Reliability and Validity of the Emergency Severity Index for Pediatric Triage Debbie A. Travers, PhD, RN, Anna E. Waller, ScD, Jessica Katznelson, MD, and Robert Agans, PhD Abstract Objectives: The Emergency Severity Index (ESI) triage algorithm is a five-level triage acuity tool used by emergency department (ED) triage nurses to rate patients from Level 1 (most acute) to Level 5 (least acute). ESI has established reliability and validity in an all-age population, but has not been well studied for pediatric triage. This study assessed the reliability and validity of the ESI for pediatric triage at five sites. Methods: Interrater reliability was measured with weighted kappa for 40 written pediatric case scenar- ios and 100 actual patient triages at each of five research sites (independently rated by both a triage nurse and a research nurse). Validity was evaluated with a sample of 200 patients per site. The ESI rat- ings were compared with outcomes, including hospital admission, resource consumption, and ED length of stay. Results: Interrater reliability was 0.77 (95% confidence interval [CI] = 0.76 to 0.78) for the scenarios (n= 155 nurses) and 0.57 (95% CI = 0.52 to 0.62) for actual patients (n= 498 patients). Inconsistencies in triage were noted for the most acute and least acute patients, as well as those less than 1 year of age and those with medical (rather than trauma) chief complaints. For the validity cohort (n= 1,173 patients), outcomes differed by ESI level, including hospital admission, which went from 83% for Level 1 patients to 0% for Level 5 (chi-square, p < 0.0001). Nurses from dedicated pediatric EDs were 31% less likely to undertriage patients than nurses in general EDs (odds ratio [OR] = 0.31, 95% CI = 0.14 to 0.67). Conclusions: Reliability of the ESI for pediatric triage is moderate. The ESI provides a valid stratifica- tion of pediatric patients into five distinct groups. We found several areas in which nurses have difficulty triaging pediatric patients consistently. The study results are being used to develop pediatric-specific ESI educational materials to strengthen reliability and validity for pediatric triage. ACADEMIC EMERGENCY MEDICINE 2009; 16:843–849 ª 2009 by the Society for Academic Emergency Medicine Keywords: triage, pediatrics D uring emergency department (ED) triage, patients are briefly assessed by nursing staff and prioritized for treatment. A primary goal of triage is to separate patients needing immediate care from those who can wait or be assigned to less acute care areas of the ED, such as fast track. The American College of Emergency Physicians and the Emergency Nurses Association have recommended that EDs use a reliable and valid five-level triage acuity system to priori- tize patients for care in the ED. 1 The Emergency Severity Index (ESI) triage system is one such system. 2 It is a five-level triage acuity tool that is used to assign a triage level from 1 to 5 (1 is most acute and 5 is least acute). ESI users consider only the urgency of the patients’ condition for ESI Levels 1 and 2. For patients not meeting ESI 1 or 2 criteria, users predict ED resource utilization to assign levels ESI 3, 4, or 5. Multiple studies of general ED populations have shown that the ESI is a valid triage tool, with good in- terrater reliability and the ability to predict ED resource ª 2009 by the Society for Academic Emergency Medicine ISSN 1069-6563 doi: 10.1111/j.1553-2712.2009.00494.x PII ISSN 1069-6563583 843 From the Schools of Nursing (DAT) and Medicine (DAT, AEW, JK) and the Gillings School of Global Public Health (RA), Uni- versity of North Carolina at Chapel Hill, Chapel Hill, NC. Received January 15, 2009; revisions received April 17 and May 1, 2009; accepted May 8, 2009. Presented at the Society for Academic Emergency Medicine annual meeting, San Francisco, CA, May 18–21, 2006; and the American College of Emergency Physicians scientific assembly, New Orleans, LA, October 2006. DT is a member of the Emergency Severity Index (ESI) Triage Research Team, LLC, the entity that holds the copyright for the triage algorithm evaluated in this manuscript. The ESI algo- rithm, video training materials, and implementation handbook are disseminated at no charge to individual health care profes- sionals through the Agency for Healthcare Research and Quality (http://www.ahrq.gov/research/esi). DT has not received any personal compensation from use of the ESI algorithm or train- ing materials (nor has the University of North Carolina received any royalties). Address for correspondence and reprints: Debbie A. Travers, PhD, RN; e-mail: dtravers@email.unc.edu.