ObjectiveTo establish combined detection factor model in the diagnosis of patients with hepatitis B cirrhosis complicated with primary liver cancer based on logistic regression analysis, drawing receiver operating characteristic (ROC) curve to analyze the application value of the model. MethodsA total of 150 patients with hepatitis B cirrhosis admitted to our hospital from November 2017 to March 2020 were selected, and they were divided into liver cancer group (n=57) and non-liver cancer group (n=93) according to the presence of complicated with primary liver cancer. The general clinical data of the two groups were collected. Logistic regression analysis was used to construct the combined detection factor model of patients with hepatitis B cirrhosis complicated with primary liver cancer, and the ROC curve was drawn to analyze the application value of the model. ResultsThere were no statistically significant differences in gender, age, complicated with hypertension and hyperlipidemia, and HBeAg positive rate between the liver cancer group and the non-liver cancer group (P＞0.05). There were statistically significant differences in the state of complicated with diabetes and non-alcoholic fatty liver, and family history of liver cancer, conditions of tobacco dependence, alcohol dependence, antiviral treatment, HBV-DNA positive rate, as well as serum ferritin, alpha-fetoprotein, Golgi glycoprotein 73, adenosine deaminase, glutamyl transpeptidase, 5′-nucleotidase levels, etc., between the liver cancer group and non-liver cancer group (P＜0.05). Multivariate logistic regression analysis results showed that diabetes, family history of liver cancer, serum ferritin, alpha-fetoprotein, and glutamyl transpeptidase were independent risk factors for patients with hepatitis B cirrhosis complicated with primary liver cancer. Diabetes, family history of liver cancer, serum ferritin, alpha-fetoprotein, and glutamyl transpeptidase were used as covariates X1, X2, X3, X4, X5, and the combined detection factor model expression interpreted as: Logit(P)=-8.527+2.113X1+1.772X2+0.571X3+0.338X4+0.975X5. The combined detection factor model was used to predict the sensitivity (80.70%), specificity (86.02%), accuracy (84.00%), and the area under the curve (0.859, in range of 0.677 to 0.953) of hepatitis B cirrhosis patients complicated with primary liver cancer. ConclusionThe data of the combined detection factor model established based on logistic regression analysis is easily accessible and the diagnostic efficiency is high. It can enable the patients with hepatitis B cirrhosis complicated with primary liver cancer to obtain early diagnosis and treatment, so as to improve the prognosis of the patients.

无