Background

Every day, hundreds of people are injured by firearms in the United States; between 2009 and 2017, an annual average of 85,694 people were treated in the emergency department (ED) for firearm-related injuries¹. In 2019, firearms accounted for 39,707 U.S. deaths, a rate of 12.1 people per 100,000²; in comparison, the rate of motor vehicle traffic deaths in the same year was 11.5 per 100,000³. Males account for 86% of U.S. firearm-related deaths and injuries, particularly those between the ages of 15 and 34 (39% of total)^4,5. Most patients who present alive to the ED for a firearm-related injury have been wounded in an assault (49.5%) or unintentionally (35.3%). Attempted suicides account for 5.3% of injuries⁶, though they are likely underreported due to stigma and fear of reprisal or institutionalization⁷.

The medical costs of firearm injuries can be substantial for the injured. Ranney et al. demonstrated that the average per member initial ED healthcare cost for patients with firearm injuries who were discharged from the ED was $5,686, while the healthcare cost for a hospital admission, inclusive of an ED visit, for persons with firearm injuries requiring a hospital stay was $70,644⁸. Compared to the six months before a firearm injury, the average per patient cost in the six months after it (including costs from the day of the injury) increased by 347% for those discharged from the ED and 2,138% for those hospitalized⁸.

Current literature on the costs of firearm injuries focuses primarily on the relatively immediate medical and societal costs of firearm deaths. There are few studies examining healthcare costs following a nonfatal firearm injury, and even fewer using adjudicated healthcare claims data. This study examined healthcare costs in the six months preceding and following more than 2,500 nonfatal firearm injury events (FIEs) among commercially insured patients using insurance claims data, which captures the total payer cost of care for firearm injuries more completely than hospital charges alone.

Findings

Substantial healthcare costs persist following a firearm injury event

Within our study population, most of the cost for a nonfatal firearm injury event^† (FIE) occurred during the initial treatment of the injury (i.e., the admission or the ED visit following the injury), with an average cost of approximately $30,121 per patient (all costs reported in 2019 dollars). However, healthcare costs in the six months following the initial treatment of the nonfatal FIE remain at higher levels than healthcare costs in the six months preceding the FIE (Figure 1). The average cost of care for the six months following the FIE was 2.7^* times the average of the six months preceding it (Figure 1). When the population is split into those who were admitted during their FIE treatment and those who had an ED or hospital observation (Obs) visit that did not result in an admission, the differences in the initial FIE treatment cost are substantial ($65,568, admitted for FIE, versus $5,648, ED/Obs only for FIE). The patients who were admitted due to an FIE also have higher longer-term healthcare costs, with average costs in the six months after the FIE at 3.1^* times the average costs in the six months preceding it; for those with only an ED or hospital observation stay, the ratio is 2.0^*.

While inpatient costs, including the initial treatment of the FIE, are the largest contributor to the total cost of care, higher healthcare costs persist in the six months following the FIE across all service categories. For example, the cost of outpatient facility care was 3.1^* times higher for the total population in the six months after the FIE compared to the six months preceding, and professional/other medical services were 2.2^* times higher (Figure 2).

The increases in average cost of care for the six months following the FIE compared to the six months preceding it were driven largely by the select service categories shown in Figure 3.

Young adult men more likely to sustain a nonfatal firearm injury than women

The observed rate of any FIE in our 2019 adult commercially insured database was 10.8 per 100,000. In our 2017-2019 study population, men experienced more than six^* times the number of nonfatal FIEs than women overall (Figure 4), with the highest number of male FIEs (45%) sustained by those aged 18 to 29 (Appendix Figure 7). Most diagnosed nonfatal FIEs were classified in our commercial data source as Unintentional (70.5%), followed by Attempted Homicide (24.5%), with Attempted Suicide being the lowest (5.0%). See Appendix, Figure 7).

