Study of the process of secondary failure of sulphonylurea by a Markov model.

PURPOSE: The study was to reflect and forecast the evolutive tendency and influence factors of secondary failure of sulphonylurea (SFS) changing with time by using a Markov (MKV) model in the elderly diabetic population in Shanghai. METHODS: A total of 549 patients with elderly diabetes mellitus (DM) were enrolled and grouped in the study. A door-to-door retrospective epidemiological survey was used to collect data. The MKV model was used to assess the process and influence factors of SFS and the MKV process decision support system was adopted to calculate state probability of the MKV process. RESULTS: The rate of SFS in the group of all cases, FPG < or = 10 mmol . L(-1) before treatment and FPG > 10 mmol . L(-1) before treatment, taking single type of sulphonylurea (SU) and taking two types of SU and over respectively was 9.11%, 3.55%, 11.03%, 8.54% and 11.21%. The years of changing into the state of secondary failure in half patients was 5 years, 11-12 years, 4 years, 5 years, 4 years, respectively in the following groups: all cases, FPG < or = 10 mmol . L(-1) before treatment and FPG > 10 mmol . L(-1) before treatment, taking single type of SU and taking two types of SU and over. CONCLUSIONS: A MKV model could predict the long-term evolutive process of SFS by a short-term observation; the speed of SFS was related to the degree of DM patients' condition, patients with higher glucose levels prior to treatment would develop SFS faster; but we cannot postpone the development of secondary failure by exchanging SU types.