Compliance in microcomputer-assisted conventional insulin therapy: computer simulation study results

Abstract
Compliance in diabetes self-management is a complex issue. It involves the interdependent daily actions of self-measurement of blood glucose and adherence to a prescribed schedule of daily activities. This impacts strongly on lifestyle because it necessitates precise meal timing as well as control of size and carbohydrate content. We sought to identify how strongly relaxing the lifestyle constraints per se would impact the ability to achieve improved metabolic control. To isolate these effects from those that result from poor measurement compliance, we used a computer simulator called OMNI et al. Furthermore, to standardize the "clinical" therapy, a second microprocessor device called an "Insulin Dosage Computer" was used to adjust insulin doses based on the usual clinical practice of four times a day precibal blood glucose measurements. Ten type 1 diabetic patients were stimulated and each followed for 120 simulated days. In each such subject, the simulation was repeated three times to include three different levels of lifestyle compliance ranging from excellent to poor. In all three protocols, starting from a level of poor control of diabetes, mean blood glucose values were significantly improved without significant differences after 120 days of computer-simulated treatment. Only the standard deviations, expressing the fluctuations of blood glucose and hence its stability, increased with decreasing lifestyle compliance. This computer simulation predicts that consistent self-monitoring of four blood glucose values per day is the cornerstone of diabetic self-control and that the use of these data according to a standardized therapeutic algorithm for insulin adjustment may successfully stabilize even patients with poor lifestyle compliance. Clinical studies must follow.