Markov models for patients with bladder cancer

Bladder cancer is one of the most common cancers in the world. Cigarette smoking is the major risk factor affecting bladder cancer. The aim of this study is to show how Markov process may be used to estimate the expected survival before cystectomy and understand bladder cancer prognosis. The 193 patient who have undergone transurethral resection were attendend to our study from Ankara Atatürk Education and Research Hospital. Data were collected and analyzed using IBM SPSS Statistics 20.0 and MS-Excel 2007. Fundemantal Matrix Solution, Markov Cohort Approach and Monte Carlo Simulation are used. Accuracy rate was presumed p<0.05 when result was criticizing. According to Fundemantal Matrix Solution, patients at high risk will undergo cystectomy after 5 years. The expected survival is 6 years for non-smoker patients, in Cohort approach. Patients who smoke less than 30 packets of cigarette per year will undergo cystectomy after 36 years according to Monte Carlo simulation. This thesis is one of first studies using Markov process in medical decision making in Turkey. We found that the use of cigarette is associated with a significant increase in the expected survival before cystectomy.