Table 4.
Characteristics of studies included in the update (see Table S4, online resource 1 for details)
| Study, ref, design, LoE, comments | Activation criterion / comparison, N, main results |
|---|---|
| Vital signs | |
| Systolic blood pressure | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
SBP < 90 mmHg (N = 11, 212) • Accurate prediction of mortality: 29.6% |
| Brown 2016 [26], prognostic cross-sectional study, LoE: 2b |
Step 1 or Step 2 NTTP criteriaausing SBP < 90 mmHg vs. Step 1 or Step 2 NTTP criteria using SBP < 110 mmHg (N = 1,555,944 overallb) • Undertriagec reduction by substituting an SBP < 110 mmHg: 4.4% for geriatric cohort; 4.3% for adult cohort • Overtriagec increase by substituting an SBP < 110 mmHg: 4.3% for geriatric cohort; 5.3% for adult cohort |
| Damme 2016 [28], prognostic cross-sectional study, LoE: 2b |
SBP ≤ 110 mmHg (N = 81) vs. SBP > 110 mmHg (N = 206) • Significantly more ICU admissions, longer ICU length of stay, more ventilator days, more packed red blood cells and higher ISS for SBP ≤ 110 mmHg |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b, underpowered |
SBP < 90 mmHg (N = 9) • Accurate prediction of ISS > 15: 56% • Accurate prediction of need for emergency procedurec: 44% |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
SBP per 5 mmHg, (N = 387 overallb) • Lower need for resuscitative care (not significant) |
| Hasler 2011 [32], prognostic cross-sectional study, LoE: 2b |
Comparison of SBP intervals in blunt major trauma patients (N = 47,927 overallb) • Significant increase in mortality in patients with SBP < 110 mmHg, linear trend |
| Hasler 2012 [33], prognostic registry study, LoE: 2b |
Comparison of SBP intervals in penetrating major trauma patients (N = 3444 overall b ) • Significant increase in mortality in patients with SBP < 110 mmHg, significant linear trend |
| Hranjec 2012 [35], prognostic cross-sectional study, LoE: 2b |
Comparison of SBP intervals (N = 57,973 overallb) • Significantly higher mortality for SBP 0–60 mmHg and 60–90 mmHg vs. 90–120 mmHg (reference) • Significantly lower mortality for SBP 120–150 mmHg and 150–180 mmHg vs. 90–120 mmHg (reference) |
| Singh 2014 [42], prognostic cross-sectional study, LoE: 2b |
SBP < 90 mmHg vs. SBP > 90 mmHg (N = 9860 overall b ) • Significantly higher mortality for SBP < 90 mmHg |
| Tignanelli 2018 [44], prognostic cross-sectional study and comparative registry study, LoE: 2b |
SBP ≤ 90 mmHg (N = 1346) • Accurate prediction of need for emergency procedurec: 63% |
| Diastolic blood pressure | |
| Singh 2014 [42], prognostic cross-sectional study, LoE: 2b |
DBP < 60 mmHg vs. DBP > 60 mmHg (N = 9860 overall b ) • Significantly higher mortality for DBP < 60 mmHg |
| Heart rate | |
| Singh 2014 [42], prognostic cross-sectional study, LoE: 2b |
HR > 120 bpm vs. HR < 120 bpm (N = 9860 overall b ) • Significantly higher mortality for HR > 120 bpm |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b, underpowered |
HR > 130 bpm (N = 3) • Accurate prediction of ISS > 15: 0% • Accurate prediction of need for emergency procedurec: 67% |
| Airway obstruction, stridor | |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Airway obstruction, stridor (N = 4) • Accurate prediction of ISS > 15: 50% • Accurate prediction of need for emergency procedurec: 75% |
| Respiratory rate | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Respiratory rate < 9 or > 29/min (N = 3207) • Accurate prediction of mortality: 45.3% |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Respiratory rate > 30/min (N = 14) • Accurate prediction of ISS > 15: 71% • Accurate prediction of need for emergency procedurec: 21% |
| Saturation of peripheral oxygen | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Saturation of peripheral oxygen < 90% (N = 9484) • Accurate prediction of mortality: 31.5% |
| Glasgow Coma Scale | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
