An increase in use of metformin and long-acting insulin was seen in elderly Medicare patients from 2007 to 2014, but more expensive and complicated regimens remained common, a recent study indicated.
Researchers performed a retrospective cohort study of Medicare claims from 2007 to 2014 in traditional Medicare beneficiaries who had ever received a diagnosis of diabetes. Main outcome measures were first choice of diabetes medication for those who had not previously taken medication and patterns in which medications were added. In addition, researchers looked at use of diabetic testing supplies, use of statins and angiotensin-converting enzyme inhibitors/angiotensin-receptor blockers, and spending. The results were published online Feb. 9 by Journal of General Internal Medicine.
From 2007 to 2014, the percentage of beneficiaries with a diabetes diagnosis increased from 28.7% to 30.2%. Metformin as initial therapy increased from 50.2% in 2007 to 70.2% in 2014, and use of new classes of diabetes medications, such as dipeptidyl peptidase-4 (DPP-4) inhibitors and sodium-glucose cotransporter 2 (SGLT2) inhibitors, also increased, from 4.7% to 7.9%. Long-acting sulfonylureas as initial therapy decreased from 16.6% to 8.2% over the same time period, as did thiazolidinediones, from 11.3% to 1.5%. In patients who were prescribed insulin for the first time, long-acting regimens became more common, increasing from 38.9% in 2007 to 56.8% in 2014. However, many patients continued to receive short-acting or combination regimens, especially if they were older or sicker.
Diabetic testing supplies were often prescribed for testing more than once per day, especially in patients taking short-acting insulin. Use of statins and angiotensin-converting enzyme inhibitors/angiotensin-receptor blockers increased over the study period. An increase was seen in mean total cost of diabetes medications per patient during the study period from $880 to $1,406 per year, related to increased use of high-cost therapies. However, median plan costs and beneficiary out-of-pocket costs decreased, as did both mean and median total costs of diabetic testing supplies.
The researchers noted that their study relied on administrative claims data, that they did not have information on patients' functional status and frailty or duration of diabetes, and that prescription data for Medicare Part D were not available for all patients. However, they concluded that while treatment of Medicare patients with diabetes has improved substantially, expensive and complicated regimens are still prescribed, and glucose self-monitoring is often overused. “Substantial opportunity remains to simplify treatment regimens and testing, making it easier for patients to manage their medications, contributing to cost savings, and improving the value of diabetes treatment for both Medicare beneficiaries and elderly patients in general,” the authors wrote.