, 14, 142-148.

Nichol, G., Detsky, A. S., Stiell, I. G., Orourke, K., Wells, G., & Laupacis, A. (1996). Effectiveness of emergency medical services for victims of out-of-hospital cardiac arrest: a metaanalysis. Annals of Emergency Medicine, 27, 700-710.

Nicholl, J., & Turner, J. (1997). Effectiveness of a regional trauma system in reducing mortality from major trauma: before and after study. BMJ, 315, 1349-1354.

Pitetti, R., Glustein, J. Z., & Bhende, M. S. (2002). Prehospital care and outcome of pediatric out-of-hospital cardiac arrest. Prehospital Emergency Care, 6, 283-290.

Rainer, T. H., Houlihan, K. P., Robertson, C. E., Beard, D., Henry, J. M., & Gordon, M. W. (1997). An evaluation of paramedic activities in prehospital trauma care. Injury, 28, 623-627.

Sethi, D., Kwan, I., Kelly, A. M., Roberts, I., & Bunn, F. (2001). Advanced trauma life support training for ambulance crews. Cochrane Database of Systematic Reviews, 2, CD003109.

Shuster, M., Keller, J., & Shannon, H. (1995). Effects of prehospital care on outcome in patients with cardiac illness. Annals of Emergency Medicine, 26, 138-145.

Shuster, M., & Shannon, H. S. (1994). Differential prehospital benefit from paramedic care. Annals of


Soo, L. H., Gray, D., Young, T., Huff, N., Skene, A., & Hampton, J. R. (1999). Resuscitation from out-of-hospital cardiac arrest: is survival dependent on who is available at the scene? Heart, 81, 47-52.

Sukumaran, S., Henry, J. M., Beard, D., Lawrenson, R., Gordon, M. W., O'Donnell, J. J., & Gray, A. J. (2005). Prehospital trauma management: a national study of paramedic activities. Emergency

Medicine Journal, 22, 60-63.

Svensson, L., Karlsson, T., Nordlander, R., Wahlin, M., Zedigh, C., & Herlitz, J. (2003a). Implementation of prehospital thrombolysis in Sweden: components of delay until delivery of treatment and examination of treatment feasibility. International Journal of Cardiology, 88, 247-256.

Svensson, L., Karlsson, T., Nordlander, R., Wahlin, M., Zedigh, C., & Herlitz, J. (2003b). Safety and delay time in prehospital thrombolysis of acute myocardial infarction in urban and rural areas in Sweden. American Journal of Emergency Medicine, 21, 263-270.

Vertesi, L., Wilson, L., & Glick, N. (1983). Cardiac arrest: comparison of paramedic and conventional ambulance services. Canadian Medical Association Journal, 128, 809-812.

Weston, C. F. M., Jones, S. D., & Wilson, R. J. (1997). Outcome of out-of-hospital cardiorespiratory arrest in south Glamorgan. Resuscitation, 34, 227-233.

White, R. D., Asplin, B. R., Bugliosi, T. F., & Hankins, D. G. (1996). High discharge survival rate after out-of-hospital ventricular fibrillation with rapid defibrillation by police and paramedics. Annals of Emergency Medicine, 28, 480-485.


136

APPENDIX 5: EXCLUDED RETRIEVED PAPERS: COMPARISON OF

OUTCOMES AMONGST CREWS THAT DO AND DO NOT PERFORM

RAPID SEQUENCE INTUBATION AND/OR THORACOSTOMY


Arntz, H. R., Oeff, M., Willich, S. N., Storch, W. H., & Schroder, R. (1993). Establishment and results of an EMT-D program in a two-tiered physician-escorted rescue system. The experience in Berlin, Germany. Resuscitation, 26, 39-46.


Bateman, C. (2005). Saving lives: Who picks up the tab? South African Medical Journal Suid-

Afrikaanse Tydskrif Vir Geneeskunde, 95, 545-550.

Baxt, W. G., & Moody, P. (1983). The impact of a rotorcraft aeromedical emergency care service on trauma mortality. JAMA, 249, 3047-3051.

Baxt, W. G., & Moody, P. (1987). The impact of advanced prehospital emergency care on the mortality of severely brain-injured patients. Journal of Trauma-Injury Infection & Critical Care, 27, 365-369.

Benagh, J. (2005). EMS in South Africa: a passionate few provide care to the masses. Journal of

Emergency Medical Services, 30, 78-80,84-93, .

Bernard, A., Handel, D., & Locasto, D. (2006). Prehospital rapid sequence intubation: a review of the literature for paramedics. Emergency Medical Services, 35, 77-81, 102.

Bjerre, S. K., Hansen, T. M., Melchiorsen, H., & Christensen, E. F. (2002). Prehospital treatment of patients with acute exacerbation of chronic pulmonary disease - before and after introduction of a mobile emergency care unit. Ugeskrift for Laeger, 164, 1349-1352.

Bjorklund, P., & O'Rourke, M. F. (1984). Pre-hospital emergency care: evaluation of an Australian system. Australian & New Zealand Journal of Medicine, 14, 419-423.

Blumen, I. J., & Gordon, R. S. (1989). Taking to the skies. Emergency, 21, 32-38.

Brismar, B., Alveryd, A., Johnsson, O., & Ohrvall, U. (1986). The ambulance helicopter is a prerequisite for centralised emergency care. Acta Chirurgica Scandinavica - Supplementum, 530, 89-93.

Burillo-Putze, G., Duarte, I. H., & Alvarez Fernandez, J. A. (2001). Helicopter emergency medical service in Spain. Air Medical Journal, 20, 21-23.




Surgery, 63, 790-797.


137

Cannell, H., Silvester, K. C., & O'Regan, M. B. (1993). Early management of multiply injured patients with maxillofacial injuries transferred to hospital by helicopter. British Journal of Oral &

Maxillofacial Surgery, 31, 207-212.

Celli, P., Fruin, A., & Cervoni, L. (1997). Severe head trauma. Review of the factors influencing the prognosis. Minerva Chirurgica, 52, 1467-1480.



Cline Jr, C. T., Smith, S., & Davenport, P. (2006). Carilion Life-Guard: a quarter century of air medical service. Air Medical Journal, 25, 118-121.

Cocanour, C. S., Fischer, R. P., & Ursic, C. M. (1997). Are scene flights for penetrating trauma justified? Journal of Trauma-Injury Infection & Critical Care, 43, 83-86; discussion 86-88.

