MODEL COX PROPORTIONAL HAZARD UNTUK MENGIDENTIFIKASI FAKTOR RISIKO KEMATIAN PADA PASIEN PRIMARY BILIARY CIRRHOSIS

Ulfa Rusyada (Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Lambung Mangkurat)
Reny Maharani (Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Lambung Mangkurat)
Samsul Arifin (Program Studi Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam, Universitas Lambung Mangkurat)

Jurnal Gaussian
Vol 15, No 1 (2026): Jurnal Gaussian

Primary biliary cirrhosis (PBC) is a progressive chronic liver disease associated with a high risk of mortality. This study aims to identify clinical factors influencing the survival of PBC patients using the Cox Proportional Hazard (Cox PH) model. The Cox PH model was selected because it is appropriate for censored survival data and offers the advantage of estimating the effects of multiple covariates without requiring a specific baseline hazard distribution. The study used data from 312 PBC patients treated at the Mayo Clinic. All variables were evaluated using the Log-Rank test and satisfied the proportional hazard assumption. The results indicate that ascites (HR = 3.951), high copper levels (HR = 1.964), high bilirubin levels (HR = 3.789), and advanced disease stage (HR = 9.222) significantly increase the risk of mortality, while higher cholesterol levels reduce the risk (HR = 0.392). These findings demonstrate that the Cox PH model is effective for identifying significant clinical risk factors and providing interpretable survival analysis for PBC patients, which may support clinical decision-making and patient management.