The primary goal of treatment for diabetes, as for many chronic conditions, is to lower rates of death and complications from the disease. But an equally important goal in medicine, one recognized over 2 millennia ago, is to “do no harm,” and sometimes this balance between help and harm can be difficult to accomplish.

In this week’s issue of JAMA, a theme issue on diabetes, a viewpoint discusses the overtreatment of diabetes in certain patient populations—that is, the overuse of blood sugar–lowering medications to the point at which average blood sugar levels may be too low and cause more harm than benefit. The authors refer back to a study published about 6 months ago in JAMA Internal Medicine, in which researchers looked at administrative data from a large population of Veterans Affairs patients and found surprisingly high rates of overtreatment.

The JAMA Internal Medicine study focused on patients with diabetes who were most susceptible to hypoglycemia (low blood sugar) caused by diabetes overtreatment, which included patients older than 75 years of age, those who had poor kidney function, and those who had a diagnosis of cognitive impairment or dementia. Overtreatment among these patients was defined as having a hemoglobin A_1c (HbA_1c) level (a measure of average blood sugar values over a 3-month period) that was less than 6.0% (roughly corresponding to a blood sugar of 126 mg/dL), less than 6.5% (blood sugar of 140 mg/dL), or less than 7.5% (blood sugar of 154 mg/dL).

The researchers found that rates of overtreatment were 10.1% for HbA_1c less than 6.0%, 25.2% for HbA_1c less than 6.5%, and 44.3% for HbA_1c less than 7.0%. However, no information on actual rates of hypoglycemia or the complications thereof (such as emergency department visits or hospitalizations) was collected.

Author Patrick O’Malley, MD, MPH, discusses his views on the implications of this study with news@JAMA.

news@JAMA: The study showed that 44% of VA patients with diabetes and risk factors for serious hypoglycemia may be getting overtreated to some degree. In your article, you state that these results are “sobering.” Why do you think overtreatment happens so often?

Dr O’Malley: In general, overtreatment is a multifactorial issue related to the culture of care. In medicine, physicians always want to do something, to take something, such as a blood sugar value, from abnormal to normal. We rely on clinical evidence for these treatment goals, but sometimes, in the absence of good clinical evidence, we must instead rely on biologic plausibility, and extrapolate data from one population to another. This extrapolation often happens with the elderly, who are not included as participants in many clinical trials on which many general clinical guidelines and recommendations are based. That’s why overtreatment tends to happen in the elderly, not just for diabetes, but for other diseases as well.

Another issue is the use of quality metrics, or A_1c in the case of diabetes, with the notion of “lower is better” driving this culture of aggressive care. The VA system in particular is very good at driving such quality metrics to success, which probably leads to higher rates of overtreatment. It’s basically a case of good intentions and good systems gone awry.

news@JAMA: It seems that since elderly individuals are underrepresented in several of the major diabetes trials, evidence on the benefits vs harms of intensive glucose control might be difficult to interpret, and the ideal A_1c level for elderly people is still a mystery. Is there a bottom line that can be drawn from these studies?

Dr O’Malley: Yes, the bottom line is that for anyone who has a life expectancy greater than 15 years, an A_1c goal of less than 7 is reasonable. For all other people, which includes elderly individuals as well as younger individuals with multiple comorbidities and lower life expectancies, an A_1c goal of less than 7 might be more harmful than beneficial. Providers should be tailoring A_1c goals to the individual, and life expectancy should be a major part of that equation.

news@JAMA: For individuals with a life expectancy of less than 15 years, what is a reasonable A_1c goal? Is there an upper limit?

Dr O’Malley: There’s not a whole lot of great evidence to support a specific upper limit, but in general, a blood glucose consistently over 200 is where people start running into problems with immune system dysfunction and increased risk of infections. This corresponds to an A_1c of about 9.

news@JAMA: What can doctors and hospital systems do to address the issue of overtreatment?

Dr O’Malley: On a systems level, we have to be smarter about choosing metrics, and smarter about individualizing metrics to individual patients—older patients should have different metrics than younger ones. On an individual physician level, we should strive to be less pharmacologically oriented, especially in the elderly. There is a lot that therapeutic lifestyle changes can achieve. We also tend to be overly aggressive with monitoring, ordering A_1c tests too frequently, and telling our patients to check blood sugars four times a day. Checking blood sugars three times a week is plenty for most stable patients with diabetes.

news@JAMA: What about for patients who might be skeptical or resistant toward the idea of withdrawing or decreasing the intensity of care?

Dr O’Malley: It’s true that this might be a major barrier, as patients are sometimes reluctant to do this, or confused as to why the same doctor who told them to intensify their diabetes control is now telling them to scale back. This is why it’s so important to have a trusting and invested doctor-patient relationship.

Sometimes as doctors, in an effort to stay as up to date and proactive as possible, we change our culture toward treatment before the evidence is truly there, and this ends up harming patients, wasting resources, and engendering distrust among patients toward their physicians and the healthcare system as a whole. Another example of this is what happened with hormone replacement therapy for menopausal women back in the ’90s—we were so quick to act on the benefits of hormone replacement before the long-term evidence was available and then realized it was actually more harmful than beneficial.