Background: Two systematic reviews evaluating influenza vaccine efficacy in healthy children have recently been published. Although quantitative summary estimates were similar, authors' conclusions were quite contrasting. We carried out another meta-analysis reevaluating study inclusion criteria and using metaregression techniques in addition to sensitivity and subgroups analyses to evaluate potential sources of heterogeneity of efficacy estimates, including methodologic quality of studies.
Methods: Only randomized clinical studies assessing the efficacy of influenza vaccine in healthy children/adolescents (age < or =18 years) for preventing naturally occurring influenza and/or acute otitis media cases were included. Summary estimates of effect were obtained using a random effects model. The methodologic quality of each study was assessed using 3 systems: Chalmers scale, Jadad scale and Schulz components (randomization, allocation concealment and double-blinding).
Results: The overall vaccination efficacy was 36% (95% confidence interval: 31-40%) against clinically diagnosed illnesses (evaluated by 19 randomized clinical studies for a total of 247,517 children); 67% (51-78%) against laboratory-confirmed cases (18 trials, n = 8574); and 51% (21-70%) against acute otitis media (11 trials, n = 11,349). Significant sources of between-study heterogeneity were participants' age and study quality both directly correlated with the efficacy. When the analysis was performed excluding USSR studies, the overall efficacy of the vaccine in preventing clinical cases substantially increased (from 36% to 61%).
Conclusions: These findings may indicate that the vaccine efficacy might be greater than the overall estimates. Although no safety and cost considerations are addressed in this analysis, the present findings support vaccination as a possible option for the prevention of influenza in healthy children and adolescents.