T. Byrd1, L. Lu1, S. Boland1, D. Silver1, M. Neal1, J.B. Brown1 1University of Pittsburgh, Surgery, Pittsburgh, PA, USA
Introduction: Both distance from a trauma center and social determinants of health (SDoH) have each been shown to impact outcomes after injury. However, little is known regarding if and how proximity and SDoH interact, or whether one mediates the effect of the other on outcomes. Additionally, measures of institutional racism have been demonstrated to be associated with health outcomes across multiple disciplines; however, no studies have evaluated the impact on trauma patients. Our objective was to evaluate the relative contributions of trauma center proximity and SDoH to in-hospital mortality after trauma.
Methods: A retrospective cohort study was performed including scene patients at a quaternary level 1 trauma center between 2018–2023. Transport driving distance to the trauma center was calculated using GIS network analysis. We evaluated 2 measures of SDoH linked to patient zip codes: The area depravation index (ADI) and a redlining index constructed using multi-level modeling of home mortgage loan data. Patient-level multivariable logistic regression determined the association of in-hospital mortality with distance from trauma center and ADI, as well as tested the interaction between these variables. We adjusted for age, injury severity score, mechanism of injury, and systolic blood pressure. A zip-code level spatial autoregressive model determined the association between neighborhood mortality rate and redlining index while accounting for spatial correlation of mortality and redlining in neighborhoods. Finally, causal mediation analysis was performed to quantify the proportion of the effect of distance on mortality mediated by ADI and vice versa.
Results: A total of 9,592 patients were included. An increase in the odds of mortality was associated with increasing distance per mile (aOR 1.008; 95%CI 1.003-1.013, p<0.001) and increasing (more disadvantaged) ADI aOR 1.069; 95%CI 1.04-1.0137, p=0.036) at the patient-level. The interaction between distance and ADI was not significant (p=0.194). Residence in a redlined neighborhood (Figure) was positively associated with mortality rates (coef 0.027; 95%CI 0.005, 0.049). Mediation analysis demonstrated ADI mediated 5% of the effect of distance on mortality while distance mediated 11% of the effect of ADI on mortality.
Conclusions: Trauma center proximity and SDoH, including measures of structural racism, appear to contribute independently to trauma mortality. Both distance and SDoH should be considered separately when evaluating outcomes after injury. Additional efforts should be taken to understand how social policy impact health outcomes.