Collier, J. (2006). Air Evac Services: improving Arizona's health over 36 years. Air Medical Journal, 25, 196-199.

Conroy, M. B., Rodriguez, S. U., Kimmel, S. E., & Kasner, S. E. (1999). Helicopter transfer offers a potential benefit to patients with acute stroke. Stroke, 30, 2580-2584.


Davis, D. P., Ochs, M., Hoyt, D. B., Bailey, D., Marshall, L. K., & Rosen, P. (2003). Paramedic-administered neuromuscular blockade improves prehospital intubation success in severely head-injured patients. Journal of Trauma-Injury Infection & Critical Care, 55, 713-719.

Davis, D. P., Pettit, K., Rom, C. D., Poste, J. C., Sise, M. J., Hoyt, D. B., & Vilke, G. M. (2005a). The safety and efficacy of prehospital needle and tube thoracostomy by aeromedical personnel. Prehospital Emergency Care, 9, 191-197.

Davis, D. P., Vadeboncoeur, T. F., Ochs, M., Poste, J. C., Vilke, G. M., & Hoyt, D. B. (2005b). The association between field Glasgow Coma Scale score and outcome in patients undergoing paramedic rapid sequence intubation. Journal of Emergency Medicine, 29, 391-397.

Einav, S., Donchin, Y., Weissman, C., & Drenger, B. (2003). Anesthesiologists on ambulances: where do we stand? Current Opinion in Anaesthesiology, 16, 585-591.

Eisen, J. S., & Dubinsky, I. (1998). Advanced life support vs basic life support field care: an outcome study. Academic Emergency Medicine, 5, 592-598.


Erich, J. (2003). Road trauma. Emergency Medical Services, 32, 55-56, 58, 60 passim.

Falcone, R. E., Herron, H., Dean, B., & Werman, H. (1996). Emergency scene endotracheal intubation before and after the introduction of a rapid sequence induction protocol. Air Medical Journal, 15, 163-167.



138


Giraud, F., Rascle, C., & Guignand, M. (1996). Out-of-hospital cardiac arrest. Evaluation of one year of activity in Saint-Etienne's emergency medical system using the Utstein style. Resuscitation, 33, 19-27.


Gomes, E., Araujo, R., Soares-Oliveira, M., & Pereira, N. (2004). International EMS systems: Portugal. Resuscitation, 62, 257-260.

Guly, U. M., Mitchell, R. G., Cook, R., Steedman, D. J., & Robertson, C. E. (1995). Paramedics and technicians are equally successful at managing cardiac arrest outside hospital. BMJ, 310, 1091-1094.

Hachinski, V. (2001). Sky doc. CMAJ Canadian Medical Association Journal, 165, 1621-1622.



Hillis, M., Sinclair, D., Butler, G., & Cain, E. (1993). Prehospital cardiac arrest survival and neurologic recovery. Journal of Emergency Medicine, 11, 245-252.

Holcomb, J. B., Niles, S. E., Miller, C. C., Hinds, D., Duke, J. H., & Moore, F. A. (2005). Prehospital physiologic data and lifesaving interventions in trauma patients. Military Medicine, 170, 7-13.

Isakov, A. P. (2006). Souls on board: helicopter emergency medical services and safety. Annals of


Isenberg, D. L., & Bissell, R. (2005). Does advanced life support provide benefits to patients?: A literature review. Prehospital & Disaster Medicine, 20, 265-270.

Kuper, P. (2005). Austin-Travis County STAR flight. Air Medical Journal, 24, 192-194.

Ladwig, K. H., Schoefinius, A., Danner, R., Gurtler, R., Herman, R., Koeppel, A., & Hauber, P. (1997). Effects of early defibrillation by ambulance personnel on short- and long-term outcome of cardiac arrest survival: the Munich experiment. Chest, 112, 1584-1591.

Lechleuthner, A., Bouillon, B., Neugebauer, E., Mennigen, R., & Tiling, T. (1994). Prehospital chest tubes incidence and analysis of iatrogenic injuries in the Emergency Medical-Service Cologne. Theoretical Surgery, 9, 220-226.


Care, 49, 584-599.

Little, M. (1994). Another kind of flying doctor. Medical Journal of Australia, 160, 214-216.

Lubin, J. S., Delbridge, T. R., Cole, J. S., Nicholas, D. H., Fore, C. A., & Wadas, R. J. (2005). EMS and emergency department physician triage: injury severity in trauma patients transported by helicopter. Prehospital Emergency Care, 9, 198-202.

Lyons, T. J., & Connor, S. B. (1995). Increased flight surgeon role in military aeromedical evacuation. Aviation Space & Environmental Medicine, 66, 927-929.

Macintyre, I. (1994). Another kind of flying doctor. BMJ, 309, 1745-1746.


139

Mackenzie, C. F., Shin, B., & Matjasko, M. J. (1987). Physicians on aeromedical teams. JAMA-Journal

of the American Medical Association, 258, 2377-2378.

Macrae, D. J. (1994). Paediatric intensive care transport. Archives of Disease in Childhood, 71, 175-178.

Maio, R. F., Green, P. E., Becker, M. P., Burney, R. E., & Compton, C. (1992). Rural motor vehicle crash mortality: the role of crash severity and medical resources. Accident Analysis &


Martin, S. K., Shatney, C. H., Sherck, J. P., Ho, C. C., Homan, S. J., & Neff, J. (2002). Blunt trauma patients with prehospital pulseless electrical activity (PEA): poor ending assured. Journal of


Massarutti, D., Trillo, G., Berlot, G., Tomasini, A., Bacer, B., D'Orlando, L., Viviani, M., et al. (2006). Simple thoracostomy in prehospital trauma management is safe and effective: a 2-year experience by helicopter emergency medical crews. European Journal of Emergency

Medicine, 13, 276-280.



Munford, B., & Manning, R. (1994). Paramedic helicopter retrieval of trauma patients. Australian and

New Zealand Journal of Surgery, 64, 640-641.

Nguyen-Van-Tam, J. S., Dove, A. F., Bradley, M. P., Pearson, J. C., Durston, P., & Madeley, R. J. (1997). Effectiveness of ambulance paramedics versus ambulance technicians in managing out of hospital cardiac arrest. Journal of Accident & Emergency Medicine, 14, 142-148.


Nicholl, J., & Turner, J. (1997). Effectiveness of a regional trauma system in reducing mortality from major trauma: before and after study. BMJ, 315, 1349-1354.

