Examining Heat-Related Emergency Department Visits in North Carolina, 2009-2017

Abstract

Background: Climactic changes in heat are expected to affect patterns of heat-related illness (HRI) in North Carolina. In order to understand patterns in HRI over time, we assessed the frequency and distribution of HRI emergency department (ED) visits across North Carolina during 2009-2017. Methods: Using a previously validated ICD-9/ICD-10 and chief complaint-based case definition, we identified visits from the North Carolina Disease Event Tracking and Epidemiologic Collection Tool (NC DETECT) from January 1, 2009 to December 31, 2017. The analysis used a generalized linear model for time series of counts with R package tscount. A poisson model was used with response variable as number of daily HRI ED visits. We assessed whether number of peaks (i.e., aberrations) increased annually from 2009 to 2017. Finally, to assess whether length of the heat season increased annually 2009 to 2017, the duration between date of the first peak and last peak were compared annually. Results: Most emergency department visits occurred during summer months. Monthly distributions for each year were similar. During 2010 and 2015, the monthly highs occurred slightly earlier than other years. In the Poisson model, controlling for maximum monthly temperature, the coefficient for year was positive with a significant p-value, suggesting an expected increase in the number of HRI ED visits over time. No significant distribution in the number of peaks by year or duration of heat season was observed. Conclusions: The model demonstrated a significant increase in the number of HRI ED visits over time. The results of this study reinforce the strong relationship between increasing temperatures and HRI, and suggests that as temperatures continue to increase, the incidence of HRI will likewise increase.