Abstract

BACKGROUND: Pneumococcal conjugate vaccines (PCVs) introduced in childhood national immunization programs lowered vaccine-type invasive pneumococcal disease (IPD), but replacement with non-vaccine-types persisted throughout the PCV10/13 follow-up period. We assessed PCV10/13 impact on pneumococcal meningitis incidence globally. METHODS: The number of cases with serotyped pneumococci detected in cerebrospinal fluid and population denominators were obtained from surveillance sites globally. Site-specific meningitis incidence rate ratios (IRRs) comparing pre-PCV incidence to each year post-PCV10/13 were estimated by age (<5, 5-17 and ≥18 years) using Bayesian multi-level mixed effects Poisson regression, accounting for pre-PCV trends. All-site weighted average IRRs were estimated using linear mixed-effects regression stratified by age, product (PCV10 or PCV13) and prior PCV7 impact (none, moderate, or substantial). Changes in pneumococcal meningitis incidence were estimated overall and for product-specific vaccine-types and non-PCV13-types. RESULTS: Analyses included 10,168 cases <5 y from PCV13 sites and 2849 from PCV10 sites, 3711 and 1549 for 5-17 y and 29,187 and 5653 for ≥18 y from 42 surveillance sites (30 PCV13, 12 PCV10, 2 PCV10/13) in 30 countries, primarily high-income (84%). Six years after PCV10/PCV13 introduction, pneumococcal meningitis declined 48-74% across products and PCV7 impact strata for children <5 y, 35-62% for 5-17 y and 0-36% for ≥18 y. Impact against PCV10-types at PCV10 sites, and PCV13-types at PCV13 sites was high for all age groups (<5 y: 96-100%; 5-17 y: 77-85%; ≥18 y: 73-85%). After switching from PCV7 to PCV10/13, increases in non-PCV13-types were generally low to none for all age groups. CONCLUSION: Pneumococcal meningitis declined in all age groups following PCV10/PCV13 introduction. Plateaus in non-PCV13-type meningitis suggest less replacement than for all IPD. Data from meningitis belt and high-burden settings were limited.