No End in Sight
It has been 16 years since the results of the landmark Diabetes Control and Complications Trial (DCCT) were published. Despite its continuing legacy of proof that maintaining blood glucose levels as close to normal as possible reduces the risk of diabetes complications, today less than half of people with diabetes are reaching target blood glucose levels, according to most estimates. Health-care providers and researchers continue to struggle with how to improve those numbers.
Recalling the DCCT
Researchers began recruiting participants for the DCCT in 1983. They signed up 1,441 people with Type 1 diabetes, roughly half within five years of their diagnoses and the rest within 15 years of diagnosis. The subjects were randomly assigned to either conventional insulin treatment or intensive treatment. Those on intensive treatment took three or more insulin injections each day or used an insulin pump and monitored their blood glucose levels three or four times a day. The goal for the intensive treatment group was to keep their glycosolated hemoglobin, or HbA1c levels (a measure of blood glucose control over two to three months) at or below the top of the normal range: 6.05%. In actuality they achieved a median HbA1c of 7.2%. The study’s subjects were followed for six and a half years, and the publication of the results in late 1993 changed the standards of Type 1 diabetes management forever.
The research team had hoped to be able to demonstrate a 30% to 40% reduction in complications among those on intensive treatment. What they actually saw was so significant that they ended the study early and advised the subjects on conventional treatment to switch to intensive treatment.
Those who were within five years of diagnosis and had been assigned to intensive treatment experienced a 76% reduction in the risk of diabetic retinopathy, or eye disease, a 34% reduction in the risk of microalbuminuria (an early stage of diabetic nephropathy, or kidney disease), and a 69% reduction in the risk of diabetic neuropathy, or nerve disease. Results for those within 15 years of diagnosis and on intensive treatment were less dramatic but still significant: a risk reduction of 54% for retinopathy, 43% for microalbuminuria, and 57% for neuropathy. In both groups there was a 41% reduction in the risk of cardiovascular disease.
The drawbacks for the intensive treatment group were a three-times higher incidence of severe hypoglycemia (very low blood glucose levels), and an average weight gain of 4.6 kilograms (about 10 pounds).
The study also answered the question of whether or not there is a point below which a lower HbA1c does not yield additional benefit. The answer was that there is no “glycemic threshold”: The closer to normal the better.
The DCCT ended in 1993, but more than 1,370 of its subjects were then enrolled in a long-term follow-up study called Epidemiology of Diabetes Interventions and Complications (EDIC). The news from EDIC in the past few years has been almost too good to be true: The benefit of intensive blood glucose management continues even after treatment becomes less monitored and less rigorous.
A report from the EDIC group in 2002 found that the people from the intensive treatment arm of the DCCT still had a 62% reduction in the risk of retinopathy seven years after the end of the DCCT. Long-term results for nephropathy were published in 2003, covering eight years of follow-up, and members of the intensive treatment group still enjoyed a 59% reduction in the risk of microalbuminuria. More striking still, that group saw an 84% reduction in the risk of developing clinical albuminuria, a more advanced stage of kidney disease. The latest data for neuropathy were published in 2006, and showed that eight years after the DCCT ended, the participants who had been on intensive treatment experienced a 64% reduction in the risk of neuropathy.
The DCCT studied Type 1 diabetes, but the vast majority of cases of diabetes are Type 2. However, the value of intensive therapy was demonstrated for Type 2 diabetes by the United Kingdom Prospective Diabetes Study (UKPDS), published in 1998. The UKPDS covered a 20-year period and included more than 5,100 people with Type 2 diabetes in England, Northern Ireland, and Scotland.
The UKPDS showed that intensive treatment of Type 2 diabetes, using sulfonylurea drugs (oral medicines that increase the secretion of insulin by the pancreas) or insulin, reduced microvascular complications (primarily retinopathy and nephropathy) by 25%. Another group of people in the study who were overweight and received intensive treatment with metformin (an oral medicine that suppresses glucose production by the liver and enhances the body’s sensitivity to insulin) saw a 32% decrease in diabetes complications.
Where are we now?
Despite the overwhelming evidence of the importance of “tight” blood glucose control in people with Type 1 and Type 2 diabetes, it is still far from the norm. In a study published in the January/February 2006 issue of Annals of Family Medicine, Dr. Stephen J. Spann and his colleagues reported on a review of charts of people with Type 2 diabetes from four large primary-care-based research networks. They found that only 40.5% of these people had HbA1c values below 7%. Moreover, 31.3% of HbA1c values were above 8%.
These findings echo those of the National Health and Nutrition Examination Survey (NHANES) from 1999–2000, in which only 37% of participants were found to have achieved the target goal of an HbA1c level less than 7%. Another 37.2% of the participants had an HbA1c above 8%. It is notable that these percentages did not change significantly from the earlier NHANES covering 1988–1994.
Not even a majority of people who get their diabetes care in specialized clinics at academic medical centers are reaching target HbA1c values. A 2002 review from 30 academic medical centers in the United States found that while nearly all patients with Type 1 or Type 2 diabetes had received an HbA1c test within the previous year, only 34% were at 7% or less.
“The reason we did our study,” said Dr. Spann, who is Chairman of Family and Community Medicine at Baylor College of Medicine in Houston, Texas, “is that while we understand that improved glycemic control is better, we struggle with how to make it happen in primary care. We wondered what we could learn about elements that might predict good glycemic control.”
“We didn’t really find anything that was absolutely predictive of good control,” he said. “We did learn that the level of complexity rises as you start adding the other important treatment targets, such as blood pressure and cholesterol. It’s hard enough to hit one target, and even harder to achieve targets simultaneously for multiple risk factors.”
