Sleep-related fatal vehicle accidents: Characteristics of decisions made by multidisciplinary investigation teams

Abstract

Objectives: To analyze factors that explain the attribution of crash causes as sleep-related by accident investigators. Design: Analysis of national database of fatal road accidents studied in depth. All nonprofessional nonintoxicated car drivers responsible for a fatal accident from 1991 to 2001 were included (N = 1464). Setting: Finland, with approximately 5.1 million inhabitants and 2.3 million motor vehicles. Participants: N/A. Interventions: N/A. Measurements: Comprehensive database recorded by multidisciplinary investigation teams, with specific emphasis on the availability of sleep-related driver variables and sleep-related causal decisions by teams. Results: Injury severity, age, and marital status of the responsible car driver were related to the proportion of missing data in fatigue-related variables in the database (sleeping time, time awake, lifetime mileage). While there were differences between investigation teams and their activities, a series of logistic regression models showed that the lack of relevant variables in the database did not affect the proportion of accidents attributed to falling asleep (10% of cases) or as having fatigue-related causal factors (an additional 5% of the cases). The accident type (head-on and running-off versus other) and road conditions (dry or wet versus icy or snowy pavement) predicted the investigation teams' attribution of sleep-related causes in all models. Conclusions: Multidisciplinary teams' attribution of sleep-related causal factors were rather stable, comprising 10% to 15% of the cases investigated, independent of the availability of specific sleep-related information.