Management of newly treated diabetes in Medicare beneficiaries with and without heart failure.

BACKGROUND

Several diabetes mellitus (DM) therapies are associated with worse heart failure (HF) outcomes, yet limited data exist characterizing routine DM management based on HF status.

HYPOTHESIS: DM medications prescribed for patients with HF will differ in meaningful ways from patients without HF, and co-morbidities will impact the choice of medications prescribed.

METHODS

Using Medicare fee-for-service claims data, we identified patients with newly treated DM between 2008 and 2011 and used multivariable logistic regression to assess associations between baseline HF status and DM medication use. We used the cumulative incidence function to describe outcomes at 1 year by HF status and DM medication use.

RESULTS

Of 35 603 patients with newly treated DM, 8965 (25.2%) had HF at baseline. Patients with HF had greater comorbidity, including renal dysfunction. After adjustment for baseline covariates, patients with HF had higher odds of initiation on insulin monotherapy (odds ratio: 1.72, 95% confidence interval: 1.59-1.87, P < 0.001) and lower odds of initiation on metformin (odds ratio: 0.66, 95% confidence interval: 0.62-0.71, P < 0.001) compared with patients without HF. Cumulative incidence of 1-year mortality was higher among patients with HF than among those without (22.4% vs 6.4%) and highest among those prescribed insulin.

CONCLUSIONS

Patients with HF were more likely to be initiated on insulin monotherapy than were patients without HF, who were more likely to be initiated on metformin. Studies of comparative effectiveness and safety are needed for DM treatment options in patients with comorbid HF.