Meteorological associations with out-of-hospital cardiac arrest: a national population-based time-series analysis

Abstract

Objectives: Meteorological factors may influence cardiovascular emergency incidence, but comprehensive national evidence for out-of-hospital cardiac arrest (OHCA) associations remains limited. We investigated meteorological associations with OHCA occurrence using complete national population data. Study design: Population-based time-series retrospective, non-interventional analysis. Methods: We conducted a population-based time-series analysis using the Hungarian National Ambulance Service registry from November 2018 to December 2023. After excluding COVID-19 disruption period, 114830 OHCA cases across 1584 days were analysed. Meteorological parameters included temperature, wind speed, atmospheric pressure, humidity, and air quality. Associations were assessed using negative binomial regression models with temporal lag structures (0–3 days). We used a rolling 30-day z-score to detect outlier days with high OHCA cases and identified their unique weather conditions. Machine learning validation was performed with XGBoost and SHAP interpretation. Results: Daily OHCA incidence averaged 60⋅9 ±14⋅3 cases, peaking in winter (17⋅8 % higher than summer, p < 0⋅001). Each 1 ◦C temperature decrease was associated with a 1⋅4 % increase in daily OHCA incidence (IRR 0⋅986). Wind speed demonstrated inverse association (7⋅9 % decrease in OHCA incidence per-IQR effect; IRR 0⋅928). The highest-incidence days saw 31⋅9 % more cases, equivalent to 19 additional cases daily, linked to adverse weather. Conclusion: Meteorological factors demonstrate strong, predictable associations with OHCA incidence, with extreme weather increasing rates by nearly one-third. The 3-day lag patterns enable weather-based early warnings, supporting the integration of meteorological data into emergency response to reduce preventable deaths.