Optimum doctor-visit strategy for Hepatitis B patients using Markov decision processes

• Main Authors: Guo-Jhih Li, 李國誌
• Other Authors: Hsueh-Yi Lu
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/04745933202938536499

碩士 === 國立雲林科技大學 === 工業工程與管理研究所碩士班 === 100 === Hepatitis B patients may cause cirrhosis or liver cancer when the time goes by. Because hepatitis is an asymptomatic disease, the most important and efficient way to avoid cirrhosis or liver cancer is a good strategy of doctor-visit to monitor the progress of the disease. In order to let patient receive medical treatment immediately when hepatitis B becomes cirrhosis, doctors may consider increase the hostipal visit frequencies. On the other hand, patients may consider cutting down the number of visits or increase the revisit intervals to minimize the cost for medical examinations. In this study, we use Markov decision process model to optimize the doctor-visit strategy in various stages of the hepatitis B patients. We developed a framework to evaluate a series of decisions for the visit time intervals based on the stochastically progression process to find a trade-off setting with considering minimizing the visit frequency as well as shorten the non-diagnosis window period to optimize patients’ HBV disease monitoring policy. In this study, we use the risk status indicators of hepatitis patients and created an algorithm to find out the trend of value in patients test data. We set the different cost parameter and delayed diagnosis penalty cost to help model find the optimal doctor-visit strategy. The results indicate that the optimal doctor-visit strategy will have significant difference when the costs changed and found that it has different doctor-visit strategy among different patient groups. The results also showed that this study helps hepatitis B patients to get disease tracking efficiently and it will let the health care resources gets the better utilization.