Evaluating risk factors associated with fatalities among powered two-wheeler crashes in the United States

The objectives of this study are to: (1) evaluate the fatality risk of powered two-wheeler (PTW) riders in collisions with various types of opponent vehicles; (2) estimate the likelihood of different impact orientations between PTWs and other vehicles involved, and (3) develop an initial crash-based speed–fatality prediction model specific to the United States that controls for vehicle orientation. Data was extracted from the National Highway Traffic Safety Administration’s Crash Reporting Sampling System and the Fatality Analysis Reporting System covering 2017–2021. This data was used to estimate the fatality risks associated with various types of opponent vehicles involved in PTW crashes. The vehicles involved in each crash were coupled based on the vehicles’ impact locations, and the effects of different vehicle orientations on fatality were estimated using odds ratio analysis. Multivariate logistic regressions were used to model the relationship between impact speed and fatality risk for front-end collisions involving PTWs. Crashes involving buses and heavy trucks posed a significantly higher risk to PTW riders, with the fatality risk being four times greater compared to collisions with passenger vehicles. This risk varied based on impact orientation; frontal collisions with the front or sides of the opposing vehicle were the most dangerous, with fatality odds approximately four times higher than rear-end impacts. The speed-fatality prediction model showed the fatality risk increased with higher PTW travel speed while accounting for the expected higher fatality risk in crashes involving older riders, heavy vehicles, or un-helmeted riders. The significant influence of opponent vehicle type on fatality risk in crashes involving PTWs highlights the need for further investigation into vehicle-specific crash prevention and mitigation strategies, especially for light and heavy trucks. Similarly, the high variability in fatality odds across different crash configurations underscores the importance of integrating impact orientation surrogate variables into injury prediction models. The speed–fatality prediction model developed in this study provides a foundational framework for evaluating the effectiveness of advanced rider assistance systems and other safety interventions that reduce crash speed. Future research should explore the benefits of such measures through simulation and real-world testing.