LAS VEGAS – Although inpatient glucose levels in critically ill adults are generally kept between 140 and 180 mg/dL, Toronto investigators have found that it might be best to keep pediatric burn patients between 130 and 140 mg/dL.
Morbidity and mortality outcomes were better in that range when 760 children – the majority boys under 10 years old burned over half their bodies, usually by flame – were followed for 2 months after ICU admission.
"We used about 300,000 glucose measurements" during the study "to determine the ideal glucose target. For burn patients, the range of 130-140 mg/dL should be targeted. You avoid hyperglycemia as well as hypoglycemia. [Also,] a major factor to be considered" in burns "is that protein glycosylation starts at around 150 mg/dL; we burn surgeons try to be below that," said lead researcher Dr. Marc Jeschke, director of the burn center at Toronto’s Sunnybrook Health Sciences Centre.
Dr. Jeschke’s study isn’t the first to suggest that range for critically ill children. "There seems to be a signal [across] studies that this is the range we should target in order to see the benefits of glucose control," he said (e.g., J. Pediatr. 2009;155:734-9).
It’s been known that burns cause a hyperglycemic response, especially those in excess of 40% of the body. When not reined in, "you lose more grafts, you have more infections, and you [are more likely to] die," Dr. Jeschke said at the annual meeting of the Surgical Infection Society.
In its investigation, however, his team also found that, as in critically ill adults, hypoglycemia must also be avoided in pediatric burn patients.
Eighty-five patients had one hypoglycemia episode in the study, defined as blood glucose below 60?mg/dL, and 107 had two or more. The remaining 568 children had no episodes.
Twenty-one percent of patients who had hypoglycemic episodes – versus 6.5% of those who did not – developed sepsis; 47.9%, versus 10%, developed multiple organ failure, and a quarter died. Mortality was 3.3% in the nonhypoglycemic group. The differences were statistically significant.
"We [also] found that hypoglycemic patients are more inflammatory and hypermetabolic," Dr. Jeschke said.
Hypoglycemia was associated with larger burns and more inhalation injuries, so the team did a propensity analysis comparing 166 children who had hypoglycemic episodes to 166 with similar injury severities who did not.
Among those matched patients, children were 2.67 more likely to die (95% confidence interval 1.15-6.20) if they had one hypoglycemic episode, 5.58 more likely to die if they had two (95% CI 2.26-13.81), and 9.25 times more likely if they had three (95% CI 4.30-19.88).
Hypoglycemia during sepsis or at time of death was excluded in the analysis. Even so, "we don’t know" if hypoglycemia was the cause of death or a marker for another fatal process, Dr. Jeschke noted.
Whatever the case, the results indicate that just as in sick adults, "glycemic control in [pediatric] burns is an integral part of good clinical outcomes," he said.
Dr. Jeschke said he has no conflicts of interest.