Discussion

Nonfatal firearm injuries carry a substantial cost burden, both in the initial treatment of the injuries and in the subsequent six months. In our study, the average cost for the initial treatment of a nonfatal FIE was around $30,121 per patient, and per patient per month (PPPM) costs were 2.7 times larger in the six months following the initial treatment than in the six months preceding it ($2,824 vs. $7,674). Certain healthcare services, like surgical care, both inpatient (IP) and outpatient (OP), IP medical admissions, ED and observation stays, physical therapy (PT)/occupational therapy (OT)/speech therapy (ST), psychiatric treatment, and home healthcare costs were the primary drivers of these increases. In a commercially insured population, most of these costs are borne by the payer (e.g., health plan or employer) while patients bear some portion of these costs through copays, deductibles, and coinsurance.

Given the direct medical costs of nonfatal firearm injuries, healthcare stakeholders have an interest in supporting further research to mitigate firearm injury risk.

Further research is needed to understand the full scope and costs of nonfatal firearm injuries. Research questions of note that can be informed by large administrative claims data sets include:

• How could predictive analytics be used to identify firearm injury risk?
• How does the incidence and cost burden of nonfatal firearm injuries vary by payer type (e.g., Medicare, Medicaid, commercial, etc.)?
• How do health outcomes and costs of nonfatal firearm injuries vary by population (e.g., geography, demographics, or racial or socioeconomic groups)?
• How do nonfatal firearm injuries affect the health outcomes and costs of family members?
• What is the long-term cost burden of nonfatal firearm injuries beyond the first six months following the event?
• What segments of the healthcare delivery system bear the burden of incidence and cost of nonfatal firearm injuries over the long term (e.g., long-term care facilities)?

Data source and methodology

Data source

Milliman Consolidated Health Cost Guidelines Sources Database (CHSD)+

The Milliman CHSD+ database contains proprietary historical claims experience from several of Milliman’s Health Cost Guidelines™ (HCG) data contributors. The database contains annual enrollment and paid medical and pharmacy claims for over 50 million commercially insured individuals covered by the benefit plans of large employers, health plans, and governmental and public organizations nationwide. Other groups available in the database include Medicare Advantage and Medicaid. We used data years 2017 to 2019 for this analysis.

Methodology

We identified commercially insured patients with firearm-related initial encounters using ICD-10-CM diagnosis codes in the 2017-2019 CHSD+ data set. Study patients reported at least six months of continuous enrollment both before and after the initial encounter. The FIE was identified as the first inpatient or ED claim with a firearm-related initial encounter ICD-10-CM diagnosis code (code list available upon request). If a member had more than one FIE, only the first was included in the analysis. Patients were stratified into four age groups (under 18, 18-29, 30-49, and 50-64 years), by firearm injury types (suicide, homicide, unintentional, legal, and undetermined), sex (male and female), and site-of-service for their FIE (admitted for an inpatient stay for their FIE, or ED/Observation visit only). Claims with undetermined and legal injury types (about 3%) were included in the unintentional cohort. Claims coded with multiple firearm injury types were assigned to a single category using the following hierarchy: suicide, homicide, unintentional. Enrollment and PPPM costs were assessed over the six months prior, six months following, and during the initial firearm injury treatment event. Allowed costs were trended monthly to July 1, 2019, at an annual rate of 2.5%.

To account for high-cost and low-cost outliers, we winsorized the population at the 5th and 95th percentile of total allowed costs by FIE site-of-service from the six months pre-FIE to six months post-FIE. We note that the cost distribution for patients with FIEs is skewed right when all patients are included, resulting in a mean cost that is larger than the median cost. Winsorizing at the 5th and 95th percentiles therefore lowers the total average allowed cost. Patients with total allowed costs below the 5th percentile and above the 95th percentile for each site-of-service were excluded from figures throughout the body of the paper. The winsorization thresholds are shown in Figure 5.

We used t-tests to determine statistical significance between pre-measurements and post-measurements within cohorts (admitted for FIE and ED/Obs only) and did not compare between cohorts. Comparisons within cohorts reduces the likelihood of confounding factors. We did not attempt to control for potential confounding variables that may impact differences in cost between the periods before and after an FIE other than the FIE itself.

