Influenza Attack Rates and Vaccine Effectiveness by Prior-Season Infection and Vaccination: A 17-Season Analysis.

Abstract

Vaccine effectiveness (VE) is commonly interpreted as a measure of protection but may be influenced by differences in baseline risk across populations. The effects of prior-season influenza infection and vaccination on both attack rates (ARs) and VE have not been fully characterized. We analyzed 148,108 person-seasons from a 17-season prospective, community-based cohort study in Japan (2002-2003 through 2018-2019). Influenza infection was confirmed using rapid antigen testing. ARs among current-season unvaccinated individuals (a proxy for baseline risk) and vaccinated individuals were compared across groups defined by prior-season infection and vaccination. Adjusted odds ratios (aORs) and relative aORs (rORs) for current-season vaccination were estimated using multivariable logistic regression, with laboratory-confirmed influenza A infection (yes/no) as the outcome. Among current-season unvaccinated individuals, ARs were higher in those with prior-season vaccination than in those without (14.0% vs 8.6% in children; p<0.001). ARs were similar regardless of prior-season infection but higher with prior-season vaccination. Current-season vaccination reduced odds of influenza A infection (aOR, 0.52; 95% CI, 0.47-0.57). This reduction was more pronounced among individuals with prior-season infection (aOR, 0.35; 95% CI, 0.25-0.49) and less pronounced among those with prior-season vaccination (aOR, 0.59; 95% CI, 0.50-0.70). Protection remained across all groups (aOR <1). Prior-season infection was associated with enhanced VE of current-season vaccination and similar ARs among vaccinated individuals, whereas prior-season vaccination was associated with attenuated but effective VE. These findings highlight the importance of interpreting AR alongside VE, as ARs reflect differences in baseline risk.