How tight is too tight?

Regarding the debate on tight glycemic control published in the June issue of Canadian Family Physician,1 I think it might be worthwhile to look at the hemoglobin (HbA_1c) levels the studies referred to actually achieved and reported, rather than what their targets were.

The achieved levels of HbA_1c for the UKPDS (United Kingdom Prospective Diabetes Study) follow-up were 8.5% (conventional) versus 7.9% (intensive) in the insulin-sulphonylurea group and 8.9% (conventional) versus 8.4% (intensive) in the metformin group. There were better cardiovascular outcomes in the intensive groups, meaning those who achieved HbA_1c levels of 8.4% in the metformin group or 7.9% in the insulin-sulphonylurea group. This does not mean that going below 7.9% will result in better outcomes—we simply do not know.

The HbA_1c levels achieved in the ADVANCE (Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation) study were 6.5% versus 7.3%. The decrease in end points was mainly driven by improvements in nephropathy.

In the VADT (Veterans Affairs Diabetes Trial), achieved HbA_1c levels were 6.9% versus 8.4%. There were no differences in outcomes between the 2 groups.

In the ACCORD (Action to Control Cardiovascular Risk in Diabetes) study, levels were 6.4% versus 7.5%. There was increased mortality in the tight control group.

These studies seem to indicate that better glycemic control improves outcomes—up to a point. Where that point is is open to debate. The average HbA_1c level in those studies seems to indicate that an appropriate HbA_1c level is somewhere between 7% and 8%. The current average HbA_1c level for my practice is 7.5%; if I systematically target patients with levels above 7% for intensification, I will drive my average down. I do not think any of the studies above tell me to do that for my practice. It seems to me that I should make a systematic effort to reduce HbA_1c levels in the individual patient level to below 8%; at the practice level, the average level should be between 7% and 8%. I would have to target patients with the highest levels of HbA_1c, as they will benefit the most—perhaps leaving those with levels between 7% and 8% alone—to get to results similar to those of the UKPDS.

We can now start to translate evidence from individual patient care to care of a practice population, which is a different way of looking at evidence-based medicine. I do not think clinicians should go beyond the evidence, and right now evidence does not seem to support a goal of having an overall HbA_1c practice average of 7% or less. What that means in terms of individual patient goals needs to be reviewed.

We need a very clear and evidence-based definition of what “tight glycemic control” is; authors of guidelines might wish to revisit their current recommendations.

• Copyright© the College of Family Physicians of Canada