Trade-offs in treatment alternatives for non-insulin-dependent diabetes mellitus.

In order to evaluate the choice among alternatives quantitatively, a formal decision model must be developed. Yet the literature does not currently include the information required to develop this model. Data tell us very little about what factors should be considered in the choice of treatment for NIDDM patients. Table 1 summarizes some of the advantages and disadvantages of insulin, sulfonylureas, and dietary treatments. Insulin may have the greatest effect upon blood glucose, but may also be associated with the greatest likelihood of nuisance for the patient. At the other extreme, dietary treatment may be safe, but may have a low probability of achieving long-term blood glucose control. There is remarkably little in the literature that considers nuisance factors for the patient, minor but persistent side effects, or the likelihood of other physical changes such as weight gain. We know even less about how to integrate preferences for benefits and side effects into a comprehensive decision. Although some profiles of laboratory results clearly dictate a treatment protocol, there is considerable variability in the treatment options for a large number of NIDDM patients. Consider, for example, the patient who has a fasting blood glucose of 250 mg/dl but no symptoms. There may be several treatment alternatives. Yet the chances of therapeutic success could be influenced by the patient's concern about being dependent upon medication, willingness to comply with life-style changes, and fear of using needles. We suggest that the patient must be active in negotiating the choice of treatment, and that patient preferences for expected outcomes, side effects, and nuisance factors need to be considered.(ABSTRACT TRUNCATED AT 250 WORDS)