Optimizing Mass Casualty Triage: Using Discrete Event Simulation to Minimize Time to Resuscitation

Abstract

BACKGROUND: Urban areas in the US are increasingly focused on mass casualty incident (MCI) response. We simulated prehospital triage scenarios and hypothesized that using hospital-based blood product inventories for on-scene triage decisions would minimize time to treatment. STUDY DESIGN: Discrete event simulations modeled MCI casualty injury and patient flow after a simulated blast event in Boston, MA. Casualties were divided into moderate (Injury Severity Score 9 to 15) and severe (Injury Severity Score >15) based on injury patterns. Blood product inventories were collected from all hospitals (n = 6). The primary endpoint was the proportion of casualties managed with 1:1:1 balanced resuscitation in a target timeframe (moderate, 3.5 U red blood cells in 6 hours; severe, 10 U red blood cells in 1 hour). Three triage scenarios were compared, including unimpeded casualty movement to proximate hospitals (Nearest), equal distribution among hospitals (Equal), and blood product inventory–based triage (Supply-Guided). RESULTS: Simulated MCIs generated a mean ± SD of 302 ± 7 casualties, including 57 ± 2 moderate and 15 ± 2 severe casualties. Nearest triage resulted in significantly fewer overall casualties treated in the target time (55% vs Equal 86% vs Supply-Guided 91%, p < 0.001). These differences were principally due to fewer moderate casualties treated, but there was no difference among strategies for severe casualties. CONCLUSIONS: In this simulation study comparing different triage strategies, including one based on actual blood product inventories, nearest hospital triage was inferior to equal distribution or a Supply-Guided strategy. Disaster response leaders in US urban areas should consider modeling different MCI scenarios and casualty numbers to determine optimal triage strategies for their area given hospital numbers and blood product availability.