Many Patients With Type 2 Diabetes Stop Meds After a Year

Almost two fifths of people with type 2 diabetes discontinue their second-line medication within a year of starting, reveals an analysis that suggests the risk may be the greatest with glucagon-like peptide-1 (GLP-1) receptor agonists.

The research, published in the American Journal of Management Care on December 12, looked at prescribing patterns of more than 80,000 individuals who started a prescription for one of five classes of second-line diabetes medications.

“Discontinuation is bad,” commented first author David T. Liss, PhD, Division of General Internal Medicine, Northwestern University Feinberg School of Medicine, Chicago, Illinois, in a release.

“It is common in all five types of medications, but we see significantly more in those prescribed GLP-1 receptor agonists,” he said, adding: “Presumably, the doctor is saying: ‘You need to start a new medication to control your type 2 diabetes,’ and then within a year, half of them just stop and don’t start another one, and that’s not a good thing.”

Liss said the findings “highlight the need for new prescribing approaches and to better understand the barriers patients face when taking these medications, to ultimately reduce wasting patients’ time, clinicians’ time, and the health system’s money.”

Considering the increased discontinuation rates with GLP-1 receptor agonists vs sulfonylureas, Liss told Medscape Medical News the latter “are certainly a low-cost option.”

Moreover, dose titration with GLP-1 receptor agonists “can be challenging for patients.”

Rates Lower When Prescribed by Endocrinologist

The study also indicated there were lower rates of drug discontinuation when endocrinologists made the prescription.

Liss said “it’s reasonable to assume that the endocrinologist had particular expertise in both the benefits and the drawbacks of these medicines.”

“Perhaps they were able to leverage some of that expertise to then have more informed conversations with the patients” for shared decision-making.

Liss emphasized, however, that for their analysis, they relied on insurance claims data, which “do not contain contextual details, like a patient’s reasoning for discontinuation…so we’re left to speculate a bit.”

Approached for comment, Marilyn Tan, MD, Clinical Associate Professor of Medicine at Stanford University School of Medicine, Stanford, California, said “the discontinuation rates are higher than in my own clinical experience,” although she also pointed to the lower rates when the prescription came from an endocrinologist.

“It really depends on individual patients. Most of my patients have been quite motivated to stay on GLP-1 receptor agonists for the weight loss and cardiovascular benefits, if they can afford to stay on them and can tolerate them.”

Tan told Medscape Medical News there are, in fact, “many reasons” why patients may have difficulties with medication adherence, such as “busy schedules, a high pill burden, medication side effects, perceived lack of efficacy, and, importantly, costs.”

She noted that patients with diabetes also often have “multiple medications and complicated medication schedules,” so the overall regimen “can become very complicated and burdensome.”

Rodolfo J. Galindo, MD, Director, Comprehensive Diabetes Center at the University of Miami Health System, Miami, Florida, commented that, “unfortunately, this is not new in clinical practice,” as previous studies have revealed “high rates of drug discontinuation with diabetes medications.”

Indeed, a recent study by Galindo and colleagues of cross-sectional data from the 2010-2019 Medical Expenditure Panel Surveys indicated that discontinuation rates with weight-reducing medications was nearly twice that of weight-inducing drugs.

He told Medscape Medical News that the “correct answer” to the question as to why patients discontinue medication is “usually multifactorial.”

“Some patients stop medication based on side effects, albeit not very frequently in clinical trials, but many people stop GLP-1 receptor agonists also because of supply issues, lack of insurance coverage, and switching to other agents.”

Galindo explained that the demand GLP-1 receptor agonists means they are being used off-label, which is affecting supplies, and their high cost is an issue. Moreover, insurance coverage for them is not consistent in the United States.

“They are covered today,” he said, “and tomorrow they say you need a prior authorization” to prescribe them. “That is very relevant” to the question of drug discontinuation, and “it is alarming.”

Galindo also pointed out that the time period covered by the study is before the 2018 Standards of Medical Care in Diabetes by the American Diabetes Association, which shifted the emphasis of management away from A1c levels to “cardiovascular and then cardiorenal complications.”

The study is, therefore, “5 years old and so much as changed over the last 2 or 3 years.”

The authors say recent data shows 77% of patients with type 2 diabetes initiating medication start with metformin, “reflecting treatment patterns largely in accordance with clinical guidelines.”

“However, to achieve long-term glycemic control, most patients ultimately require combination therapy with other antidiabetic medication classes,” they continue, and the options for second-line therapy are “increasing rapidly.”

With “limited data” on changes in second-line antidiabetes medication use, the team set out to investigate further by examining the period after the introduction of sodium-glucose cotransporter 2 (SGLT2) inhibitors.

Discontinuation Most Common Among GlP-1 Agonists

They conducted a retrospective analysis of adults enrolled in a commercial or Medicare Advantage health plan and who initiated an index, second-line antidiabetes prescription between January 1, 2014, and June 30, 2017.

The index medications had to come from five classes:

• Sulfonylureas
• Dipeptidyl peptidase 4 (DPP4) inhibitors
• SGLT2 inhibitors
• GLP-1 receptor agonists
• Thiazolidinediones

The individuals were required to have preexisting type 2 diabetes, at least one pharmacy claim for metformin during the 6 months before the index date, and a pharmacy claim on the index date that included an index medication, at a supply of at least 28 days.

The researchers identified 82,624 adults with type 2 diabetes who met the inclusion criteria, of whom 54.0% were men. The majority (57.0%) had a non-Medicare commercial insurance plan, and 61.5% had no diabetes complications.

Sulfonylureas were the most commonly prescribed index medication, in 51.0% of cases, followed by DPP4 inhibitors (24.0%), SGLT2 inhibitors (11.6%), GLP-1 receptor agonists (8.1%), and thiazolidinediones (5.3%).

During the follow-up, 63.6% of patients had a treatment modification. The most common was treatment discontinuation, in 38.6%, while 19.8% had a treatment intensification, and 5.2% switched their treatment class.

The researchers report that discontinuation was most common among patients initially prescribed a GLP-1 receptor agonist, at 50.3% vs 36.6% with sulfonylureas, 39.5% with DPP4 inhibitors, 39.4% with SGLT2 inhibitors, and 34.2% with thiazolidinediones.

Multivariate Cox proportional analysis revealed that patients were significantly more likely to discontinue GLP-1 receptor agonists, at a hazard ratio of 1.28, and DPP4 inhibitors, at a hazard ratio of 1.07, than sulfonylureas.

The team also found that compared with sulfonylureas, patients were significantly more likely to switch from all other medication classes and less likely to intensify their medication.

Individuals aged 18-44 years were more likely to discontinue, switch, and intensify their medication than older people, regardless of prescription, and men were at a lower risk for treatment modification of any kind than women.

Receiving the prescription from an endocrinologist was associated with a lower risk for discontinuation and a higher likelihood of intensification than that from a family medicine or internal medicine physician.

Funding support and data access for this work were provided through a grant to Northwestern University from UnitedHealth Group, the company whose insurance arm provided coverage to the members under study.

Liss declares no relevant relationships.

Other authors declare relationships with Novo Nordisk, Eli Lilly and Company, and UnitedHealth Group.