CHICAGO – Glycemic overtreatment of high-risk diabetic patients is rampant within the Veterans Affairs health care system, according to a national study.
Moreover, because many of these diabetic veterans who are at high risk for serious hypoglycemia are also Medicare eligible, it’s quite likely that overtreatment is a common problem in the Medicare population as well.
"I think these findings are directly relevant to Medicare," Dr. Leonard M. Pogach said in presenting the VA study results at the annual scientific sessions of the American Diabetes Association.
Current performance measures do not assess potential overtreatment of high-risk diabetic patients in either the VA or Medicare populations. But as a result of the VA study findings, joint federal efforts are underway to address this shortcoming, according to Dr. Pogach, who is national program director for endocrinology and diabetes at the Veterans Health Administration and professor of medicine at the New Jersey Medical School, Newark.
He cited as a major impetus for the VA study an eye-opening 2011 report by investigators at the Centers for Disease Control and Prevention that identified insulin and sulfonylurea drugs as the combined number-two cause of emergency hospitalizations for adverse drug events in the United States, second only to warfarin (N. Engl. J. Med. 2011;365:2002-12).
As part of the Choosing Wisely campaign, the American Geriatrics Society recommends that medications other than metformin not be routinely used to lower hemoglobin A1c below 7.5% in patients older than 65 years. The group further recommends an HbA1c target of 8%-9% for those with serious comorbid conditions.
Similarly, the ADA now recommends an HbA1c target of 7.5%-8.0% or slightly more in patients at increased risk for serious hypoglycemia or with reduced life expectancy, rather than its former universal goal of less than 7.0%.
For purposes of the VA study, Dr. Pogach and coinvestigators defined "high-risk" diabetes patients as those receiving insulin and/or sulfonylurea therapy and who are at least 70 years of age, have renal impairment as reflected in a serum creatinine level greater than 1.7 mg/dL, or have been diagnosed with cognitive impairment or dementia.
During the study year of 2009, a total of 285,476 of 652,738 VA patients, or 44%,with diabetes on insulin and/or sulfonylurea therapy qualified as high risk, based upon that definition. They received treatment in 139 VA facilities in 21 regions nationally.
Overall, 48% of these high-risk diabetes patients were likely being overtreated, as reflected in an HbA1c below 7.0%. Moreover, one-quarter of the high-risk group had an HbA1c below 6.5%, and 10% had an HbA1c of less than 6.0%, Dr. Pogach reported.
He added that these results likely underestimate the true extent of the glycemic overtreatment problem, because he and his coinvestigators defined "high risk" quite conservatively – based simply upon advanced age, renal dysfunction, and cognitive impairment.
Had they included other reasonable criteria – specifically, diminished life expectancy; stroke and other major neurologic disorders; cardiovascular disease; major depression; alcohol and/or drug abuse; and advanced diabetic complications – the proportion of the 652,738 diabetic VA patients on insulin and/or sulfonylurea therapy who would have qualified as being at high risk for serious hypoglycemia would have climbed from 44% all the way up to 71%.
The variation in overtreatment rates from region to region within the VA system was fairly tight. The range was much greater among facilities within a given regional district, where overtreatment rates varied from a low of 37% at a select few facilities to as high as 63%.
"We were able to identify several facilities with statistically remarkably lower rates of overtreatment than in the larger regions they lie in," Dr. Pogach noted. "Those are the sites where we might want to do site visits and qualitative studies to try to figure out what made them different. What is it about their culture or their patients? What happened in those places that we might subsequently want to replicate?"
Of note, the VA never adopted the one-size-fits-all goal of an HbA1c below 7.0% that the ADA recommended until recently. For more than a decade, VA clinical practice guidelines have included stratified glycemic targets based on comorbidities and life expectancy.
One audience member took issue with Dr. Pogach’s use of the word "overtreatment."
"Overtreatment is really a somewhat pejorative term," he argued. "I think you really have to have data showing that we are harming these people, not in terms of the surrogate outcome of hypoglycemia, but in real negative consequences."
Dr. Pogach was quick to rebut.
"I think the published results of the ACCORD and ADVANCE studies demonstrate that there’s a very strong association between cardiovascular morbidity and mortality and self-reported hypoglycemia, with adjusted odds ratios of about two and one-half," he explained. "I don’t think those data will ever be beat. We’re never going to have a randomized trial. So, I think the association with adverse outcomes is very clear, although we don’t know if hypoglycemia is the proximate cause or not."