The simple multivariable model for predicting liver fibrosis in Vietnamese male adults: a combination of Bayesian model averaging and stepwise method

Background Liver fibrosis is a significant health burden in Vietnamese male adults, driven by high rates of hepatitis B and hepatitis C, excessive alcohol consumption, and genetic and environmental factors. Despite progress in diagnostic tools, there is a pressing need for cost-effective screening methods tailored to this high-risk group, particularly in resource-limited settings. Methods This study enrolled 952 Vietnamese male adults over 40 years old undergoing FibroScan, excluding those with conditions affecting test accuracy. Data on demographics, clinical history, and anthropometrics were collected, and fibrosis stages were classified using the METAVIR system. Model development combined Bayesian model averaging and forward stepwise methods, with predictive performance validated via receiver operating characteristic (ROC) analysis and area under the curve (AUC) estimation in the R environment. Results Among 952 male participants, the prevalence of liver fibrosis was 19.9%, with most cases classified as mild (F1). Multivariate analysis identified significant risk factors, including advanced age (odds ratio (OR) = 1.6; 95% confidence interval (CI) [1.02–2.51]), alcohol abuse (OR = 4.44; 95% CI [2.65–7.42]), hepatitis B (OR = 6.76; 95% CI [3.14–14.54], hepatitis C (OR = 33.04; 95% CI [5.26–207.42]), family history of cirrhosis (OR = 16.14; 95% CI [3.28–79.55]), and hepatic steatosis (OR = 4.02; 95% CI [2.57–6.28]). The predictive model demonstrated good discriminative performance with an AUC of 0.769 (95% CI [0.734–0.800]) and showed satisfactory calibration through bootstrap resampling, indicating close agreement between predicted and observed risks. Conclusion The current prevalence of liver fibrosis among Vietnamese male adults was found to be 19.9%, and the developed risk prediction model effectively identifies high-risk individuals, enabling early diagnosis and targeted prevention, particularly in resource-limited settings. However, the lack of external validation and the sample restricted to Vietnamese male adults limit the generalizability of the model, which should be further evaluated in other populations.