Nocera, A., & Dalton, A. M. (1994). Disaster alert! The role of physician-staffed helicopter emergency medical services. Medical Journal of Australia, 161, 689-692.

Oppe, S., & De Charro, F. T. (2001). The effect of medical care by a helicopter trauma team on the probability of survival and the quality of life of hospitalised victims. Accident Analysis &


Orlando, R., Schwartz, R., Lee, M., & Jacobs, L. (1987). The role of the flight physician in helicopter critical care transport. Critical Care Medicine, 15, 367-367.



Osterwalder, J. J. (2003). Mortality of blunt polytrauma: a comparison between emergency physicians and emergency medical technicians - prospective cohort study at a level I hospital in eastern Switzerland. Journal of Trauma-Injury Infection and Critical Care, 55, 355-361.



140


Rivkind, A. I., & Almogy, G. (2006). Civilian use of helicopters improves access to trauma care and increases chances of survival. Israel Medical Association Journal: IMAJ, 8, 56-57.

Rodenberg, H. (1992a). Effect of aeromedical aircraft on care of trauma patients - evaluation using the revised trauma score. Southern Medical Journal, 85, 1065-1071.

Rodenberg, H. (1992b). The revised trauma score: a means to evaluate aeromedical staffing patterns. Aviation Space & Environmental Medicine, 63, 308-313.



Sethi, D., Kwan, I., Kelly, A. M., Roberts, I., & Bunn, F. (2001). Advanced trauma life support training for ambulance crews. Cochrane Database of Systematic Reviews, 2, CD003109.

Shufflebarger, C., & Townsend, R. (1987). Physicians on aeromedical teams. JAMA-Journal of the

American Medical Association, 258, 2378-2378.


Shuster, M., & Shannon, H. S. (1994). Differential prehospital benefit from paramedic care. Annals of


Slagt, C., Zondervan, A., Patka, P., & de Lange, J. J. (2004). A retrospective analysis of the intubations performed during 5 years of helicopter emergency medical service in Amsterdam. Air Medical

Journal, 23, 36-37.


Soo, L. H., Gray, D., Young, T., Huff, N., Skene, A., & Hampton, J. R. (1999). Resuscitation from out-of-hospital cardiac arrest: is survival dependent on who is available at the scene? Heart, 81, 47-52.

Spangler, D. E., Jr., Rogers, W. J., Gore, J. M., Griffith, M., Maske, L. E., Morgan, T. E., & Corrao, J. (1991). Early tPA treatment and aeromedical transport of patients with acute myocardial infarction. Journal of Interventional Cardiology, 4, 81-89.

Stauffer, U. G. (1995). Surgical and critical care management of children with life-threatening injuries: the Swiss experience. Journal of Pediatric Surgery, 30, 903-910.




Surgery, 33, 1388-1392.


141

Svensson, L., Karlsson, T., Nordlander, R., Wahlin, M., Zedigh, C., & Herlitz, J. (2003a). Implementation of prehospital thrombolysis in Sweden: components of delay until delivery of treatment and examination of treatment feasibility. International Journal of Cardiology, 88, 247-256.

Svensson, L., Karlsson, T., Nordlander, R., Wahlin, M., Zedigh, C., & Herlitz, J. (2003b). Safety and delay time in prehospital thrombolysis of acute myocardial infarction in urban and rural areas in Sweden. American Journal of Emergency Medicine, 21, 263-270.



van Wijngaarden, M., Kortbeek, J., Lafreniere, R., Cunningham, R., Joughin, E., & Yim, R. (1996). Air ambulance trauma transport: a quality review. Journal of Trauma-Injury Infection & Critical

Care, 41, 26-31.

Vertesi, L., Wilson, L., & Glick, N. (1983). Cardiac arrest: comparison of paramedic and conventional ambulance services. Canadian Medical Association Journal, 128, 809-812.

Vilke, G. M., Hoyt, D. B., Epperson, M., Fortlage, D., Hutton, K. C., & Rosen, P. (1994). Intubation techniques in the helicopter. Journal of Emergency Medicine, 12, 217-224.

Weston, C. F. M., Jones, S. D., & Wilson, R. J. (1997). Outcome of out-of-hospital cardiorespiratory arrest in south Glamorgan. Resuscitation, 34, 227-233.

White, R. D., Asplin, B. R., Bugliosi, T. F., & Hankins, D. G. (1996). High discharge survival rate after out-of-hospital ventricular fibrillation with rapid defibrillation by police and paramedics. Annals of Emergency Medicine, 28, 480-485.


Zalstein, S., & Cameron, P. A. (1997). Helicopter emergency medical services: their role in integrated trauma care. Australian & New Zealand Journal of Surgery, 67, 593-598.


142

APPENDIX 6: EXCLUDED RETRIEVED PAPERS: OUTCOMES BY

TIME FROM AMBULANCE CALL OUT TO EMERGENCY

DEPARTMENT DELIVERY

Acosta, J. A., Hatzigeorgiou, C., & Smith, L. S. (2006). Developing a trauma registry in a forward deployed military hospital: preliminary report. Journal of Trauma-Injury Infection & Critical

Care, 61, 256-260.


Anderson, T. E., Rose, W. D., & Leicht, M. J. (1987). Physician-staffed helicopter scene response from a rural trauma center. Annals of Emergency Medicine, 16, 58-61.

Andrews, C. N., Kobusingye, O. C., & Lett, R. (1999). Road traffic accident injuries in Kampala. East

African Medical Journal, 76, 189-194.

Arreolarisa, C., Mock, C. N., Padilla, D., Cavazos, L., Maier, R. V., & Jurkovich, G. J. (1995). Trauma care systems in urban latin-america - the priorities should be prehospital and emergency room management. Journal of Trauma-Injury Infection and Critical Care, 39, 457-462.


Battistella, F. D., Nugent, W., Owings, J. T., & Anderson, J. T. (1999). Field triage of the pulseless trauma patient. Archives of Surgery, 134, 742-745; discussion 745-746.

Bickell, W. H., Wall, M. J., Pepe, P. E., Martin, R. R., Ginger, V. F., Allen, M. K., & Mattox, K. L. (1994). Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. New England Journal of Medicine, 331, 1105-1109.

Bjerre, S. K., Hansen, T. M., Melchiorsen, H., & Christensen, E. F. (2002). Prehospital treatment of patients with acute exacerbation of chronic pulmonary disease - Before and after introduction of a Mobile Emergency Care Unit. Ugeskrift for Laeger, 164, 1349-1352.

