Loss of glycemic control in patients with type 2 diabetes mellitus who were receiving initial metformin, sulfonylurea, or thiazolidinedione monotherapy

Study Objectives. To examine glycosylated hemoglobin (A1C) values longitudinally in patients who newly started metformin, sulfonylurea, or thiazolidinedione monotherapy; in a subset of patients whose A1C values were 7% or greater before starting therapy (baseline) and who achieved A1C goal (A1C < 7%) during therapy, rates of secondary failure (i.e., A1C value returned to >= 7% during therapy) were compared for each drug. Design. Four-year retrospective analysis. Data Source. Administrative database from a large health care plan. Patients. Patients who filled at least one prescription for metformin (5453 patients), sulfonylurea (2373), and thiazolidinedione (1590) therapy, respectively, between January 1, 2001, and March 31, 2004, were enrolled. Measurements and Main Results. Patients' demographic and clinical characteristics, baseline A1C values, changes in A1C values (last available result during follow-up minus baseline value), and A1C values before and after the addition of an antidiabetic drug other than the index drug (therapy intensification) were documented. Mean age was 50.7 years; 5027 (53.4%) were men. Mean baseline A1C value was 8.4%, and about 70% of patients had an A1C value of 7% or greater before starting therapy Mean follow-up was 1.9 years, and mean decrease in A1C values was 1.47% (to 6.91%). The probabilities of attaining A1C goals were similar for patients receiving metformin, sulfonylurea, or thiazolidinedione therapy The rate of therapy intensification among patients taking metformin (24.7%) was lower than that of patients taking a sulfonylurea (30.1%, p<0.001) but similar to that of those taking a thiazolidinedione (24.6%). Secondary failure occurred in 36.3% of patients; mean time from the start of therapy to its failure was about 1.51 years. Patients receiving a sulfonylurea were 1.25 (95% confidence interval [CI] 1.05-1.50) times more likely than patients receiving metformin to experience secondary failure, whereas failure rates were similar for thiazolidinediones and metformin (odds ratio 0.78, 95% CI 0.62-0.99). Conclusion. In the subset of patients assessed for secondary failure, although treatment initially reduced A1C values, more than one third experienced failure. Real-world studies of AIC goal attainment must follow patients on a long-term basis to assess the maintenance of glycemic control over time.