All-Cause and Cause-Specific Mortality Attributable to Seasonal Influenza: A Nationwide Matched Cohort Study

Abstract

Background: Although influenza poses substantial mortality burden, most studies have estimated excess mortality using time-aggregated data. Here, we estimated mortality risk and population attributable fraction (PAF) attributed to seasonal influenza using individual-level data from a nationwide matched cohort.

Methods: Individuals with influenza during four consecutive influenza seasons (2013-2017) (n = 5,497,812) and 1:4 age- and sex-matched individuals without influenza (n = 20,990,683) were identified from a national health insurance database. The endpoint was mortality within 30 days after influenza diagnosis. All-cause and cause-specific mortality risk ratios (RRs) attributed to influenza were estimated. Excess mortality, mortality RR, and PAF of mortality were determined, including for underlying disease subgroups.

Results: Excess mortality rate, mortality RR, and PAF of all-cause mortality were 49.5 per 100,000, 4.03 (95% confidence interval [CI], 3.63-4.48), and 5.6% (95% CI, 4.5-6.7%). Cause-specific mortality RR (12.85; 95% CI, 9.40-17.55) and PAF (20.7%; 95% CI, 13.2-27.0%) were highest for respiratory diseases. In subgroup analysis according to underlying disorders, PAF of all-cause mortality was 5.9% (95% CI, 0.6-10.7%) for liver disease, 5.8% (95% CI, 2.9-8.5%) for respiratory disease, and 3.8% (95% CI, 1.4-6.1%) for cancer.

Conclusion: Individuals with influenza had a 4-fold higher mortality risk than individuals without influenza. Preventing seasonal influenza may lead to 5.6% and 20.7% reductions in all-cause and respiratory mortality, respectively. Individuals with respiratory disease, liver disease, and cancer may benefit from prioritization when establishing influenza prevention strategies.