Prevalence of maternal colonisation with group B streptococcus: a systematic review and meta-analysis

Abstract

The most important risk factor for early-onset (babies younger than 7 days) invasive group B streptococcal disease is rectovaginal colonisation of the mother at delivery. We aimed to assess whether differences in colonisation drive regional differences in the incidence of early-onset invasive disease. We did a systematic review of maternal group B streptococcus colonisation studies by searching MEDLINE, Embase, Pascal Biomed, WHOLIS, and African Index Medicus databases for studies published between January, 1997, and March 31, 2015, that reported the prevalence of group B streptococcus colonisation in pregnant women. We also reviewed reference lists of selected studies and contacted experts to identify additional studies. Prospective studies in which swabs were collected from pregnant women according to US Centers for Disease Control and Prevention guidelines that used selective culture methods were included in the analyses. We calculated mean prevalence estimates (with 95% CIs) of maternal colonisation across studies, by WHO region. We assessed heterogeneity using the I(2) statistic and the Cochran Q test. 221 full-text articles were assessed, of which 78 studies that included 73 791 pregnant women across 37 countries met prespecified inclusion criteria. The estimated mean prevalence of rectovaginal group B streptococcus colonisation was 17·9% (95% CI 16·2-19·7) overall and was highest in Africa (22·4, 18·1-26·7) followed by the Americas (19·7, 16·7-22·7) and Europe (19·0, 16·1-22·0). Studies from southeast Asia had the lowest estimated mean prevalence (11·1%, 95% CI 6·8-15·3). Significant heterogeneity was noted across and within regions (all p≤0·005). Differences in the timing of specimen collection in pregnancy, selective culture methods, and study sample size did not explain the heterogeneity. The country and regional heterogeneity in maternal group B streptococcus colonisation is unlikely to completely explain geographical variation in early-onset invasive disease incidence. The contribution of sociodemographic, clinical risk factor, and population differences in natural immunity need further investigation to understand these regional differences in group B streptococcus maternal colonisation and early-onset disease. None.