Abstract:
Primary Biliary Cirrhosis (PBC) is a dreadful rare disease as it affects an important human organ, the liver. This study was conducted by survival analysis on Primary Biliary Cirrhosis patients in the United States using prerecorded data from 310 patients who were suffering from PBC. The Follow-up time, Status at the end, and seventeen other factors were analyzed statistically to identify the risk factors for Primary Biliary Cirrhosis. Baseline characteristics were calculated and a nonparametric model (Kaplan Meier), a semi-parametric model (Cox Proportional Hazards), and two parametric models (Exponential and Weibull) were developed for analyzing mortality and survival appropriately. Area under the Receiver Operator Characteristics curve was used for comparing the four models and the model which produced the highest Area under the Receiver Operator Characteristics curve was identified as the best fitted model. A nomogram was constructed for the graphical prediction of mortality or survival appropriately. Baseline characteristics showed that Age, Serum Bilirubin, Albumin, Urine Copper, Alkaline Phosphatase, SGOT level, Platelets, Prothrombin Time, Presence of Ascites, Presence of Hepatomegaly, Presence of Spiders, Presence of Edema, and Histologic Stage of Disease are significant factors on PBC. The Cox Proportional Hazards model could be identified as the best-fitted model from among the four fitted models, according to which Age, Serum Bilirubin, Albumin, and Urine Copper were the risk factors of Primary Biliary Cirrhosis. Moreover, Age and Serum Bilirubin were the most significant risk factors as they emerged in all fitted models.
