Comparing Ugandan motorcycle taxi driver estimations of injury incidence to District-level injury surveillance data as a proxy to determine factors influencing risk perception.

Abstract

Introductionroad traffic incidents (RTIs) are a leading cause of death among young people, disproportionately affecting low- and middle-income countries (LMICs), where motorcycle taxis disproportionately contribute to injury. Though driver behavior has been identified as the most important factor in RTIs, the factors that influence risk perception, which affect driver behavior, have not been well-studied in LMICs and may inform future strategies to limit risky behavior.MethodsUgandan motorcycle taxi drivers (n=117) were surveyed on personal characteristics and experiences, ranking apparent risk of select injury conditions. Rankings were then compared against the actual frequency of corresponding District-level injury surveillance data for the same injury conditions to investigate the accuracy of respondent risk perception. Personal characteristics were then regressed against the perceived risk of certain injury classification rankings to investigate possible factors influencing rankings.Resultsover 26 months, 21,253 injury-related events were recorded in Iganga District, of which 7,424 patient encounters (34.93%) were related to RTI. Ugandan motorcycle taxi drivers tended to over-estimate the risk associated with their profession, but correctly classified the three most common injuries. Regression analyses revealed personal characteristics including personal exposure to RTIs (B=0.037, t=2.035, p=0.044) and years of experience (B=0.026, t=1.828, p=0.070) predicted perceived risk.ConclusionUgandan motorcycle taxi drivers accurately predict the risks associated with their profession. The perception of these risks may be affected by years of experience and previously witnessed RTIs. Further empirical investigation is required to document all key motives and perspectives of drivers as factors that influence risk perception and subsequent behavior in LMICs and may inform future strategies to limit risky behavior.