[Skip to Navigation]
Sign In
Views 1,164
Citations 0
Invited Commentary
Public Health
July 28, 2021

Using State Hospitalization Databases to Improve Firearm Injury Data—A Step in the Right Direction

Author Affiliations
  • 1Department of Surgery, Medical University of South Carolina, Charleston, South Carolina
JAMA Netw Open. 2021;4(7):e2115807. doi:10.1001/jamanetworkopen.2021.15807

The study by Hsu and colleagues1 used a state hospitalization database to determine the rate of nonfatal firearm injury hospitalizations in New York State from 2005 to 2016. They found the overall rate of nonfatal firearm injury hospitalizations to be 18.4 hospitalizations per 100 000 population, noting significant county-level differences in incidence based on demographics, urbanicity, and changes in incidence over time. The study by Hsu et al1 contributes to an increasing effort to improve the quality of firearm injury data in the US. Firearm injury data have been problematic, as there is no nationwide comprehensive public health data set that provides accurate surveillance or robust descriptions of nonfatal firearm injuries. Many of the limitations of existing data sources are highlighted in the recent report, “First report of the expert panel on firearms data infrastructure: the state of firearms data in 2019,” published by the NORC at the University of Chicago.2 Many of the data sets used to determine national estimates of firearm injury and hospitalization are flawed by using small sample sizes to generate probability estimates, undersampling trauma centers, which treat most firearm injuries in the US, and using inaccurate or miscategorized causes of injury in administrative coding.

While we are fortunate to have accurate fatal firearm injury data reported by the Centers for Disease Control and Prevention’s Web-based Injury Statistics Query and Reporting System and National Violent Death Reporting System, which provide robust data to describe the risks and circumstances surrounding such injuries, the lack of simultaneous systems and accurate rates of nonfatal firearm injury is very limiting, given that most firearm injuries due to assault and unintentional firearm injuries are nonfatal.3 To inform our understanding of the burden of firearm injury and develop and evaluate strategies to control it, we must have access to epidemiologic data for both fatal and nonfatal firearm injuries.

Some states have invested in robust hospitalization databases or state inpatient databases to track hospital encounters for all payers, which provides an opportunity to assess detailed state-level firearm hospitalization rates. The study by Hsu and colleagues1 used state-level data from the New York Statewide Planning and Research Cooperative System to report rates of hospital encounters for nonfatal firearm injuries from 2005 to 2016. Although perhaps limited by the use of administrative and billing data and minimal reporting on intent, such a system allows for assessment of incidence over time with the ability to stratify rates based on demographics and geographic areas (ie, counties). These data can then be used to assess associations between firearm injury rates and county-level characteristics, such as income, which Hsu et al1 explored. They found notable fluctuations in the rates based on the year, with significantly higher rates for men, Black individuals, and individuals living in large metropolitan areas and counties with lower median household incomes. Although the intent behind the injury was not reported and is a limitation here, other studies have reported the fatality rate at 5% for unintentional injury and 20% for assaults, compared with 85% for suicide attempts, leading the reader to infer that most of the reported injuries reported in the database were due to assault, especially since children younger than 15 years were excluded.3 The findings of Hsu et al1 are reflective of what is reported in the literature regarding populations that are more likely to experience firearm injuries, including Black men and individuals in urban areas. They also found that 10.7% of individuals who have experienced an initial firearm injury experience a second firearm injury.3,4

Hsu et al1 observed significant variation in firearm injury incidence and changes in incidence over time among counties. This is important to note, and this is a valuable aspect of state databases or any data set that has the ability to report the county, city, or zip code of residence or location of injury. As more data emerge about the association of social determinants of health, such as unemployment, income inequality, social capital, social mobility, and community investment in social welfare, with firearm injury risk it is imperative that such investigations assess how these differ and change over time relative to firearm injury incidence, as such changes may be associated with difference between communities and provide more opportunity for targeted prevention efforts at the macro level.5 While only income level was explored in the study by Hsu et al,1 county- or zip code–level linkages to other social determinants of health or indices of disparity that are more inclusive of multiple elements of disadvantage or inequality could be further explored to better understand these differences appreciated at the county level, or even at neighborhood level, given that these may vary dramatically within counties.

