A Nationwide Investigation of Microscopic and Macroscopic Factors and Screening Counties with Respect to Fatal Crashes Due to Drowsiness

[background/purpose]

Drowsy driving has been considered a serious problem in traffic safety because it makes driver less attentive, slows reaction time and affects a driver’s capability to make decisions. The objectives of this study are: 1) discovering contributing factors for fatal crashes due to drowsiness both at microscopic and macroscopic levels; and 2) screening counties with respect to drowsiness related fatal crashes.

[methods]

Binary logistic and negative binomial models were developed at the microscopic and macroscopic levels, respectively, to uncover factors for fatal crashes due to drowsiness using FARS data (2007-2009). Subsequently, a nationwide county-level screening was conducted to identify counties with drowsy driving problems.

[results/outcomes]

The microscopic model identified that heavy trucks on the high-speed roadway segments with the less number of lanes in dark conditions are more likely to be involved in drowsiness related fatal crashes. On the other hand, the macroscopic modeling revealed county-level factors such as demographic (age, race, etc.) and socio-economic factors (commuting characteristics, industry, etc.). The top 5% of counties with respect to drowsy driving problems were identified by the screening technique. Among these counties, the largest number of counties is located in Texas (35%), and followed by California (8%), and Utah (6%).

[conclusions]

This study identified the contributing factors for drowsiness related fatal crashes at macro- and micro-level. The key findings from this study are expected to be useful for effective nation-wide strategic plans to alleviate drowsy driving relevant factors. Furthermore, the screening results where to focus with priority at the macroscopic level.