CHICAGO -- Those daily glucose-monitoring finger sticks may be overkill for many patients with type 2 diabetes, reported investigators here.
CHICAGO, June 28 -- Those daily glucose-monitoring finger sticks may be overkill for many patients with type 2 diabetes, reported investigators here.
In a one-year study, glycosylated hemoglobin levels were similar among patients whose only glucose monitoring testing was an HbA1c every three months compared with patients who followed one of two frequent self-monitoring regimens, reported Andrew J. Farmer, M.D., of the University of Oxford in England, and colleagues in British and U.S. centers.
Results were no better for patients who followed an intensive self-monitoring regimen than those who followed a less-demanding one, Dr. Farmer said in a late-breaking studies session at the American Diabetes Association meeting.
"Patients, clinicians and policy-makers will need to look at the results to reach decisions about appropriate use of self-monitoring of blood glucose technology," he said. "The results of this trial will add to the evidence available to make the decisions."
He cautioned, however, that all patients with type 1 diabetes, and patients with type 2 who require insulin injections, should adhere to self-monitoring guidelines, as tight glycemic control with insulin, plus lifestyle interventions, have been conclusively shown to reduce the occurrence of microvascular and macrovascular complications of diabetes.
John Buse, M.D., Ph.D., of the University of North Carolina at Chapel Hill, the president of the ADA, commented that he had no major quibble with Dr. Farmer's results. His own research, he said, had found that self monitoring was not as beneficial as had been anticipated, and he suspects that "a lot of glucose self-monitoring in the U.S. is not cost-effective. A lot of patients check more frequently that they should."
Self-monitoring can be useful for newly diagnosed type 2 patients to determine the proper dose of drugs, but the U.S. medical system, as opposed to the British system, works against frequent visits to a physician for HbA1c testing, Dr. Buse pointed out.
Self-monitoring is a good tool, Dr. Buse said, and it can work. Yet he cites the practice of many patients who check their blood one or more or times a day and don't write the readings down, or who keep them faithfully but don't report them to their doctors.
"How useful it is depends on how you use it," he added.
Dr. Farmer said previous studies of self-monitoring of blood glucose have been multifactorial, confounding the relative contributions of self-monitoring or other interventions on risk-factor reduction in patients with non-insulin dependent diabetes.
The Oxford investigators designed a trial to see whether standard care, self-monitoring of blood glucose alone, or self monitoring in combination with diabetes education (aimed at incorporating HbA1c results into self-care could) have an effect on overall glycemic control in patients who do not require exogenous insulin.
They recruited 453 adults with type 2 diabetes from 48 family practices. The mean patient age was 65.7 years, and 57% of the participants were male. They were randomly assigned to either:
The primary study outcome was HbA1c at one year, adjusted for baseline values.
A total of 57 patients (13%) were lost to follow-up at one year. The loss of patients was distributed even across the three groups, and the analysis was by intention-to-treat.
The authors found that two-thirds of patients who followed the less-intensive self-monitoring regimen and half of those following the more intensive plan had monitored their blood glucose levels at least twice per week over the past 12 months.
The mean HbA1c at baseline was similar across the three groups at 7.5%, above the less than 7% value recommended by the American Diabetes Association., and at one year there were still no significant differences among the groups. The mean difference in HbA1c between controls and patients in the less-intensive monitoring groups was 0.14% in favor of monitoring, and between controls and the hard-core self monitors, the mean difference was 0.17%, also in favor of monitoring. These differences were not statistically significant, however (P=0.12).
"A difference of -0.5% or more would generally be needed for a therapy to be considered clinically effective, and we powered the study to be able to see a difference that large if it existed, but we did not get it," said Dr. Farmer.