JAMA Intern Med 2014 Oct 27;
Initial metformin yielded a lower likelihood of needing a second agent compared with other drug classes.
Metformin is recommended by several organizations and guidelines as first-line therapy in patients with newly diagnosed type 2 diabetes, but clinicians often start with other classes of hypoglycemics. In this retrospective cohort study, researchers used pharmacy and claims data from a large national health insurance plan to assess the consequences of first-line treatment in about 15,000 patients who initiated therapy with metformin (58% of patients), sulfonylureas (23%), dipeptidyl peptidase 4 inhibitors (13%), or thiazolidinediones (6%). Follow-up was approximately 1 year.
In adjusted analyses, patients who initiated therapy with drug classes other than metformin were about 60% more likely to require a second agent, either another oral drug or insulin (37% vs. 24%). The second drug was most commonly metformin. Initial sulfonylurea use was associated with excess risk for new adverse cardiovascular events and for emergency department or hospital admissions for hypoglycemia.
Limitations of this study include its retrospective nature and the researchers’ inability to adjust for certain potentially important variables, such as glycosylated hemoglobin and serum creatinine levels. Nevertheless, the results reinforce recommendations for using metformin as first-line therapy for patients with new diagnoses of type 2 diabetes; they also add to increasing concern about using sulfonylureas. The finding that metformin was the initial choice in fewer than 60% of patients suggests that we need to develop approaches to change physician behavior.