Risk of Hospitalization for Heart Failure Associated with Thiazolidinedione Therapy: A Medicaid Claims–Based Case-Control Study


College of Pharmacy, University of Illinois at Chicago, 833 South Wood Street, Chicago, IL 60612.


Study Objectives. To determine, in patients with type 2 diabetes mellitus, whether an association exists between thiazolidinedione therapy or other diabetes therapies and hospital admission for heart failure.

Design. Retrospective case-control study.

Data Source. Oregon Medicaid claims database.

Patients. A total of 288 case patients and 1652 control patients.

Measurements and Main Results. Case patients were defined as any patients hospitalized for heart failure, controls as any patients with a hospital claim for a condition other than heart failure. Controls were matched by age and sex in a 6:1 ratio. Exposure to a thiazolidinedione or other antihyperglycemic drug was assessed 60 days before the first hospitalization. This was used to construct an odds ratio of exposure in case patients compared with controls using a multivariate logistic regression model and controlling for potential confounders. Charlson comorbidity index scores and frequency of diabetes-related office visits were significantly higher for case patients than for controls. The unadjusted and adjusted odds ratios for exposure to a thiazolidinedione were 1.71 (95% confidence interval [CI] 1.24–2.36) and 1.37 (95% CI 0.98–1.92), respectively. The unadjusted and adjusted odds ratios for exposure to insulin in patients hospitalized for heart failure were 1.68 (95% CI 1.27–2.22) and 1.25 (95% CI 0.92–1.69), respectively. The unadjusted and adjusted odds ratios for exposure to a combination of insulin and a thiazolidinedione in case patients were 1.81 (95% CI 1.14–2.86) and 1.35 (95% CI 0.84–2.18), respectively. No association with hospitalization for heart failure was found for patients exposed to a sulfonylurea, metformin, or α-glucosidase inhibitor.

Conclusion. This study showed a likely association between thiazolidinedione therapy and hospitalization for heart failure within 60 days of the prescription date. A similar trend occurred with insulin therapy alone and with the combination of thiazolidinedione and insulin, but not with other oral antihyperglycemics. When thiazolidinediones are prescribed, careful consideration should be given to patients with known heart failure. In addition, all patients should be educated regarding heart failure and monitored for signs and symptoms.