Current diabetes quality indicators do not reflect considerations of clinical complexity, hypoglycemia risk, and treatment burden in identifying optimal therapeutic regimens and targets.

Based on existing evidence and input from patients and clinician, payer, and quality measurement experts, we developed a metric that incorporates HbA1c level, comorbidity burden, and number/type of glucose lowering medications (Table) to classify patients as potentially overtreated, undertreated, or appropriately treated. Measure was tested in a national cohort of commercial and Medicare Advantage enrollees with diabetes using the OptumLabs® Data Warehouse.

Among 215,429 adults ≥18 years (mean 65 years, 49.6% female, 59.4% white, 4.2% type 1 diabetes, 12.1% clinically complex, 52.7% with HbA1c <7%), 71.0% were treated appropriately, 21.3% undertreated, and 7.7% overtreated in 2015. Compared with current measures (HbA1c <8% for all), 4.4% low complexity and 12.6% high complexity patients were reclassified as no longer appropriately treated. Rates of hypoglycemia-related ED/hospitalization visits in 2016 were 7.0, 12.4, and 22.9 per 1000 persons/year for those appropriately, under-, and over-treated in 2015. Rates of hyperglycemia-related visits were 2.5, 12.2, and 4.6.

Implementation of this measure to minimize over- and under-treatment through a patient-centered approach to glucose lowering therapy is planned.


R.G. McCoy: None. K.J. Lipska: Consultant; Self; Health Services Advisory Group (HSAG). Research Support; Self; National Institute on Aging. Other Relationship; Self; Centers for Medicare and Medicaid Services. H. Van Houten: None. N. Shah: None.


American Association of Retired Persons; OptumLabs; National Institute of Diabetes and Digestive and Kidney Diseases (K23DK114497)

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