Occult Hepatitis B Virus Infection Among HBsAg-Negative Maintenance Haemodialysis Patients in Malaysia: Association with Anti-HBs Serostatus

Abstract Background Occult Hepatitis B infection (OBI)—defined as the presence of replication-competent HBV DNA in HBsAg-negative individuals—poses a severe transmission and reactivation risk in haemodialysis units. This study aims to investigate the prevalence and clinical characteristics of detectable HBV DNA among HBsAg-negative maintenance haemodialysis patients, particularly those with anti-HBc positivity. Methods A retrospective observational study was conducted involving 2,146 patients who attended the dialysis unit at Universiti Malaya Medical Centre (UMMC) from 1 January 2019 to 30 June 2024. Patients established on maintenance haemodialysis were included, while those with HBsAg positivity or missing data were excluded. Patient demographics, clinical variables, and biochemical information were extracted from the Electronic Medical Record (EMR). Data collection was conducted using REDCap version 14.3.13, and statistical analysis was performed using SPSS version 29.0 and Python version 3.10.16. A Firth penalised multivariable binary logistic regression model was constructed utilising OBI status as the dependent variable to identify independent clinical predictors, addressing potential small-sample bias inherent in standard logistic regression. Results Among 2,146 screened patients, 1,635 patients were included. Anti-HBc IgG positivity was identified in 23.5% (N = 384) of the cohort. HBV DNA testing was not performed in 146 patients under the testing practices prevailing during the study period. Within the tested core-positive subgroup (N = 238), overall OBI prevalence was 9.2% (N = 22). In patients with positive HBV DNA, 54.6% were concurrently Anti-HBs IgG positive (≥ 10 IU/L), and 18.2% possessed an Anti-HBs IgG level ≥ 100 IU/L. The multivariable Firth penalised logistic regression model revealed that while an Anti-HBs titre ≥ 10 IU/L was a significant negative predictor of OBI (OR 0.18, 95% CI 0.07–0.50, p = 0.001), it failed to exclude active viraemia entirely. Conclusion Our study identified an overall OBI prevalence of 9.2% in a tested Malaysian ESKD cohort. Firth penalised logistic regression model utilising OBI status as the dependent variable confirmed that relying solely on Anti-HBs IgG to infer viral clearance is inadequate in this immunocompromised population, especially given that over half (54.6%) of OBI patients maintained Anti-HBs IgG positivity.