Blackwell, T. H., & Kaufman, J. S. (2002). Response time effectiveness: comparison of response time and survival in an urban emergency medical services system. Academic Emergency Medicine, 9, 288-295.

Bottiger, B. W., Grabner, C., Bauer, H., Bode, C., Weber, T., Motsch, J., & Martin, E. (1999). Long term outcome after out-of-hospital cardiac arrest with physician staffed emergency medical services: the Utstein style applied to a midsized urban/suburban area. Heart, 82, 674-679.

Brathwaite, C. E. M., Rosko, M., McDowell, R., Gallagher, J., Proenca, J., & Spott, M. A. (1998). A critical analysis of on-scene helicopter transport on survival in a statewide trauma system. Journal of Trauma-Injury Infection and Critical Care, 45, 140-144.

Brison, R. J., Davidson, J. R., Dreyer, J. F., Jones, G., Maloney, J., Munkley, D. P., Oconnor, M., et al. (1992). Cardiac-arrest in Ontario - circumstances, community response, role of prehospital defibrillation and predictors of survival. Canadian Medical Association Journal, 147, 191-199.

Bunch, T. J., West, C. P., Packer, D. L., Panutich, M. S., & White, R. D. (2004). Admission predictors of in-hospital mortality and subsequent long-term outcome in survivors of ventricular fibrillation out-of-hospital cardiac arrest: a population-based study. Cardiology, 102, 41-47.


143

Burger, T., Meyer, F., Tautenhahn, J., & Halloul, Z. (1999). Ruptured infrarenal aortic aneurysm--a critical evaluation. Vasa, 28, 30-33.


Caldwell, M. A., Froelicher, E. S., & Drew, B. J. (2000). Prehospital delay time in acute myocardial infarction: an exploratory study on relation to hospital outcomes and cost. American Heart

Journal, 139, 788-796.


Surgery, 63, 790-797.

Chappell, V. L., Mileski, W. J., Wolf, S. E., & Gore, D. C. (2002). Impact of discontinuing a hospital-based air ambulance service on trauma patient outcomes. Journal of Trauma-Injury Infection

& Critical Care, 52, 486-491.



Citerio, G., Galli, D., Cesana, G. C., Bosio, M., Landriscina, M., Raimondi, M., Rossi, G. P., et al. (2002). Emergency system prospective performance evaluation for cardiac arrest in Lombardia, an Italian region. Resuscitation, 55, 247-254.

Citerio, G., Galli, D., & Pesenti, A. (2006). Early stroke care in Italy - a steep way ahead: an observational study. Emergency Medicine Journal, 23, 608-611.

Clevenger, F. W., Yarbrough, D. R., & Reines, H. D. (1988). Resuscitative thoracotomy: the effect of field time on outcome. Journal of Trauma-Injury Infection & Critical Care, 28, 441-445.

Cooke, W. H., Salinas, J., Convertino, V. A., Ludwig, D. A., Hinds, D., Duke, J. H., Moore, F. A., et al. (2006). Heart rate variability and its association with mortality in prehospital trauma patients. Journal of Trauma-Injury Infection & Critical Care, 60, 363-370; discussion 370.

Cornwell, E. E., 3rd, Belzberg, H., Hennigan, K., Maxson, C., Montoya, G., Rosenbluth, A., Velmahos, G. C., et al. (2000). Emergency medical services (EMS) vs non-EMS transport of critically injured patients: a prospective evaluation. Archives of Surgery, 135, 315-319.

Cunningham, P., Rutledge, R., Baker, C. C., & Clancy, T. V. (1997). A comparison of the association of helicopter and ground ambulance transport with the outcome of injury in trauma patients transported from the scene. Journal of Trauma-Injury Infection & Critical Care, 43, 940-946.

Dash, H. H., & Kaul, N. (2004). Prehospital care of the head trauma patients. Journal of

Anaesthesiology Clinical Pharmacology, 20, 333-337.

de Vreede-Swagemakers, J. J., Gorgels, A. P., Dubois-Arbouw, W. I., Dalstra, J., Daemen, M. J., van Ree, J. W., Stijns, R. E., et al. (1998). Circumstances and causes of out-of-hospital cardiac arrest in sudden death survivors. Heart, 79, 356-361.

Dean, N. C., Haug, P. J., & Hawker, P. J. (1988). Effect of mobile paramedic units on outcome in patients with myocardial infarction. Annals of Emergency Medicine, 17, 1034-1041.

Demetriades, D., Chan, L., Cornwell, E., Belzberg, H., Berne, T. V., Asensio, J., Chan, D., et al. (1996). Paramedic vs private transportation of trauma patients - effect on outcome. Archives of

Surgery, 131, 133-138.


144

Devlin, J., Wendon, J., Heaton, N., Tan, K. C., & Williams, R. (1995). Pretransplantation clinical status and outcome of emergency transplantation for acute liver-failure. Hepatology, 21, 1018-1024.

Di Bartolomeo, S., Sanson, G., Michelutto, V., Nardi, G., Burba, I., Francescutti, C., Lattuada, L., et al. (2004). Epidemiology of major injury in the population of Friuli Venezia Giulia-Italy. Injury, 35, 391-400.

Di Bartolomeo, S., Sanson, G., Nardi, G., Scian, F., Michelutto, V., & Lattuada, L. (2001). Effects of 2 patterns of prehospital care on the outcome of patients with severe head injury. Archives of

Surgery, 136, 1293-1300.

Di Chiara, A., Chiarella, F., Savonitto, S., Lucci, D., Bolognese, L., De Servi, S., Greco, C., et al. (2003). Epidemiology of acute myocardial infarction in the Italian CCU network: the BLITZ study. European Heart Journal, 24, 1616-1629.

Eisen, J. S., & Dubinsky, I. (1998). Advanced life support vs basic life support field care: an outcome study. Academic Emergency Medicine, 5, 592-598.


Eng Hock Ong, M., Chan, Y. H., Anantharaman, V., Lau, S. T., Lim, S. H., & Seldrup, J. (2003). Cardiac arrest and resuscitation epidemiology in Singapore (CARE I study). Prehospital

Emergency Care, 7, 427-433.

Engdahl, J., Bang, A., Lindqvist, J., & Herlitz, J. (2000). Can we define patients with no and those with some chance of survival when found in asystole out of hospital? American Journal of

Cardiology, 86, 610-614.

