DIAS-NIDDM - A MODEL-BASED DECISION-SUPPORT SYSTEM FOR INSULIN DOSE ADJUSTMENT IN INSULIN-TREATED SUBJECTS WITH NIDDM

A decision support system has been developed, Diabetes Insulin Advisor y System for patients with non-insulin dependent diabetes mellitus (DI AS-NIDDM), assisting in the adjustment of insulin doses in insulin-tre ated subjects. DIAS-NIDDM uses a causal probabilistic network (CPN) mo del of carbohydrate metabolism to make stochastic predictions of blood glucose (BG) excursions. The CPN model is an extension of an existing model with an added component representing endogenous insulin secreti on. A linear relationship between BG and insulin concentration due to BG stimulated insulin secretion is assumed. Model parameters (pancreat ic sensitivity, insulin sensitivity, and time-to-peak of NPH insulin) are estimated by Bayesian probability updating from patient's specific data (food intake, insulin doses, BG measurements) recorded over a pe riod of 4 days. The estimated parameters allow the system to be potent ially used as a diagnostic tool to identify abnormalities of carbohydr ate metabolism: impaired insulin secretion, insulin resistance and the severity of the impairments. DIAS-NIDDM was used to predict patient-s pecific BG profiles and advise on insulin doses during a pilot study i n eight patients with NIDDM of whom five were treated with insulin. Co mpared to the administered insulin amount, daily insulin amount advise d by DIAS-NIDDM was similar (within 4 U) in three patients, higher by 20% (19 U) in one patient and lower by 40% (18 U) and 50% (11 U) in tw o patients, respectively. The inter-day coefficient of variation of th e daily insulin advice suggests that, at least according to DIAS-NIDDM criteria, day-to-day adjustment of insulin doses is necessary to main tain optimum control. (C) 1998 Elsevier Science Ireland Ltd. All right s reserved.