Search | Search by Center | Search by Source | Keywords in Title
Kerr EA, Holleman R, Klamerus ML, Saini SD, Min L, Vijan S, Sussman JB, Hofer TP. Failing to Choose Wisely: Lack of Treatment De-intensification in Older Patients with Diabetes. Paper presented at: Society of General Internal Medicine Annual Meeting; 2014 Apr 24; San Diego, CA.
Background: It is increasingly recognized that efforts to intensively improve hemoglobin A1c (A1c) and blood pressure (BP) control in patients with diabetes sometimes unintentionally result in overtreating patients who are unlikely to benefit. Recent guidelines and initiatives like Choosing Wisely recommend less aggressive treatment for older patients with limited life expectancy (LE), since such treatment is unlikely to improve outcomes and may cause harm. This suggests that medication de-intensification is indicated among many patients with limited LE. Yet, we know little about how often providers decrease treatment intensity among patients who are unlikely to benefit. We examined how often the number or dose of medications are decreased in older patients with diabetes and lower than recommended levels of A1c or BP, and whether medication de-intensification is more common in patients with lower LE. Methods: Using the Veterans Health Administration's (VHA) Corporate Data Warehouse, we identified all patients with diabetes, age 70 and older, receiving primary care in VHA during 2012. We then identified the patients' last A1c and BP during 2012, and their prescribed medications at the time of this index A1c or BP. Patients with A1c < 6.5% and on hypoglycemic medications (other than metformin alone) were considered eligible for de-intensification. Patients with BP < 130/65 and on BP medications (other than low dose angiotensin converting enzyme inhibitor (ACEI) alone or low dose angiotensin receptor blockade (ARB) alone) were similarly eligible. We determined the proportion of patients who had medications discontinued or had the dose decreased in the 6 months after the index A1c or BP. Using logistic regression, we examined the association between limited LE, based on the Charlson comorbidity index, and de-intensification, controlling for number of medications the patient was taking at the time of the index A1c or BP. Results: 500,742 patients with diabetes, age 70+, were receiving primary care in VHA during 2012. Among the 114,411 with A1c < 6.5%, 51% were on no medications, 15% were on metformin only, 8% were on insulin only, 14% were on sulfonylureas only, and 12% were on other or combination therapy. Among the 36,686 eligible for deintensification, 8,466 (23%) were de-intensified within 6 months. Patients with LE < 5 years were modestly more likely to be de-intensified than those with LE > 10 years (predicted probability 29% vs 19%, p < 0.001). Among the 66,750 patients with a BP < 130/65, 10% were on no medications, 6% were on low dose ACEI or ARB only, 22% were on 1 BP medication, and 62% were on 2 or more BP medications. Among those eligible for de-intensification, 9,883 (21%) were de-intensified. Patients with LE < 5 years were somewhat more likely to be de-intensified than those with LE > 10 years (predicted probability 24% vs 17%, p < 0.001). Conclusions: Despite increasing awareness that overtreatment of diabetes may be harmful in older adults, less than one quarter of patients treated to levels significantly lower than recommended had medications de-intensified. While patients with limited LE were more likely to be de-intensified than those in good health, the majority of eligible older patients, even those with limited LE, continued to receive care that is of low value or potentially harmful. Future initiatives that seek to help patients and providers choose wisely will need to address this significant clinical inertia for de-intensification of treatment.