Esposito, T. J., Maier, R. V., Rivara, F. P., Pilcher, S., Griffith, J., Lazear, S., & Hogan, S. (1995). The impact of variation in trauma care times: urban versus rural. Prehospital & Disaster Medicine, 10, 161-166; discussion 166-167.

Farooq, M. M., Freischlag, J. A., Seabrook, G. R., Moon, M. R., Aprahamian, C., & Towne, J. B. (1996). Effect of the duration of symptoms, transport time, and length of emergency room stay on morbidity and mortality in patients with ruptured abdominal aortic aneurysms. Surgery, 119, 9-14.

Fredriksson, M., Herlitz, J., & Engdahl, J. (2003). Nineteen years' experience of out-of-hospital cardiac arrest in Gothenburg - reported in Utstein style. Resuscitation, 58, 37-47.


Grossman, D. C., Hart, L. G., Rivara, F. P., Maier, R. V., & Rosenblatt, R. (1995). From roadside to bedside: the regionalization of trauma care in a remote rural county. Journal of Trauma-Injury

Infection & Critical Care, 38, 14-21.

Guglin, M. E., Wilson, A., Kostis, J. B., Parrillo, J. E., White, M. C., & Gessman, L. J. (2004). Immediate and 1-year survival of out-of-hospital cardiac arrest victims in southern New Jersey: 1995-2000. Pacing & Clinical Electrophysiology, 27, 1072-1076.

Hackbarth, R. M., Rzeszutko, K. M., Sturm, G., Donders, J., Kuldanek, A. S., & Sanfilippo, D. J. (2002). Survival and functional outcome in pediatric traumatic brain injury: a retrospective review and analysis of predictive factors. Critical Care Medicine, 30, 1630-1635.

Handel, D. A., Gallo, P., Schmidt, M., Bernard, A., Locasto, D., Collett, L., & Lindsell, C. J. (2005). Prehospital cardiac arrest in a paramedic first-responder system using the Utstein style. Prehospital Emergency Care, 9, 398-404.


145

Harley, B. J., Beaupre, L. A., Jones, C. A., Dulai, S. K., & Weber, D. W. (2002). The effect of time to definitive treatment on the rate of nonunion and infection in open fractures. Journal of

Orthopaedic Trauma, 16, 484-490.

Hata, N., Kobayashi, N., Imaizumi, T., Yokoyama, S., Shinada, T., Tanabe, J., Shiiba, K., et al. (2006). Use of an air ambulance system improves time to treatment of patients with acute myocardial infarction. Internal Medicine, 45, 45-50.

Haveman, J. W., Karliczek, A., Verhoeven, E. L. G., Tielliu, I. F. J., de Vos, R., Zwaveling, J. H., van den Dungen, J., et al. (2006). Results of streamlined regional ambulance transport and subsequent treatment of acute abdominal aortic aneurysms. Emergency Medicine Journal, 23, 807-810.

Hayashi, Y., Hiraide, A., Morita, H., Shinya, H., Nishiuchi, T., Mukainaka, S., Kai, T., et al. (2002). An analysis of time factors in out-of-hospital cardiac arrest in Osaka Prefecture. Resuscitation, 53, 121-125.

Herlitz, J., Bang, A., Gunnarsson, J., Engdahl, J., Karlson, B. W., Lindqvist, J., & Waagstein, L. (2003). Factors associated with survival to hospital discharge among patients hospitalised alive after out of hospital cardiac arrest: change in outcome over 20 years in the community of Goteborg, Sweden. Heart, 89, 25-30.

Herlitz, J., Ekstrom, L., Wennerblom, B., Axelsson, A., Bang, A., & Holmberg, S. (1994). Survival in patients found to have ventricular-fibrillation after cardiac-arrest witnessed outside hospital. European Heart Journal, 15, 1628-1633.

Herlitz, J., Ekstrom, L., Wennerblom, B., Axelsson, A., Bang, A., & Holmberg, S. (1996). Type of arrhythmia at EMS arrival on scene in out-of-hospital cardiac arrest in relation to interval from collapse and whether a bystander initiated CPR. American Journal of Emergency Medicine, 14, 119-123.

Herlitz, J., Engdahl, J., Svensson, L., Angquist, K. A., Young, M., & Holmberg, S. (2005). Factors associated with an increased chance of survival among patients suffering from an out-of-hospital cardiac arrest in a national perspective in Sweden. American Heart Journal, 149, 61-66.

Hill, D. A., Delaney, L. M., & Duflou, J. (1996). A population-based study of outcome after injury to car occupants and to pedestrians. Journal of Trauma-Injury Infection & Critical Care, 40, 351-355.

Hodgetts, T. J., Brown, T., Driscoll, P., & Hanson, J. (1995). Pre-hospital cardiac arrest: room for improvement. Resuscitation, 29, 47-54.

Hollenberg, J., Bang, A., Lindqvist, J., Herlitz, J., Nordlander, R., Svensson, L., & Rosenqvist, M. (2005). Difference in survival after out-of-hospital cardiac arrest between the two largest cities in Sweden: a matter of time? Journal of Internal Medicine, 257, 247-254.

Honigman, B., Rohweder, K., Moore, E. E., Lowenstein, S. R., & Pons, P. T. (1990). Prehospital advanced trauma life support for penetrating cardiac wounds. Annals of Emergency Medicine, 19, 145-150.

Hu, S. C., Kao, W. F., Tsai, J., Chern, C. H., Yen, D., Lo, H. C., & Lee, C. H. (1996). Analysis of prehospital ALS cases in a rural community. Chung Hua i Hsueh Tsa Chih - Chinese Medical

Journal, 58, 171-176.

Jennings, P., & Pasco, J. (2001). Survival from out-of-hospital cardiac arrest in the Geelong region of Victoria, Australia. Emergency Medicine (Fremantle, WA ), 13, 319-325.

Jones, J. B., Leicht, M., & Dula, D. J. (1998). A 10-year experience in the use of air medical transport for medical scene calls. Air Medical Journal, 17, 7-11; discussion 11-12.


146

Karch, S. B., Graff, J., Young, S., & Ho, C. H. (1998). Response times and outcomes for cardiac arrests in Las Vegas casinos. American Journal of Emergency Medicine, 16, 249-253.

