• influenza vaccination;
  • booster dose;
  • elderly;
  • hemagglutination titers;
  • morbidity

OBJECTIVES: To determine whether assessing seroprotection after influenza vaccine and administering booster vaccination where not achieved reduces hospitalization and death. To estimate the overall seroprotection rate of influenza vaccine.

DESIGN: A two-arm, partially blind, randomized, multicenter, parallel-group, controlled trial.

SETTING: Twenty-six care homes in three South London boroughs in fall 2004.

PARTICIPANTS: Two hundred seventy-seven elderly permanent care home residents meeting eligibility criteria.

INTERVENTION: Postvaccination blood samples were randomized to booster evaluation or no booster evaluation (control). If evaluation revealed inadequate seroprotection, a booster vaccine was administered.

MEASUREMENTS: Primary outcome was hospitalization to end April 2005; secondary outcomes were death, antibiotic use, and seroprotection.

RESULTS: Sixty percent of the controls and 41% of the booster evaluation group responded to routine vaccination. Booster vaccination where indicated increased seroprotection rates in the booster evaluation group to 66%. Treatment groups did not differ in any outcome measures in the intention-to-treat analysis (hospitalization odds ratio=1.02, 95% confidence interval=0.55–1.87). There was a tendency towards greater differences between groups in the per-protocol analysis than in the intention-to-treat analysis, particularly regarding seroprotection rates. The same effect was observed in the a priori exploratory analysis of residents not seroprotected after routine vaccination alone.

CONCLUSION: In a year without circulating influenza, there is no clinical benefit of administering a booster vaccine if routine trivalent vaccination fails to result in seroprotection. Hemagglutination titers rose in two strains postbooster vaccination but fell against the novel strain, Wyoming. The benefit of such a booster strategy when influenza is prevalent thus remains unc ertain.