Factor predictor for spontaneous rupture of hepatocellular carcinoma at Lopburi Cancer Hospital

Abstract

This predictive correlational study aimed to 1) identify factors associated with liver rupture of hepatocellular carcinoma (HCC) and 2) develop a prediction model for the risk of liver rupture. The study was conducted in two phases. Phase 1 involved a retrospective review of the medical records of 349 patients with HCC admitted to Lopburi Cancer Hospital between October 1, 2015, and September 30, 2019, to identify potential predictors of liver rupture. Phase 2 evaluated the predictive performance of the developed model using medical records of 20 patients with HCC admitted between October 1, 2019, and September 30, 2023. The research instrument was a medical record data extraction form developed from a comprehensive literature review. The collected variables included demographic characteristics, clinical information, and potential predictors of liver rupture, namely tumor size, child–pugh score, total bilirubin level, alpha-fetoprotein level, pain score, and obesity. Content validity was assessed by three experts, yielding an Item-Objective Congruence index of .95. Data was analyzed using descriptive statistics and binary logistic regression analysis.                     

The results revealed that significant predictors of liver rupture were a pain score of 1–4 points (Adj.OR = 11.73) and a tumor size greater than 10 cm (Adj.OR = 10.51). The predictive model demonstrated an optimal cut-off score of 4.97, with a sensitivity of 85.71% and a specificity of 77.31%. When validated using an independent dataset, the model achieved a sensitivity of 70.00% and a specificity of 80.00%.