Prevalence and Serological Characterisation of Weakly or Variably Expressed B Antigen (Weak B Subgroups) Among Blood Donors and Patients: A Systematic Review and Meta-Analysis
by Sayantani Ghosh Hazra
Published: June 25, 2026 • DOI: 10.51244/IJRSI.2026.1306000118
Abstract
Background: The weak B subgroups — B3, Bx, Bm and Bel — carry reduced or atypical B antigen and react only weakly, or not at all, with anti-B, making them a recognised cause of ABO grouping discrepancy and of potential mistyping with transfusion-safety implications. Their reported frequency is inconsistent and has never been systematically synthesised. Methods: Following PRISMA 2020, PubMed, Embase, Scopus, Web of Science and Google Scholar were searched for studies reporting the frequency or serological characterisation of weak/variant B. Two reviewers independently screened and extracted data and appraised quality using the Joanna Briggs Institute checklist for prevalence studies. Study-level prevalence was expressed per 100,000 with exact (Clopper–Pearson) 95% confidence intervals (CIs); a random-effects (Freeman–Tukey) proportion meta-analysis was planned where three or more comparable cohorts were available. Results: In an Eastern Indian cohort of 84,534 donors, weak B occurred in 9 — 10.65 per 100,000 (95% CI 4.87–20.21), about 1 in 9,400 — ordered B3 > Bm > Bx = Bel; older Indian data report rarer figures (~1 in 24,000). A Chinese cohort of 2,945,643 samples reported all ABO subgroups at 31.47 per 100,000 (95% CI 29.48–33.56) and estimated that 27.81% were missed by routine testing. With only one comparable weak-B denominator, pooling criteria were not met; per-study estimates and a structured narrative synthesis are presented, and across the literature weak B characteristically presents as a forward/reverse discordance resolved by adsorption–elution, saliva testing and molecular analysis. Conclusion: Weakly expressed B antigen is rare (~1 in 9,000–24,000 donors) but routinely under-detected by low-sensitivity forward grouping. A definitive pooled prevalence awaits the full search; meanwhile, forward and reverse grouping should agree, confirmed by sensitive and molecular methods, before any B-containing group is reported. A practical resolution algorithm is provided to support transfusion services.