Should patients with type 2 diabetes test their blood glucose (BG) levels?
The American Diabetes Association recommends BG testing for all patients. But British researchers Andrew Farmer and colleagues announced at the June meeting of the American Diabetes Association that BG monitoring in type 2s resulted in no change in hemoglobin A1c results and suggested that the costs would be better spent “supporting other health-related behaviors.”
The announcement set off a feeding frenzy of articles in the news media as well as Internet health blogs, most of them with headlines saying testing in type 2s was not useful. For example: “Not Using Insulin? You May Not Need Home Glucose Tests” or “Blood Glucose Monitoring for Non-Insulin Users Shows No Benefits.”
Like most media soundbites, or “blogbites,” many of the stories were brief oversimplifications of Farmer’s study, and they didn’t point out any of the caveats from the full paper in the British Medical Journal (BMJ).
Farmer’s study, called the DiGEM study, isn’t the first to tackle this question. There have been numerous previous studies, some showing a significant benefit of testing and some showing none. Farmer’s group pointed out the flaws in previous studies and said that their study was sufficiently large and well controlled to show a significant A1c change of 0.5 or greater.
However, anyone taking the time to read the full study will see that the results were much more limited than as reported in the press. For example, the study concluded that testing had no benefit “in reasonably well controlled” type 2s not using insulin. That’s because the mean A1c of the participants was about 7.5.
A more serious flaw, in my opinion, was the omission of a statement in a 2004 paper describing how the DiGEM study was going to be conducted, noting that “This trial is mainly generalizable to that group of patients willing to be randomised to no self-testing. lt will be limited in its ability to inform management of people who are enthusiastic about regular meter use.”
This caveat was never mentioned in the 2007 BMJ paper reporting the results. That means that even a person reading the full 2007 BMJ paper very carefully would not come across this statement. And how many readers (except compulsive diabetes bloggers) when reading a news story saying that type 2s don’t need to test anymore are apt to go on the Internet and dig out an old study reporting nothing but methods?
This caveat is because the study excluded all those who were already testing their BG regularly. So basically, the study recruited people who weren’t particularly motivated to test their BG and who probably weren’t particularly motivated to do much self-care of their diabetes at all!
The “intensive treatment” group were instructed to test three times a day, two days a week. Even with this relatively low testing frequency, only 52% maintained this level of testing throughout the 12-month study, again suggesting a lack of interest in self-care.
