Quality indicators are used to measure the quality of health care delivered to patients. Quality indicators are used extensively in the VA health system, and efforts are underway in Medicare to tie reimbursement levels to performance on quality indicators.

The motivations for using quality indicators are guided by the best of intentions. There are many problems with the quality of health care in the US, and quality indicators aim to improve this care. When put to their best use, quality indicators can improve care.

However, a recent commentary in JAMA from our UCSF colleagues, geriatricians Sei Lee and Louise Walter raise serious concerns about unintended harms from quality indicators. Lee and Walter make a compelling arguement that quality indicators, when used indiscriminantly, can actually harm the quality of care provided to the older persons. This is particularly true for the most frail and vulnerable elders.

How is it possible that something designed to improve care can actually be dangerous to the patients they are supposed to help?

A key issue is that quality indicators almost always promote more medical intervention and more medical intervention is not always better. This is especially the case in frail older persons, where the risks of treatments often exceeds the benefits.

Lee and Walter illustrates this problem with the HEDIS quality indicator for hypertension. This indicator measures the proportion of patients with high blood pressure who have a blood pressure of less then 140/90. In most cases, this is a reasonable target. However, in some frail older patients, the focus on blood pressure targets need to be balanced by concerns about side effects. This includes sometimes debilitating orthostatic hypotension (a drop in blood pressure when standing that can lead to dizziness and falls) and side effects of adding additional medicines in a patient with an already very full pillbox.

A doctor who decides it is in a frail older patient’s best interest to allow the blood pressure to go over the HEDIS target will appear to be providing poor quality of care. In contrast, a doctor who blindly aims for the HEDIS target in all patients will look very good. Lee and Walter note that HEDIS should balance its hypertension quality indicator by measuring consequences of overtreatment like orthostatic hypotension and syncope.

Lee notes a similar problem with many quality indicators for glucose control in patients with diabetes. Virtually all measure the proportion of patients with glucose levels below a specified target. But virtually none measure complications of aggressive blood sugar control like hypoglycemia. This lack of balance can be dangerous in the older patient, in whom the benefits of tight blood sugar control may be small, but the consequences of hypoglycemia can be catastrophic.

Quality indicators, including HEDIS measures, have undoubtedly improved the health of many patients. But the lack of balance in these measures has likely harmed the health of some frail older patients. Quality indicators need to balance their focus on more medical intervention by acknowledging that more medical intervention is sometimes harmful.

It is time for the quality improvement community to start taking concerns about potential downsides of quality indicators seriously.

Ken Covinsky is Professor of Medicine, University of California, San Francisco who blogs at GeriPal.

Submit a guest post and be heard on social media’s leading physician voice.