|The findings were published online December 17 in Diabetes Care by Ram D Pathak, MD, an endocrinologist at the Marshfield Clinic, Marshfield, Wisconsin and the University of Wisconsin School of Medicine and Public Health, Madison, and colleagues.
Using electronic health records for an insured cohort of nearly one million adults with diabetes seen during 20052011, the majority of which was type 2 diabetes, the authors found rates of severe hypoglycemia requiring medical intervention of 1.4 to 1.6 per 100 person-years.
This is in contrast to the 0.3 to 1.0 rates seen in the standard-care arms of randomized clinical trials such as Action to Control Cardiovascular Risk in Diabetes (ACCORD), Action in Diabetes and Vascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE), and Outcome Reduction With Initial Glargine Intervention (ORIGIN).
Rates of severe hypoglycemia were particularly high among older patients and those with chronic kidney disease, congestive heart failure, cardiovascular disease, depression, and higher HbA1c levels. Severe hypoglycemia was also more common among patients taking insulin, insulin secretagogues, and beta-blockers.
"Strategies that reduce the risk of hypoglycemia in high-risk patients are needed. These data can inform the development of clinical management strategies to more effectively reduce the occurrence of severe hypoglycemia in community-treated patients," Dr Pathak and colleagues write.
Asked to comment, Kasia J Lipska, MD, an endocrinologist at Yale University, New Haven, Connecticut, who has studied hypoglycemia in the elderly, told Medscape Medical News, "The rates of severe hypoglycemia are usually much lower in randomized clinical trials than in real-world populations….That's most likely because patients recruited for clinical trials are typically healthier and less likely to experience hypoglycemia compared with the complex patients we see in clinical practice."
She added, "I worry that clinicians underappreciate the risk of hypoglycemia among people with type 2 diabetes and the impact it has on people's lives."
Severe Hypoglycemia Rises With Comorbidities
The investigators used 20052011 figures for 917,440 adult patients with diabetes from the Surveillance, Prevention, and Management of Diabetes Mellitus (SUPREME-DM) DataLink, the second-largest insured diabetes patient cohort with data from electronic health records in the United States (after the Veterans Administration).
The database doesn't distinguish between type 1 and type 2 diabetes, with the latter presumed to account for about 95% of the total. The prevalence of comorbidities ranged from 4.9% with congestive heart failure to 18.7% for cardiovascular disease.
Severe hypoglycemia was defined by primary or secondary hypoglycemia-related ICD-9 codes from an emergency-department visit or inpatient stay. Annual rates of severe hypoglycemia ranged from 1.59 per 100 person-years in 2006 to 1.37/100 person-years in 2010, with no significant trend by year and no difference by sex (P = .15). Severe hypoglycemia was most common in both the younger and older age groups, ranging from 1.06 to 1.35/100 person-years among those aged 20 to 44 and from 2.34 to 2.90/100 person-years among those aged 75 and older.
Rates in the youngest age group (2044) were higher than those among the 45- to 64-year-olds (0.931.02), likely reflecting a higher prevalence of type 1 diabetes in this younger group, the authors note.
Individuals with chronic kidney disease, congestive heart failure, and/or cardiovascular disease had between four and eight times greater rates of severe hypoglycemia than did those with diabetes but without those comorbidities (P<.001 for all).
In addition, patients with depression at study entry had an approximately 50% greater risk for severe hypoglycemia compared with those who were not depressed (P < .001).
Patients with two or more episodes of severe hypoglycemia per year tended to have more comorbidities than did those with one or fewer episodes of severe hypoglycemia.
Higher HbA1c, Beta-blocker Use Also Associated With Hypoglycemia
Patients with higher HbA1c levels also had higher rates of severe hypoglycemia (P < .01), ranging from 0.86 to 1.20 per 100 person-years with HbA1c levels below 7% to 2.44 to 2.77 per 100 person-years for HbA1c of 9% or higher.
This may seem surprising, but there are several possible explanations, Dr Lipska noted.
For one, the HbA1c value preceded the hypoglycemia, which may have resulted from treatment intensification. Second, glycemic variability may be higher among those in poor control, predisposing them to extremes in both directions. Thirdly, intermittent compliance with glucose-lowering therapy could contribute to both poor control and hypoglycemia.
"Looking at subsequent trajectories of glucose control may provide more information going forward," she said.
Not surprisingly, patients filling prescriptions for insulin had the highest rates of severe hypoglycemia, 10- to 12-fold greater than those on nonsecretagogue glucose-lowering drugs and three to five times higher than those on secretagogues (eg, sulfonylureas).
Beta-blocker usage was also significantly associated with severe hypoglycemia, with rates of 1.65 to 3.29/100 person-years vs 1.23 to 1.45/100 person-years for nonusers (P < .001).
Beta-blockers can suppress the symptoms of hypoglycemia, but a previous analysis found no increased risk of hypoglycemia with cardioselective beta-blocker usage. Beta-blocker type was not assessed in the current study, but confounding by comorbidities is likely, Dr Pathak and colleagues note.
"Treating physicians may need to take beta-blocker therapy into consideration when selecting glucose-lowering therapy and the HbA1c target. For many patients with diabetes and CVD, hypoglycemia concerns related to beta-blockers may be outweighed by their cardiovascular benefits," they write.
|Better Surveillance Systems Needed for Hypoglycemia
Ascertaining hypoglycemia using claims or EHR data is very difficult, Dr Lipska pointed out, adding, "This is something that I think needs urgent attention. Hypoglycemia is common, dangerous, and potentially avoidable, but we have no systems of surveillance for hypoglycemia."
She praised the investigators for "a great job compiling multiple sources of data and using a list of potential ICD-9 codes for hypoglycemia," but she noted that information about events occurring outside the hospital system is still missing.
And she said that by capturing hypoglycemic events that occurred during hospital stays, "we can't easily attribute them to ambulatory management of diabetes. Interventions to reduce inpatient hypoglycemia will need to differ from those targeting hypoglycemia in the outpatient setting, so looking at these events separately might be helpful as well."
This project was supported by the Agency for Healthcare Research and Quality. Dr Pathak has no relevant financial relationships; disclosures for the coauthors are listed in the article. Dr Lipska has no relevant financial relationships.