Introduction
Despite advances in automobile safety, motor vehicle collisions (MVC) represent one of the most common mechanisms of major traumatic injury in the world. Triage guidelines exist to inform emergency personnel on criteria to expedite transfer from the scene to a major trauma center. However, field triage is currently not significantly informed by the damage patterns to the vehicle itself. Given this, we sought to study associations of the damaged vehicle to patient injury.
Methods
We conducted a cohort study using novel linkage of publicly available police crash reports (UD10) and Michigan Trauma Quality Improvement Program (MTQIP) trauma registry data. We compared outcomes of the patients evaluated for traumatic injuries by the details of their MVC, including vehicle age, type of collision, restraint use, and airbag deployment, controlling for patient factors. Multivariable logistic regressions were used to identify characteristics predictive for severe injury, defined as an Injury Severity Score > 15 (ISS15), the need for higher levels of care (operating room, ICU, or transfer to another hospital), and mortality.
Results
There were 15,775 UD10 and MTQIP linked case. This group contained 42.7% males with mean age 42.1 years (SD 19.8 years). Several vehicular features were associated with overall outcomes after MVC. In multivariable regression, the vehicle type and crash pattern were the strongest predictive factors for ISS15, need for higher level care, and mortality. All-terrain-vehicle (ATV) and motorcycle collisions (MCC) (ORs 1.74 and 2.32, p<0.001) were the vehicle types with the highest increased odds for ISS15, while only MCC was associated with higher odds of mortality (OR 2.41, p<0.001). ATV (OR 1.52, p=0.01) and MCC (OR 1.87, p<0.001) were the vehicle types associated with higher odds for higher levels of care. Angled, head on, single vehicle collisions had higher odds of ISS15 (OR 1.18, 1.70, and 1.45; p=0.02, p<0.01, and p<0.01). Head on and single vehicle collisions also had higher odds of mortality (OR 1.82 and 1.36, p<0.01 and p=0.04) and needing higher level of care (OR 1.48 and 1.40, p<0.01 and p<0.01).
Tables & Figures
Logistic Regressions
