Current Issue Pic Top Harvard Health Policy Review Current Issue Top
Current Issue Pic Middle About Us Fill Current Issue Bottom
Contact Us Fall 2001; Volume 2, Number 2
Feature: Violence and Healthcare

Linking Data to Save Lives: Recent Progress in Establishing a National Violent Death Reporting System

Deborah Azrael, Ph.D., Catherine Barber, MPA, and James Mercy, PhD

page 1 | page 2

U.S. homicides and suicides has been hampered by the paucity of information about the characteristics of these events that can be derived from publicly available data. Researchers attempting to evaluate policies aimed at reducing violence or better understanding its etiology have generally relied on the limited outcome data that are available from the National Vital Statistics Mortality System and the Federal Bureau of Investigation’s Uniform Crime Reports (UCR) or on costly case control studies with limited generalizability. As a consequence, public discourse on violence prevention often takes place in an information vacuum and the development of effective prevention programs and policies has been inhibited.

In many other public health arenas, researchers rely on surveillance systems to provide them with rich data with which to characterize the epidemiology of particular conditions and to assess the effectiveness of interventions in ameliorating problems. Such surveillance systems are at the very heart of the public health enterprise, yet the U.S. currently lacks a state-based, national surveillance system for this leading cause of death.1

Establishing a National Violent Death Reporting System

The enormous benefits of a state-based, national reporting system to track the incidence and characteristics of health conditions have been well-established in other areas, such as infectious disease and motor vehicle safety. For example, the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System (FARS) collects over 125 pieces of information on each of the approximately 40,000 motor vehicle-related fatalities that occur annually - including features of the crash, the vehicles involved, the road and weather conditions, and the people involved. FARS has been used to evaluate the effectiveness of speed limit laws, minimum age drinking laws, air bag and seat belt effectiveness, and numerous other state and federal initiatives over the past 3 decades.2-23 Our ability to maintain, refine or change policies based on the results of these analyses is clearly one of the reasons that the U.S. motor vehicle fatality rate has fallen dramatically over the past 20 years.

The utility of FARS and other surveillance systems has led many to call for the creation of a comparable reporting system for violent deaths. In 1999, for example, the Institute of Medicine (IOM) called for a national data system for homicides and suicides to provide objective data with which to monitor trends and evaluate the effectiveness of prevention programs and policies.24 These calls have not gone unheeded: in his recent National Strategy for Suicide Prevention: Goals and Objectives for Action, the Surgeon General set as an objective “[implementation of]a national violent death reporting system that includes suicides and collects information not currently available from death certificates.”25

Preliminary Efforts to Establish a National Reporting System
Efforts to establish a national reporting system for violent injuries have a history going back more than 10 years. Beginning in the mid 1990’s, CDC, which had started to focus on violence as a public health problem, supported the development of violence-related injury surveillance systems at the national, state, and local level, culminating in their funding 7 states to conduct local surveillance in 1994.26

In 1999, convinced that basic data collection was a federal responsibility, but aware that federal efforts to establish a national surveillance system were stalled, 6 private foundations (under the leadership of the Joyce and Soros Foundations), funded the Harvard Injury Control Research Center (HICRC) at the Harvard School of Public Health to establish the National Firearm Injury Statistics System as a method of jump-starting the federal initiative. Originally conceived as a firearm injury surveillance project, the scope of the pilot was soon expanded to include all violent deaths (homicides and suicides by any means, and unintentional firearm injuries) and renamed the National Violent Injury Statistics System, NVISS.

The NVISS Pilot

NVISS is working with 13 state and local grantees to design and pilot this reporting system, which in many respects mirrors the Fatality Analysis Reporting System (FARS) for motor vehicle-related deaths. The NVISS model builds on models developed by the Medical College of Wisconsin and other sites and links information from four major reporting sources: death certificates, coroner/medical examiner reports, police Supplementary Homicide Reports (and, in some jurisdictions, incident reports), and crime laboratory data. The system serves to combine previously collected but unlinked data from these sources and thus provides rich, heretofore unavailable, information about the circumstances of suicides and homicides (includng those that are school-related, domestic violence-related and drug-related). For example, information is collected about the relationship between victims and suspects in violent crimes, and about characteristics of the events. These include where they took place, whether or not the victim was intoxicated, and if the victim or perpetrator was a minor, whether or not their access to a weapon was authorized. The heart of the linkage is a set of uniform data elements used by all pilot sites to collect over 50 pieces of information about each incident in their jurisdiction that results in a violent death.

NVISS funds programs conducting statewide surveillance in Connecticut, Kentucky, Maine, Maryland, Michigan, Utah and Wisconsin and local surveillance in Allegheny County and Bethlehem, Pennsylvania; Miami-Dade County, Florida; metropolitan Atlanta, San Francisco and Youngstown, Ohio. Together, the programs developed uniform data elements for fatal violent injuries that are outlined in a detailed manual, and NVISS has developed incident-based relational software for the system. The sites are currently piloting the fatality reporting system, collecting data on an estimated 4,000 violent deaths in 2001.

The Expert Meeting on Firearm Injury Reporting

The system being developed and tested by NVISS has served as a useful starting point for planning a national violent death reporting system, a process that was moved significantly forward in May of 2000 when the Joyce Foundation and the Harvard School of Public Health co-sponsored an Expert Meeting on Firearm Injury Reporting in Washington, DC, to discuss the development of a national reporting system for firearm injuries. The goal of the Expert Meeting was to move the federal government toward implementing a firearm injury reporting system by bringing together key representatives of the federal agencies concerned with firearm and other violence-related deaths, the fields of injury and surveillance, and public health and public safety.

A key outcome of the meeting was a consensus document outlining a set of 10 core principles of a national reporting system for violent deaths. Among the principles agreed upon at the meeting was that the system should be state-based and administered at the national level by the CDC. Participants also agreed that the scope of the system should include all violent deaths (as well as unintentional and undetermined firearm deaths), with data coming from multiple criminal justice and public health sources at a level of detail comparable to that available through the FARS system.

page 1 | page 2
Spring 2001, Volume 2, Number 1
Table of Contents
Editor's Note
Features: Violence and Healthcare
Gun Violence
Health Highlights
In Focus

about us | links | contact us | subscribe | epihc