Karch, S. B., Lewis, T., Young, S., & Ho, C. H. (1995). Surgical delays and outcomes in patients treated with pneumatic antishock garments: a population-based study. American Journal of


Kearney, P. A., Terry, L., & Burney, R. E. (1991). Outcome of patients with blunt trauma transferred after diagnostic or treatment procedures or four-hour delay. Annals of Emergency Medicine, 20, 882-886.

Kelly, A. M., Kerr, D., Patrick, I., & Walker, T. (2003). Call-to-needle times for thrombolysis in acute myocardial infarction in Victoria. Medical Journal of Australia, 178, 381-385.

Kuisma, M., & Jaara, K. (1997). Unwitnessed out-of-hospital cardiac arrest: is resuscitation worthwhile? Annals of Emergency Medicine, 30, 69-75.


Liberman, M., Mulder, D., Lavoie, A., Denis, R., & Sampalis, J. S. (2003). Multicenter Canadian study of prehospital trauma care. Annals of Surgery, 237, 153-160.


Care, 49, 584-599.

Lorenzi, L., Kerr, M. E., Yonas, H., Alexander, S., & Crago, E. (2003). Influence of delaying treatment after symptoms develop from subarachnoid hemorrhage: a preliminary analysis. Journal of

Neuroscience Nursing, 35, 210-214.

Lyon, R. M., Cobbe, S. M., Bradley, J. M., & Grubb, N. R. (2004). Surviving out of hospital cardiac arrest at home: a postcode lottery? Emergency Medicine Journal, 21, 619-624.

Mashiko, K., Otsuka, T., Shimazaki, S., Kohama, A., Kamishima, G., Katsurada, K., Sawada, Y., et al. (2002). An outcome study of out-of-hospital cardiac arrest using the Utstein template - a Japanese experience. Resuscitation, 55, 241-246.

Massarutti, D., Trillo, G., Berlot, G., Tomasini, A., Bacer, B., D'Orlando, L., Viviani, M., et al. (2006). Simple thoracostomy in prehospital trauma management is safe and effective: a 2-year experience by helicopter emergency medical crews. European Journal of Emergency

Medicine, 13, 276-280.

Maynard, C., Althouse, R., Olsufka, M., Ritchie, J. L., Davis, K. B., & Kennedy, J. W. (1989). Early versus late hospital arrival for acute myocardial infarction in the Western Washington thrombolytic therapy trials. American Journal of Cardiology, 63, 1296-1300.

McMillan, R. W., Landreneau, M. D., McCormick, G. M., & McDonald, J. C. (1992). Major vascular injuries of the torso. Southern Medical Journal, 85, 375-377.

McNicholl, B. P. (1994). The golden hour and prehospital trauma care. Injury, 25, 251-254.

Mitchell, R. G., Brady, W., Guly, U. M., Pirrallo, R. G., & Robertson, C. E. (1997). Comparison of two emergency response systems and their effect on survival from out of hospital cardiac arrest. Resuscitation, 35, 225-229.

Moore, W., Kee, F., Evans, A. E., McCrum-Gardner, E. E., Morrison, C., & Tunstall-Pedoe, H. (2005). Pre-hospital coronary care and coronary fatality in the Belfast and Glasgow MONICA populations. International Journal of Epidemiology, 34, 422-430.


147

Mullins, R. J., & Mann, N. C. (1999). Population-based research assessing the effectiveness of trauma systems. Journal of Trauma-Injury Infection & Critical Care, 47, 3 Suppl, S59-66.

Nakae, H., & Wada, H. (2002). Characteristics of burn patients transported by ambulance to treatment facilities in Akita Prefecture, Japan. Burns, 28, 73-79.

Nakstad, A. R., Sorebo, H., Heimdal, H. J., Strand, T., & Sandberg, M. (2004). Rapid response car as a supplement to the helicopter in a physician-based HEMS system. Acta Anaesthesiologica


Nathens, A. B., Brunet, F. P., & Maier, R. V. (2004). Development of trauma systems and effect on outcomes after injury. Lancet, 363, 1794-1801.


Norris, R. M., Wong, P. S., Dixon, G., Morris, N., Penny, W. J., El Gaylani, N., Thomas, A., et al. (1998). Effect of time from onset to coming under care on fatality of patients with acute myocardial infarction: effect of resuscitation and thrombolytic treatment. Heart, 80, 114-120.

Olson, D. W., LaRochelle, J., Fark, D., Aprahamian, C., Aufderheide, T. P., Mateer, J. R., Hargarten, K. M., et al. (1989). EMT-defibrillation: the Wisconsin experience. Annals of Emergency

Medicine, 18, 806-811.

Pamerneckas, A., Macas, A., Vaitkaitis, D., Vaitkaitis, A., & Gudeniene, R. (2003). Golden hour - early postinjury period. Medicina (Kaunas), 39, 845-851.

Papadimitriou, D. G., Mathur, M. N., & Hill, D. A. (1994). A survey of rural road fatalities. Australian

& New Zealand Journal of Surgery, 64, 479-483.

Pell, J. P., Sirel, J. M., Marsden, A. K., Ford, I., & Cobbe, S. M. (2001). Effect of reducing ambulance response times on deaths from out of hospital cardiac arrest: cohort study. British Medical

Journal, 322, 1385-1388.

Petri, R. W., Dyer, A., & Lumpkin, J. (1995). The effect of prehospital transport time on the mortality from traumatic injury. Prehospital & Disaster Medicine, 10, 24-29.


Pompilio, G., Spirito, R., Alamanni, F., Agrifoglio, M., Polvani, G., Porqueddu, M., Reali, M., et al. (2001). Determinants of early and late outcome after surgery for type A aortic dissection. World Journal of Surgery, 25, 1500-1506.

Pons, P. T., Haukoos, J. S., Bludworth, W., Cribley, T., Pons, K. A., & Markovchick, V. J. (2005). Paramedic response time: does it affect patient survival? Academic Emergency Medicine, 12, 594-600.

Poon, W. S., & Li, A. K. (1991). Comparison of management outcome of primary and secondary referred patients with traumatic extradural haematoma in a neurosurgical unit. Injury, 22, 323-325.

Qureshi, A. I., Kirmani, J. F., Sayed, M. A., Safdar, A., Ahmed, S., Ferguson, R., Hershey, L. A., et al. (2005). Time to hospital arrival, use of thrombolytics, and in-hospital outcomes in ischemic stroke. Neurology, 64, 2115-2120.



