Invitation letters were sent to a convenience sample of 20 primary and 5 secondary schools. In 3 schools that agreed to participate, letters were distributed to the parents of 2190 children, and 54 households were enrolled. Fifteen hundred invitations were sent to households of children who are members of a local birth cohort, and 51 households were enrolled. A further 14 households were enrolled through personal referral.
A total of 119 households were enrolled and randomized. Subjects and household contacts in the TIV and control groups had similar baseline characteristics. Children from 2 households did not receive the intervention and withdrew from the study: 1 child with a history of epileptic seizures was assessed by the study nurse to be contraindicated against giving blood specimens and receiving vaccination at the time of presentation, and blood specimens could not be obtained from another child. The households of 11 of 117 children who received the intervention did not complete the study. Following the principle of intention‐to‐treat, we included all 119 households in the primary analyses.
A single dose of TIV led to substantial and statistically significant increases in antibody titers to the seasonal strains among study subjects. Children who received TIV had statistically significant but limited geometric mean increases in antibody titers to pandemic A/H1N1 virus following receipt of TIV, and all post‐vaccination titers were below 1:40. No serious adverse reactions were observed, and pain at the injection site was the only adverse event for which there was a statistically significant difference between arms.
Children who received TIV had lower rates of serologically confirmed seasonal influenza infection, compared with the placebo group, although reductions were not statistically significant. By the end of the study period, 8% and 7% of the children in the TIV group had serologically confirmed infection with seasonal influenza A/H1N1 and A/H3N2, compared with 21% and 12%, respectively, in the control group. Children who received TIV had a non–statistically significant higher rate of pandemic A/H1N1 infection during the summer season, compared with children who received placebo. After adjusting for potential cross‐reactions, we estimated that 31% of children who received TIV were infected with pandemic A/H1N1, compared with 12% of children who received placebo. No differences were observed in the rates of RT‐PCR–confirmed influenza, ILI, or acute respiratory illness during the study. Among those who had serologically confirmed pandemic A/H1N1 virus infection, no difference in ILI rates was observed between study subjects who received TIV and those who received placebo.