Major article Using electronic medical records to increase the efficiency of catheter-associated urinary tract infection surveillance for National Health and Safety Network reporting John Shepard MBA, MHA a, *, Eric Hadhazy MS b , John Frederick BA c , Spencer Nicol BA d , Padmaja Gade BS a , Andrew Cardon BA d , Jorge Wilson BA, MS a , Yohan Vetteth BA a , Sasha Madison MPH, CIC e a Department of Clinical Business Analytics, Stanford Hospital and Clinics, Stanford, CA b Department of Quality, Patient Safety, and Effectiveness, Stanford Hospital and Clinics, Stanford, CA c Department of Hospital Epidemiology, Veterans Administration, New York, NY d Health Catalyst, Salt Lake City, UT e Infection Prevention and Control Department, Stanford Hospital and Clinics, Stanford, CA Key Words: Infection control surveillance Electronic surveillance Automated surveillance CAUTI CAUTI surveillance Cost reduction Background: Streamlining health careeassociated infection surveillance is essential for health care fa- cilities owing to the continuing increases in reporting requirements. Methods: Stanford Hospital, a 583-bed adult tertiary care center, used their electronic medical record (EMR) to develop an electronic algorithm to reduce the time required to conduct catheter-associated urinary tract infection (CAUTI) surveillance in adults. The algorithm provides inclusion and exclusion criteria, using the National Healthcare Safety Network definitions, for patients with a CAUTI. The algo- rithm was validated by trained infection preventionists through complete chart review for a random sample of cultures collected during the study period, September 1, 2012, to February 28, 2013. Results: During the study period, a total of 6,379 positive urine cultures were identified. The Stanford Hospital electronic CAUTI algorithm identified 6,101 of these positive cultures (95.64%) as not a CAUTI, 191 (2.99%) as a possible CAUTI requiring further validation, and 87 (1.36%) as a definite CAUTI. Overall, use of the algorithm reduced CAUTI surveillance requirements at Stanford Hospital by 97.01%. Conclusions: The electronic algorithm proved effective in increasing the efficiency of CAUTI surveillance. The data suggest that CAUTI surveillance using the National Healthcare Safety Network definitions can be fully automated. Copyright Ó 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. State and national mandates require the reporting of select health careeassociated infections (HAIs) to the Center of Disease Control and Prevention’s (CDC) National Health Safety Network (NHSN). 1-3 Catheter-associated urinary tract infections (CAUTIs) are HAIs reported for both accreditation purposes and regulatory re- quirements. 4-6 CAUTIs are considered the most common HAI in the United States, associated with increased health care costs and mortality, 7 and identifying CAUTIs is costly and time-consuming for infection prevention departments. 1,6,8-11 Given the time and costs associated with HAI surveillance, many facilities are looking to their electronic medical record (EMR) to provide a more efficient process. The purpose of this study was to develop an accurate and completely automated electronic algo- rithm, using our EPIC 2012 EMR, which eliminates the cost of CAUTI surveillance. The primary study outcome was the number of urine cultures in which the electronic algorithm could automatically identify the presence or absence of a CAUTI. METHODS Stanford Hospital, a 583-bed tertiary care center, assigned a multidisciplinary team of front-line providers, infection pre- ventionists, and clinical informaticists to create an automated electronic algorithm and data reporting system to increase the * Address correspondence to John Shepard, MBA, MHA, 940 Oakes St, East Palo Alto, CA 94303. E-mail address: [email protected](J. Shepard). This study was funded through hospital quality improvement initiatives. Conflict of interest: None to report. Contents lists available at ScienceDirect American Journal of Infection Control journal homepage: www.ajicjournal.org American Journal of Infection Control 0196-6553/$36.00 - Copyright Ó 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.ajic.2013.12.005 American Journal of Infection Control 42 (2014) e33-e36
Transcript
lable at ScienceDirect
American Journal of Infection Control 42 (2014) e33-e36
Contents lists avai
American Journal of Infection Control
journal homepage: www.aj ic journal .org
American Journal of Infection Control
Major article
Using electronic medical records to increase the efficiencyof catheter-associated urinary tract infection surveillancefor National Health and Safety Network reporting
John Shepard MBA, MHA a,*, Eric Hadhazy MS b, John Frederick BA c, Spencer Nicol BA d,Padmaja Gade BS a, Andrew Cardon BAd, Jorge Wilson BA, MS a, Yohan Vetteth BA a,Sasha Madison MPH, CIC e
aDepartment of Clinical Business Analytics, Stanford Hospital and Clinics, Stanford, CAbDepartment of Quality, Patient Safety, and Effectiveness, Stanford Hospital and Clinics, Stanford, CAcDepartment of Hospital Epidemiology, Veterans Administration, New York, NYdHealth Catalyst, Salt Lake City, UTe Infection Prevention and Control Department, Stanford Hospital and Clinics, Stanford, CA
Key Words:Infection control surveillanceElectronic surveillanceAutomated surveillanceCAUTICAUTI surveillanceCost reduction
* Address correspondence to John Shepard, MBA, MAlto, CA 94303.
