Hypoglycemia Exceeds Hyperglycemia Admissions for Elderly

Miriam E. Tucker

July 01, 2013

CHICAGO — Among Medicare beneficiaries with diabetes, hospital admissions for hypoglycemia now outpace those for hyperglycemia, a new claims data analysis shows.

The findings were presented last week at the American Diabetes Association (ADA) Scientific Sessions by Kasia Lipska, MD, from the department of internal medicine, section of endocrinology, at Yale University School of Medicine, New Haven, Connecticut.

As a result of increased efforts to intensify glycemic control over the past decade, the proportion of older persons meeting the HbA1c goal of less than 7% has improved, but "there's been a flip, with hypoglycemia now clearly exceeding hyperglycemia admission rates," Dr. Lipska said in her presentation.

"We've reduced glucose levels over time, including among older adults. But there are some data that suggest that very intensive glucose control may increase mortality and is also associated with increased hypoglycemia… So, there are benefits but also some suggested risk."

Session moderator Hermes Florez, MD, PhD, MPH, interim chief of the division of gerontology and geriatric medicine at the University of Miami Miller School of Medicine, Florida, told Medscape Medical News that the data point to "the need to increase awareness and individualize therapy for diabetes management, particularly in the elderly," including evaluation of risk factors for hypoglycemia such as depression, cognitive decline, reduced appetite, and drug interactions.

Dr. Florez said that Dr. Lipska's data are consistent with issues discussed in the ADA's recent consensus statement on diabetes in older adults, for which he served on the writing panel.

Dr. Lipska told Medscape Medical News that the findings also call into question the use of HbA1c as a measure for judging the performance of physicians and healthcare systems, since it tends to promote a "one-size-fits all" treatment that may lead to hypoglycemia.

"[HbA1c] has an important role in the clinical care of patients with diabetes. I just think that on a system level, when you drive people to specifically meet that target, it has the potential to do harm. It doesn't take into account that you have to individualize, or how you lower the blood glucose… I think it's too narrow."

"Flip" Seen Between 1999 and 2011

Data for nearly 34 million Medicare fee-for-service beneficiaries aged 65 years and older seen during 1999 and 2011 were analyzed, with more than 350 million patient-years of observation. A secondary analysis included information on diabetes prevalence from the Behavioral Risk Factor Surveillance System (BRFSS).

In all, there were 302,095 hospitalizations for hyperglycemia, which included uncontrolled diabetes, diabetic ketoacidosis, and hyperosmolarity. Hypoglycemia admissions totaled 429,850, with specific diagnoses including hypoglycemic coma or "shock" and "poisoning by insulin or antidiabetic agents."

From 1999 to 2011, the rate per 100,000 total patient-years of admissions for hyperglycemia dropped by 39%, from 114 to 70. Such hospitalizations were much more frequent among the older elderly (aged 75 and older) in 1999, but by 2011 the rates by age had equalized.

Distinct disparities were seen by race, with nearly 4-fold higher rates of admission for hyperglycemia among black patients, at all time points.

For hypoglycemia, the overall rate per 100,000 patient-years rose from 94 in 1999 to a peak of 130 in 2007, then dropped slightly to 105 in 2011. In 2011, admission rates among beneficiaries aged 75 and older were approximately double those of younger beneficiaries aged 65 to 74.

Dr. Florez said that these data are consistent with his experience, having received consults for hypoglycemia among elderly patients with diabetes who are taking sulfonylureas — particularly the older-generation glyburide — or who are on "sliding-scale" insulin regimens. "These are treatment modalities that increase the risk of hypoglycemia and therefore should be avoided," he told Medscape Medical News.

Large racial disparities were seen with hypoglycemia as well, again with nearly quadrupled admission rates among blacks compared with whites.

The racial-disparity finding is consistent with previous data showing that minority populations may be at higher risk for poor glycemic control compared with white non-Hispanics, Dr. Florez noted.

And Dr. Lipska told Medscape Medical News that it likely relates to poorer quality and/or access to care in black compared with white patients, "but we need to do more to understand these disparities and improve care."

Because the Medicare claims database did not include data on diabetes prevalence, the results were plotted against the BRFSS age-based data to show the adjusted rate per 100,000 patient-years with diabetes. Here, the "flip" is clearly seen: In 1999, admissions for hyperglycemia and hypoglycemia were 820 and 676, respectively. In 2011, those rates were 367 and 612, respectively.

Reasons and Implications

In addition to the increased intensity of glycemic-control efforts in recent years, other possible reasons for the shift might include changes in the use of glucose-lowering medications, as well as differences in criteria used for hospital admission, including increased use of "observation stays" of less than 24 hours that are not counted as admissions, Dr. Lipska said

Indeed, because admission for both hyper- and hypoglycemia is relatively rare, these data were capturing the most severe cases, she told Medscape Medical News.

Clearly, she said in her presentation, these findings suggest the need to focus increased attention on reducing hypoglycemia among black patients and the older elderly.

More broadly, the data support the ADA's new focus on individualization of treatment goals, spelled out in its April 2012 position statement, she said.

In that document, the authors state, "Importantly, utilizing the percentage of diabetic patients who are achieving an HbA1c < 7.0% as a quality indicator, as promulgated by various healthcare organizations, is inconsistent with the emphasis on individualization of treatment goals."

Subsequent to that, the Centers for Medicare and Medicaid Services issued quality performance standards for accountable-care organizations , in which a goal of less than 8% is given as the performance measure, with explanatory text discussing the need for individualization.

Dr. Lipska believes that, while more "reasonable" than a target of less than 7%, the metric is still too narrow.

"The proportion of people that meet a performance measure like A1c doesn't capture the clinical impact on the population… A narrow measure like this will guide therapy one way, but I think we really need to look more broadly at how patients actually do and whether they have more adverse effects from the treatment strategies we're using."

Neither Dr. Lipska nor Dr. Florez has reported any relevant financial relationships . Dr. Lipska blogs for Medscape Medical News about diabetes meetings

Scientific Sessions of the American Diabetes Association. Abstract 274-OR, presented June 24, 2014.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.