Cluster-randomized Trial of Adjuvanted Versus Nonadjuvanted Trivalent Influenza Vaccine in 823 US Nursing Homes
Kevin W. McConeghy, H. Edward Davidson, David H. Canaday, Lisa Han, Elie Saade, Vince Mor, Stefan Gravenstein
Abstract
BACKGROUND: Influenza leads in preventable infection-related hospitalization in nursing home (NH) residents. The adjuvanted trivalent influenza vaccine (aTIV) is more immunogenic than similarly dosed nonadjuvanted trivalent influenza vaccine (TIV), and observational studies suggest aTIV better prevents hospitalizations in older adults. We prospectively tested this in an NH setting. METHODS: NHs with ≥50 long-stay residents aged ≥65 years were randomized to offer aTIV or TIV for residents for the 2016-2017 influenza season. Using intent-to-treat resident-level analysis with Cox proportional hazards regression models adjusted for clustering by facility and a priori baseline covariates (eg, age, heart failure, and facility-level characteristics), we assessed relative aTIV:TIV effectiveness for hospitalization (ie, all-cause, respiratory, and pneumonia and influenza [P&I]). RESULTS: We randomized 823 NHs, housing 50 012 eligible residents, to aTIV or TIV. Residents were similar between groups by age (mean, ~79 years), heart failure, lung disease, and influenza and pneumococcal vaccine uptake, except aTIV homes housed fewer Black residents (14.5 vs 18.9%). Staff vaccine uptake was similar (~55%). P&I and all-cause resident hospitalization rates were lower (adjusted HR [aHR], .80 [95% CI, .66-.98; P = .03] and .94 [.89-.99; P = .02], respectively) for aTIV versus TIV, while the respiratory hospitalization rate was similar, in a season where vaccine effectiveness was considered poor. CONCLUSIONS: aTIV was more effective than TIV in preventing all-cause and P&I hospitalization from NHs during an A/H3N2-predominant season when TIV was relatively ineffective. CLINICAL TRIALS REGISTRATION: NCT02882100.