GCS score < 9 (N = 15,099) • Accurate prediction of mortality: 37.5% Drop in GCS of 2 points or more (N = 3706) • Accurate prediction of mortality: 12.9% |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
GCS score < 13 (N = 87) • Accurate prediction of ISS > 15: 38% • Accurate prediction of need for emergency procedurec: 32% |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
Initial GCS score per increment of 1 (N = 387 overall b ) • No detectable difference in need for resuscitative carec |
| Hranjec 2012 [35], prognostic cross-sectional study, LoE: 2b |
Comparison of different motor GCS scores (N = 57,973 overalllb) • Significantly higher mortality for motor GCS scores of 1 and 2–5 compared to 6 |
| Tignanelli 2018 [44], prognostic cross-sectional study and comparative registry study, LoE: 2b |
GCS score < 9 (N = 2475) • Accurate prediction of need for emergency procedurec: 92% |
| Body temperature | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Hypothermia < 35 °C (N = 3040) • Accurate prediction of mortality: 28.9% |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Hypothermia (core temperature < 32 °C) (N = 11) • Accurate prediction of ISS > 15: 27% • Accurate prediction of need for emergency procedurec: 18% |
| Hranjec 2012 [35], prognostic cross-sectional study, LoE: 2b |
Comparison of temperatures (N = 57,973 overalllb) • Significantly higher mortality for temperatures between 65-97.7 °[F] vs. 97.7-101.3 °[F] • Higher mortality for temperatures above 101.3 °[F] compared to 97.7-101.3 °[F] (not significant) |
| Shock index | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Shock index > 0.9 (N = 17,720) • Accurate prediction of mortality: 17.9% |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
Shock index per increment of 0.1 (N = 387 overall b ) • Significantly higher need for resuscitative carec |
| Singh 2014 [42], prognostic cross-sectional study, LoE: 2b |
Shock index < 0.5 vs. >0.5 • No detectable difference for mortality Shock index < 0.9 vs. >0.9 • No detectable difference for mortality |
| Modified shock index | |
| Singh 2014 [42], prognostic cross-sectional study, LoE: 2b |
Modified shock index < 0.7 vs. >0.7 • Significantly higher mortality for modified shock index < 0.7 Modified shock index < 1.3 vs. >1.3 • Significantly higher mortality for modified shock index > 1.3 |
| Lactate | |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
Point-of-care lactate ≥ 2.5 mmol/L (N = 228) Need for resuscitative carec associated with a 1-mmol/L difference in point-of-care lactates: • Higher within the range of < 2.5 (not significant) • Significantly higher within the range of 2.5–3.9 • Not detectable within the range of ≥ 4.0 |
| St. John 2018 [43], secondary analysis of a prognostic cross-sectional study, LoE: 3b↓ |
Prehospital lactate (N = 314 overall b ) Need for resuscitative carec associated with a 1 mmol/L difference in prehospital lactate concentration: • Higher within the range of < 2.5 (not significant) • Significantly higher within the range of 2.5-4.0 • Significantly higher within the range of ≥ 4.0 |
| Type and extent of injuries | |
|---|---|
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Flail chest (N = 2) • Accurate prediction of ISS > 15: 50% • Accurate prediction of need for emergency procedurec: 0% Unstable fracture of the pelvis / fracture in two or more long bones (N = 5) • Accurate prediction of ISS > 15: 40% • Accurate prediction of need for emergency procedurec: 0% Injury in two or more body regions (head/neck/chest/abdomen/pelvis/femur/back) (N = 61) • Accurate prediction of ISS > 15: 15% • Accurate prediction of need for emergency procedurec: 13% Paralysis (N = 10) • Accurate prediction of ISS > 15: 80% • Accurate prediction of need for emergency procedurec: 10% Penetrating injury of the head/neck/chest/abdomen/pelvis/groin/back (N = 5) • Accurate prediction of ISS > 15: 0% • Accurate prediction of need for emergency procedurec: 60% Second-degree or third-degree burn injury > 15% of body surface (N = 5) • Accurate prediction of ISS > 15: 40% • Accurate prediction of need for emergency procedurec: 60% Burn injury with inhalation injury (N = 5) • Accurate prediction of ISS > 15: 40% • Accurate prediction of need for emergency procedurec: 40% |
