The unrecognized burden of patient-level social determinants of health on health outcomes following traumatic injury.

Abstract

Social determinants of health (SDOH) and their relationship to patient outcomes have become center-points in efforts to improve healthcare. Most studies rely on aggregated datasets and proxy measures, often using geographic location to assess SDOH, but this approach is prone to ecological fallacy. Using patient-level data, this study sought to analyze the correlation between SDOH and specific outcomes in trauma. Patient demographics, social work interview data and trauma admission data were combined for the years 2019-2020 at a large urban academic level 1 trauma center. Missing data were estimated using multiple imputation. Post-operative outcomes were studied including mortality, major complication, readmission, length of stay (LOS), and discharge location. Multivariable analysis was performed by creating models that included the SDOH. 868 patients with social work evaluations were identified. The mean age was 60.8 years, 62.2 % were male and 77.4 % were white. Financial concerns was the most common SDOH in this population (33.4 %). In multivariable analysis, financial concerns and interpreter requirement were associated with major complication and mortality. Psychiatric history was associated with major complications. Alcohol use was associated with major complications and increased LOS. Male sex, black race, and motor vehicle collision were also associated with an increased LOS. At the patient level, certain demographics and SDOH are predictors of negative outcomes. Early identification of social barriers to discharge in trauma patients may help facilitate disposition, reduce LOS, and help identify patients at risk for post-operative complications and mortality. The mechanism by which SDOH affect post-operative outcomes should be further evaluated.