Accident Analysis and Prevention 37 (2005) 661–667 Firearm storage practices and rates of unintentional firearm deaths in the United States Matthew Miller * , Deborah Azrael, David Hemenway, Mary Vriniotis Harvard Injury Control Research Center, Harvard School of Public Health, 677 Huntington Ave., 3rd Fl., Boston, MA 02115, USA Received 14 February 2005; accepted 24 February 2005 Abstract Background: The American Academy of Pediatrics and the American Medical Association recommend storing firearms unloaded and locked up to minimize the chance of injury. Although these recommendations appeal to common sense, no study has yet addressed whether firearm storage practices influence the risk of unintentional firearm injury. Methods: Negative binomial regression analyses were used to assess the cross sectional relation between firearm storage practices and rates of unintentional firearm death in the United States, controlling for rates of firearm prevalence, poverty and urbanization. Recently available state-level measures of household firearm prevalence and firearm storage practices were obtained from the 2002 Behavioral Risk Factor Surveillance System. Unintentional firearm death counts and population data came from the National Center for Health Statistics. Results: Independent of firearm prevalence, urbanization, and poverty, a disproportionately large share of unintentional firearm fatalities occurred in states where gun owners were more likely to store their firearms loaded, the greatest risk occurring in states where loaded firearms were more likely to be stored unlocked. Conclusion: Our findings provide empirical support for recommendations issued by the AMA and the AAP that firearms should be stored unloaded and locked, and suggest that promoting safer storage practices could save many lives. © 2005 Elsevier Ltd. All rights reserved. Keywords: Firearms; Mortality; Unintentional injury; Storage practices 1. Introduction The circumstances involved in unintentional firearm fatal- ities are not well characterized. Although we know that 90% of the 11,509 Americans who died from unintentional firearm injuries between 1991 and 2000 were male and a quarter were under age eighteen (CDC, 2003b), we know little about the type of firearm involved or how the offending firearm had been stored. Existing studies suggest that routine handling may play a role in both non-fatal (Stucky and Loder, 1991; Kennedy et al., 1993; Payne et al., 1993; Annest et al., 1995; Go DeVivo et al., 1995; Sinauer et al., 1996) and fatal (Morrow * Corresponding author. Tel.: +1 617 432 1459; fax: +1 617 432 3699. E-mail addresses: mmiller@hsph.harvard.edu (M. Miller), azrael@hsph.harvard.edu (D. Azrael), hemenway@hsph.harvard.edu (D. Hemenway), mvriniot@hsph.harvard.edu (M. Vriniotis). and Hudson, 1986; Cole and Patetta, 1988; Carter, 1989; Wintemute et al., 1989; Grossman et al., 1999) events. For example, among children most fatal events occurred when children were playing with guns they had found, and in at least twenty percent of the cases the child who fired the gun did not know it was loaded (Wintemute et al., 1987, 1989). Of note, firearms owned by relatives or friends living in house- holds other than the victim’s accounted for half (49%) of all unintentional injuries and deaths (Grossman et al., 1999), suggesting that an ideal measure of firearm exposure cap- tures firearm-related characteristics (e.g. storage practices) not only in the victim’s home, but also in the community. Two empirical studies (a case–control study (Wiebe, 2003) and an ecologic study (Miller et al., 2001)) have evaluated whether household firearm ownership is related to the rate of unintentional firearm deaths in the United States. Both found a strong relation between household firearm ownership and rates of unintentional firearm deaths. To our knowledge no 0001-4575/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.aap.2005.02.003