| Lin 2012 [39], prognostic cross-sectional study, LoE: 2b |
Two or more long bone fractures (humerus, radius, ulna, femur, tibia, fibula) (N = 37) • Overtriagec: 29.7% • Significant prediction of ISS ≥ 25 • Significant prediction of need for an emergency operation |
| Tignanelli 2018 [44], prognostic cross-sectional study and comparative registry study, LoE: 2b |
Central gunshot wound (N = 1931) • Accurate prediction of need for emergency procedurec: 67% |
| Mechanism of injury | |
|---|---|
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Ejected from vehicle (N = 6) • Accurate prediction of ISS > 15: 67% • Accurate prediction of need for emergency procedurec: 0% Death of another passenger in the vehicle (N = 5) • Accurate prediction of ISS > 15: 40% • Accurate prediction of need for emergency procedurec: 0% Trapped in wreck (N = 9) • Accurate prediction of ISS > 15: 33% • Accurate prediction of need for emergency procedurec: 11% Pedestrian or cyclist hit by motor vehicle (N = 15) • Accurate prediction of ISS > 15: 13% • Accurate prediction of need for emergency procedurec: 13% Fall from > 5 m (N = 20) • Accurate prediction of ISS > 15: 50% • Accurate prediction of need for emergency procedurec: 15% Avalanche accident (N = 1) • Accurate prediction of ISS > 15: 0% • Accurate prediction of need for emergency procedurec: 0% |
| Matsushima 2016 [40], prognostic cross-sectional study, LoE: 2b |
Motor vehicle intrusion (MVI) and age ≥ 65 years (N = 288) • Significantly higher mortality for MVI and age ≥ 65 years Motor vehicle intrusion (MVI) and male sex (N = 2259) • No significantly higher mortality for MVI and male sex Motor vehicle intrusion (MVI) and no airbag deployment (N = 300) • No significantly lower mortality for MVI and no airbag deployment Motor vehicle intrusion (MVI) and use of seat belt (N = 3254) • No significantly lower mortality for MVI and use of seat belt Motor vehicle intrusion (MVI) and heart rate > 100 bpm (N = 1175) • Significantly higher mortality for MVI and heart rate > 100 bpm Motor vehicle intrusion (MVI) and SBP < 110 mmHg (N = 251) • Significantly higher mortality for MVI and SBP < 110 mmHg |
| Interventions | |
|---|---|
| Airway assistance | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Advanced airway management (N = 22,771) • Accurate prediction of mortality: 27.0% |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
Any airway / bag valve mask attempted (N = 387overall b ) • Significantly higher need for resuscitative carec |
| Hranjec 2012 [35], prognostic cross-sectional study, LoE: 2b |
Mechanical ventilation (N = 57,973 overallb) • Significantly higher mortality in ventilated patients |
| Lin 2012 [39], prognostic cross-sectional study, LoE: 2b |
Active airway assistance beyond supplemental O 2 (N = 40) • Overtriagec: 15% • Significant prediction of ISS ≥ 25 • Significant prediction of emergency operation |
| Tignanelli 2018 [44], prognostic cross-sectional study and comparative registry study, LoE: 2b |
Intubation (N = 3459) • Accurate prediction of need for emergency procedurec: 100% |
| Other interventions | |
| Bieler 2021 [25], prognostic cross-sectional study, LoE: 2b |
Cardiopulmonary resuscitation (N = 3162) • Accurate prediction of mortality: 76.2% Insertion of a chest tube (N = 8823) • Accurate prediction of mortality: 23.0% Administration of catecholamines (N = 13,150) • Accurate prediction of mortality: 35.7% |
| Combined criteria | |
|---|---|
