The number of deaths attributable to any given EHE is generally calculated using statistical estimates of expected and excess mortality (Henderson et al., 2021). Under ideal circumstances, the number of excess deaths estimated for an EHE would be consistent with the number of deaths in provincial and national vital statistics records with an ICD-10 code of X30 or T67 during the same period. Furthermore, there would be perfect agreement between the EHE deaths identified by vital statistics, probabilistic methods, and enhanced surveillance when multiple approaches are available.
This case study highlights the value of having vital statistics records with the X30 code while demonstrating that different approaches to attributing individual EHE mortality produce somewhat discordant results. We found that 102 of 353 individual deaths were attributed by at least one of three approaches, and approximately 50 of these deaths were excess based on statistical estimates. There were 36 deaths with an X30 code, 49 attributed by the probabilistic method, and 58 identified through enhanced surveillance. The overlap between two or more approaches was limited to 32 deaths, and 70 were only identified by a single approach. We also found discrepancies in the types of deaths identified by the probabilistic and enhanced surveillance approaches, suggesting that both may have missed some important indicators of deaths due to EHE.
Each approach has strengths and limitations. It is critically important to have appropriate ICD-10 codes reflected in national vital statistics records so that deaths due to EHE can be identified in the same way as deaths due to most other causes in Canada. However, cause of death information is subject to significant delays in vital statistics data and cannot facilitate the near-real-time analyses possible with the probabilistic and enhanced surveillance methods described here. The ideal future state may combine these types of approaches to ensure that rapid assessments are possible, and that vital statistics records appropriately reflect individual deaths due to EHE.
The experience in BC during the 2021 heat dome provides a framework on which a more systematic future approach could be modelled. Early in the event, the BC Coroners Service (BCCS) sent a reminder to clinicians that any death with extreme heat as a possible contributor should be reported for investigation. Soon after the event, the BCCDC estimated 740 excess deaths during the EHE (Henderson et al., 2021) and conducted epidemiologic analyses on the deaths most likely due to the extreme heat (Henderson et al., 2022). Since then, BCCS, BC Vital Statistics, and BCCDC have worked together to ensure that the mortality impacts of the EHE are well understood and that most deaths will be appropriately attributed. To date, the BCCS has confirmed at least 619 heat-related deaths that will be recorded with X30 or T67 in the provincial vital statistics data (BC Coroners Service, 2022).
More generally, Environment and Climate Change Canada provides hot weather notifications across the country. Such notifications could be used to trigger regional public health reminders for clinicians and coroners to consider the role of high temperatures in any deaths they attend. If heat may have been a causal or contributing factor, noting it on the certificate of death creates the opportunity for future review and appropriate attribution. A death simply cannot be coded as X30 or T67 by vital statistics agencies unless this information is included on the certificate of death they receive. Coroners, clinicians, public health, vital statistics, and other agencies could then work together after the event to review specific cases, especially those where the role of heat is less clear. This type of systematic and cooperative approach would facilitate improved understanding of EHE and the burden of mortality, risk factors, and potentially effective interventions in Canada.