Barriers to tight control
There are many reasons tight blood glucose control is not more widespread, despite its clear benefits. “Some people have access problems,” Dr. Spann says, “such as insurance that doesn’t cover extensive treatment. It also is difficult to get many people to a point where they’re willing and able to make lifestyle changes to improve their control.”
Since diabetes is a progressive condition, it can become increasingly difficult to achieve treatment targets. Several studies have noted that the percentage of people achieving targets diminishes with duration of diabetes.
What’s more, the “system” of medical care may be a problem in itself, according to Dr. Spann and others. The way primary care is organized and reimbursed by insurance companies probably serves as a barrier to achieving optimal blood glucose control for many people. Dr. Spann points to the “Chronic Care Model” developed by Dr. Ed H. Wagner at Improving Chronic Illness Care (ICIC), a national program of the Robert Wood Johnson Foundation based in Seattle, as a way to change the system.
“I think that the proponents of the Chronic Care Model are right on,” Dr. Spann says. “There have been a number of studies that have shown that practices that have most of the elements of the model in place do better in getting more of their patients to treatment targets.” Those elements include electronic registries that track and report important patient information, multidisciplinary treatment teams, and care management. “The doctor can’t do it all,” Dr. Spann says. “We know that nurses and diabetes educators probably do better health education than we do.”
One health system that has adopted the Chronic Care Model and applied it to diabetes is MaineHealth in Portland, Maine. In conjunction with its physician–hospital organization, the MMC PHO, MaineHealth has developed an Internet-based Clinical Improvement Registry for primary-care doctors. The doctors enter information about their patients, and the system provides current clinical data at each visit, including the latest laboratory results, status of key tests (for example, HbA1c and blood pressure), and problems to be addressed. The system can also be used to generate notices for patients and doctors, to remind them of overdue examinations or laboratory tests. Data may be viewed for each patient or for a doctor’s overall practice.
The MMC PHO also employs 14 Chronic Illness Care Managers, who are embedded in primary-care practices as part of the patient care team (they manage not just diabetes, but asthma and heart failure). These nurse specialists provide patients with intensive education and motivational support, even paying home visits if that will help.
The most recent data from the program covers the experience of 15 primary-care practices over a period from the program’s inception to the end of its first year. Before the program existed, 80% of people with diabetes had received an HbA1c test within the past year. After a year, 93% of people had received one. The percentage of people with HbA1c values less than 7% rose from 41% to 49%—a 20% increase. The percentage of people with HbA1c values above 8% decreased from 31% to 24%, and the percentage of people with HbA1c values above 9.5% decreased from 13% to 9%. There were similar results in measures of LDL (or “bad”) cholesterol and blood pressure.
“This is not a question of bad doctors or bad patients,” says Larry Anderson, M.D., Senior Director for Quality Improvement and Medical Affairs at the MMC PHO. “It is a question of a care model that is focused on illness instead of prevention, and systems that have been created that don’t accommodate a change in focus. We’re changing the focus, including offering financial incentives for physicians whose patients do better.”
What does it mean to you?
The HbA1c test is considered the gold standard for the evaluation of blood glucose control. While the test has been in use for roughly 30 years, widely accepted target levels are a relatively recent development.
Before the DCCT, the American Diabetes Association’s (ADA) Standards of Medical Care (which is considered the primary source of recommendations for effective treatment), did not recommend a target HbA1c level. In response to the DCCT, the ADA in 1995 set a target of less than 7%, and set 8% as the “action” limit, above which additional intervention should be undertaken. The action limit was removed in the 2003 recommendations, effectively indicating that any level above 7% was cause for action.
The trend toward recommending lower blood glucose levels continued in 2004, when the ADA stated that “more stringent” targets could be considered for individuals. In 2006, that position was clarified further to indicate that while the HbA1c goal for people in general is less than 7%.
The ADA recommends that people meeting their treatment goals have an HbA1c test at least twice a year, and that those whose therapy has changed or who are not meeting their goals be tested four times a year.
It’s important not just to have the test, but to understand and be familiar with the results. A pair of recent studies found that only about a quarter of people with diabetes knew their most recent HbA1c value.
There is a growing recognition that the approach to diabetes management needs an overhaul to reflect the condition’s many challenges. An international task force called the Global Partnership for Effective Diabetes Management recently published a series of recommendations for Type 2 diabetes that includes aggressive interventions for any HbA1c level that stays above 6.5% for a six-month period. Combinations of oral diabetes drugs and the addition of insulin should be prescribed sooner, according to the group. Moreover, the group recommends pursuing cholesterol and blood pressure targets with equal aggressiveness, and implementing a multidisciplinary team approach to help people take control of their diabetes.
Meanwhile, the good news about tight blood glucose control keeps coming in. A DCCT/EDIC study published in December 2005 demonstrated that the risk of cardiovascular events was reduced by 42% in people who had been in the intensive treatment arm of the DCCT—even a dozen years after the study ended. The findings extended earlier observations that the intensive control group had slower progression of carotid intima–media thickness and coronary artery calcification, both signs of developing atherosclerosis (hardening of the arteries).
“The bottom line is that this is very difficult,” Dr. Spann says. “As our study and others have shown, even with good practices with good physicians, and even with motivated patients, it’s so hard that only 40% of people hit the targets. But there’s really solid evidence that hitting those targets prevents blindness, amputations, and kidney disease. That’s a significant payoff—and a reason to keep trying.”