Elevated Baseline Cortisol Levels Are Predictive of Bad Outcomes in Critically Ill Children.

OBJECTIVE: The definition of an adequate adrenal response in critically ill children continues to be controversial. We aimed to evaluate the cortisol levels at baseline and after adrenocorticotropin (ACTH) stimulation and determine their association to clinical outcome of critically ill children. METHODS: All children who underwent an ACTH test in the pediatric intensive care unit (PICU) in a tertiary medical center between 2006 and 2013 were included in the study. Data on age, sex, diagnosis, vasoactive-inotropic score, length of pediatric intensive care unit stay, and mortality were obtained. Laboratory variables included hematologic and chemistry data, arterial lactate, and total plasma cortisol levels at baseline and after ACTH stimulation. RESULTS: Ninety-nine patients (61 males; median [range] age, 2 [0-204] months) were enrolled. The mortality rate of children with a baseline cortisol level of 600 nmol/L or greater was 36% (12/33 patients) versus 18% (12/66 patients) for children with a baseline cortisol level of less than 600 nmol/L (odds ratio, 2.6 [95% confidence interval, 1-6.6]; P = 0.05). There was a positive correlation between baseline cortisol and lactate levels (r = 0.40, P < 0.0001), vasoactive-inotropic scores (r = 0.24, P = 0.02), and mortality (P = 0.05). There was no correlation between peak cortisol measured at the ACTH test or the delta increment of cortisol from baseline and mortality. CONCLUSIONS: A high baseline cortisol level in critically ill children was associated with more severe illness, higher lactate level, and a higher mortality rate. Routine baseline cortisol assessment is recommended to identify patients at high mortality risk.