• 
• 

Three Part Question

In [adults sustaining traumatic injury] does [mechanism of injury or anatomical injury or physiological derangement] identify patients requiring [treatment at a major trauma centre]?

Clinical Scenario

You are first on scene to a road traffic collision (RTC) involving a 32 year old male who has crashed his motorbike at 30mph. He is haemodynamically normal but complains of pain in his right chest and right hip and is unable to walk. You wonder whether the optimum management of this patient would be to bypass the nearby trauma unit for direct transfer to the closest major trauma centre.

Search Strategy

Medline 1946 – May Week 3, 2016. [(trauma*.ti,ab OR exp WOUNDS AND INJURIES/ OR exp MULTIPLE TRAUMA/) AND (exp ADULT/) AND (mechanism.ti,ab OR anatom*.ti,ab OR “vital sign*”.ti,ab OR physiolog*.ti,ab) AND ((“trauma cent*”.ti,ab OR MTC.ti,ab OR “trauma unit”.ti,ab OR “emergency medical service”.ti,ab OR paramedic.ti,ab) AND (triage.ti,ab OR transfer.ti,ab))]

Search Outcome

The search produced 241 articles. Following a review of title and abstract, 26 papers were identified to review in full. Following full text review 9 were removed as not directly relevant to the research question. The remaining 17 papers are presented below (Table 1).