148

Regel, G., Lobenhoffer, P., Grotz, M., Pape, H. C., Lehmann, U., & Tscherne, H. (1995). Treatment results of patients with multiple trauma: an analysis of 3406 cases treated between 1972 and 1991 at a German Level I Trauma Center. Journal of Trauma-Injury Infection & Critical

Care, 38, 70-78.

Rogers, F. B., Simons, R., Hoyt, D. B., Shackford, S. R., Holbrook, T., & Fortlage, D. (1993). In-house board-certified surgeons improve outcome for severely injured patients: a comparison of two university centers. Journal of Trauma-Injury Infection & Critical Care, 34, 871-875; discussion 875-877.

Rosengart, M. R., Smith, D. R., Melton, S. M., May, A. K., & Rue, L. W. (1999). Prognostic factors in patients with inferior vena cava injuries. American Surgeon, 65, 849-855.

Ruchholtz, S., Waydhas, C., Lewan, U., Piepenbrink, K., Stolke, D., Debatin, J., Schweiberer, L., et al. (2002). A multidisciplinary quality management system for the early treatment of severely injured patients: implementation and results in two trauma centers. Intensive Care Medicine, 28, 1395-1404.

Rudehill, A., Bellander, B. M., Weitzberg, E., Bredbacka, S., Backheden, M., & Gordon, E. (2002). Outcome of traumatic brain injuries in 1,508 patients: impact of prehospital care. Journal of

Neurotrauma, 19, 855-868.

Rutledge, R., Smith, C. Y., & Azizkhank, R. G. (1994). A population-based multivariate analysis of the association of county demographic and medical system factors with per capita pediatric trauma death rates in North Carolina. Annals of Surgery, 219, 205-210.

Safford, S. D., Hayward, T. Z., Safford, K. M., Georgiade, G. S., Rice, H. E., & Skinner, M. A. (2002). A cost and outcomes comparison of a novel integrated pediatric air and ground transportation system. Journal of the American College of Surgeons, 195, 790-795.

Sammel, N. L., Taylor, K., Selig, M., & O'Rourke, M. F. (1981). New South Wales intensive care ambulance system: outcome of patients with ventricular fibrillation. Medical Journal of

Australia, 2, 546-550.

Sampalis, J. S., Denis, R., Frechette, P., Brown, R., Fleiszer, D., & Mulder, D. (1997a). Direct transport to tertiary trauma centers versus transfer from lower level facilities: impact on mortality and morbidity among patients with major trauma. Journal of Trauma-Injury Infection & Critical

Care, 43, 288-295; discussion 295-286.

Sampalis, J. S., Tamim, H., Denis, R., Boukas, S., Ruest, S. A., Nikolis, A., Lavoie, A., et al. (1997b). Ineffectiveness of on-site intravenous lines: is prehospital time the culprit? Journal of

Trauma-Injury Infection & Critical Care, 43, 608-615; discussion 615-607.

Sanson, G., Di Bartolomeo, S., Nardi, G., Albanese, P., Diani, A., Raffin, L., Filippetto, C., et al. (1999). Road traffic accidents with vehicular entrapment: incidence of major injuries and need for advanced life support. European Journal of Emergency Medicine, 6, 285-291.

Scarpelini, S., de Andrade, J. I., & Dinis Costa Passos, A. (2006). The TRISS method applied to the victims of traffic accidents attended at a tertiary level emergency hospital in a developing country. Injury, 37, 72-77.


Schwartz, M. L., Sharkey, P. W., & Andersen, J. A. (1991). Quality assurance for patients with head-injuries admitted to a regional trauma unit. Journal of Trauma-Injury Infection and Critical

Care, 31, 962-967.


149

Shackford, S. R., Mackersie, R. C., Davis, J. W., Wolf, P. L., & Hoyt, D. B. (1989). Epidemiology and pathology of traumatic deaths occurring at a Level I Trauma Center in a regionalized system: the importance of secondary brain injury. Journal of Trauma-Injury Infection & Critical Care, 29, 1392-1397.

Shatney, C. H., Homan, S. J., Sherck, J. P., & Ho, C. C. (2002). The utility of helicopter transport of trauma patients from the injury scene in an urban trauma system. Journal of Trauma-Injury



Stockinger, Z. T., & McSwain, N. E., Jr. (2004). Prehospital endotracheal intubation for trauma does not improve survival over bag-valve-mask ventilation. Journal of Trauma-Injury Infection &

Critical Care, 56, 531-536.




Surgery, 33, 1388-1392.

Talving, P., Palstedt, J., & Riddez, L. (2005). Prehospital management and fluid resuscitation in hypotensive trauma patients admitted to Karolinska University Hospital in Stockholm. Prehospital & Disaster Medicine, 20, 228-234.

Tanigawa, K., Tanaka, K., & Shigematsu, A. (2000). Outcomes of out-of-hospital ventricular fibrillation: their association with time to defibrillation and related issues in the defibrillation program in Japan. Resuscitation, 45, 83-90.



van Alem, A. P., de Vos, R., Schmand, B., & Koster, R. W. (2004). Cognitive impairment in survivors of out-of-hospital cardiac arrest. American Heart Journal, 148, 416-421.

Vukmir, R. B. (2004). Witnessed arrest, but not delayed bystander cardiopulmonary resuscitation improves prehospital cardiac arrest survivial. Emergency Medicine Journal, 21, 370-373.

Vukmir, R. B. (2006). Survival from prehospital cardiac arrest is critically dependent upon response time. Resuscitation, 69, 229-234.

Waalewijn, R. A., de Vos, R., & Koster, R. W. (1998). Out-of-hospital cardiac arrests in Amsterdam and its surrounding areas: results from the Amsterdam resuscitation study (ARREST) in 'Utstein' style. Resuscitation, 38, 157-167.

Wang, H. E., Min, A., Hostler, D., Chang, C. C. H., & Callaway, C. W. (2005). Differential effects of out-of-hospital interventions on short-and long-term survival after cardiopulmonary arrest. Resuscitation, 67, 69-74.

Weston, C. F., & McCabe, M. J. (1992). Audit of an emergency ambulance service: impact of a paramedic system. Journal of the Royal College of Physicians of London, 26, 86-89.

Weston, C. F., Wilson, R. J., & Jones, S. D. (1997). Predicting survival from out-of-hospital cardiac arrest: a multivariate analysis. Resuscitation, 34, 27-34.