E-mail address: [email protected] (J. ShepardThis study was funded through hospital quality iConflict of interest: None to report.
0196-6553/$36.00 - Copyright � 2014 by the Associahttp://dx.doi.org/10.1016/j.ajic.2013.12.005
Background: Streamlining health careeassociated infection surveillance is essential for health care fa-cilities owing to the continuing increases in reporting requirements.Methods: Stanford Hospital, a 583-bed adult tertiary care center, used their electronic medical record(EMR) to develop an electronic algorithm to reduce the time required to conduct catheter-associatedurinary tract infection (CAUTI) surveillance in adults. The algorithm provides inclusion and exclusioncriteria, using the National Healthcare Safety Network definitions, for patients with a CAUTI. The algo-rithm was validated by trained infection preventionists through complete chart review for a randomsample of cultures collected during the study period, September 1, 2012, to February 28, 2013.Results: During the study period, a total of 6,379 positive urine cultures were identified. The StanfordHospital electronic CAUTI algorithm identified 6,101 of these positive cultures (95.64%) as not a CAUTI,191 (2.99%) as a possible CAUTI requiring further validation, and 87 (1.36%) as a definite CAUTI. Overall,use of the algorithm reduced CAUTI surveillance requirements at Stanford Hospital by 97.01%.Conclusions: The electronic algorithm proved effective in increasing the efficiency of CAUTI surveillance.The data suggest that CAUTI surveillance using the National Healthcare Safety Network definitions can befully automated.
Copyright � 2014 by the Association for Professionals in Infection Control and Epidemiology, Inc.Published by Elsevier Inc. All rights reserved.
State and national mandates require the reporting of selecthealth careeassociated infections (HAIs) to the Center of DiseaseControl and Prevention’s (CDC) National Health Safety Network(NHSN).1-3 Catheter-associated urinary tract infections (CAUTIs) areHAIs reported for both accreditation purposes and regulatory re-quirements.4-6 CAUTIs are considered the most common HAI in theUnited States, associated with increased health care costs andmortality,7 and identifying CAUTIs is costly and time-consuming forinfection prevention departments.1,6,8-11
HA, 940 Oakes St, East Palo
).mprovement initiatives.
tion for Professionals in Infection C
Given the time and costs associated with HAI surveillance, manyfacilities are looking to their electronic medical record (EMR) toprovide a more efficient process. The purpose of this study was todevelop an accurate and completely automated electronic algo-rithm, using our EPIC 2012 EMR, which eliminates the cost of CAUTIsurveillance. The primary study outcome was the number of urinecultures in which the electronic algorithm could automaticallyidentify the presence or absence of a CAUTI.
METHODS
Stanford Hospital, a 583-bed tertiary care center, assigned amultidisciplinary team of front-line providers, infection pre-ventionists, and clinical informaticists to create an automatedelectronic algorithm and data reporting system to increase the
ontrol and Epidemiology, Inc. Published by Elsevier Inc. All rights reserved.
J. Shepard et al. / American Journal of Infection Control 42 (2014) e33-e36e34
J. Shepard et al. / American Journal of Infection Control 42 (2014) e33-e36 e35
efficiency of CAUTI surveillance and reporting. Using the CDC def-initions published in January 2013,12 the team designed an algo-rithm that provides inclusion and exclusion criteria for patientsmeeting the criteria for a CAUTI. The team visually displayed thealgorithm logic using Microsoft PowerPoint, and the algorithmwascoded in Microsoft Management Studio, using standard querylanguage (SQL) to extract the data from the Epic 2012 Claritydatabase. Approximately 10 iterations of the algorithm designwereperformed before the algorithm was deemed suitable for testing.
