||Negative impact of prior influenza vaccination on current influenza vaccination among people infected and not infected in prior season: A test-negative case-control study in Japan
Saito, Nobuo ,
Komori, Kazuhiro Suzuki, Motoi ,
Morimoto, Kounosuke ,
Kishikawa, Takayuki ,
Yasaka, TakahiroAriyoshi, Koya
693 , 2017-01-23 , Elsevier Ltd
Background Accumulating evidences indicate that repeated influenza vaccination has negative impact on the vaccine effectiveness (VE). However no published studies considered past influenza infection when assessing the VE of repeated vaccination. Methods Prospective surveillance was conducted from 2009 to 2012 at a community hospital on a small island in Japan. The study included all outpatients with an influenza-like illness (ILI) who attended the hospital, and a rapid diagnostic test (RDT) was used to diagnose influenza A/B infection. The VE of trivalent inactivated influenza vaccine (TIV) against medically attended influenza A (MA-fluA) was estimated using a test-negative case-control study design. The influence of TIV in the prior season on VE in the current season was investigated in the context of MA-fluA during the prior season. Results During the three influenza seasons, 5838 ILI episodes (4127 subjects) were analysed. Subjects who had an episode of MA-fluA in the prior season were at a significantly lower risk of MA-fluA in the current season (adjusted odds ratio: 0.38, 95% CI: 0.30–0.50). The overall adjusted VE was 28% (95% CI, 14–40). VE was substantially lower in subjects vaccinated in the prior season compared to those who had not been vaccinated in prior season (19%; 95% CI: 0–35 vs 46%; 95% CI: 26–60, test for interaction, P value <0.05). In subjects who did not have MA-fluA in the prior season showed the attenuation of VE due to repeated vaccination (13%; 95% CI: −7 to 30 vs 44%; 95% CI: 24–59, test for interaction, P < 0.05). However this effect was not detected in subjects who had contracted MA-fluA in the prior season. Conclusions Negative effects of repeated vaccination were significant among those without history of MA-fluA in the prior season.