A. Barone-Camp1,3, K. Dickinson1,3, D. Garofalo1, V. McCarthy1,2,3, C. Heron1, Q.W. Myers1,2,3, C.G. Velopulos1,2,3 1University of Colorado School of Medicine, Department Of Surgery, Aurora, CO, USA 2Firearm Injury Prevention Initiative, Aurora, CO, USA 3Injury and Violence Prevention Center, Aurora, CO, USA
Introduction: While the built environment and its relationship to social determinants of health is being increasingly explored in surgery, little is known of the impact of place of injury on outcomes. While the Social Vulnerability Index (SVI) is designed to quantify vulnerability based on census tract of home address, and is generally applied to people or communities, questions have been raised about the impact of social vulnerability of the neighborhood where one is injured. In this study we assign SVI to both the individual (iSVI) and the place of injury (pSVI) to examine the contribution of each.
Methods: We retrospectively studied MVA patients > 18 years in our Level 1 Trauma Registry from 1/1/2017-12/31/2022. We assigned SVI scores using patient address and injury location. We utilized both quartiles and dichotomized SVI (>75%ile vs <75%ile) in our analysis. The primary outcomes were death in the emergency department (ED), inpatient mortality, hospital length of stay (LOS), and readmission. Secondary outcomes were ED to in-hospital destinations and discharge disposition. We used chi-square tests for categorical variables, pooled sample t-tests and Kruskall-Wallis Rank sum tests for continuous variables.
Results: Our sample included 2,313 patients with complete home address information. Of those, 1,750 also had information about their injury location. Whether using iSVI or pSVI, patients with high and low vulnerability had no difference in age, shock index, inpatient mortality, admission to ICU, or discharge location. Blood alcohol levels were different for high SVI whether using iSVI or pSVI (mean 73 to 98.2, p<.001 and mean 70.4 to 98.2, p<.001). While there was no difference in ED deaths or LOS between high and low iSVI, there were significantly more ED deaths for high pSVI (4.4% vs. 1.8%, p=.001), and LOS was slightly shorter (3 [1.0-9.0] vs 4 [1.0-11.0], p=.002). Comparing high iSVI vs. high pSVI, there were no differences in any of the outcomes, with no change in the magnitude or direction of the associations. Sensitivity analysis using quartiles instead of dichotomized SVI yielded similar results.
Conclusion: As there was minimal to no difference between iSVI and pSVI for in-hospital outcomes and discharge disposition, the addition of pSVI does not significantly change our understanding of how vulnerability impacts our patients once admitted. However, the finding of more ED deaths for high vulnerability places of injury versus low vulnerability places suggests that environment-specific characteristics of injury location may affect how pre-hospital care is delivered and progressed to definitive care. Next steps are to assess for differences in scene time and time to definitive care, number of interventions attempted/performed, and EMS providers’ perceived ability to care for the patient in high vulnerability areas.