Progression to insulin for patients with diabetes mellitus using the Texas Medicaid database

Patients newly diagnosed with type 2 diabetes mellitus generally initiate therapy with either metformin [Met] or a sulfonylurea [SU] drug, followed by the addition of a second agent (Met, an SU drug, or a thiazolidinedione [TZD] drug) if the diabetes is not well controlled. If necessary, the usual third line of treatment is the addition of insulin.

The purpose of our study was to compare the progression to insulin among 3 cohorts receiving the oral antidiabetic (OAD) drug combinations Met/SU, Met/TZD, or SU/TZD.

This study used data from the Texas Medicaid database. The date of addition of a second OAD was considered a patient's index date and patients were followed for up to 5 years. Cox proportional hazards regression compared the progression to first insulin use among cohorts, using the Met/SU cohort as the reference group, while adjusting for demographics, comorbidities, and propensity scores.

A total of 4083 patients were included in the study (Met/SU = 2872, Met/TZD = 438, and SU/TZD = 773). Insulin was added to the medication regimen of patients by the end of follow-up in 19.7% of the Met/SU cohort, 17.6% of the Met/TZD cohort, and 26.3% of the SU/TZD cohort. The adjusted Cox proportional model estimated that patients in the SU/TZD cohort had a 40% higher probability of progression to insulin than patients in the Met/SU cohort (odds ratio [OR] = 1.40; 95% CI, 1.19-1.64), whereas there was no significant difference between the Met/TZD and Met/SU cohorts (OR = 0.85; 95% CI, 0.67-1.08).

It appears that mechanism of action may play a role in progression to insulin for concomitant OAD agents. A slower progression to insulin was seen for patients receiving a paired sensitizer regimen (ie, Met/TZD) compared with those receiving a secretagogue sensitizer regimen (ie, SU/TZD).