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Long et al. 1986 United States	2511 trauma patients over a 16 month period	Retrospective cohort study	Correlation of the pre-hospital trauma score with the ISS	Prolonged extrication >20 minutes and death in same vehicle associated with ISS >16	Single centre. Use of ISS as an output measurement. Not all patients had a trauma score calculated in the pre-hospital setting.
			Accuracy of mechanism of injury at predicting ISS >15
Simon et al. 1994 United States	Review of 1235 consecutive trauma team activations at a single hospital	Retrospective cohort study	To provide a secondary triage tool to reduce un-necessary trauma team activation.	Mechanism of injury alone has a PPV of 38% for serious injury. By introducing the vehicular checklist PPV increases to 61%. Abnormal physiology created false positives in uninjured patients.	List of life-saving interventions not exhaustive and not described in full. Both head on collision >30mph & vehicular intrusion had same proportion of minimally injured and severely injured patients, making it impossible to draw solid conclusions.
Esposito et al. 1995 United States	5028 patients with pre-hospital criterion for treatment at trauma centre. Data collected from 222 pre-hospital provider agencies & 53 hospitals.	Prospective cohort study	Identify indicators of major trauma victims using ISS and mortality for each pre-hospital criterion: anatomical, physiological, mechanism of injury & clinical gestalt. High Yield >30%, Intermediate 20-30%, Low <20% (all with ISS >15)	Only 60% cases included had a single pre-hospital criteria. Pedestrian struck, prolonged pre-hospital time & abnormal physiology (SBP<90mmHg, 10>RR>29, GCS<13) associated with high yield ISS>15 Ejection & Vehicle deformity associated with intermediate yield. Fall >6m and clinician gestalt associated with low yield.	ISS only recorded for 45%, therefore large amount of missing data. ISS has been shown to not fully correlate with the resource requirements of a trauma patient
Cooper et al. 1995 United States	Questionnaires given to 112 Emergency Medical Services Personnel conveying trauma patients to a single centre.	Survey study	To identify the PPV of mechanism of injury in isolation at predicting trauma centre need.	26% patients conveyed due to mechanism of injury alone. Mechanism of injury in isolation has a low PPV.	Study period not defined
Wuerz et al. 1996 United States	333 patients transported by helicopter to a level 1 trauma centre	Case series	Performance characteristics of physiological criteria	Using ISS>15, physiological criteria under-triaged 44.3% (n=67) and mortality 16.1% (n=5)	Low median study age (26 years, IQR 19-42). Limited to helicopter transport only.
Schoettker et al. 2001 Switzerland	Comparison of patients involved in RTC (cars) with ejected (n=71) vs non-ejected (n=539). patients from RTC	Prospective cohort study	Type of injury	No obvious evidence found	Outcome measures not matched between all groups. Patients who died had their ISS excluded from the study median calculation.
			Pre-hospital vital signs	43.7% (n=31) ejected had GCS <8
			Hospital diagnosis	(not included in analysis)
			ISS (median)	Ejected = 17 vs non-ejected = 9
			Need for ICU	ICU admission in 34% (n=21) of ejected patients
			Need for life-saving surgery	38% of ejected patients (n=24)
			Outcome	24% mortality (n=17) in ejected patients
			Situational criteria performance	Physiological criteria under-triaged 67 patients – addition of situational criteria reduced this by 86.6% (n=58)
Holcomb et al. 2005 United States	Helicopter transport of 216 patients to major trauma centre (2001-7) and who required admission to hospital.	Cohort study	Life-saving intervention both pre-hospital and in-hospital	48/114 patients with pre-hospital HR >100 required life-saving intervention. 90% of patients with pre-hospital capillary refill >2s required life-saving intervention (OR 17.43). 73% with GCS motor score <6 and 37% with RR >24 required life-saving intervention. 87% with pre-hospital SBP <90 required life-saving intervention (OR 16.81). Using logistical regression patients with GCS motor score <6 and SBP<90mmHg have a 95% probability of requiring a life-saving intervention vs 21% in patients with GCS motor score 6 and SBP >90mmHg.	Convenience sampling. Limited analysis of life-saving interventions (only in-hospital and 5 pre-hospital listed), with no specific differentiation. No differentiation between which life-saving intervention is associated with which physiological derangement, and which of these life-saving interventions were performed in the pre-hospital environment.
Holcomb et al. [B] 2005 United States	Helicopter transfer of 793 pre-hospital trauma patients.	Retrospective cohort study	Correlation of physiological signs and need for life-saving intervention. Additional review of vital signs measured either manually or through automation.	Verbal and motor component of GCS, along with radial pulse, had the greatest predictive power at predicting the need for life-saving intervention.	Head injured patients (AIS > 3) were excluded retrospectively. Large amount of missing data, resulting in final analysis set of n=381 (48%).
