Association between cold climates and in-hospital mortality of acute heart failure: An observational study using multilevel analysis

Cold climates have been reported to increase mortality in heart failure patients. Colder regions in Japan often overlap with rural areas where cardiologist availability is limited. However, the relationship between residing in colder regions and mortality from acute heart failure (AHF), while considering the availability of cardiologists, remains unknown. We investigated whether the cold climates were associated with in-hospital mortality after adjusting for patient factors and cardiologist availability. This cross-sectional observational study identified 38,668 AHF patients of 546 acute care hospitals from all Japanese 47 prefectures. Hospital latitude was used as a surrogate for ambient temperature, categorized into colder (>37°N), moderate (34°N-37°N), and warmer (<34°N) climate regions. Hospitals were divided into 4 groups based on the number of cardiologists per facility: none, 1 to 4, 5 to 9, and 10+ cardiologists. Three-level logistic regression analyses were performed to evaluate the associations between latitude and mortality, with 3 models examined (Model 1: null model, Model 2: patient-level factors added, and Model 3: a hospital-level factor added to Model 2) using bootstrapping methods. Model 3 showed higher odds ratio in colder regions (adjusted odds ratio 1.20, 95% confidence intervals 1.06-1.33) and lower odds ratio in warmer regions (adjusted odds ratio 0.88, 95% confidence intervals 0.79-0.98). Colder climates were found to be a risk factor of in-hospital mortality of AHF patients, even after adjusting for patient severity and cardiologist supply. Local climates should be considered when predicting patient outcomes and evaluating hospital performance.