Clinical Infectious Diseases, 2019. [Epub ahead of print]
Segaloff HE, Leventer-Roberts M, Riesel D, Malosh RE, Feldman BS, Shemer-Avni Y, Key C, Monto AS, Martin ET, Katz MA.
Background: Influenza vaccine effectiveness (VE) varies by season, circulating influenza strain, age, and geographic location. There have been few studies of influenza VE among hospitalized children, particularly in Europe and the Middle East.
Methods: We estimated VE against influenza hospitalization among children six months to eight years at Clalit Health Services hospitals in Israel in the 2015-16, 2016-17, and 2017-18 influenza seasons using the test-negative design. Estimates were computed for full and partial vaccination.
Results: We included 326 influenza-positive cases and 2821 influenza-negative controls (140 cases and 971 controls from 2015-16, 36 cases and 1069 controls from 2016-17, and 150 cases and 781 controls from 2017-18). Over all seasons, VE was 53.9% for full vaccination (95% CI:(38.6,68.3)), and 25.6% for partial vaccination (95% CI:(-3, 47)). In 2015-16, most viruses were influenza A(H1N1) and vaccine lineage-mismatched influenza B/Victoria; VE for fully vaccinated children was statistically significant for influenza A (80.7%, 95% CI:(40.3,96.1)) but not B (23.0%, 95% CI:(-38.5, 59.4)). During 2016-17, influenza A(H3N2) predominated, and VE was (70.8%, 95% CI:(17.4, 92.4)). In 2017-18, influenza A(H3N2), H1N1 and lineage-mismatched influenza B/Yamagata co-circulated; VE was statistically significant for influenza B (63.0% 95% CI: (24.2,83.7)), but not A (46.3%, 95% CI:(-7.2, 75.3)).
Conclusions: Influenza vaccine was effective in preventing hospitalizations among fully vaccinated Israeli children over three influenza seasons, but not among partially vaccinated children. There was cross-lineage protection in a season where the vaccine contained B/Victoria and the circulating strain was B/Yamagata, but not in a season with the opposite vaccine-circulating strain distribution.