Comparisons of rosiglitazone versus pioglitazone monotherapy introduction and associated health care utilization in Medicaid-enrolled patients with type 2 diabetes mellitus.

BACKGROUND: Outcomes in patients with type 2 diabetes mellitus (DM) can differ based on the antidiabetic medication that is used. Thiazolidinediones (TZDs) are a newer class of agents used for the treatment of type 2 DM. No previous study has compared health care utilization associated with the 2 TZDs on the market.

OBJECTIVE: The objective of this study was to compare health care utilization and costs associated with initiation of treatment with either rosiglitazone or pioglitazone by Medicaid-enrolled patients with type 2 DM.

METHODS: This was a retrospective data analysis comparing cohorts of patients with type 2 DM starting a new antidiabetic medication in terms of hospitalizations, emergency department visits, outpatient physician visits, and health care costs reimbursed by the North Carolina Medicaid program. The perspective adopted in this analysis was that of the third-party payer (ie, the North Carolina Medicaid program). Patients starting rosiglitazone between July 1, 2001, and June 30, 2002, were compared with patients starting pioglitazone during the same period. The patients were followed up for 30 months to examine the difference in health care utilization over time. Multivariate regression techniques were employed for comparisons between the 2 different antidiabetic therapies.

RESULTS: A total of 1705 patients with type 2 DM were identified and included in the final cohort. There were 660 patients (mean [SD] age, 49.0 [10.2] years) in the rosiglitazone arm and 1045 patients (mean [SD] age, 49.1 [10.5] years) in the pioglitazone arm. Multivariate analysis showed that the rosiglitazone monotherapy group was associated with a 12.2% decrease in the mean number of hospitalizations, a 10.4% decrease in the mean number of emergency department visits, and a 7.3% decrease in total health care costs compared with the pioglitazone monotherapy group (all, P < 0.05). This study only looked at patients who used the same drug for the entire follow-up period. It did not account for drug switching or addition of a new drug to an existing therapy.

CONCLUSIONS: Introduction of rosiglitazone was associated with a decreased number of hospitalizations, emergency department visits, and total health care costs compared with pioglitazone. The utilization of oral antidiabetic agents, with documented clinical and economic benefits, should continue to be advocated to reduce avoidable medical care utilization and to improve patient outcomes in this population.