The team identified that all patients with a CAUTI must have apositive urine culture, and that it is possible to have more than 1CAUTI in a single patient admission. Given these parameters, theautomated electronic algorithmwas designed to review all positiveurine cultures collected in the hospital and to assign each urineculture to 1 of 2 categories, definite CAUTI or not a CAUTI. Unfor-tunately, the Epic 2012 EMR used at Stanford Hospital could notcapture all of the data required for a completely automated algo-rithm. The EMR did not capture electronically if a patient had uri-nary urgency with no other recognized cause except urinary tractinfection (UTI), frequency with no other recognized cause exceptthe UTI, dysuria with no other recognized cause except the UTI,suprapubic tenderness with no other recognized cause except theUTI, or costovertebral angle pain or tenderness. These data ele-ments are required for the use of a completely automated algorithmfor CAUTI surveillance.12
Given the limitations of the data, the team tailored the auto-mated algorithm to be compatible with the Stanford Hospital EMR.For this study, the revised algorithmdthe Stanford Hospital Elec-tronic CAUTI Algorithm (SHECA)dwas used to review all of thepositive urine cultures collected in the hospital between September1, 2012, and February 28, 2013. On review, the SHECA placed eachpositive urine culture in 1 of 3 categories: definite CAUTI, not aCAUTI, and possible CAUTI. The cultures deemed a possible CAUTIwere those cultures that infection preventionists at Stanford Hos-pital still needed to conduct a chart review on, because the SHECAdid not have sufficient data to identify the culture as a definiteCAUTI or not a CAUTI. These results were then used to evaluate thechange in the efficiency of CAUTI surveillance from use of theSHECA.
To validate the algorithm, 2 clinical informaticists conducted aretrospective electronic chart review to identify all positive urinecultures reported by the Stanford Hospital’s laboratory betweenSeptember 1, 2012, and February 28, 2013. A random sample ofapproximately 150 positive urine cultures was then blindlyreviewed by both the SHECA and 2 infection preventionists toidentify whether they met the NHSN definition of a CAUTI. Theinfection preventionists and clinical informaticists then reviewed arandom sample of positive urine cultures from every algorithmscore, to ensure that each patient had the appropriate algorithmscore and that the algorithm score was accurately designating thepatient as having or not having a CAUTI. No discrepancies in resultsfrom the SHECA and the validation teamwere identified. This studywas ruled exempt by the internal review board. Details of the al-gorithm design are available on request.
RESULTS
Using the SHECA, we retrospectively analyzed all positive urinecultures collected at Stanford Hospital between September 1, 2012,and February 28, 2013. During this period, 6,379 positive urinecultures were recorded; of these, the SHECA identified 6,101(95.64%) as not a CAUTI because the positive culture did not meetthe NHSN criteria; 191 (2.99%) as a possible CAUTI because therewere insufficient data to verify whether the culture was or was nota CAUTI; and 87 (1.36%) as a definite CAUTI (Fig 1). Overall, the
electronic algorithm reduced CAUTI surveillance requirements atStanford Hospital by 97.01%.
DISCUSSION
Infection prevention surveillance is a very time-consuming andresource-intensive process. It is estimated that 45% of infectionpreventionists’ time is spent on surveillance and analysis.13 Withthe ever-increasing reporting requirements for infection preven-tion departments2,14,15 comes the need to maximize the efficiencyof surveillance and analysis. As a result of the mass adoption ofEMRs in recent years, hospitals now have the ability to leveragetheir data to increase the efficiency of HAI surveillance.
During the course of this study, we made minor changes to ourEpic 2012 EMR, but we were not able to make all of the changesneeded to implement the automated surveillance algorithm. Forthis reason, we were able to reduce the level of CAUTI surveillanceonly by approximately 97%. However, our data suggest that healthcare facilities could reconfigure their EMRs to implement acompletely automated electronic algorithm for CAUTI surveillance.
Limitations of this study include the level of configurability inour facility’s EMR. To fully test the reliability of our automated al-gorithm, we would need to implement the algorithm at anotherfacility that has an appropriately configured EMR. Given this limi-tation, we were unable to verify whether our algorithm couldcompletely automate CAUTI surveillance. Future studies shouldinvestigate the development and implementation of similar algo-rithms for other HAIs.