| Lin 2012 [39], prognostic cross-sectional study, LoE: 2b |
Consciousness: BMR < 5 or paralysis or suspicion of spinal cord injury or loss of sensation or GCS score ≤12 (N = 128) • Overtriagec: 29.7% • Significant prediction of ISS ≥ 25 • Significant prediction of need for an emergency operation Circulation: no radial pulse and sustained heart rate ≥ 120 bpm, or SBP ≤ 90 mmHg (N = 63) • Overtriagec: 31.7% • Significant prediction of ISS ≥ 25 • Significant prediction of need for an emergency operation Cutaneous: deep penetrating injury to head, neck & torso, amputation at or proximal to wrist or ankle (N = 139) • Overtriagec: 52.5% • No significant prediction of ISS ≥ 25 • Significant prediction of need for an emergency operationc Two or more level two criteria met (age > 55 years, respiratory rate ≥ 30/min, BMR = 5, sustained heart rate = 120 bpm, long bone fracture sustained in a motor vehicle collision or fall ≥ 10 feet, major degloving injury, or major flap avulsion > 5 inches, or gunshot wound to the extremities) (N = 44) • Overtriagec: 72.7% • No significant prediction of ISS ≥ 25 |
| Dehli 2016 [30], prognostic cross-sectional study, LoE: 3b↓, underpowered |
Traumatic amputation or crush injury above wrist/ankle (N = 1) • Accurate prediction of ISS > 15: 0% • Accurate prediction of need for emergency procedurec: 100% |
| Cull 2019 [27], prognostic cross-sectional study, LoE: 2b |
Trauma activation prediction models (mathematical equations including SBP, PR, RR, GCS) for falls, gunshot wounds and stab wounds, (N = 157,164 overallb) • Accurate prediction of trauma activation level: approximately 52% (gunshot wounds) 59% (falls) 75% (stab wounds) |
| Guyette 2015 [31], prognostic cross-sectional study, LoE: 2b |
Point-of-care lactate ≥ 2.5 mmol/L and SBP 91–100 mmHg (N = 93) • Accurate prediction of need for resuscitative carec: 22.6% |
| Heindl 2021 [34], prognostic cross-sectional study, LoE: 2b |
GoR A criteria (N = 32) vs. GoR B criteria (N = 84) (according to the German Polytrauma Guideline of 2016) vs. GoR 0 criteria (N = 48) (TTA based on the emergency physician’s assessment alone) • Significantly higher mortality for GoR A criteria • More emergency interventions for GoR A criteria |
| Kalkwarf 2021 [37], prognostic cross-sectional study, LoE: 3b↓, post-hoc analysis |
≥ 2 ABC criteria (N = 25) (penetrating trauma, heart rate > 120 bpm, SBP < 90 mmHg, positive abdominal FAST) vs. <2 ABC criteria (N = 266) • Significantly higher mortality with ≥ 2 criteria met |
| Lehmann 2009 [38], prognostic cross-sectional study, LoE: 2b |
Institution’s current triage system based on three steps (vital signs and level of consciousness, anatomy of injury, biomechanics of injury and other risk factors) vs. simplified triage protocol using four variables (SBP < 100 mmHg, GCS < 14, altered respirations, penetrating truncal injury) (N = 244 overallb) • Significantly better prediction of need for emergency intervention with simplified criteria; significant differences between different steps of current criteria • Negative predictive value of need for emergency intervention: 99.6% with current system, 96% with simplified protocol • Positive predictive value of 21% with current system and 58% with simplified protocol |
| Shawhan 2015 [41], prognostic cross-sectional study, LoE: 2b |
Level 1 d (N = 89) vs. level 2 e (N = 146) activation criteria vs. level 3 (trauma consultation) (N = 225) • Better prediction of patients requiring ICU admission for level 1 compared to level 2 • Better prediction of patients requiring urgent intervention for level 1 compared to levels 2 or 3 |
| Adult vs. geriatric patients / comparison of age groups | |
|---|---|
| Brown 2016 [26], prognostic cross-sectional study, LoE: 2b |
