Table 2.

Performance characteristics of prehospital triage tools from a subset of the highest quality articles

Type	Study	Tool name	Pop	Primary research question	Major findings
General	Meisel 2009	PEAR	A	To validate PEAR for predicting hospital admission using routinely collected out-of-hospital information	AUC for combined cohort was 0.83 for all admissions and 0.72 for ICU admissions. n = 1102
	Hoikka 2018	NEWS	All	To examine the accuracy of the prehospitally implemented NEWS in predicting 1-day and 30-day mortalities in an unselected EMS population	The high-risk NEWS group (score ≥ 7) had sensitivities for 1-day and 30-day mortalities of 0.801 (CI 0.74–0.86) and 0.42 (CI 0.38–0.47), respectively. n = 12,426
	Leeies 2017	CTAS	A	To prospectively evaluate CTAS interrater reliability between EMS providers and ED triage nurses	Interrater reliability κw = 0.437 (p < 0.001, 95% CI 0.421–0.452). n = 14,378
	Magnusson 2019	RETTS-p	P	To evaluate agreement between the EMS field assessment using RETTS-p and the hospital diagnosis of emergent condition	Sn 66.7% and Sp 67.0%, with under-triage rate 33%, over-triage rate 33.3%. n = 716
Stroke	Helwig 2019	LAMS	A	To compare LAMS to Mobile Stroke Unit (MSU) in accurately triaging patients to the appropriate stroke hospital (CSC vs PSC)	An accurate triage decision was reached for 69.8% in the LAMS group and 100% in the MSU group (difference, 30.2%; 95% CI, 17.8%-42.5%; P < 0.001). n = 116
	Carrera 2019	RACE	A	To revalidate RACE after its region-wide implementation in Catalonia	RACE ≥ 5 showed Sn 84%, Sp 60%, AUC 0.77, for detecting LVO. N = 1822
	Jumaa 2019	RACE	A	To report performance characteristics of RACE for LVO eligible for mechanical thrombectomy	A RACE cut-off point of ≥ 5 had Sn 77%, Sp 75%, PPV 0.97, NPV 0.25, accuracy 75.3% (95%CI 73.1–77.4). n = 1147
Trauma triage	Brown 2011	FTDS	A	To analyze whether trauma center need was accurately predicted solely by the physiologic (PHY) and anatomic (ANA) criteria of the FTDS	Application of only the PHY and ANA criteria identifies trauma center need with Sn 49%, Sp 78% and undertriage rate of 51%. Mechanism of injury and special considerations criteria play an important role in minimizing under-triage rates. n = 1,086,764
	Newgard 2011	FTDS	All	To evaluate diagnostic performance of FTDS for identifying major trauma (Injury Severity Score [ISS] ≥ 16)	Sn 85.8% (95% CI 85.0–86.6) and Sp 68.7% (95% CI 68.4–68.9). n = 122,345
	Barnett 2013	FTDS	All	To describe the use of field triage criteria by EMS personnel in the Western United States	The three most common criteria cited (of 33 in use) were EMS provider judgment, age < 5 or > 55 years, and GCS < 14. n = 46,414
	Davidson 2014	FTDS	All	To determine the likelihood of serious trauma based on vehicle damage sustained in a crash as described in Step 3 of the FTDS	Crash characteristics that predict severe injury included intrusion of greater than 12 inches (PPV of 10.4%; 95% CI, 9.5–11.3) and steering wheel collapse (PPV of 25.7%; 95% CI, 23.0–28.4%). n = 85,761
	Lerner 2017	FTDS	P	To determine the change in under- and over-triage rates when the 2011 Field Triage Guidelines are compared to the 2006 and 1999 versions	Applying the 1999, 2006, or 2011 Guidelines to the EMS interview data the over-triage rate was 32.6%, 27.9%, and 28.0%, respectively. The under-triage rate was 26.5%, 35.1%, and 34.8%, respectively. The 2011 Guidelines resulted in an 8.2% (95% CI 0.6–15.9%) absolute increase in under-triage and a 4.6% (95% CI 2.8–6.3%) decrease in over-triage compared to 1999 Guidelines. n = 5594
	Ardolino 2015	PTS	P	To assess performance of English pediatric prehospital trauma triage tools	East Midlands (18%), North West (21%) and Northern (19%) tools had the best over-triage rates. All had under-triage rates of 0%. n = 2934
	Cox 2012	VSTTC	A	To evaluate performance of the Victorian prehospital trauma triage criteria in discriminating for major trauma	Sn 95.3%, Sp 62.7%, under-triage rate 4.7%, and over-triage rate 37.3% for major trauma. n = 45,332
	vanLaarhoven 2014	Dutch	A	To evaluate the protocol's ability to identify severely injured adult trauma patients (ISS >  = 16)	Sn 89.1%, Sp 60.5%, PPV 26.5%, NPV 97.2%, undertriage rate 10.9%, overtriage rate 39.5%. n = 1607
Trauma for helicopter EMS	Brown 2012	FTDS	All	To determine which FTDS criteria can be used by field EMS providers to predict which trauma patients would benefit from helicopter transport	Odds of increased survival to discharge by helicopter transport found in following conditions: GCS < 14 (aOR 1.22); respiratory rate < 10 or > 29 (aOR 1.32); penetrating injury (aOR 1.40), age > 55 (aoR 1.15). n = 258,387
	Brown 2017	AMPT	A	To validate the effectiveness of the AMPT score to identify patients with a survival benefit from helicopter EMS (HEMS)	For AMPT score ≥ 2, HEMS increases odds of in-hospital survival by 6.7% (ARR 1.067; 95% CI 1.040–1.083, p < 0.001). n = 222,827
Traumatic brain injury	Fuller 2016	HITS-NS	A	To determine the accuracy of the HITS-NS triage rule for identifying patients with significant TBI	Sn 28.3% and Sp 94.4% for significant TBI. n = 3828

PEAR, Philadelphia EMS Admission Rule; NEWS, National Early Warning Score; CTAS, Canadian Triage and Acuity Scale; RETTS-p, Rapid Emergency Triage and Treatment System-pediatrics; LAMS, Los Angeles Motor Scale; RACE, Rapid Arterial Occlusion Evaluation; FTDS, Field Triage Decision Scheme; PTS, Pediatric trauma score, pediatric triage tape, East Midlands, London, Northwest, Northern, Southwest London, Wessex tools; VSTTC, Victorian state prehospital trauma triage criteria; Dutch, Dutch Field Triage Protocol; AMPT, Air Medical Prehospital Triage; HITS-NS, Head Injury Transportation Straight to Neurosurgery; Pop, population; A, Adult; P-Pediatric; GCS, Glasgow Coma Scale; Sn, sensitivity; Sp, specificity; AUC, area under the curve; CI, confidence interval; aOR, adjusted odds ratio; PPV, positive predictive value; NPV, negative predictive value