CONCLUSION
Our study suggests that health care facilities, regardless of theEMR in use, available data, or available funds, can effectivelyleverage their current EMR infrastructure by investing in low-costsolutions to greatly increase the efficiency and effectiveness ofHAI surveillance. The health care environment is requiring hospi-tals to reduce costs and improve quality; infection prevention de-partments are focused on achieving both. However, growingsurveillance requirements hinders infection preventionists fromdevoting time to valuable clinical interventions. The design pro-vided by our team allows for facilities of all levels to increase theefficiency of HAI surveillance, freeing up more time for imple-menting much-needed quality interventions.
References
1. Stevenson KB, Khan Y, Dickman J, Gillenwater T, Kulich P, Myers C, et al.Administrative coding data, compared with CDC/NHSN criteria, are poor in-dicators of health care-associated infections. Am J Infect Control 2008;36:155-64.
2. Julian KG, Brumbach AM, Chicora MK, Houlihan C, Riddle AM, Umberger T, et al.First year of mandatory reporting of healthcare-associated infections, Penn-sylvania: an infection control-chart abstractor collaboration. Infect ControlHosp Epidemiol 2006;27:926-30.
3. Reagan J, Hacker C. Laws pertaining to healthcare-associated infections: a re-view of 3 legal requirements. Infect Control Hospital Epidemiol 2012;33:75-80.
4. Meddings JA, Reichert H, Rogers MA, Saint S, Stephansky J, McMahon LF. Effectof nonpayment for hospital-acquired, catheter-associated urinary tract infec-tion: a statewide analysis. Ann Intern Med 2012;157:305-12.
5. Gould CV, Umscheid CA, Agarwal RK, Kuntz G, Pegues DA. Healthcare InfectionControl Practices Advisory Committee. Guideline for prevention of catheter-associated urinary tract infections 2009. Infect Control Hosp Epidemiol 2010;31:319-26.
7. Klevens RM, Edwards JR, Richards CL Jr, Horan TC, Gaynes RP, Pollock DA, et al.Estimating health careeassociated infections and deaths in US hospitals, 2002.Public Health Rep 2007;122:160-6.
J. Shepard et al. / American Journal of Infection Control 42 (2014) e33-e36e36
8. Sherman ER, Heydon KH, St John KH, Teszner E, Rettig SL, Alexander SK, et al.Administrative data fail to accurately identify cases of healthcare-associatedinfection. Infect Control Hosp Epidemiol 2006;27:332-7.
9. Tambyah PA, Knasinski V, Maki DG. The direct costs of nosocomial catheter-associated urinary tract infection in the era of managed care. Infect ControlHosp Epidemiol 2002;23:27-31.
10. Haley RW, Culver DH, White JW, Morgan WM, Emori TG, Munn VP, et al.The efficacy of infection surveillance and control programs in pre-venting nosocomial infections in US hospitals. Am J Epidemiol 1985;121:182-205.
11. Wright SB, Huskins WC, Dokholyan RS, Goldmann DA, Platt R. Administrativedatabases provide inaccurate data for surveillance of long-term centralvenous catheter-associated infections. Infect Control Hosp Epidemiol 2003;24:946-9.
12. Centers for Disease Control and Prevention. April 2013 CDC/NHSN protocolcorrections, clarification, and additions: catheter-associated urinary tractinfection (CAUTI) event. Available from: http://www.cdc.gov/nhsn/pdfs/pscManual/7pscCAUTIcurrent.pdf. Accessed January 7, 2014.
13. Stone PW, Dick A, Pogorzelska M, Horan TC, Furuya EY, Larson E. Staffing andstructure of infection prevention and control programs. Am J Infect Control2009;37:351-7.
14. McKibben L, Horan TC, Tokars JI, Fowler G, Cardo DM, Pearson ML, et al.Guidance on public reporting of healthcare-associated infections: recommen-dations of the Healthcare Infection Control Practices Advisory Committee.Infect Control Hosp Epidemiol 2005;26:580-7.
15. Stone PW, Horan TC, Shih HC, Mooney-Kane C, Larson E. Comparisons of healthcareeassociated infections identification using two mechanisms for publicreporting. Am J Infect Control 2007;35:145-9.