This study by Hsu et al1 joins an evolving effort to use or improve existing data sources to better capture firearm injuries. The recently published RAND Corporation database of hospitalizations for firearm injury6 attempts to determine state firearm injury hospitalization rates between 2000 and 2016, using state inpatient databases, data from the Healthcare Cost and Utilization Project, and other supplementary sources, as not all states have dedicated state inpatient databases.6 It is limited by the fact that state databases are not uniform, and the analysis does not include fatal firearm injuries that result in death prior to seeking medical attention, nor does it include individuals with gunshot injuries who are treated in and released from emergency departments (EDs). However, the ability to extrapolate firearm injury hospitalizations in the US is valuable to assess trends in firearm injury and what potential factors could impact them from state to state, such as firearm legislation, firearm ownership, and community social determinants. As more states invest in and propagate violence prevention and intervention strategies, including hospital-based violence intervention programs and community violence interruption programs, state-level hospitalization rates will help assess outcomes related to such public health strategies. However, the ability to report rates at the county level will be the most helpful to assess outcomes with local interventions if they are made available in only a few communities within states.

A notable difference is reported between the study by Hsu et al1 and the RAND Corporation database,6 as the RAND database reported the rate of hospitalization for nonfatal firearm injury in New York state to be 0.84 hospitalizations per 10 000 population (ie, 8.4 hospitalizations per 100 000 population), compared with 18.4 hospitalizations per 100 000 population reported by Hsu and colleagues. This discrepancy could be influenced by the additional years included in the RAND analysis between 2000 and 2004, but it is more likely owing to the ability of Hsu and colleagues1 to include gunshot wound ED discharges in the reported hospitalization incidence. As suggested by Fowler et al,3 ED discharges may actually account for almost half of hospital-treated gunshot wounds, which may explain the more than 2-fold difference between the RAND Corporation report6 and the study by Hsu et al.1 This calls attention to the importance of establishing and reporting on comprehensive firearm injury databases that include admission and ED discharges and combining nonredundant fatal and nonfatal reports to truly understand the burden of firearm injuries and death, especially because states with high rates of firearm suicide may have a lower hospitalization rate owing to the high fatality rate of firearm suicide attempts.3

The ultimate goals for a firearm injury surveillance system is to count every single firearm injury, fatal and nonfatal, identify the setting and characteristics of treatment, and to further describe the injuries, risk factors, intent, circumstances, and outcomes to fully inform prevention, intervention, and evaluation efforts. Further work to link data sets and investment and infrastructure to mature our injury and violence surveillance systems must be done to accomplish these goals.

Back to top
Article Information

Published: July 28, 2021. doi:10.1001/jamanetworkopen.2021.15807

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Hink AB. JAMA Network Open.

Corresponding Author: Ashley B. Hink, MD, MPH, Department of Surgery, Medical University of South Carolina, 96 Jonathan Lucas St, Charleston, SC 29425 (hink@musc.edu).

Conflict of Interest Disclosures: None reported.

Hsu  YT, Chen  YW, Chang  DC,  et al.  Annual incidence of hospitalization for nonfatal firearm-related injuries in New York from 2005 to 2016.   JAMA Netw Open. 2021;4(7):e2115713. doi:10.1001/jamanetworkopen.2021.15713Google Scholar
NORC at the University of Chicago. First report of the expert panel on firearms data infrastructure: the state of firearms data in 2019. Accessed May 16, 2021. https://www.norc.org/PDFs/Firearm%20Data%20Infrastructure%20Expert%20Panel/State%20of%20Firearms%20Research%202019.pdf
Fowler  KA, Dahlberg  LL, Haileyesus  T, Annest  JL.  Firearm injuries in the United States.   Prev Med. 2015;79:5-14. doi:10.1016/j.ypmed.2015.06.002PubMedGoogle ScholarCrossref
Kao  AM, Schlosser  KA, Arnold  MR,  et al.  Trauma recidivism and mortality following violent injuries in young adults.   J Surg Res. 2019;237:140-147. doi:10.1016/j.jss.2018.09.006PubMedGoogle ScholarCrossref
Kim  D.  Social determinants of health in relation to firearm-related homicides in the United States: a nationwide multilevel cross-sectional study.   PLoS Med. 2019;16(12):e1002978. doi:10.1371/journal.pmed.1002978PubMedGoogle Scholar
Smart  R, Peterson  S, Schell  TL, Kerber  R, Morral  AR.  Inpatient Hospitalizations for Firearm Injury: Estimating State-Level Rates From 2000 to 2016. RAND Corporation; 2021. doi:10.7249/TL-A243-3
Limit 200 characters
Limit 25 characters
Conflicts of Interest Disclosure

Identify all potential conflicts of interest that might be relevant to your comment.

Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.

Err on the side of full disclosure.

If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.

Not all submitted comments are published. Please see our commenting policy for details.

Limit 140 characters
Limit 3600 characters or approximately 600 words