A recent paper, published in the New England Journal of Medicine asked the question: does treating elevated blood sugar in children with critical illness help or hurt? We have asked the same question – regarding adult patients – previously on this blog.
To get you up to speed I will reprint part of that earlier entry:
“In 2001, a paper by van den Berghe and colleagues was published in the New England Journal of Medicine. It described a trial of intensive blood sugar control in critically ill patients and reported improved survival with intensive glucose treatment using insulin.
This study led to a substantial increase in aggressive hyperglycemia treatment known as “tight glycemic control” in the intensive care unit.
In 2009, another paper refuted the results of the first. The NICE SUGAR study enrolled 6000 critically ill patients, randomizing 3000 of them to tight glycemic control. The key result:
Mortality was higher (27.5% v. 24.9%) in the intensive insulin treatment arm.”
Now, in medicine, 2009 was something like a million years ago, and you might be forgiven for thinking that the NICE SUGAR study would be the final word on the subject. However, leaving high blood sugar alone – permissive hyperglycemia – is a heretical concept, and one that has not died easily. So, in my hospital, blood sugars are normalized, though perhaps not as aggressively as in the pre- NICE SUGAR era.
To be sure, the NICE SUGAR study did not involve children. A subsequent study examined children after cardiac surgery – and did not show a benefit for normalization of blood sugar in that pediatric population. But non-surgical pediatric patients had not been studied. To the authors of the current study, that necessitated doing a randomized controlled trial of aggressive insulin treatment of hyperglycemia in sick kids.
Researchers enrolled 713 critically ill children and randomly assigned them to a lower target blood sugar group who received more insulin (doing more something) and higher target group who received less insulin (doing more nothing). The main outcome measure was ICU-free days – i.e. time not in the ICU up to day 28. This outcome was similar in both groups. Mortality was also similar.
But, this study was stopped early, because the interim analysis determined a low likelihood of benefit from giving more insulin and a high risk of harm. Indeed, in the enrolled group, children receiving more insulin had “higher rates of health care–associated infections” (12 of 349 patients [3.4%] vs. 4 of 349 [1.1%], P= 0.04) in the group receiving less insulin. Not surprisingly, trying to aggressively normalize blood sugar resulted in higher rates of severe hypoglycemia (a blood glucose level below 40 mg/dl. (18 patients [5.2%] vs. 7 [2.0%], P = 0.03).
It is very unlikely that further pediatric trials on aggressive treatment of hyperglycemia will be performed. Why? Because we cannot make a case for equipoise, the state of scientific uncertainty between two possible treatments – that would make it ethical to undertake a similar trial. This is especially so for a study of kids: children are a vulnerable population unable to provide informed consent for themselves.
There are two points to made about these results, one clinical and one theoretical. Clinically, I will expect that ill children especially will not undergo aggressive insulin therapy for high blood sugars in the hospital. At least I hope so. There comes a time when antiquated ideas in medicine need to die. That time is now.
The second point is a reappraisal of the role of hyperglycemia in critical illness. It is my view that hyperglycemia is an adaptation that confers a survival benefit on average – making patients better able to meet the challenge of their critical illness. That view is supported by the recent NEJM study and the previous NICE SUGAR study.
Evolutionary medicine is useful as far as it leads to distinct and better outcomes than conventional, non-evolutionary, approaches in medicine. In the case of normalizing high blood sugars in critical illness, evolutionary medicine can help doctors identify adaptation where previously they saw only pathology. This evolutionary perspective will help save the lives of children by speeding the adoption of better, less aggressive treatments, now strongly supported by the best quality evidence.