Limitations of diabetes pharmacotherapy: results from the Vermont Diabetes Information System study

BACKGROUND There are a wide variety of medications available for the treatment of hyperglycemia in diabetes, including some categories developed in recent years. The goals of this study were to describe the glycemic medication profiles in a cohort of adult patients enrolled in primary care, to compare the regimens with measures of glycemic control, and to describe potential contraindicated regimens. METHODS One thousand and six subjects with diabetes cared for in community practices in the Northeast were interviewed at home at the time of enrollment in a trial of a diabetes decision support system. Laboratory data were obtained directly from the clinical laboratory. Current medications were obtained by direct observation of medication containers by a research assistant. RESULTS The median age of subjects was 63 years; 54% were female. The mean A1C was 7.1%, with 60% of subjects in excellent glycemic control (A1C < 7%). Ninety percent of patients were taking 2 or fewer medications for glycemic control, with a range of 0 to 4 medications. Insulin was used by 18%. As the number of diabetes medications increased from 0 to 4, the A1C increased from 6.5% to 9.2% (p < 0.001). The association between glycemic control and number of glycemic medications was confirmed using logistic regression, controlling for potential confounders. Almost 20% of subjects on metformin or thiazolidenediones had potential contraindications to these medications. CONCLUSION Patients with diabetes cared for in primary care are on a wide variety of medication combinations for glycemic control, though most are on two or fewer medications. A greater number of diabetes medications is associated with poorer glycemic control, reflecting the limitations of current pharmacotherapy. One quarter of patients are on glycemic medications with potential contraindications.