Suman Sudha Tirkey1, Rakesh Kumar Verma2
BACKGROUND
Hyperglycaemia is a common occurrence in children with critical illness. Several
studies relate hyperglycaemia occurring during intensive care unit (ICU) admission
to be associated with increased mortality. We wanted to evaluate the incidence of
hyperglycaemia and its association with mortality among critically ill children.
METHODS
A prospective observational study was conducted among critically ill children
admitted to the Paediatric Intensive Care Unit (PICU) of a tertiary care centre,
Raipur, Chhattisgarh in central India, from 1st May 2016 to 31st October 2016. All
patients aged 1 month to 14 years who were admitted in PICU during this study
period (N = 113) were included. Children who were known cases of diabetes
mellitus, hepatic failure or renal failure requiring dialysis and children who left
against medical advice (LAMA) or died within 24 hours of admission were excluded
(N = 13). Thus 100 children were included in the study.
RESULTS
In the study population, incidence of hyperglycaemia was 60 %; overall 37 (37
%) died and the mortality rate was significantly higher (46.6 % vs. 19.4 %) in
children with hyperglycaemia than in children without hyperglycaemia. Nonsurvivors
had higher mean blood glucose levels at 48 hours (218.35 ± 87.42 mg
/ dL) than survivors (141.12 ± 55.26 mg / dL) (P < .001). Peak blood glucose
(218.35 ± 87.42 mg / dL vs. 141.12 ± 55.26 mg / dL), need for mechanical
ventilation (54.5 % vs. 27.3 %), need for inotropes (76.4 % vs. 23.5 %) and
Paediatric Risk of Mortality Score (PRISM) III (16.25 ± 5.46 vs. 9.06 ± 4.35) were
significantly higher in non-survivors than in survivors. On regression analysis,
blood sugar at 24 hours and duration of stay were found to be significant.
CONCLUSIONS
In this study, in the PICU, the mortality rate was significantly higher in children
with hyperglycaemia than in children without hyperglycaemia. Non-survivors had
significantly higher mean blood glucose levels at 48 hours than survivors. Peak
blood glucose, need for mechanical ventilation, need for inotropes and PRISM III
scores were significantly higher in non-survivors than in survivors.
