Background: Influenza outbreaks in nursing homes pose a threat to frail residents and occur even in vaccinated populations. We conducted a pragmatic cluster-randomized trial comparing adjuvanted trivalent influenza vaccine (aTIV) versus trivalent influenza vaccine (TIV). Here, we report an exploratory analysis to compare the effect of aTIV versus TIV on facility-reported influenza outbreaks.

Methods: Nursing homes were randomized to offer older residents either aTIV or TIV for the 2016-17 influenza season. The impact of the intent-to-treat vaccine assignment was evaluated for the total number of outbreaks reported from November-March. We collected data according to standard CDC definitions for both suspected outbreaks and those with a laboratory-confirmed case in nursing homes, and adjusted for facility-level vaccination rates and resident characteristics.

Results: Of 823 randomized nursing homes, 777 (aTIV, n=387; TIV, n=390) reported information on influenza outbreaks. The treatment groups had similar characteristics at baseline except for race/ethnicity: homes assigned to TIV had a higher percentage of African-American residents (18.0% versus 13.7%). There were 133 versus 162 facility-reported suspected influenza outbreaks in aTIV versus TIV facilities respectively, of these 115 versus 140 were laboratory confirmed. The aTIV group experienced a 17% reduction in suspected (rate ratio, RR, 0.83, 95% confidence interval, CI: 0.65, 1.05) and laboratory-confirmed influenza outbreaks (RR 0.83, 95%CI: 0.63, 1.06). Covariate adjustment increased the estimated reduction for suspected outbreaks to 21% (RR 0.79, 95%CI: 0.61, 0.99) and 22% for laboratory confirmed (RR 0.78, 95%CI: 0.60, 1.02).

Conclusions: In an exploratory analysis of a cluster-RCT we observed 17-21% fewer outbreaks with aTIV than TIV