Evaluation of a Canadian heat illness-related emergency department visit surveillance system

OBJECTIVES: In the summer of 2021, the Canadian province of British Columbia (B.C.) experienced an unprecedented extreme heat event, which resulted in 619 heat-related deaths across the province (Henderson et al., 2022). This evaluation aims to assess the sensitivity and specificity of a surveillance algorithm to identify heat illness-related emergency department (ED) visits during this period, as well as to explore the impact of adjusting the case definition to enhance algorithm performance. METHODS: A chart review of ED visits across a regional health authority in B.C. during the 2021 period of extreme heat was conducted. This was considered the most accurate information to which the algorithm was compared. Sensitivity and specificity were calculated for the original algorithm, as well as for an expanded algorithm including heat-related illnesses and the following additional diagnoses: syncope, altered level of consciousness, acute kidney injury, acute renal failure, and general weakness. RESULTS: The original algorithm focusing on explicitly recorded heat illnesses has a sensitivity of 59% and a specificity of 100%. The expanded algorithm produced a sensitivity of 80% and a specificity of 96%. CONCLUSION: Surveillance sensitivity for heat-related illness may be substantially impacted by the inclusion of diagnoses not explicitly naming effects of heat. Other jurisdictions could consider expanding heat-related illness definitions to include additional heat-associated diagnoses to gain a more comprehensive understanding of the impact of heat on health to support monitoring and action.