We compile citations and summaries of about 400 new articles every week.
Email Signup | RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article


Kellermann AL, Rivara FP, Rushforth NB, Banton JG, Reay DT, Francisco JT, Locci AB, Prodzinski J, Hackman BB, Somes G. New Engl. J. Med. 1993; 329(15): 1084-1091.


Department of Internal Medicine, University of Tennessee, Memphis.


(Copyright © 1993, Massachusetts Medical Society)






BACKGROUND. It is unknown whether keeping a firearm in the home confers protection against crime or, instead, increases the risk of violent crime in the home. To study risk factors for homicide in the home, we identified homicides occurring in the homes of victims in three metropolitan counties. METHODS. After each homicide, we obtained data from the police or medical examiner and interviewed a proxy for the victim. The proxies' answers were compared with those of control subjects who were matched to the victims according to neighborhood, sex, race, and age range. Crude and adjusted odds ratios were calculated with matched-pairs methods. RESULTS. During the study period, 1860 homicides occurred in the three counties, 444 of them (23.9 percent) in the home of the victim. After excluding 24 cases for various reasons, we interviewed proxy respondents for 93 percent of the victims. Controls were identified for 99 percent of these, yielding 388 matched pairs. As compared with the controls, the victims more often lived alone or rented their residence. Also, case households more commonly contained an illicit-drug user, a person with prior arrests, or someone who had been hit or hurt in a fight in the home. After controlling for these characteristics, we found that keeping a gun in the home was strongly and independently associated with an increased risk of homicide (adjusted odds ratio, 2.7; 95 percent confidence interval, 1.6 to 4.4). Virtually all of this risk involved homicide by a family member or intimate acquaintance. CONCLUSIONS. The use of illicit drugs and a history of physical fights in the home are important risk factors for homicide in the home. Rather than confer protection, guns kept in the home are associated with an increase in the risk of homicide by a family member or intimate acquaintance.

VioLit summary:

The goal of this study by Kellermann et al. was to examine the relationship between owning a gun and the level of protection or risk this provided in the home.

A quasi-experimental research design was employed for this study. Homicides that were included in the study were those that involved the residents of King County, Washington, or Shelby County, Tennessee between August 23, 1987 and August 23, 1992. Cuyahoga County, Ohio, was also included but only covered January 1, 1990, to August 23, 1992. Deaths were included if they were judged to be homicides or if assault-related injuries had resulted in death within three months after the injury. Homicides which included victims under the age of twelve were excluded. Homicides were included if they occurred in the home or in some location adjacent to a home. Murder-suicides were included if the victim of the homicide was older than the perpetrator; these were entered as one event. Multiple homicides were also entered as one event with data recorded on only the oldest victim. Information concerning the homicides themselves was gathered from reports made at the scene by either the medical examiner or the police. These data were supplemented by newspapers accounts, obituaries, and calls to funeral homes. A list of names of persons closest to the victim was also gathered at the scene for later interviews. These interviews were treated as proxies for the victims. A matched control group for case subjects was also gathered to match the subject in gender, race, age, and neighborhood. Random selection of matching households in neighborhoods was used. The adult in the first matching household was approached to do the interview. The authors reported that 84% of the controls were a the closest possible match for case subjects. Both proxies and controls were offered ten dollars for their participation. The interviews for both groups were identical. The authors stated that they used items from the Short Michigan Alcoholism Screening Test, Hollingshead-Wilson 2 factor index of social position, and a poll of gun ownership. There were no other details given about the questionnaire instrument. Homicide scene reports were used to describe the study population, and the interviews were used as a risk assessment. Chi square and conditional logistic regression were used to analyze the data.

It was discovered that 1,860 homicides occurred in the study counties during the time of the study. After subjecting these to the study criteria, 420 were left for study. 63.1% of the victims were men. 50.9% occurred during a quarrel or romantic triangle. 7.6% were related to drug deals; 21.9% were committed during the commission of another crime, and 13.3% had no discernable motive except homicide. 76.7% of victims knew the offender. Blacks comprised 61.9% of victims. Handguns were used in 42.9% of the homicides while rifles (2.4%), shotguns (3.6%), and unknown firearms (1.0%) were used less frequently. Alcohol was more frequently consumed by households of case subjects and case subjects than controls; the case subjects/households also showed more behavioral correlates of alcoholism and illicit drug use than controls. Previous episodes of violence were reported more frequently by case subjects' households; 31.8% reported physical fights versus 5.7% of controls. One or more members of the case households were also more likely to have been arrested or to have been involved in a physical fight outside the home than controls. Guns were kept in 45.4% of the homes of the case subjects while only 35.8% of the control subjects did. Case households were significantly more likely to have a handgun than control households. The logistic regression model included the variables of home rented, case subject or control lived alone, any household member ever hit or hurt in a fight in the home, any household member ever arrested, any household members used illicit drugs, and one or more guns kept in the house. The presence of one or more firearms in the home was strongly associated with an increased risk of homicide in the home. Stratified analyses revealed that the connection between guns and homicide in the home was present within gender, racial, and age categories. Gun ownership was most strongly associated with homicide at the hands of family or other intimates but not for acquaintances or strangers. There was no evidence for protective benefits from gun ownership found from any subgroup. Homicide by firearms was the only kind of homicide significantly related to gun ownership. Households where someone had previously been hit or hurt in a fight in the home was also strongly associated with homicide. This relationship survived controls for effects of gun ownership and the other variables in the model. This relationship held even in stratified analyses. The increased risk was accounted for by the strong association with family and intimate homicide. The use of illicit drugs was found to be an important risk factor for homicide.

The authors stated that the association between guns and homicide risk in the home should serve as a basis for discouragement for people to acquire and keep guns at home. Society, it was argued, should spend money to improve intervention in battering relationships. Additionally, it was argued that health care professionals and the criminal justice system should work harder to identify and protect victims of domestic violence in order to provide intervention which may prevent a later homicide.

(CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado)

Firearms in Home
Firearms Ownership
Firearms Violence
Firearms Death
Firearms Homicide
Family Environment
Homicide Risk Factors


All SafetyLit records are available for automatic download to Zotero & Mendeley