If PCV has not been recommended, the control group could be given placebo, provided it is ethically acceptable in the trial population. If a placebo is not acceptable, a non-pneumococcal control vaccine should be sought. Preferably, it should be a vaccine already registered, rather than an investigational one. Optimally, the non-pneumococcal control vaccine should not impact the microbiota of the upper respiratory tract as interactions between different bacterial occupying the same ecological niche have been observed [12]. If the use of a non-pneumococcal control vaccine is

not an acceptable Etoposide purchase approach, the presently used (licensed) pneumococcal vaccine may serve as an active control. The main points in choosing the control vaccine are summarised in Table 1. We consider the statistical power of VEcol studies for showing either the efficacy against Venetoclax all vaccine-type (VT) acquisition or serotype-specific efficacy

against acquisition of individual serotypes. The estimation method is based on a cross-sectional sample under the assumption of no efficacy on duration [1] and [10]. Based on the scenarios presented in the previous section, we discuss the following two alternatives regarding the control vaccine: (A) A control vaccine with known zero (biological) efficacy against the pneumococcal colonisation endpoint; Controlled trials. Alternative A leads to a standard superiority trial with a non-active control.

Here, the statistical power is defined as the probability for the lower bound of the confidence interval for VEacq to exceed 0 under the alternative hypothesis, i.e. when VEacq is at least D (the smallest meaningful efficacy). The choice of D can be based on the herd immunity threshold, that is, a level of direct protection against colonisation which would induce significant indirect protection in the population. Theoretical modelling suggests that even 50% efficacy (VEacq) could be enough for herd immunity, if the coverage of vaccination in the infant programme is high [13]. Fig. 2 presents the power of a controlled study under scenario A for different much values of the sample size (number of individuals per study group) and the hypothesised efficacy (D). For example, a group size of 300 is enough to obtain 80% power, if the vaccine efficacy against vaccine-type acquisition is 50%. The results are essentially similar under high (left panel) or moderate (right panel) overall rate of pneumococcal acquisition. Head-to-head trials. Under alternative B, the investigational vaccine’s effect is measured against an active pneumococcal vaccine. The hazard rate ratio (investigational vs.