Residence-Based Geospatial Analysis of Severe Traumatic Brain Injury in a Regional Neurotrauma System

Severe traumatic brain injury (sTBI) creates a substantial burden on neurotrauma systems. While geospatial analyses of TBI have focused on injury location, less is known about how patients' residential environments contribute to regional sTBI burden. This study aimed to identify residence-based spatial clustering of sTBI across Long Island and characterize community-level factors associated with excess incidence.

We conducted a retrospective cohort study of adults (≥ 18 years) with sTBI (Glasgow Coma Scale ≤ 8) admitted to a Level 1 trauma center (2010-2021). Patient-reported home ZIP codes were mapped to ZIP Code Tabulated Areas (ZCTAs) and linked to U.S. Census American Community Survey data. Spatial autocorrelation was assessed using Global Moran's I and Local Indicators of Spatial Association (LISA). Standardized incidence ratios (SIRs) accounted for population differences. Community-level predictors of elevated SIR were evaluated using beta regression and multivariable models.

Among 608 patients, raw sTBI counts demonstrated significant spatial clustering (Moran's I = 0.228, p < 0.05). Population-adjusted analyses identified 28 ZCTAs with significantly elevated SIRs. SIR-based spatial analysis demonstrated significant clustering (Moran's I = 0.321, p < 0.05), with five high-high LISA clusters concentrated in eastern Long Island. Higher proportions of residents aged ≥ 65 and unemployment were associated with increased SIR, while lower proportions of married residents also predicted elevated incidence.

Residence-based sTBI incidence is spatially clustered independent of population size and associated with specific social vulnerabilities. These findings identify communities disproportionately contributing to neurosurgical trauma burden and inform trauma system planning, prevention, and resource allocation.