Kann et al. 2007 Denmark	Consecutive injured patients presenting to single hospital over 6 months. 848 patients included; 242 trauma team activations.	Prospective cohort study	Evaluation of rates of over-triage (inappropriate trauma team activation) from mechanism of injury in isolation.	5/606 without trauma team activation had ISS>15. ISS < 15 for 78% trauma team activations. 60 patients had single mechanism of injury criterion responsible for trauma team activation. 92% had ISS<15. High speed RTC (>40mph) in isolation associated with ISS>15 in only 7%.	Reported that combination of abnormal physiology & mechanism of injury improve reliability of trauma team activation associated with ISS > 15 however results not documented.
Haan et al 2009 United States	Mechanism of injury – vehicle rollover only. 569 patients not meeting other trauma triage criteria.	Retrospective cohort study	Hospital admission	35% of patients admitted	The surrogates used are not objectively transferable to a definition of trauma centre need.
			Surgical intervention	6 patients (1%) required urgent surgery <12 hours after admission. Further 124 (21.7%) required surgery at later date
			ICU admission	8 patients (1.4%) required ICU admission.
Brown et al. 2011 United States	1,086,764 patients from the National Trauma Databank.	Retrospective cohort study	Trauma centre need defined as ISS>15, ICU admission, urgent surgery Logistic regression analysis to identify individual factors linked with trauma centre need.	Physiological criteria outperform anatomical criteria at determining trauma centre need (sensitivity 32% vs 26%, specificity 91% vs 86%). Anatomical criteria best at predicting need for surgical intervention. Physiological criteria best at predicting ISS>15, but poor at predicting surgical need. Flail chest greatest predictor of trauma centre need. In order to reduce under-triage, mechanism of injury required in addition to anatomical & physiological criteria.	Retrospective database review with limited variables leading to ambiguity. Selection bias present due to database skewed to major trauma centres.
Lerner et al. 2011 United States	Major trauma centre transfer on mechanism of injury alone (not fulfilling anatomical & physiological criteria). 9,483 patients with mechanism of injury of assault, motor vehicle crash, fall or pedestrian/ cyclist struck.	Prospective cohort study	Mechanism of injury as an indicator for major trauma centre need	2,363 fulfilled mechanism of injury criteria. 9% (n=204) defined as requiring trauma centre need. Sensitivity 39.7%. LR >5 mechanism of injury predictors were death of another occupant; fall >20 feet; >20min extrication.	Use of interviews to determine mechanism of injury.
Stuke et al 2013 United States	Two groups: assessment of anatomical & physiological criteria (n= 6584) vs mechanism of injury alone (n=3315).	Retrospective cohort study	Discharge status 6 hours after ED admission.	55% (n=3613) of anatomical & physiological patients admitted. 45% (n=2971) of mechanism of injury patients admitted. Removing mechanism of injury as sole admission criteria would have resulted in 2700 fewer admissions.	Emphasis on mechanism of injury and reduction in overtriage.
Lerner et al. 2013 United States	11,892 patients brought to trauma centres by Emergency Medical Services. 1,274 excluded due to meeting other trauma triage criteria.	Prospective cohort study	Anatomical injury as criteria for trauma centre need. Trauma centre need defined as ISS>15, need for non-orthopaedic surgery within 24 hours, ICU admission or death prior to discharge.	Anatomical criteria had 38% sensitivity and 91% specificity at predicting trauma centre need. Flail chest, paralysis, ≥2 long bone fractures & amputation had LR >5 for trauma centre need. 503 patients missed by anatomical & physiological criteria. 41% (n=204) would have been captured by mechanism of injury.	EMS perceived anatomical injury compared with ICD-9-CM codes attributed by billing teams. Subsequent to the study ‘flail chest’ was removed from guidelines & replaced with chest wall instability/deformity.
Stuke et al. [B] 2013 United States	Using mechanism of injury alone, 3,569 patients transported to trauma centre.	Retrospective cohort study	Mechanism of injury as an indicator for major trauma centre need. Trauma centre need defined as ISS >15, ED transfusion, ICU admission, laparotomy/thoracotomy/ vascular/surgery < 24 hours, pelvic fracture, >2 proximal long bone fractures or neurosurgical intervention.	23% (n=821) with mechanism of injury required trauma centre need. LRs >5 defined as death in same passenger compartment; ejection from vehicle; > 20mins extrication; fall >20feet; pedestrian thrown/run over.	Missing data due to retrospective database review. From 2006 patients with mechanism of injury in isolation were transferred to Major Trauma Centre only at the patient’s request.
Potter et al. 2013 United Kingdom	171 patients with ISS>15 presenting to a single major trauma centre.	Retrospective cohort study	Sensitivity of Wessex Triage Tool (WTT) in identifying ISS>15 patients	Wessex triage tool demonstrates sensitivity of 53%. Performance reduced in older population with low energy trauma.	Missing data due to retrospective database review
Davidson et al. 2014 United States	85,761 patients involved in motor vehicle collisions.	Retrospective cross-sectional study	PPV of anatomical markers, physiological markers and mechanism of injury as indicators of severe trauma, defined as ISS >15.	Physiological criteria - PPV 20.8% Anatomical criteria - PPV 48.5% Mechanism of injury criteria - PPV 9.7%	Use of ISS >15 was used to define trauma centre need and not resource requirement.

