Moderate Glucose Target May Reduce Deaths : Moderate blood glucose control was linked to a significant 40% reduced mortality after CABG.

The study by Dr. Ailawadi and his associates included patients at the University of Virginia during 1995–2008 who underwent CABG and either had known type 2 diabetes preoperatively or developed perioperative hyperglycemia, defined as a blood glucose level of 126 mg/dL or higher at their first postoperative glucose measurement or a mean glucose of at least 126 mg/dL during the first 3 days after surgery. Their average age was 64 years, and about 70% were men.

During the first 3 postoperative days, average blood glucose levels were 118.9 mg/dL in the tight-control group, 152.5 mg/dL in the moderate-control group, and 214.6 mg/dL in the liberal-control group.

Unadjusted mortality rates during hospitalization were 2.9% in the tight-control group, 2.0% in the moderate-control group, and 3.4% in the liberal-control group.

The unadjusted rates of the combined complication end point were 19.4%, 11.1%, and 14.2% in the three groups, respectively.

Multivariate analysis that controlled for clinical and demographic differences revealed that moderate glucose control was linked to a significant 40% reduced mortality and a 30% reduced rate of complications compared with the liberal-control group.

The tight-control group had a 50% reduced mortality compared with the liberal-control group, but the difference was not significant.

Patients in the tight-control group had the same rate of complications as those in the liberal-control group, said Dr. Ailawadi.

'We're not saying that you can be lax about glucose, but that you don't need to be as tight as 80–110 mg/dL.”

Source DR. AILAWADI

My Take

Study Did Not Prove Moderate Better Than Tight Glucose Control

Neither the title of this paper nor the conclusions are in any way supported by the data. Nor is it possible that any difference between the tight- and moderate-control groups, if it did actually exist, could have been statistically detected in this study.

Due to the lack of sufficient numbers of patients in the tight-control group, the study was markedly underpowered to detect any difference between the tight- and moderate-control groups. The power of this study to detect an absolute 1% reduction in mortality between the two groups was only 3%. Furthermore, these were entirely retrospective data from an administrative database, not a clinical database. Clinical outcomes are assumed by interpolating coding information.

What this report has shown is that moderate control is better than no control at all, a finding supported by 15 years of published literature. Moderate control reduced mortality by 40% compared with no control. Interestingly, the point estimate for tight control in the multivariate mortality analysis shows a 50% reduction in mortality compared with no control, but with not enough patients to bring the point estimate to statistical significance. To imply that moderate control is superior to tight control when they weren't even directly compared is either wishful thinking or misleading marketing rhetoric that is not supported by the statistical data.

I have similar concerns about the morbidity conclusions. Of the five major complications included in the morbidity end point, three—stroke, prolonged ventilation, and reoperation—have never been shown to be associated with or caused by glycemia in cardiac surgery patients. Thus, glycemic control is unrelated to the complications studied.