Identifying heat-related diagnoses in emergency department visits among adults in Chicago: a heat-wide association study

2026-01-23

AI Summary

This study addresses the limitations of traditional research, which is often constrained by predefined diagnostic categories and thus unable to comprehensively identify emergency conditions associated with extreme heat. Leveraging over 900,000 adult emergency department visits in Chicago from 2011 to 2023, the authors conduct the first systematic screening across the full diagnostic spectrum—encompassing 1,803 ICD codes—for heat-related emergencies. Employing a two-stage analytical strategy, they first apply quasi-Poisson regression for initial screening, followed by distributed lag nonlinear models (DLNMs) integrated with a time-stratified case-crossover design to assess both same-day and short-term cumulative effects of extreme high temperatures. Beyond confirming established associations such as heat stroke and dehydration, the study uncovers previously underrecognized heat-sensitive conditions, including hypotension, edema, acute kidney injury, and multiple injuries, substantially expanding the understanding of health risks posed by extreme heat.

Technology Category

Machine Learning: Multi-class/Multi-label Learning & Extreme ClassificationKnowledge Representation and Reasoning: Diagnosis and Abductive ReasoningHumans and AI: Brain-Sensing and Analysis

Application Category

Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphsSearch and Retrieval-Augmented AI: Web evaluation methodologies and metricsSecurity and Privacy: Large-scale security measurements

Abstract

Extreme heat is an escalating public health concern. Although prior studies have examined heat-health associations, their reliance on restricted diagnoses and diagnostic categories misses or misclassifies heat-related illness. We conducted a heat-wide association study to identify acute-care diagnoses associated with extreme heat in Chicago, Illinois. Using 916,904 acute-care visits -- including emergency department and urgent care encounters -- among 372,140 adults across five healthcare systems from 2011-2023, we applied a two-stage analytic approach: quasi-Poisson regression to screen 1,803 diagnosis codes for heat-related risks, followed by distributed lag non-linear models in a time-stratified case-crossover design to refine the list of heat-related diagnoses and estimate same-day and short-term cumulative odds ratios of acute-care visits during extreme heat versus reference temperature. We observed same-day increases in visits for heat illness, volume depletion, hypotension, edema, acute kidney failure, and multiple injuries. By analyzing the full diagnostic spectrum of acute-care services, this study comprehensively characterizes heat-associated morbidity, reinforcing and advancing existing literature.

Problem

• heat-related illness
• emergency department visits
• extreme heat
• diagnosis misclassification
• heat-health association

Innovation

• heat-wide association study
• distributed lag non-linear model
• case-crossover design
• extreme heat
• diagnostic spectrum