Nomogram for predicting recurrent bleeding risk in patients with viral hepatitis-related cirrhosis in China

BACKGROUNDUpper gastrointestinal bleeding (UGIB) is a life-threatening complication in cirrhosis, with recurrent rates up to 80% within the first year.Existing scores underperform in this population and lack etiology specificity.A simple, targeted predictive tool is urgently needed. AIMTo develop a nomogram to predict the recurrent bleeding risk in patients with viral hepatitis-related cirrhosis after their first episode of UGIB. METHODSWe retrospectively analyzed data from 206 patients with viral hepatitis-related cirrhosis who experienced their first UGIB between January 1, 2014, and December 31, 2018, at Qingdao Municipal Hospital.All predictor variables were collected at baseline.Multivariate logistic regression was applied to identify the risk factors for recurrent bleeding, which were then used to construct the nomogram.Predictive accuracy was assessed using the concordance index (C-index) and the area under the receiver operating characteristic curve, while calibration was evaluated using a calibration plot.Clinical utility was assessed using decision curve analysis and clinical impact curve.Internal validation was performed using 1000 bootstrap resamples. RESULTSThe cohort was stratified into recurrent bleeding (n = 108) and non-recurrent bleeding (n = 98) groups.Multivariate analysis identified esophageal and gastric varices, ascites, portal vein diameter, and portal vein thrombosis as significant predictors of recurrent bleeding.Based on these factors, an exploratory nomogram was constructed, with a C-4 / 39 index of 0.731 (95%CI: 0.661-0.802)and a bias-corrected C-index of 0.716 in internal validation.No external validation was performed. CONCLUSIONIn this single-center retrospective cohort, the exploratory nomogram showed acceptable discrimination for recurrent bleeding in patients with viral hepatitis-related cirrhosis after their first UGIB.However, these findings are hypothesis-generating only.External validation in independent, multicenter cohorts is required before the model can be considered for clinical use.