Design questions for Streptococcus pneumoniae vaccine trials with a colonisation endpoint

Evaluation of vaccine efficacy for protection against colonisation (VE_col) with Streptococcus pneumoniae and other bacterial pathogens is often based on a cross-sectional study design, in which only one nasopharyngeal sample is obtained per study subject. Here we investigate the feasibility of this study design by investigating a number of practical design problems. Specific questions are related to the timing of colonisation measurement with respect to the time of vaccination, the adjustment for the within-host replacement of vaccine-type colonisation by the non-vaccine type pneumococci, and the impact of multiple serotype colonisation on VE_col estimation. We also discuss the issue of choosing the control vaccine, including comparison of two active pneumococcal vaccines, as well as the sample size and the statistical power of colonisation endpoint trials. In addition, the statistical design with the specific aim to include information about VE_col in the licensure process of new pneumococcal vaccine products is discussed.