Evaluation of adrenal function in critically ill children

ObjectiveThere is no consensus on adequate adrenal response to critical illness. We aimed to evaluate adrenal function in critically ill children and its association with clinical outcome. We hypothesized that salivary cortisol would be a more appropriate tool to evaluate adrenal function in critically ill children. MethodsThis was a prospective cohort study. The concentrations of serum total and salivary cortisol were measured in 34 critically ill children before and after stimulation with 250g adrenocorticotropic hormone (ACTH), and values were compared to a control group of healthy children (n=15). Association between outcome and adrenal insufficiency defined by an increment in serum cortisol 250nm (9g/dl) post-ACTH was assessed. ResultsSerum total and salivary cortisol concentrations pre- and post-ACTH were significantly higher in patients, and they were correlated at baseline (r=067; P<00001) and after ACTH (r=041; P=002). The incidence of adrenal insufficiency was 323%. This group had higher Paediatric Risk of Mortality III score (P=004) but Paediatric Logistic Organ Dysfunction and vasoactive inotropic scores, duration of mechanical ventilation and length of paediatric intensive care unit and hospital stay were not significantly different compared with those with an increment >250nm (9g/dl) post-ACTH. An inverse correlation between salivary cortisol post-ACTH and vasoactive inotropic score (r=-056; P=00008) was observed. A salivary cortisol concentration post-ACTH of 226nm (82g/dl) had a sensitivity of 79% and a specificity of 62% to discriminate need for vasoactive or inotropic support (area under receiver operating characteristic (ROC) curve 074). ConclusionAdrenal insufficiency defined by the delta criterion' was not associated with outcome. A post-ACTH salivary cortisol of 226nm (82g/dl) may be suggestive of an insufficient adrenal response to critical illness.