0909 Tight glucose control in paediatric ICU patients

While studies in adult ICU populations might not be absolutely conclusive, it is probably fairly widely accepted that it is beneficial for patients to have their glucose levels controlled within a tight range if/when their medical and/or surgical conditions warrant management in an intensive care unit. Little has been done in this regard in the paediatric ICU population, and randomised controlled studies are lacking or absent. This deficiency prompted a group in Belgium to study their patients, resulting in a paper that appeared in a recent edition of the Lancet. Seven hundred subjects were enrolled (383 children ≥1 yr and 317 <1 yr) and randomised within the latter two groups to receive insulin aimed at maintaining blood glucose levels in age appropriate ranges for the duration of ICU stay (conventional therapy of <11.9mmol/l for both age groups vs 2.8-4.4 for infants and 3.9-5-6 in children). Primary endpoints were CRP as a marker of inflammation and duration of ICU stay. Mean glucose levels in the tightly controlled groups were 4.8mmol/l for infants and 5.3 for children vs levels in the conventionally-treated of 6.4mmol/l for infants and 8.3 for children. Mostly-asymptomatic hypoglycaemia occurred in 25% of the tightly controlled subjects. In terms of the study objectives, duration of stay in ICU was 5.5 days vs 6.2 (p<0.017) and 37.8% of intensively treated subjects had stays of >3 days vs 47.0% of those treated conventionally (p<0.013). Monitoring of CRP showed significant differences between the groups, starting on day 2 and diverging even more thereafter. A number of secondary endpoints also showed an advantage for tightly-controlled subjects; these included a significant difference in mortality (9 deaths in the tightly-controlled group vs 20 in the conventionally-treated: 3 vs 6%; p<0.038). There was also a diminished requirement for pressor support, and troponin levels were lower in cardiac patients. The latter point brings out a shortcoming in terms of generalisability of the results in that 75% of patients were post-surgery for congenital heart disease. These findings are consistent with many emanating from adult ICU’s, but again need to be confirmed, particularly in other diagnostic groups (i.e. without the cardiac emphasis). Concern has been expressed about the frequency and possible consequences of hypoglycaemia with tight control, as well as the need to only conduct such studies in environments that are able to monitor intensively and react appropriately to swings in the glucose levels.

Read more:
Lancet 2009; 373: 547-56 and 520-1
Paediatrics 2006; 118: 173-9
Pediatr Crit Care Med 2008; 9: 147-52