Physiologic Step 1 or anatomic Step 2 NTTP criteria using SBP < 90 mmHg vs. physiologic Step 1 or anatomic Step 2 NTTP criteria using SBP < 110 mmHg, (N = 1,555,944 overalllb) • Undertriagec reduction by substituting an SBP < 110 mmHg: 4.4% for geriatric cohort; 4.3% for adult cohort • Overtriagec increase by substituting an SBP < 110 mmHg: 4.3% for geriatric cohort; 5.3% for adult cohort |
| Hranjec 2012 [35], prognostic cross-sectional study, LoE: 2b |
Comparison of different age groups (N = 57,973 overallb) • Significantly higher mortality for > 85 years vs. 65 years (reference) • Significantly higher mortality for 65–75 years vs. 65 years (reference) |
| Ichwan 2014 [36], prognostic cross-sectional study, LoE: 2b |
Standard adult triage criteriafvs. geriatric triage criteriag(age ≥ 70) (N = 101,577 overallb): • Geriatric triage criteria more sensitive for prediction of ISS > 15 compared to standard adult triage criteria for both, adults and geriatrics • Geriatric triage criteria less specific for prediction of ISS > 15 compared to standard adult triage criteria for both, adults and geriatrics |
| Wermann 2011 [45], prognostic cross-sectional study, LoE: 3b↓, indirectness |
GCS = 14 (geriatric patients > 70 years) vs. GCS = 13 (adult patients) (N = 90,597 overallb): • Significantly higher mortality in geriatric patients SBP 91–100 mmHg (geriatric patients > 70 years) vs. SBP 81–90 mmHg (adult patients): • Trend towards comparable mortality Significantly higher mortality in geriatric patients (> 70 years) vs. adult patients on the basis of: • fall with traumatic brain injury, pedestrian struck by vehicle, multiple body system injuries, fracture of humerus or femur from motor vehicle Non-significant trend towards higher mortality in geriatric patients (> 70 years) vs. adult patients on the basis of: • fall with traumatic chest injury or traumatic spinal cord injury |
| Matsushima 2016 [40], prognostic cross-sectional study, LoE: 2b |
Motor vehicle intrusion in patients < 18 years vs. patients between 19 and 64 years vs. patients ≥ 65 years, (N = 3998 overallb) • Significantly higher mortality in patients ≥ 65 years of age |
| Staffing– training | |
|---|---|
| Daurka 2015 [29], RCT, LoE: 2b↓, unclear risk of bias, underpowered |
Pelvic training + an introduction to the ABC algorithm (N = 11) vs. pelvic training alone (N = 9) • The ABC teaching concept yielded improvements in coagulopathy assessment and management, urological injury, bowel injury / open fracture assessment, and appropriate prioritisation |
a Step 1 NTTP criteria (GCS score ≤ 13, SBP < 90 mm Hg, respiratory rate [RR] < 10 or RR > 29), Step 2 criteria (penetrating injury, flail chest, open skull fracture, ≥ 2 proximal long bone fractures, pelvic fracture, crush injury, amputation, paralysis); boverall: Number of cases only available for the entire study population and not for individual activation criteria c for definitions of undertriage, overtriage and emergency procedures see Table S4, Online Resource 1; d level 1: hypotension (SBP ≤ 90 mmHg), GCS < 13 (currently), penetrating injury to neck, chest, or abdomen, altered respirations or intubation in the field, proximal extremity amputation, multiple incoming patients with severe injuries; e level 2: GCS 13–14, pulse > 12, mangled extremity or distal amputation, age > 65 years + mechanism, neurologic deficit, burns > 20% BSA or inhalation, multiple long bone fractures or mangled extremity, flail chest, peritonitis on abdominal exam, pregnancy; f SBP < 90 mmHg, or radial pulse absent with carotid pulse present, GCS ≤ 13, fractures of ≥ 2 proximal long bones; g SBP < 100 mmHg or radial pulse absent with carotid pulse present, GCS ≤ 14 in trauma patient with a known or suspected traumatic brain injury, fracture of one proximal long bone sustained in a motor vehicle crash, injury to ≥ 2 body regions, pedestrian struck by motor vehicle, fall from any height including standing falls with evidence of traumatic brain injury
For abbreviations and acronyms see list included