Mapping Overdose Risk in Real Time: A Risk Terrain Modeling Analysis of 911 Calls in Detroit, 2022-2024.
Journal of public health management and practice · 2025-09-19 · Observational study
Abstract
CONTEXT: Drug overdose deaths in the United States remain a leading cause of preventable mortality. Existing data systems, such as vital statistics and hospital records, often suffer from reporting delays and limited geographic resolution, hindering timely public health responses. OBJECTIVES: To identify high-risk locations for overdose-related emergency calls in Detroit, Michigan, using Risk Terrain Modeling (RTM) and publicly available 911 call data from 2022 to 2024. DESIGN: A retrospective geospatial analysis using RTM was conducted to evaluate the spatial relationship between overdose incidents and built environment features. SETTING: City of Detroit, Michigan, USA. PARTICIPANTS: Emergency call data for overdose-related incidents (N = 18 034) were analyzed. No individual-level data were used. INTERVENTION: No intervention was implemented. The study employed RTM as a geospatial method to identify environmental risk factors and predict high-risk locations for overdose events. MAIN OUTCOME MEASURE: Relative Risk Scores (RRS) generated from RTM to quantify overdose risk across 250 × 250 m grid cells in Detroit. RESULTS: Overdose-related emergency calls were spatially concentrated. RTM identified 8 significant risk factors, including ATMs, retail locations, and religious organizations. Relative Risk Scores ranged from 1 to 142.5 (mean = 9.77, SD = 8.55), with 2.7% of locations classified as very high risk. CONCLUSIONS: RTM applied to 911 call data offers a timely, place-based approach to identifying overdose risk. Public health agencies may use this method to prioritize harm reduction strategies and allocate resources more effectively.