Comment(s)

There is some evidence supporting the use of anatomical injury, mechanism of injury and physiological markers in the pre-hospital triage of injured patients. However, there is a paucity of high quality evidence in this field. Much of the literature is derived from the retrospective analysis of trauma registry databases, with the associated limitations of incomplete data capture and non-standardised outcome variables, often leading to contradictory results between studies. Additionally, selection bias is evident in a number of studies; patients exhibiting physiological derangement were removed from further analysis. This may limit analysis as patients with abnormal vital signs may be more likely to have endured a significant mechanism of injury or anatomical injury. There is no consensus as to what constitutes a need for trauma centre care. It is vital that this is agreed in order to standardise outcomes by which to conduct further research. In a number of studies, the outcome measure defining trauma centre need was an Injury Severity Score (ISS) >15, which itself is a retrospective measurement. Although this is a definition of major trauma, it may not represent the optimal marker for trauma centre need and there is evidence to support a lack of correlation between ISS and the need for emergency procedures. Furthermore, statistical analysis is not standardised with a combination of performance characteristics being measured.

Editor Comment

JoT JS

Clinical Bottom Line

There is some evidence to support the use of anatomical injury, mechanism of injury and physiological parameters in predicting requirement of major trauma centre resources. However, there is no high-quality evidence supporting the use of any parameter either in isolation or in combination. Targeted research is required to establish the optimal method of pre-hospital identification of the seriously injured patient, thereby determining trauma centre need.

References

1. Long WB, Bachulis BL, Hynes GD. Accuracy and relationship of mechanisms of injury, trauma score, and injury severity score in identifying major trauma Am J Surg 1986;151:581-4
2. Simon BJ, Legere P, Emhoff T, et al Vehicular trauma triage by mechanism: avoidance of the unproductive evaluation J Trauma 1994;37:645-9
3. Esposito TJ, Offner PJ, Jurkovich GJ, et al Do pre-hospital trauma center triage criteria identify major trauma victims? Arch Surg 1995;130:171-6
4. Cooper ME, Yarbrough DR, Zone-Smith L, et al pplication of field triage guidelines by pre-hospital personnel: is mechanism of injury a valid guideline for patient triage? Am Surg 1995;61:363-67
5. Wuerz R, Taylor J, Smith JS Accuracy of trauma triage in patients transported by helicopter. Air Med J 1996;15:168-70
6. Schoettker P, Ravussin P, Moeschler O. Ejection as a key word for the dispatch of a physician staffed helicopter: the Swiss experience. Resuscitation 2001;49:169-73
7. Holcomb JB, Niles SE, Miller CC, et al. Pre-hospital physiologic data and lifesaving interventions in trauma patients. Mil Med 2005;170:7-13
8. Holcomb JB, Salinas J, McManus JM, et al. Manual vital signs reliably predict need for life-saving interventions in trauma patients. J Trauma 2005;59:821-8
9. Kann SH, Hougaard K, Christensen EF. Evaluation of pre-hospital trauma triage criteria: a prospective study at a Danish level I trauma centre. Acta Anaesthesiol Scand 2007;51:1172–7
10. Haan JM, Glassman E, Hartsock R, et al Isolated rollover mechanism does not warrant trauma center evaluation. Am Surg 2009;75:1109–11
11. Brown JB, Stassen NA, Bankey PE, et al. Mechanism of injury and special consideration criteria still matter: an evaluation of the National Trauma Triage Protocol. J Trauma 2011;70:38–44; discussion44–5
12. Lerner EB, Shah MN, Cushman JT, et al. Does mechanism of injury predict trauma center need? Prehosp Emerg Care 2011;15:518–25
13. Stuke LE, Duchesne JC, Hunt JP, et al. Mechanism of injury is not a predictor of trauma center admission Am Surg 2013;79:1149–53
14. Lerner EB, Roberts J, Guse CE, et al. Does EMS Perceived Anatomic Injury Predict Trauma Center Need? Prehosp Emerg Care 2013;17:312–6.
15. Stuke LE, Duchesne JC, Greiffenstein P, et al. Not all mechanisms are created equal. J Trauma 2013;75:140–5
16. Potter D, Kehoe A, Smith JE. he sensitivity of pre-hospital and in-hospital tools for the identification of major trauma patients presenting to a major trauma centre. J R Nav Med Serv 2013;99:16–9
17. Davidson GH, Rivara FP, Mack CD, et al. Validation of pre-hospital trauma triage criteria for motor vehicle collisions. J Trauma 2014;76:755–61