Emergency response times and crash risk: An analysis framework for Costa Rica

Abstract Background About 48% of the deaths due to motor vehicle crashes in Costa Rica occurred on the site of the accident, which could be attributed to an immediate death, or to a delayed response by the emergency systems. It is expected that improving the emergency response times would reduce crash deaths. The purpose of this work is to propose a framework for analyzing the risk of dying in a crash by considering not only the estimated emergency response time but also the number and severity of crashes in different locations, using easily collectable data. Methods The risk associated with a crash is considered the product of hazard and vulnerability. Hazard is related to the frequency and severity of crashes, while vulnerability is associated with the emergency response time. Two methods are used to estimate the risk in this paper: (a) hazard and vulnerability are assumed to be continuous variables or, (b) hazard and vulnerability are considered categorical ordered variables. Results Districts in the “very high” risk category are those with the intermedia values in both hazard and vulnerability or high values in vulnerability and not too low values in the hazard. In general, there are not districts with very high values in both hazard and vulnerability. Conclusions The continuous method has the advantage of making use of the continuous nature of the hazard, vulnerability, and risk variables proposed. On the other hand, the categorical ordered method has the advantage of being more intuitive by reducing the hazard, vulnerability and risk to a small number of classes that are easy to interpret.