Adrenal insufficiency in prolonged critical illness

Introduction Adrenal insufficiency is common in critically ill patients and affects their prognosis, but little is known about how adrenal function changes during prolonged critical illness. This study was conducted to investigate dynamic changes in cortisol levels in patients with critical illness who do not improve after treatment. Methods This observational cohort study was performed in the intensive care units of a university hospital. We studied acutely ill patients with initial cortisol level above 34 mu g/dl, but who did not improve after treatment and in whom follow-up cortisol levels were determined during critical illness. All clinical information and outcomes were recorded. Results Fifty-seven patients were included. Ten patients had follow-up cortisol levels above 34 mu g/dl, 32 patients had levels between 34 and 15 mu g/dl, and 15 patients had levels under 15 mu g/dl. Outcomes did not differ significantly among the three groups with different follow-up cortisol levels. In Cox regression analysis, those patients who survived to hospital discharge with second cortisol levels under 15 mu g/dl had a longer hospital length of stay (odds ratio = 14.8, 95% confidence interval = 2.4 to 90.0; P = 0.004). Conclusion The majority of acutely ill patients who remained in a critical condition had decreased serum cortisol levels. Depressed cortisol levels at follow up may lead to worse clinical outcomes. We propose that repeated adrenal function testing be conducted in patients with prolonged critical illness.