Intensive Glucose Control: Safe for Critically Ill Kids' Brains?
After evangelizing globally for intensive glucose control (~100 mg/dL) to be the standard of care for virtually all critically ill adults for a decade, Greet van den Berghe might be disappointed that mounting evidence shows that a fanatical approach to maintenance of normoglycemia in critical illness probably kills more people than it helps. But there's still a large, untapped market for research on intensive glucose control: children.
You'd think that kids, of all people, would like a little extra sugar. But -- as with adults in the initial trials -- preliminary research has shown improved outcomes in critically ill children who get intensive glucose control, with slight reductions in ICU stay and a 3% absolute mortality reduction. (However, kids in the control arm of that study were allowed to have glucoses of 215 mg/dL.)
With the understandable safety concerns raised by NICE-SUGAR, many parents might be fearful of their children getting intensive insulin drips, with attendant high risk for severe hypoglycemia, whose effects on developing brains is poorly understood. As if on cue, van den Berghe's group has a trial this month in JAMA showing that intensive glucose control causes no harm, and in fact is good for cognitive and motor performance of children tested 4 years after their critical illness.
What They Did
569 children, infants to 16 year olds, were followed-up 3-4 years after enrollment of 700 in a previous trial testing intensive glucose control in critical illness. Patients from both arms (tight and usual control) as well as their siblings were tested whenever possible with neurocognitive and motor function tests by blinded psychologists and a pediatrician.
What They Found
Intensive glucose control did not result in worse outcomes for children four years after treatment in an ICU. Equal proportions were alive and free of major disability in each group.
Intensive glucose control also did not cause neurologic or cognitive impairment in any measured domain, including intelligence (IQ, etc), memory, and visual-motor integration. In fact, kids treated with tight glucose control had higher performance on tests of motor coordination (reaction times, finger taps, etc), and "cognitive flexibility" using an obscure (not-Googleable) test.
After pairing those with severe hypoglycemia to nearest neighbors in a propensity analysis, severe hypoglycemia was not associated with worse outcomes either.
The child survivors of critical illness (in both study arms) performed worse on most measures than their healthy siblings.
What It Means
Children and adolescents with diabetes seem to tolerate hypoglycemia very well, and apparently the old rumors of hypoglycemia induced brain damage were greatly exaggerated. Uncontrolled and severe hyperglycemia is known to be harmful in acutely ill adults and children. It's entirely possible that intensive glucose control, perfectly performed, results in slight improvements in clinical outcomes for critically ill children.
But real world care is not, and never will be, perfect. "Hypoglycemia ... was always rapidly detected and treated" in their trial, say these authors. That's great, but I'm not surprised -- the authors' continued career success partly depended on that. But now, deploy intensive glucose control to the more than 8,000 hospitals in the developed world. Think hypoglycemia will "always" be "rapidly detected and treated"?
Authors mention the finding in NICE-SUGAR that some cases of hypoglycemia were due to "inaccurate glucose monitoring tools," citing this as evidence for intensive glucose control -- in other words, if you all would just do it right, it would work. But here in the real world, glucose monitoring tools can and will fail -- as will the people who run them -- and from a systems perspective, should be expected to. Is a "razor's edge" approach to continuous infusion of a potentially lethal drug really a great idea for a best practice for our flawed and error-prone health system?
Maybe intensive glucose control "saves lives" in these trials by increasing nursing needs so much, and requiring so much bedside attention for hypoglycemia and insulin adjustments, that these patients effectively get extra care at the expense of other patients. Call it "hypoglycemic care steal syndrome." What was the mortality of the patients in the neighboring rooms, during these trials?
Dieter Mesotten, Marijke Gielen, Greet Van den Berghe et al. Neurocognitive Development of Children 4 Years After Critical Illness and Treatment With Tight Glucose Control. A Randomized Controlled Trial. JAMA 2012; 308: 1-10 (ePub online).
keyword: pediatric critical care, peds critical care