Efficacy of different pneumococcal conjugate vaccine schedules against pneumonia, hospitalisation, and mortality: re-analysis of a randomised trial in the Gambia.

BACKGROUND: Pneumococcal conjugate vaccines (PCV) reduce disease due to Streptococcus pneumoniae. We aimed to determine the efficacy of different PCV schedules in Gambian children. METHODS: We reanalysed data from a randomised placebo-controlled trial. Infants aged 6-51 weeks were allocated to three doses of nine-valent PCV (n=8718) or placebo (n=8719) and followed until age 30 months. We categorised participants to compare: (a) a first dose at age 6 or 10 weeks, (b) intervals of 1 or 2 months between doses, and (c) different intervals between second and third doses. The primary endpoint was first episode of radiologic pneumonia; other endpoints were hospitalisation and mortality. Using the placebo group as the reference population, Poisson regression models were used with follow-up after the first dose to estimate the efficacy of each schedule and from age 6 weeks to estimate the incidence rate difference between schedules. RESULTS: Predicted efficacy in the groups aged 6 weeks (n=2467, 154 events) or 10 weeks (n=2420, 106 events) at first dose against radiologic pneumonia were 32% (95% CI 19-43%) and 33% (95% CI 21-44%), against hospitalisation 14% (95% CI 3-23%) and 17% (95% CI 7-26%), and against mortality 17% (95% CI -3 to 33%) and 16% (95% CI -3 to 32%) respectively. Predicted efficacy in the groups with intervals of 1 month (n=2701, 133 events) or 2 months (n=1351, 58 events) between doses against radiologic pneumonia were 33% (95% CI 20-44%) and 36% (95% CI 24-46%), against hospitalisation 15% (95% CI 5-24%) and 18% (95% CI 8-27%), and against mortality 17% (95% CI -2 to 33%) and 13% (95% CI -8 to 29%) respectively. Efficacy did not differ by interval between second and third doses, nor did the incidence rate difference between schedules. CONCLUSIONS: We found no evidence that efficacy or effectiveness of PCV9 differed when doses were given with modest variability around the scheduled ages or intervals between doses.