Comparative effectiveness of high-dose versus standard-dose influenza vaccination on numbers of US nursing home residents admitted to hospital: a cluster-randomised trial

Background Immune responses to influenza vaccines decline with age, reducing clinical effectiveness. We compared the effect of the more immunogenic high-dose trivalent influenza vaccine with a standard-dose vaccine to identify the effect on reducing hospital admissions of nursing home residents in the USA. Methods We did a single-blind, pragmatic, comparative effectiveness, cluster-randomised trial with a 2 × 2 factorial design. Medicare-certified nursing homes in the USA located within 50 miles of a Centers for Disease Control and Prevention influenza reporting city were recruited, so long as the facilities were not located in a hospital, had more than 50 long-stay residents, had less than 20% of the population aged under 65 years, and were not already planning to administer the high-dose influenza vaccine to residents. Enrolled nursing homes were randomised to a facility-wide standard of care for the residents of either high dose or standard dose as the vaccine for the 2013–14 influenza season and half of each group were randomly allocated to free vaccines for staff. Individual residents were included in the analysis group if they were aged 65 years or older and were long-stay residents (ie, had been in the facility 90 days or more before commencing the influenza vaccination programme). The analysts and investigators with access to the raw data were masked to study group by coding the groups until after the analyses were complete. The primary outcome was hospital admissions related to pulmonary and influenza-like illness between Nov 1, 2013, and May 31, 2014, identified from Medicare hospital claims available for residents who were without private health insurance (ie, those who were considered Medicare fee-for-service). We obtained data from the Centers for Medicare & Medicaid Services (CMS) and enrolled facilities. The analyses used marginal Poisson and Cox proportional hazards regression, accounting for clustering of residents within homes, on an intention-to-treat basis, adjusting for facility clustering and prespecified covariates. Safety data were voluntarily reported according to the standard of care. This trial is registered with ClinicalTrials.gov, number NCT01815268. Findings 823 facilities were recruited to the study between March and August, 2013, to participate in the trial, of which 409 facilities were randomised for residents to receive high-dose vaccine, and 414 facilities for residents to receive standard-dose vaccine. The facilities housed 92 269, of whom 75 917 were aged 65 years or older and 53 008 were also long-stay residents, and 38 256 were matched to Medicare hospital claims as of Nov 1, 2013. Staff vaccination rates did not differ between groups, so analyses focused on the high-dose versus standard-dose vaccine comparison. On the basis of Medicare fee-for-service claims, the incidence of respiratory-related hospital admissions was significantly lower in facilities where residents received high-dose influenza vaccines than in those that received standard-dose influenza vaccines (0·185 per 1000 resident-days or 3·4% over 6 months vs 0·211 per 1000 resident-days or 3·9% over 6 months; unadjusted relative risk of 0·888, 95% CI 0·785–1·005, 0=0·061, and adjusted relative risk 0·873, 0·776–0·982, p=0·023). Interpretation When compared with standard-dose vaccine, high-dose influenza vaccine can reduce risk of respiratory-related hospital admissions from nursing home residents aged 65 years and older. Funding Sanofi Pasteur, Swiftwater, PA, USA.