Many older adults with type 2 diabetes (T2D) are tightly controlled with high hypoglycemia risk medications and may benefit from deintensifying (stopping/reducing) or switching to newer medications. As deintensifying rarely occurs in practice, we aimed to determine how physicians make these decisions. We conducted a national survey of a random sample of US physicians in general medicine, geriatrics, and endocrinology. The survey was designed by a team of patients, caregivers, and physicians, and distributed in 3 mailing waves. Physicians were asked what action they would take in three scenarios of patients with T2D treated below individualized targets: 1) a healthy 79F with HbA1c 6.3% on a sulfonylurea; 2) a 77M with stage 4 kidney disease with HbA1c 7.3% on basal insulin; 3) a 78F with advanced dementia with HbA1c 7.7% on a sulfonylurea. Scenarios started with a base case of the patient reporting no medication issues, then varied hypoglycemia history, health, and patient preferences. Statistical comparisons used multinomial logistic regression accounting for clustering. We received 427 responses to date, response rate 27%. In the base case, 57% of physicians made no change to the high-risk medication, 16% switched, 12% decreased, 12% stopped, and 3% increased. Decreasing/stopping occurred over twice as often in scenario 1 than other scenarios (p<0.001) . In scenario variations, hypoglycemia of any severity led >95% of physicians to decrease, stop or switch the high-risk medication. Worsening health status also caused substantially more deintensification, but polypharmacy, financial barriers, and patient preference did not. Endocrinologists were 3-5 times as likely to switch medications compared to other specialties (p<0.001) . Overall, we found clinical inertia to deintensify high-risk diabetes medications, especially in medically complex patients with HbA1c >7.0%. Hypoglycemia and worsening health were strong motivators to deintensify or switch to newer medications.


S.J.Pilla: None. N.Schoenborn: None. C.Boyd: Other Relationship; Dynamed, UpToDate. S.Golden: Advisory Panel; Abbott Diabetes, Medtronic. N.N.Mathioudakis: None. N.M.Maruthur: Other Relationship; Johns Hopkins HealthCare Solutions.


National Institutes of Health (1R24AG064025 and 1K23DK128572)

Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at