ConceptsCommunity-level firearm injury surveillance: Local data for local action*,**
Introduction
In the United States, firearms are used to commit homicide more frequently than all other methods combined.1, 2 Nonfatal assaults with a gun outnumber firearm homicides by a ratio of 4:1.3 According to the US Department of Justice, firearms are used to commit more than 1 million crimes each year.4
Despite the magnitude of this problem, little is known about the epidemiology of firearm crimes and assaults.5 Many states require medical personnel to report all nonfatal shootings to law enforcement agencies, but the rate of compliance with this requirement is unknown.6 One 3-city study found that 9% of medically documented cases of gunshot injuries could not be matched with a corresponding police report.3
To enhance cooperation between community emergency departments and law enforcement agencies, we created a unified firearm injury notification system for the City of Atlanta, Georgia, and 5 surrounding counties (Fulton, DeKalb, Cobb, Gwinnett, and Clayton). The system is designed to collect reports of fatal and nonfatal shootings from area law enforcement agencies, EDs, and medical examiners; link them to form a comprehensive database; and use them to monitor patterns and trends in gun violence activity throughout the metropolitan Atlanta area.7
Section snippets
ED reporting
To facilitate reporting, a central receiving point was established in the Communications Center of the Georgia Bureau of Investigation (GBI), which is located in an Atlanta suburb. Area ED providers were asked to promptly fax a simple, 1-page form to the GBI whenever they treated a victim with a gunshot wound (Figure 1).The form includes space to record the victim's name, demographic characteristics, manner of injury, type of weapon involved in the shooting, anatomic location
Law enforcement reporting
To supplement medical reporting and provide a mechanism to ensure that each shooting was investigated, area law enforcement agencies were also asked to submit a copy of their shooting reports to the GBI. Police records almost always specified the location of the event; the setting (home, street, bar, and so forth); the presumptive manner and circumstances of the shooting; offender name, description, or both; and the hospital that received the victim. A few police departments elected to mail
Medical examiner reporting
All 5 county medical examiner offices agreed to send copies of their initial field investigations to the GBI. Medical examiner reports document all firearm deaths, whether the victim died at the scene, en route to the hospital, during inpatient stay, or weeks after hospital discharge. Medical examiner data include the decedent's name, home address, and demographic characteristics, as well as the incident address, official manner of death, type of weapon involved (when known), incident
911 calls
In 1997, we began securing monthly downloads of computer-assisted dispatch data from the City of Atlanta's 911 center. The following call types were analyzed: signal 25 (shots fired), signal 69 (person armed), and signals 50, 504, and 5025 (person shot). Because Atlanta's 911 system is “enhanced” (E-911), the caller's address is automatically displayed and recorded when a 911 call is received. Center call-takers manually verify the caller's address and the incident address, if it is different.
Data linkage
Several times each week, a staff member from the Emory Center for Injury Control picked up reports from the GBI and brought them to the Center for Injury Control for data linkage (Figure 2).If the victim's name was missing or the use of an alias was suspected, a variety of alternative identifiers were used to match ED, police, and
Dissemination of reports
Using this data set, project staff periodically calculated the frequency of shootings by manner of injury, type of firearm used, victim age, and victim-offender relationship. Temporal trends were monitored in specific jurisdictions over time. The spatial distribution of shootings was plotted with geographic information system (GIS) software (Arcview 3.1 and Arcview Spatial Analyst Extension, ESRI; Redlands, CA). Fatal and nonfatal shootings were mapped by county, jurisdiction, police zone,
System impact
Shortly after our system was established, the Atlanta Police Department created a special Guns and Violent Crime Suppression Unit to target gun violence, particularly in “hot spot” neighborhoods. Other users included the Atlanta Police Department's homicide squad, the Atlanta Police Department's gang unit, and commanders of Atlanta Police Department zones with particularly high rates of gun violence. Our data have also been used by teams of agents from the Atlanta Field Office of the Bureau of
Nonmatched cases
Periodically, law enforcement agencies submit shooting reports that cannot be matched with a corresponding ED record. Despite repeated requests for data from participating hospitals and multiple searches of Grady Hospital's ED log and trauma registry, we have been unable to locate matching medical records for several of these cases. Because many “nonmatched” police reports do specify the nature or severity of the victim's injuries, it is possible that some of these victims were not shot. Other
Lessons learned
The success of our system demonstrates that it is possible to link medical and law enforcement reports to generate an accurate and timely picture of gun violence activity at the community level. With the help of GIS technology, we can identify geographic “hot spots” of gun violence and monitor trends over time. Mapping helps law enforcement to target its efforts. Mapping also helps the police determine whether gun violence is being reduced or simply displaced to other neighborhoods.
At least 3
Implications
Rather than attempt to replicate our approach with its attendant strengths and limitations, the next logical step is to create a fully automated system. This could easily be achieved with existing, “off-the-shelf” computer technology. A small but increasing number of EDs have already transitioned to a “paperless” medical record. Hundreds more use an automated tracking system to monitor throughput times and transmit a short summary of the patient's visit to his or her primary care physician. If
Acknowledgements
We thank the hundreds of medical and public safety professionals who supported the creation and maintenance of this system. We are particularly grateful for the support of Chief Beverly Harvard and the men and women of the Atlanta Police Department. Lois Mock of the National Institute of Justice was the program officer on this grant. Chris Gundry of the Greater Atlanta Data Center contributed his GIS expertise and map production services in the production of the images included in this article.
References (12)
- et al.
Firearm injury surveillance at the local level: from data to action
Am J Prev Med
(1998) Surveillance: the key to firearm injury prevention
Am J Prev Med
(1998)- et al.
Why firearm injury surveillance?
Am J Prev Med
(1998) - et al.
Firearm-related injury surveillance: an overview of progress and the challenges ahead
Am J Prev Med
(1998) Crime in the United States: Uniform Crime Reports 1996
(1997)- et al.
Injury Mortality: National Summary of Injury Mortality Data 1988-1994
(1997)
Cited by (24)
The emerging infrastructure of US firearms injury data
2022, Preventive MedicineCitation Excerpt :Recent studies comparing police and hospital recording of gunshot wound victims suggest that police data do capture cases not included in hospital data (Magee et al., 2021; Miller et al., 2022). However, hospital data may contain shooting victims unknown to police, even in localities with mandatory reporting requirements (Kellermann et al., 2001). Police agencies with gunshot detection systems may learn of some gun crimes that would otherwise have gone unreported.
Preventing violence - Caring for victims
2007, SurgeonUnintended shootings in a large metropolitan area: An incident-based analysis
2003, Annals of Emergency MedicineCitation Excerpt :For most of the remaining cases, the only documentation available was the 1-page reporting form that lists victim age, sex, and race, as well as the manner and severity of injury. The form does not provide enough information to classify the specific circumstances of the event.10 Eighty-seven percent (187) of the 216 victims were male, 65% (141) were black, 24% (51) were white/non-Hispanic, and 8% (18) were members of other racial or ethnic groups.
Docs and cops: A collaborating or colliding partnership?
2001, Annals of Emergency Medicine
- *
System development was supported by a grant (No. 95-IJ-CX-0025) from the National Institute of Justice, US Department of Justice. Continuation funding was provided by the John D. and Catherine T. MacArthur Foundation and the Harvard Injury Control Research Center.
- **
Address for reprints: Arthur L. Kellermann, MD, MPH, Department of Emergency Medicine, Emory University School of Medicine, 1365 Clifton Road NE, Suite 6200, Atlanta, GA 30322; 404-778-2600; E-mail [email protected]