Understanding disparities in firearm mortality: The role of person- and place-based factors.

Abstract

Background: Racial and socioeconomic disparities in firearm homicide rates are well-established in the United States. However, findings have been mixed regarding disparities for in-hospital mortality among firearm injury patients. The aim of this study was to evaluate the extent of in-hospital mortality disparities and whether differences persist after adjusting for person- and place-based factors.

Methods: This retrospective analysis evaluated all pediatric and adult patients admitted to a single level I trauma center with a statewide catchment area from 2010 to 2020. Patients with assault-related firearm injuries were included; those with accidental or self-inflicted firearm injuries were excluded. The primary outcome was in-hospital mortality. Predictors included demographic (i.e., race, sex, age), socioeconomic (i.e., health insurance), injury (i.e., severity), and area-level (i.e., community distress, social vulnerability, rurality/urbanicity) characteristics.

Results: The sample consisted of 2,081 patients with assault-related firearm injuries, including 1,836 Black patients (88 %) and 1,838 males (88 %). The mean age was 32.3 (SD=11.9) years. A smaller proportion of Black (19 %) compared to White (27 %) patients had health insurance coverage. Among injury patients, there were 210 firearm deaths (10 %). In logistic regression analyses adjusting for demographic, injury, and socioeconomic characteristics, both insured patients and those with unspecified insurance status had lower risk of mortality than uninsured patients; these differences in mortality risk remained after accounting for potential survivor bias. Contrary to expectation, there were no racial differences in mortality risk. In multilevel models accounting for nesting of patients within geographic areas (i.e., zip codes, counties), differences in mortality risk by insurance status remained after accounting for community distress, social vulnerability, and rurality/urbanicity. However, racial and area-level differences in mortality risk emerged after accounting for survivor bias.

Conclusions: The present findings are consistent with research showing lower in-hospital mortality among insured compared to uninsured trauma patients. Notably, this reduced mortality risk remained after controlling for important social determinants of trauma outcomes, and extended to patients with unspecified insurance status. Future research is needed to identify person- and place-based factors that could help to explain and mitigate differences in mortality risk based on insurance status.