COLUMBUS, OHIO — Real-time continuous glucose monitoring is likely to become the standard of care for the treatment of type 1 diabetes within the next 5–10 years, but for now it's best to reserve the technology for selected patients, Dr. Irl B. Hirsch said at a meeting on diabetes sponsored by Ohio State University.
Real-time continuous glucose monitors (RT-CGMs) have been on the market only since 2006 and are not yet indicated to replace fingerstick glucose testing. Currently, the main benefits of RT-CGMs are their capability for detecting glucose trends throughout the day and night, and their ability to alert the patient when glucose levels become too high or too low.
The potential for decreasing glucose variability—both the danger of severe hypoglycemia and the damaging oxidative stress that accompanies postprandial glucose surges—is what makes the technology so promising for the future and also valuable for some patients even now in its very early stages, said Dr. Hirsch, professor of medicine and medical director of the diabetes care center at the University of Washington, Seattle.
Still, until the technology and experience with RT-CGMs improve—along with the insurance coverage—they are not for everyone, according to Dr. Hirsch. This was demonstrated in the STAR I (Successful Teens Achieving Readiness for Independence I) study, in which 138 adolescents and adults with poorly controlled type 1 diabetes (hemoglobin A1c of 7.5% or greater) despite 6 months or more of insulin pump therapy were randomized to one of two groups: One group wore the combined pump/RT-CGM device (the MiniMed Paradigm 722 system) and performed self-monitoring of blood glucose (SMBG) four or more times per day, and the other group performed SMBG while wearing the pump by itself.
At 13 weeks, mean HbA1c levels had dropped significantly and to a nearly identical degree in both groups, from 8.4% to 7.8% in the controls and from 8.5% to 7.7% in the RT-CGM group. There were no further significant drops in either group, and by week 26, both groups had a mean HbA1c of 7.8%, Dr. Hirsch first reported last summer at the American Diabetes Association's annual scientific sessions (FAMILY PRACTICE NEWS, Sept. 1, 2007, p. 15).
Although the findings might seem negative, further analysis revealed that compliance strongly predicted the results among the RT-CGM patients. With “compliance” defined as wearing the sensor 6 days per week (meaning it was possible to be more than 100% compliant) HbA1c levels among the patients with 100% compliance or greater dropped from 8.6% at baseline to 7.7% at 26 weeks. Those with 80%–100% compliance dropped similarly (from 8.4% to 7.7%), as did those with 60%–80% compliance (from 8.2% to 7.5%).
All of those reductions were significant. However, when compliance dropped below 60%, mean HbA1c actually rose slightly, but not significantly (from 9.5% to 9.6%).
“I think we have oversold this technology, which is really in its first generation,” Dr. Hirsch commented at the OSU meeting. “Patients and providers have to know how to use the technology appropriately, including putting it on the right patient.”
For example, it's unlikely that patients with very poor glucose control who are not committed to treatment goals and aren't already performing frequent fingerstick glucose testing will benefit from an RT-CGM.
“If the hemoglobin A1c is high at baseline, the patient may not be using current tools adequately. My personal opinion is that RT-CGM is a better technology for the patient who is closer to target, in order to prevent hypoglycemia. Someone with an A1c above 9%–10% probably doesn't need CGM,” he said.
Based on his experience using both the Medtronic and the DexCom Inc. RT-CGM products, Dr. Hirsch offered the following clinical “pearls” for patients' use:
▸ The more you look, the better you do. Patients who wear the sensor but don't look at it often do no better than those who don't wear it at all. “No matter which sensor you're using, you have to look at it over and over to see the trend,” he said.
▸ Trend trumps “insulin on board.” This is controversial. Insulin pumps that automatically perform bolus calculations will instruct the patient not to take any more insulin if there is already a large amount “on board.” However, if the glucose is rapidly rising for longer than 1 hour, Dr. Hirsch recommends that adult patients take at least a small bolus. “If the glucose is going up, you need insulin,” he said. For children, however, pediatric endocrinologist Dr. William Tamborlane of Yale University, New Haven, Conn., recommends waiting at least 2 hours before overriding the bolus calculator.