Clinical trials have shown that insulin analogues, such as glargine, detemir, aspart, and lispro, do not offer major advantages over human insulin products, such as neutral protamine Hagedorn (NPH) and regular human insulin, for patients with type 2 diabetes.1,2 In these studies, neither the rate of severe hypoglycemia nor the achieved glycemic control improved with insulin analogues. Results from observational studies largely confirmed these findings in clinical settings. One such study, conducted at Kaiser Permanente, showed no significant differences in health care use related to hypoglycemia or in levels of glycemic control among patients with type 2 diabetes who started a basal insulin analogue vs patients who started NPH insulin.3 However, in clinical trials, insulin analogues modestly reduced the rate of nocturnal hypoglycemia, an important outcome for patients with diabetes. Notably, the clinical trials were open label, so they do not have the advantage of blinding, and the nocturnal hypoglycemia outcome was self-reported. Therefore, these trials are subject to risk of bias.