FIGURE 2.
Forest plot of hazard ratios from longitudinal cohort studies investigating the association of diabetes mellitus with risk of progression of chronic hepatitis B infection to hepatocellular carcinoma (HCC). All studies included in meta‐analysis were cohort or nested case‐control studies reporting hazards ratios minimally adjusted for age and sex. The study by Yu et al 96 provided an odds ratio and was excluded from the meta‐analysis. Dashed vertical lines represent HR based on meta‐analysis of all studies by fixed‐effect and random‐effects models. The studies for pooling the HR had sufficient quality (quality scores ≥5). CI, confidence interval; DM, diabetes mellitus; HR, hazard ratio