Emergency medical services collisions (EMVCs) are a largely unexplored area of emergency medical services (EMS) research. Factors that might contribute to an EMVC are numerous and include use of warning lights and siren (WL&S). Few of these factors have been evaluated scientifically. Similarly, the incidence and severity of EMVCs is poorly documented in the literature. This study sought to define the incidence and severity of, and where possible, identify any contributing factors to EMVCs in a large urban system.

Retrospective study of all collisions involving vehicles assigned to the EMS Division of the Houston Fire Department in calendar year 1993. Fifty-one ambulances were operational 24 hours per day during calendar year 1993. Houston EMS received 150,000 requests for assistance, made 180,000 vehicular responses, and accrued 2,651,760 miles in 1993.

Eighty-six EMVCs were identified during the study period. The gross incidence rate was therefore 3.2 EMVC/100,000 miles driven or 4.8 collisions/10,000 responses. Of the 86 EMVCs, 74 (86%) files were complete and available for evaluation. Major collisions, determined according to injuries or vehicular damage, accounted for 10.8% of all EMVCs. There were 17 persons transported to hospitals from EMS collisions, yielding an injury incidence of 0.64 injuries/100,000 miles driven or 0.94 injuries/10,000 responses. There were no fatalities. The majority of collisions (85.1%) occurred at some site other than an intersection. There was no statistical association between occurrence at an intersection and severity, day versus night, weekend versus weekday, presence or absence of precipitation, or use of WL&S versus severity of collision. Drivers with a history of previous EMVCs were involved in 33% of all collisions. The presence of prior EMVCs was associated (p < 0.001) with the number of persons transported from the collision to a local hospital. Five drivers, all with previous EMVCs, accounted for 88.2% (15/17) of all injuries.

A few drivers with previous EMVCs account for a disproportionate number of EMVCs and nearly 90% of all injuries. This risk factor—history of previous EMVC—has not been reported in the EMS literature. It is postulated that this factor ultimately will prove to be the major determinant of EMVCs. Data collection of EMS collisions needs to be standardized and a proposed collection tool is provided.