Caveats and limitations

We used administrative claims data for this study, reflecting healthcare services paid by an insurer, which only includes diagnosed firearm injuries. Our study does not capture firearm injuries that do not result in an insurer-paid healthcare encounter or events that were not coded with a firearm injury diagnosis code. We segmented the population into firearm injury type. However, we did not provide cost analysis by firearm injury type due to sample size and credibility concerns. This study also excludes all firearm injuries that resulted in death either during initial treatment of the firearm injury or in the six months following the event. Our results have not been geographically or demographically adjusted and reflect the observed populations and geographies represented in the source data.

Milliman has developed models to estimate the values included in this report. The intent of the models was to estimate the healthcare costs associated with initial nonfatal firearm injuries. We have reviewed the models, including their inputs, calculations, and outputs, for consistency, reasonableness, and appropriateness to the intended purpose and in compliance with generally accepted actuarial practice and relevant Actuarial Standards of Practice (ASOPs, particularly ASOP 56, Modeling).

The models rely on data and information as input to the models. The results presented in this report are based on an insured sample population. Results could vary across different populations for a variety of reasons, potentially including differences in socioeconomic status, population health status, reimbursement levels, delivery systems, random variation, or other factors. The models, including all input, calculations, and output may not be appropriate for any other purpose.

The American Academy of Actuaries requires its members to identify their credentials in their work product. Harsha Mirchandani is a member of the American Academy of Actuaries and meets its relevant qualification requirements.

---

Appendix

---

^† A firearm injury event includes the first claim in the study period for an inpatient admission, ED visit, or other outpatient encounter with a firearm injury initial encounter diagnosis code. Time from initial treatment of FIE begins preceding or following the date of service for the ED visit and outpatient encounters or admission/discharge date for inpatient admissions.

* Statistically significant at α < 0.01; see Methodology section below for limitations.

References

¹ E. J. Kaufman, D. J. Wiebe, R. A. Xiong, et al. Epidemiologic Trends in Fatal and Nonfatal Firearm Injuries in the US, 2009-2017. JAMA, vol. 181, no. 2, pp. 237-244, 2021.

² Centers for Disease Control and Prevention (CDC). National Center for Health Statistics: All Injuries: Mortality: All Firearm Deaths. Retrieved June 24, 2022, from https://www.cdc.gov/nchs/fastats/injury.htm.

³ Centers for Disease Control and Prevention (CDC). National Center for Health Statistics: Accidents or Unintentional Injuries. Retrieved June 24, 2022, from https://www.cdc.gov/nchs/fastats/accidental-injury.htm.

⁴ Centers for Disease Control and Prevention (CDC). Fast Facts: Firearm Violence Prevention. Retrieved June 24, 2022, from https://www.cdc.gov/firearm-violence/data-research/facts-stats/index.html.

⁵ K. A. Fowler, L. L. Dahlberg, T. Haileyesus and J. L. Annest. Firearm injuries in the United States. Prev Med, vol. 79, pp. 5-14, 2015.

⁶ F. Gani, J. V. Sakran and J. K. Canner. Emergency Department Visits for Firearm-Related Injuries in the United States, 2006–14. Health Affairs, vol. 36, no. 10, 2017.

⁷ M. M. McClay, S. S. O'Connor, B. N. Groh, et al. Denial of Suicide Attempt Among Hospitalized Survivors of a Self-Inflicted Gunshot Wound. Psychiatric Services, vol. 69, no. 6, 2018.

⁸ M. L. Ranney, C. Herges, L. Metcalfe, et al. Increases in Actual Health Care Costs and Claims After Firearm Injury. Annals of Internal Medicine, 2020.

⁹ UC Davis Health. Facts and Figures. Retrieved June 24, 2022, from https://health.ucdavis.edu/what-you-can-do/facts.html.

¹⁰ S. A. Spitzer, D. Vail, L. Tennakoon, et al. Readmission risk and costs of firearm injuries in the United States, 2010-2015. PLOS ONE, vol. 14, no. 1, p. e0209896, 2019.

¹¹ United States Government Accountability Office (GAO). Firearm Injuries: Health Care Service Needs and Costs (GAO-21-515), 2021.