To the Editor.
—Two of the most recurrent problems in epidemiologic studies of endemic firearm-related violence in America1,2 are demonstrated in the study of firearm-related deaths in Milwaukee, Wis, by Dr Hargarten and colleagues.3The first problem is making unverified assumptions regarding facts, trends, and causal relationships that may not be accurate. Patterning a call for firearm-related data collection after motor vehicle data collection is based on the untested assumption that intentional-injury prevention can be modeled after public health approaches toward unintentional-injury prevention.The study assumed data-based public health recommendations affected American policy on motor vehicles, causing a decline in motor vehicle deaths that was "no accident."3 One could as easily explain the trends with reference to better emergency medical care since other unintentional deaths, including those involving firearms, decreased at least as fast as motor vehicle deaths.4 Or the motor vehicle decline could be associated
Blackman PH. Firearms and Fatalities. JAMA. 1996;275(22):1723. doi:10.1001/jama.1996.03530460027014