It is time to eliminate unnecessary iatrogenic hypoglycemia in older adults

Over the last several years, the Journal of the American Geriatrics Society (JAGS) and other journals have published data on the astoundingly high rates of overtreatment of diabetes in older people in general, and among those in nursing homes, with multiple serious comorbidities, and at the end-of-life.^1-8 In a previous editorial, I labeled this “low hanging ‘fruit’”⁹ and have since in Editor's Notes, lectures, and panel discussions tried to cajole JAGS readers and my colleagues to implement interventions to reduce this unnecessary and preventable practice. I am, in fact, on a “campaign” to eliminate iatrogenic hypoglycemia and actively working with the Centers for Medicare and Medicaid Services to implement strategies that will incentivize clinicians to reduce the tens of thousands of episodes of overtreatment and resultant severe iatrogenic hypoglycemia that occur annually in the United States.

In that context, I applaud Cohen and her colleagues on their report in this issue of JAGS describing a quality improvement initiative carried out in five hospitals in a health system in northern New Jersey.¹⁰ The intervention was not complicated, but it was multifaceted and interprofessional and depended substantially on electronic medical record alerts and a “Nursing Reference Sheet” containing simple yet critical guidance on insulin treatment. They analyzed data over the 3-year period from 2021 to 2023 involving 123,393 patients, 753,902 patient days, and 2,049,052 blood glucose results. All five hospitals demonstrated statistically significant reductions in severe hypoglycemic events (blood glucose <40 mg/dL) in calendar year 2023 compared with 2021. Extrapolating data from Table 2, in 2021, 5429 of 627,526 (0.8%) of blood glucose levels were <40 mg/dL, compared with 4365 of 728,714 (0.6%) in 2023—an absolute reduction of 1064 episodes of hypoglycemia. This represents a relative reduction of 25% in the percentage of blood glucoses and a 20% reduction in the number of blood glucoses in the severe hypoglycemic range. These results must be interpreted cautiously because they come from one health system, the study was not controlled, and propensity matching was not used to compare the patients involved in 2021 vs. 2023.

Irrespective of the limitations in the project design, think about the potential toll on brain health and quality of life that results from unnecessary and preventable hypoglycemia in United States in older hospitalized adults every year. There are currently 6120 hospitals in the United States. If 1064 episodes could be prevented in every five hospitals, that would amount to preventing 6120/5 = 1225 X 1064—a staggering 1.3 million episodes of severe hypoglycemia in hospitals alone. This does not include hypoglycemia of lesser magnitude that could have serious consequences and many thousands if not millions of episodes of severe hypoglycemia that result in Emergency Department visits without admission and episodes that occur in nursing homes.

Reducing unnecessary iatrogenic hypoglycemia should not be complicated. But, it will take focus and hard work in order to achieve this goal. Excellent guidance has been provided by several organizations on the management of diabetes in older adults.^11-14 Insulin is in fact not a first-line treatment for diabetes in this population and should only be prescribed when diet, exercise, and a non-sulfonylurea agent have failed to adequately control blood sugar in older people whose life expectancy exceeds the time to benefit from tight glucose control.^{15, 16}

We need more studies like the one carried out by Cohen and her colleagues with rigorous quality improvement methodology deploying person-centered interventions in larger, more diverse, and propensity-matched cohorts in different health care settings. Such studies may once and for all provide the evidence and impetus to cut down this low hanging fruit.

The author is solely responsible for the content of this editorial.

The author declares no financial conflicts of interest.

None.