Diagnostic quality of triage algorithms for mass casualty incidents

Regarding survival and quality of life, recent mass casualty incidents have once more emphasized the importance of early identification of the correct degree of injury or illness, to enable prioritizing treatment of patients and transportation to an appropriate hospital. The present study investigated international triage algorithms in terms of sensitivity (SE) and specificity (SP) as well as the process duration in a relevant emergency patient cohort. A total of 500 consecutive air rescue missions were evaluated by means of standardized patient records. Interdisciplinary classification of patients was accomplished by 19 emergency physicians. Every case was independently classified according to the triage category by at least three physicians without considering any triage algorithm. The available triage algorithms PRIOR (Primary Ranking for Initial Orientation in Emergency Medical Services), mSTaRT (modified Simple Triage and Rapid Treatment), FTS (Field Triage Score), ASAV (Amberg-Schwandorf Algorithm for Triage), STaRT (Simple Triage and Rapid Treatment), CareFlight triage and Triage Sieve were additionally carried out for each patient in a computer-based procedure, to enable calculation of test quality criteria for all procedures. The analyzed cohort had a mean age of 59 +/- 25 years (+/- SD), a National Advisory Committee for Aeronautics (NACA) score of 3.5 +/- 1.1 and consisted of 57% men. On arrival 8 patients were already deceased, consequently 492 patients were included in the analysis. The distributions of triage categories I/II/III were 10%/47%/43%, respectively. The highest diagnostic quality was achieved with START, mSTaRT, and ASAV with 78% SE and 80-83% SP. The subgroup of surgical patients achieved 95% SE and 85-91% SP. The newly established algorithm PRIOR exerted an SE of 90% but an SP of only 54% in the overall cohort thereby taking the longest overall time for decisions. Triage procedures with acceptable diagnostic quality exist to identify the most severely injured. Due to its high rate of false positive results (overtriage) in this study, the recently developed PRIOR algorithm could result in exhaustion of available resources for the severely injured and therefore to undertreatment of correctly assigned triage category I cases within mass casualty incidents. Non-surgical patients are still poorly allocated by the available algorithms. Contribution available free of charge by aEuroFree Access".