Optimizing and Cost-effective Analysis of Medication Treatment of Type 2 Diabetes for Prevention of Hypoglycemia Using Markov Decision Process

Type 2 diabetes (T2D) imposes high expenses on societies, due to its effects on individual and social functions of patients. Adverse complications of this disease can be prevented through controlling patient's blood sugar level. The purpose of this study was to optimize the medication treatment of type 2 diabetes.
In this applied survey, Markov decision process (MDP) was used to model the scheduling and sequencing of T2D medication treatments. In order to have a model close to reality, expense issues, time value of money (TVM) and inflation were considered and it was tried to plan a cost-effective treatment program. Since anti-diabetic drugs cause hypoglycemia in elderly patients, a hypoglycemia preventive approach was applied on the proposed model and its effect was analyzed. The function of proposed model and efficiency of the obtained optimal policy were compared with clinical guidelines. Moreover, in order to address the model behavior, sensitivity analysis on parameters related to treatment costs was also conducted.
The proposed model showed better performance in glycemic control compared to clinical guidelines. Treatment cost parameters affected treatment program.
In general, optimizing the medication treatment of T2D not only can improve patient's life period with quality, but also can decline high costs imposed on societies’ healthcare systems.