Is the number of traffic fatalities in American hospitals decreasing?

Fatality Analysis Reporting System (FARS) data from 1977 to 2002 record a decreasing number of traffic fatalities taken to a hospital compared with traffic fatalities not taken to a hospital. In this study, we calculated the proportions of decedents reportedly taken to a hospital each year in each state, and the proportions surviving at least 1h. We also used death certificate data from the National Center for Health Statistics (NCHS) for 1979-1999 to categorize the proportion of motor vehicle fatalities in each state by hospital patient status. The annual number of traffic fatalities decreased slightly over the period of observation. The proportion of decedents recorded in FARS as transported to a hospital fell from about 73 to 43%. However, this proportion decreased abruptly at certain times in some states, suggesting previous misclassification. The proportion surviving at least 1h remained relatively constant. NCHS data showed a decline in the proportion declared dead in hospitals from 62 to 51%, including a decline in the proportion declared dead on arrival (DOA) from 20 to 8%. Along with occasional misclassification in some states, the decrease in cases transported only to be pronounced DOA could explain why FARS data show a decrease in deaths after hospital transport.