Medical service utilization under heat stress at Expo 2025 Osaka, Kansai, Japan: A ６-month time series study

This study examined the exposure–response relationships between ambient heat and medical service utilization at Expo 2025 Osaka, Japan, from April 13 to October 13, 2025. Distributed lag non-linear models with negative binomial regression were fitted to daily counts of patient presentations (PP), heat-related patient presentations (PP-HRI), and emergency transports to the hospital (TTH), using daily maximum temperature (T max ) as the primary heat exposure with daily attendance as an offset and day-of-week adjustment (maximum lag: 5 days). The median PP rate was 7.5 per 10,000 attendees, and 97.2% of cases were managed onsite. Medical encounters exhibited daily fluctuations that tracked alongside peak summer temperatures from June to August and increased in attendance toward the end of the event period. Cumulative risks at the 75th percentile of T max (34.9 °C) were elevated relative to the median (31.2 °C) for PP (RR 1.50; 95% CI 1.35–1.67) and PP-HRI (RR 4.23; 95% CI 2.51–7.10).TTH exhibited a distinct bimodal cumulative risk pattern, reaching its secondary peak of 1.71 (95% CI: 1.19–2.46) within this range. Lag-specific risks for PP and PP-HRI were elevated at lag 0–1 and attenuated at longer lags. TTH showed a more complex lag pattern, with a modest increase at lag 5. These findings characterize the temporal dynamics and exposure–response patterns of medical encounters under extreme heat and provide a basis for medical planning at mass gatherings.