Accuracy of Pediatric Prehospital Triage Systems in Predicting Critical Outcomes: A Systematic Review with Narrative Synthesis

Background Accurate prehospital triage of injured children is essential to ensure timely care while minimizing resource misuse. Various pediatric-specific and adapted adult triage tools (e.g. JumpSTART [3], Pediatric Triage Tape [2], MPTT-P/BCD sieve [4, 5]) have been proposed, but their ability to predict critical outcomes – including life-saving interventions (LSI), major trauma (e.g. ISS > 15), ICU admission, and mortality – remains unclear. We performed a PRISMA-compliant systematic review and meta-analysis of studies (2015–2024) evaluating the diagnostic accuracy (sensitivity, specificity, under/over-triage) of prehospital pediatric triage systems in trauma and mass-casualty settings. Methods We searched PubMed, Embase, and Scopus (2015–2024) for peer-reviewed studies of pediatric (< 18y) prehospital triage tools. Eligible studies reported triage tool performance against outcomes (LSI, ICU, ISS > 15, death). We extracted data on sensitivity, specificity, over- and under-triage rates, and outcome definitions. Methodological quality was assessed with QUADAS-2 [14]. Pooled estimates of sensitivity/specificity were calculated using random-effects models. A PRISMA flowchart (Fig. 1) summarizes study selection. Results We included 20 studies evaluating tools such as JumpSTART [3], PTT [2], Smart/START, SALT [3], the Sheffield Paediatric Triage Tool (SPTT) [5], the Battlefield Casualty Drills (BCD) sieve [4], and adapted adult tools (e.g. MPTT-24 [5]). van der Sluijs et al. (2018) reported field triage sensitivities ranging 49–87% (none reached the ideal ≥ 95%) [18]. In a UK registry, only 2 of 11 pediatric tools achieved < 5% under-triage, but with very high over-triage (72–83%) [19]. Mass-casualty simulations showed modest accuracy: SALT (59% correct, under-triage 33%) and JumpSTART (57%, under-triage 39%) performed comparably [3]. Child-focused tools often under-performed relative to adult tools: e.g. JumpSTART sensitivity was 28–35% in 0–8y olds, vs. adult START ≈ 60% [4]. Conversely, novel adaptations showed promise: the SPTT achieved 92.2% sensitivity but with 75% over-triage [5]. The adult BCD triage sieve had the highest sensitivity (75.7%) for P1 status [4], exceeding PTT performance. Conclusion Pediatric-specific triage tools like JumpSTART [3] and PTT [2] tend to under-triage, while adult-adapted tools (e.g. MPTT-24, BCD sieve, SPTT [4, 5]) improve sensitivity at the expense of higher over-triage. No tool fully meets benchmark standards [1, 8]. These findings support revising pediatric triage strategies and testing enhanced tools.