150


Wisborg, T., Guttormsen, A. B., Sorensen, M. B., & Flaatten, H. K. (1994). The potential of an anaesthesiologist-manned ambulance service in a rural/urban district. Acta Anaesthesiologica


Yates, D. W., Svoboda, P., & Kantorova, I. (2002). The influence of medical care on the death rate from trauma in England and South Moravia. European Journal of Trauma, 28, 304-309.

Yates, D. W., Woodford, M., & Hollis, S. (1992). Preliminary analysis of the care of injured patients in 33 British hospitals: first report of the United Kingdom major trauma outcome study. BMJ, 305, 737-740.

Zafar, H., Rehmani, R., Raja, A. J., Ali, A., & Ahmed, M. (2002). Registry based trauma outcome: perspective of a developing country. Emergency Medicine Journal, 19, 391-394.


151

APPENDIX 7: INCLUDED PAPERS

Baxt, W. G., & Moody, P. (1987). The impact of a physician as part of the aeromedical prehospital team in patients with blunt trauma. JAMA, 257, 3246-3250.

Berns, K. S., Hankins, D. G., & Zietlow, S. P. (2001). Comparison of air and ground transport of cardiac patients. Air Medical Journal, 20, 33-36.

Biewener, A., Aschenbrenner, U., Rammelt, S., Grass, R., & Zwipp, H. (2004). Impact of helicopter transport and hospital level on mortality of polytrauma patients. Journal of Trauma-Injury


Bonatti, J., Goschl, O., Larcher, P., Wodlinger, R., & Flora, G. (1995). Predictors of short-term survival after helicopter rescue. Resuscitation, 30, 133-140.



Cameron, S., Pereira, P., Mulcahy, R., & Seymour, J. (2005). Helicopter primary retrieval: tasking who should do it? Emergency Medicine Australasia, 17, 387-391.



Clarke, J. R., Trooskin, S. Z., Doshi, P. J., Greenwald, L., & Mode, C. J. (2002). Time to laparotomy for intra-abdominal bleeding from trauma does affect survival for delays up to 90 minutes. Journal of Trauma-Injury Infection & Critical Care, 52, 420-425.

Feero, S., Hedges, J. R., Simmons, E., & Irwin, L. (1995). Does out-of-hospital EMS time affect trauma survival? American Journal of Emergency Medicine, 13, 133-135.


Frezza, E. E., & Mezghebe, H. (1999). Is 30 minutes the golden period to perform emergency room thoratomy (ERT) in penetrating chest injuries? Journal of Cardiovascular Surgery, 40, 147-151.

Gao, J. M., Gao, Y. H., Zeng, J. B., Wang, J. B., He, P., Wei, G. B., & Xiang, Z. (2006). Polytrauma with thoracic and/or abdominal injuries: experience in 1 540 cases. Chinese Journal of

Traumatology, 9, 108-114.

Garner, A., Rashford, S., Lee, A., & Bartolacci, R. (1999). Addition of physicians to paramedic helicopter services decreases blunt trauma mortality. Australian & New Zealand Journal of

Surgery, 69, 697-701.

Grzybowski, M., Zalenski, R. J., Ross, M. A., & Bock, B. (2000). A prediction model for prehospital triage of patients with suspected cardiac ischemia. Journal of Electrocardiology, 33 Suppl, 253-258.


152



Hartl, R., Gerber, L. M., Iacono, L., Ni, Q. H., Lyons, K., & Ghajar, J. (2006). Direct transport within an organized state trauma system reduces mortality in patients with severe traumatic brain injury. Journal of Trauma-Injury Infection and Critical Care, 60, 1250-1256.

Kearney, P. A., Terry, L., & Burney, R. E. (1991). Outcome of patients with blunt trauma transferred after diagnostic or treatment procedures or four-hour delay. Annals of Emergency Medicine, 20, 882-886.

Koefoed-Nielsen, J., Christensen, E. F., Melchiorsen, H., & Foldspang, A. (2002). Acute myocardial infarction: does pre-hospital treatment increase survival? European Journal of Emergency

Medicine, 9, 210-216.


Lerner, E. B., Billittier, A. J., Dorn, J. M., & Wu, Y. W. (2003). Is total out-of-hospital time a significant predictor of trauma patient mortality? Academic Emergency Medicine, 10, 949-954.

Lim, G. H., & Seow, E. (2002). Resuscitation for patients with out-of-hospital cardiac arrest: Singapore. Prehospital & Disaster Medicine, 17, 96-101.



Pepe, P. E., Wyatt, C. H., Bickell, W. H., Bailey, M. L., & Mattox, K. L. (1987). The relationship between total prehospital time and outcome in hypotensive victims of penetrating injuries. Annals of Emergency Medicine, 16, 293-297.

Phillips, R. T., Conaway, C., Mullarkey, D., & Owen, J. L. (1999). One year's trauma mortality experience at Brooke Army Medical Center: is aeromedical transportation of trauma patients necessary? Military Medicine, 164, 361-365.

Sampalis, J. S., Denis, R., Lavoie, A., Frechette, P., Boukas, S., Nikolis, A., Benoit, D., et al. (1999). Trauma care regionalization: a process-outcome evaluation. Journal of Trauma-Injury


Sampalis, J. S., Lavoie, A., Williams, J. I., Mulder, D. S., & Kalina, M. (1992). Standardized mortality ratio analysis on a sample of severely injured patients from a large Canadian city without regionalized trauma care. Journal of Trauma-Injury Infection & Critical Care, 33, 205-211; discussion 211-202.

Sampalis, J. S., Lavoie, A., Williams, J. I., Mulder, D. S., & Kalina, M. (1993). Impact of on-site care, prehospital time, and level of in-hospital care on survival in severely injured patients. Journal

of Trauma-Injury Infection & Critical Care, 34, 252-261.

Schiller, W. R., Knox, R., Zinnecker, H., Jeevanandam, M., Sayre, M., Burke, J., & Young, D. H. (1988). Effect of helicopter transport of trauma victims on survival in an urban trauma center. Journal of Trauma-Injury Infection & Critical Care, 28, 1127-1134.



153


Sloan, E. P., Callahan, E. P., Duda, J., Sheaff, C. M., Robin, A. P., & Barrett, J. A. (1989). The effect of urban trauma system hospital bypass on prehospital transport times and Level 1 trauma patient survival. Annals of Emergency Medicine, 18, 1146-1150.

Date post:	23-Dec-2015
Category:	Documents
Upload:	paulo-amaral
View:	10 times
Download:	0 times

Air Amb Evidencias

Documents