Article Text
Abstract
Background Previous studies of motorcycle crash (MC) related hospital charges use trauma registries and hospital records, and do not adjust for the number of motorcyclists not requiring medical attention. This may lead to conservative estimates of helmet use effectiveness.
Methods MC records were probabilistically linked with emergency department and hospital records to obtain total hospital charges. Missing data were imputed. Multivariable quantile regression estimated reductions in hospital charges associated with helmet use and other crash factors.
Results Motorcycle helmets were associated with reduced median hospital charges of $256 (42% reduction) and reduced 98th percentile of $32 390 (33% reduction). After adjusting for other factors, helmets were associated with reductions in charges in all upper percentiles studied. Quantile regression models described homogenous and heterogeneous associations between other crash factors and charges.
Conclusions Quantile regression comprehensively describes associations between crash factors and hospital charges. Helmet use among motorcyclists is associated with decreased hospital charges.