LAUSR

INTRODUCTION There is compelling evidence for an association between influenza epidemics and major adverse cardiovascular events. However, the role of extreme influenza epidemics as a trigger of out-of-hospital cardiac arrest (OHCA) is unclear. Thus, we evaluated the potential association between extreme influenza epidemics and incidence of OHCA. METHODS We used a quasi-experimental design with time-series analysis of national registry data for cases of OHCA from all 47 prefectures of Japan during influenza seasons between 2005 and 2014. A Poisson regression time-series model with a distributed lag non-linear model was used to estimate prefecture-specific effects of influenza epidemics on OHCA. A multivariate meta-analysis was conducted for nationally pooled estimates. RESULTS In total, 481,516 OHCAs of presumed cardiac origin were reported during the study period. The minimum morbidity percentile (MMP) was estimated as the 0th percentile for influenza incidence. The overall cumulative relative risk versus the MMP was 1.25 (95% confidence interval, 1.16-1.34) for extreme influenza epidemics (at the 99th percentile of influenza incidence). The effect of extreme influenza epidemics was significant for lag periods of 1.5-7.1 and 17.9-21 days. Multivariate random-effects meta-analysis indicated significant spatial heterogeneity among prefectures (Cochran Q test, p = 0.011; I2 = 23.2%). CONCLUSION Extreme influenza epidemics are associated with higher risk of OHCA. Our findings suggest that several weeks' prevention for extreme influenza infections should be implemented to reduce the risk of OHCA.