Prevalence of group B streptococcal colonization in the healthy non-pregnant population: a systematic review and meta-analysis

Background: Colonization and transmission precede invasive group B streptococcal (GBS) disease. Data on GBS colonization prevalence, detection methods and risk factors for carriage are relevant for vaccine development and to understand GBS pathogenesis. Objectives: To evaluate GBS colonization prevalence after the first week of life in the healthy non-pregnant population. Data sources: Pubmed/Medline, Embase, Latin American and Caribbean Health Sciences Literature, World Health Organization Library Information System, and Scopus. Search performed 12 January 2021 with search terms related to ‘GBS’ and ‘colonization, epidemiology, prevalence or screening’ without restrictions. Study eligibility criteria: All studies that reported prevalence of GBS colonization (any site) in the healthy population. Participants: All individuals (>6 days of age), with no indication of pregnancy, invasive disease or severe underlying immunological co-morbidities. Methods: Logit transformation and a random effects model (DerSimonian and Laird) were used to pool colonization estimates. Subgroup analysis and meta-regression on a priori determined subgroups were performed. Results: We included 98 studies with 43 112 participants. Our search identified 9309 studies of which 8831 were excluded based on title and abstract and 380 after reading the full text. Colonization rates varied considerably between studies (I2 = 97%), which could be partly explained by differences in culture methods (R2 = 27%), culture sites (R2 = 24%), continent (R2 = 10%) and participant's age (R2 = 6%). Higher prevalence was found with selective culture methods (19%, 95% CI 16%–23% versus non-selective methods 8%, 95% CI 6%–9%; p < 0.0001). Colonization rates were highest in rectum (19%, 95% CI 15%–24%), vagina (14%, 95% CI 12%–17%) and urethra (9%, 95% CI 5%–18%). In participants with negative rectal cultures, 7% (95% CI 5%–9%) had GBS cultured from another niche. Colonization prevalence was lower in children (6 months to 16 years; 3%, 95% CI 2%–5%) compared with adults (16%, 95% CI 14%–20%; p < 0.0001). Using selective culture methods in adults resulted in a prevalence of 26% (95% CI 19%–33%) rectal, 21% (95% CI 17%–25%) vaginal and 9% (95% CI 6%–14%) urethral colonization. Conclusion: The rectum is the most common body site colonized by GBS. The best approach to screen for any GBS colonization is to screen multiple body sites using selective culture methods.