Quantifying the Indirect Effects of Haemophilus Influenzae type b Vaccination in Children Under 5 Years-Old

Background Haemophilus influenzae type b (Hib) is a significant cause of meningitis and pneumonia among children under 5 years. A conjugate vaccine has been available since the mid-1980s, leading to disease reductions well above what would be expected from direct protection of the vaccine alone. This indicates indirect effects provide substantial protection against Hib disease and are a function of population-level vaccine coverage. Indirect effects are rarely investigated because it requires large, cluster randomized trials; instead we used pre-post vaccine introduction studies to model these effects. Methods Three methods to perform this analysis were identified in the literature. Wolfson, et al. described a method to compare the observed disease reduction to the reduction expected only from direct effects, resulting in an indirect effect multiplier based on vaccine coverage. Samandari, et al. estimated a multiplier for the number of people effectively protected by vaccination using vaccine coverage and incidence rates before and after vaccination. Lastly, Adegbola, et al. used average ages of Hib infection and vaccination to calculate an alternative estimate of direct protection from vaccination. Eleven studies were used for the Wolfson and Samandari methods and three were used for the Adegbola method. Results All three methods suggest a robust indirect effect of Hib vaccine against disease, with > 90% disease reduction predicted at only 70% coverage. Indirect effects were most influential at vaccine coverages below 40%, as vaccinating one child protected anywhere from two to six others. Direct effects dominated at vaccine coverages above 60%. Validating these results against an infectious disease theoretical framework and a study that empirically examined indirect effects on an individual level confirmed the accuracy of our results. Conclusion Predicted protection varied between the methods, but all demonstrated the importance of indirect effects at low vaccine coverage. These results can be used to better estimate the expected disease reduction prior to beginning a vaccine program and can impact policy decisions regarding vaccination. The models used in this analysis can also be applied to other vaccines, such as pneumococcal conjugate vaccine