Emory Report homepage  

March 27 , 2006
Study: Directed intervention can improve diabetes care

BY Janet Christenbury

With the rate of diabetes threatening to approach epidemic proportions, Emory researchers have found that management of diabetic patients in a primary care setting can be improved by an intervention aimed at physicians.

Study organizers individualized the interventions to physicians’ record of action when patient glucose levels were high, and featured regular feedback on their performance. Such feedback seemed to improve physicians’ use of therapy and led to lower glucose levels in their patients (glucose levels are closely linked to the likelihood of developing diabetes complications such as kidney failure or blindness).

The results of the three-year study were published in the March 13 issue of Archives of Internal Medicine.

“This study reaffirms the importance of correcting what we call ‘clinical inertia’—the failure of health care providers to intensify therapy appropriately when clinically indicated,” said Lawrence Phillips, professor of medicine in the Division of Endocrinology and senior author of the paper. “We found that feedback on performance given to primary care physicians decreased their clinical inertia, and as a result, patients’ glucose levels improved. The patients became healthier.”

The research was made possible by a unique health care partnership of generalists and specialists, aimed at improving care for patients not seen by specialists. The partnership differs from typical consultations where specialists see patients directly. Researchers focused on the primary care setting because most patients with diabetes are managed by generalists in such settings.

In the study, 345 primary care practitioners at Emory (internal medicine residents, or medical doctors still in training) were randomized either to be controls (no intervention) or to receive interventions aimed at their behavior—either computerized reminders with patient-specific recommendations for changes in therapy; physician-specific feedback on their performance given by endocrinologists in five-minute, face-to-face meetings every two weeks; or both reminders and performance feedback.

When patients’ glucose (blood sugar) levels were high during health care visits, physician behavior was categorized as “did nothing,” “did anything” (any intensification of therapy) or “did enough” (if intensification met recommendations). More than 4,000 patients participated in the study.

At baseline, physicians “did anything” for 35 percent of visits and “did enough” for 21 percent. Intervention intensity increased more in the two groups receiving feedback on performance than for the other two groups. After three years, physician behavior in the reminders and control groups returned to baseline, whereas improvement with feedback alone or feedback plus reminders was sustained.

Analysis showed that feedback on performance contributed independently to the likelihood that a physician would intensify therapy when clinically indicated, and that intensification contributed independently to improved glucose levels.

“To improve diabetes outcomes—complications, death and cost—it’s critical that we find better ways to manage the disease,” Phillips said. “This study showed the importance of recognizing the problem of clinical inertia, and developing interventions, which can help overcome this problem. The feedback on performance intervention helped providers to intensify therapy more frequently when clinically indicated, to intensify therapy enough to make a clinical difference, and to bring patients with high glucose levels back for early return visits.

“The study worked,” he said, “because it succeeded in three key dimensions: identifying the problem, using an intervention that targets the problem, and focusing on specific, important behaviors to improve.”

Diabetes is the sixth leading cause of death in the United States, and the major cause of kidney failure, blindness and nontraumatic leg amputation in adults. The disease is also a major contributor to U.S. health care costs: In 2002, Phillips said, diabetes accounted for one out of every nine health care dollars and about 30 percent of Medicare costs.

TOP