Remember when intensive glucose control in adults became all the “lifesaving” rage — and was then revealed to actually be harmful? Well, it doesn’t seem to work in kids either.
Tight glucose control in critically ill children with hyperglycemia failed to reduce ICU days as compared with a higher target glucose range, in a large randomized trial published in the New England Journal of Medicine.
Children assigned to a target glucose range of 80 to 110 mg/dL or 150 to 180 mg/dL had exactly the same median number of ICU-free days. The trial ended early because of the low likelihood that the stringent glucose control would lead to a benefit.
Worse, tight blood sugar control was associated with double the rate of severe hypoglycemia and three times higher rates of infection, particularly catheter-associated bloodstream infections. There was no apparent evidence of benefit from tight glucose control in any patient subgroup.
The trial was halted early when interim data analysis (at only ~50% of target enrollment) showed a likelihood of benefit from the intervention to be <1%, along with the suspicion of harm.
A 2009 single-center randomized trial had suggested a benefit from tight glucose control in kids undergoing cardiac surgery. Multicenter trials followed, but none demonstrated lower mortality or better outcomes compared with standard care or keeping blood sugars in a reasonable range (e.g., <180 mg/dL).
Most or all trials testing intensive glucose control have had high rates of hypoglycemia in the intervention arms. This sustains the argument that harm from hypoglycemia swamps the benefits of tight control — and if we just did insulin drips better, tight control would save lives. (Whenever you hear someone say this, ask them if they have a grant proposal under review.)
The truth is, tight glucose control can’t be done better in the community (where >90% of patients are treated) than it has been done in large multicenter trials — where it failed. Improving insulin infusions further to prevent severe hypoglycemia would likely require new continuous glucose monitoring technology that’s shown promise, but is still years away from an ICU near you. In that hypothetical future era, randomized trials could be repeated to see if perfectly executed tight glucose control can improve outcomes in the ICU. Using available methods, it’s safe to assume it doesn’t.