Road Traffic Injuries in Thailand and their Associated Factors using Conway-Maxwell-Poisson Regression Model

In road traffic accidents, the number of human injuries does not follow the traditional Poisson distribution, and often exhibits overdispersion. Based on a survey of 180 countries by the World Health Organization in 2015, Thailand is the second unsafety country for driving on roads, with 36.2 deaths per 100,000 population. In this study, a total of 20,226 road traffic accidents in Thailand in 2015 were investigated. Factors associated with the injury counts per accident were disclosed using the Conway -Maxwell-Poisson regression model. Four covariates considered were road type, road surface, road section, and festive months. The results showed that the distribution of injury count data was under-dispersed, which is rare. The number of injuries was significantly determined by road type, road section, and festive months. Driving on the Local Administration Roads was risker than on the National Highways or Rural Roads. Straight road sections were found to have a lower involvement in road accident injuries than at intersections. Our findings show that the month of the festival plays a crucial role in increasing unusual human injuries on the road. Road safety managers should pay more attention on these months, not only in Thailand but also other countries where special events occur.