Efficacy and effectiveness of trivalent inactivated influenza vaccination against confirmed influenza, acute respiratory illness, and medically attended acute respiratory illness

Abstract

BACKGROUND: Trivalent inactivated influenza vaccination (TIV) has moderate to good efficacy in preventing laboratory-confirmed influenza infections in children. Cost-effectiveness calculations also require estimates of the incidence of influenza and the efficacy of vaccination against clinical outcomes. We estimated the impact of influenza in a large household cohort in Hong Kong in 2009-2010 and the direct efficacy of influenza vaccination. METHODS: 796 children 6-17 years of age were randomized to receive either 1 dose of TIV or normal saline as placebo. Subjects and their household members were followed up for approximately a year and monitored for signs and symptoms (body temperature ≥37.8° C, chills, headache, sore throat, cough, coryza, or myalgia) of acute respiratory illness (ARI) and associated visits to outpatient healthcare services through biweekly telephone follow-up, daily symptom diary, and self-reporting to study hotline as soon as their ARI started. Nose and throat swabs were collected for reverse transcription polymerase chain reaction (RT-PCR) testing from all household members once any of the members in the same household showed at least 2 of the ARI signs or symptoms. Serology was collected periodically. RESULTS: Children who were randomized to receive TIV had lower rates of RT-PCR−confirmed (0.07 per person-year; 95% CI, 0.04-0.10) influenza virus infections compared with children who received placebo (0.11 per person-year; 95% CI, 0.08-0.16). However, the crude rates of ARI (1.81 per person-year; 95% CI, 1.68-1.94) and medically attended ARI (0.93 per person-year, 95% CI, 0.83-1.02) were not significantly different in TIV and placebo recipients. The analyses were repeated in household contacts with similar observations, after adjusting for potential confounders. Approximately 40% of unvaccinated children and 17% of unvaccinated adults had serologic evidence of influenza virus infection during follow-up. CONCLUSION: Influenza had a substantial impact on adults and children in Hong Kong during our study period. While TIV reduced the risk of laboratory-confirmed influenza virus infection in children, it had little effect on the overall rates of ARI and medically attended ARI. One possible explanation is virus interference, and analysis of noninfluenza respiratory virus infections would help to clarify the situation. Analyses that focus on the effect of vaccination on laboratory-confirmed influenza may overestimate the benefit of vaccination.