Blood glucose control in critically ill children

The physiological stress response to critical illness and trauma determines that transiently raised blood glucose (hyperglycemia) is common among those admitted to intensive care, irrespective of their diabetes status. Much research has been directed at establishing whether or not intensive insulin therapy aimed at normalizing blood glucose has benefit for these critically ill patients.

Although there have been conflicting results, most studies indicate no benefit in treating mild hyperglycemia (blood glucose in the approximate range of 6.0-8.0 mmol/L), not least because striving to maintain normal blood glucose is associated with evidential risk of reduced blood glucose (hypoglycemia).
  
Studies have predominantly focused on adult patients; the most recently published was conducted on critically ill children. This multicenter international study, called the HALF-PINT (Heart and Lung Failure – Pediatric Insulin Titration) trial, identified 713 critically ill children with hyperglycemia (blood glucose >150 mg/dL [8.3 mmol/L]) at 32 ICUs across the US and beyond.

These 713 hyperglycemic children were randomized to one of two insulin treatment regimes. Three hundred and sixty (360) were assigned to the intensive insulin treatment with target blood glucose in the range of 80-110 mg/dL (4.4-6.1 mmol/L) and 353 were assigned to less intensive insulin treatment aimed at maintaining blood glucose in the higher range, 150-180 mg/dL (8.3-10.0 mmol/L).

A common protocol for blood glucose control (timing of blood glucose measurements, insulin dosing, etc.) was adopted at all study ICUs and maintained until discharge from ICU. The primary outcome measure was the number of ICU-free days at day 28, which is the inverse equivalent of the more commonly used outcome measure: 28-day hospital mortality-adjusted ICU length of stay.

Secondary outcome measures were: 90-day mortality, severity of organ dysfunction, number of ventilator-free days, incidence of ICU-acquired infection and incidence of hypoglycemia. 

Results demonstrated clear difference in blood glucose concentration between the two groups post intervention: median time-weighted average glucose in the low target group 109 mg/dL (IQR 102-118 mg/dL), i.e. 6.1 mmol/L (IQR 5.7-6.6 mmol/L) versus 123 mg/dL (IQR 108-142 mg/dL), i.e. 6.8 mmol/L (IQR 6.0-7.9 mmol/L) in the high target group.

Not surprisingly, insulin requirements differed significantly between the two groups: 
Those in the lower glucose target group received a median insulin dose of 0.74 units/kg/day compared with 0.1 units/kg/day for those in the higher target group. Nearly all (98 %) of those in the lower target group required some insulin compared with only 66 % in the higher target group. 

All outcome measures (both primary and secondary) were, however, not significantly different between the two groups. The authors of this study were able to conclude their study has demonstrated intensive insulin therapy targeting a blood glucose of 80-110 mg/dL is no more beneficial in terms of outcome for critically ill hyperglycemic children than less intensive insulin therapy targeting a blood glucose of 150-180 mg/dL. 

The study authors had originally planned to recruit 1400 children for study but interim results from the first 713 children outlined above were sufficiently clear-cut for continued recruitment to be futile, and the study was consequently halted early.