Investigating the Different Stages in the Progress of Cirrhosis Using the Markov Model

Maryam Rashidpanah, Jamileh Abolghasemi, Mohsen Nasiri Toosi, Masoud Salehi



Multistate Markov models are frequently used for analyzing data obtained from longitudinal studies and they are typically appropriate in the study of chronic diseases such as liver cirrhosis. Ascites is the most common major complication of liver cirrhosis. This study was done to examine the importance of ascites complications in the survival analysis of patients with cirrhosis.

Materials and Methods:

In this longitudinal study, 305 patients with liver cirrhosis who had enrolled in a waiting list for liver transplantation in Imam Khomeini Hospital from May 2008 to May 2009 and had been followed up for at least 7 years, were investigated. To analyze the data and estimation of transition intensities, a 4-state Markov model (state 1: liver cirrhosis (without ascites), state 2: ascites complication, state 3: liver transplantation, and state 4: death) was defined. Finally, data analysis was performed using R statistical software.


Of the 305 patients studied, 180 (59%) were male. The mean (±standard deviation) age of the patients was 43/14( ± 8/39)  years. There were 127 patients with ascites during that period. Estimated transition intensities from liver cirrhosis and ascites state to death state were 0.0419 and 0.1731 per year, respectively, and from ascites state to liver transplantation state was 0.2936 per year. 7-year survival probabilities from cirrhosis and ascites state to death state were estimated 48% and 64%, respectively. Estimated mean sojourn times in cirrhosis and ascites state were 4.0166 and 1.912 years. Serum albumin, bilirubin, and prothrombin levels as well as age, and encephalopathy had a significant effect on the transition intensity from liver cirrhosis state to ascites state (p<0.001).


The result of this study indicates that transition from ascites state to liver transplantation or death state occurs faster than transition from liver cirrhosis state. Especially, accompaniment of ascites and encephalopathy increases the transition intensity to death state. The effect of age, and serum bilirubin and prothrombin levels on the survival of patients is important. As ascites complication has an important role in the prognosis of the survival of patients with cirrhotic, it is suggested that ascites would be considered as an effective factor in prioritization of waiting lists for liver transplantation.



Multistate Markov model, Transition intensity, Liver cirrhosis, Ascites, Liver transplantation

Full Text:


Copyright (c)