Plasma cortisol is often decreased in patients treated in an intensive care unit

Objective: To investigate the prevalence of adrenal hypofunction, as assessed by plasma cortisol (p-cortisol) and its relationship to clinical events. Design: Prospective, consecutive. Setting: General intensive care unit in a university hospital. Patients: Fifty-five patients (34 men and 21 women) were studied (surgery 40 patients, hemodialysis 5, ventilator treatment 45, sepsis 21). Methods: Morning basal levels of p-cortisol were determined. Previous reports define adrenal insufficiency to be present if p-cortisol under stressful conditions is lower than either 400 or 500 nmol/l. The tetracosactoid test (250 μg Synacthen) was performed in 16 patients and urinary 24-h excretion of cortisol in 24 (none on corticosteroid treatment). Results: Median p-cortisol was 550 nmol/l (range 20–1764). In 36 % of patients p-cortisol was lower than 400 nmol/l and in 47 % lower than 500 nmol/l. There was a significantly increased probability (P < 0.05) of p-cortisol being below 400 nmol/l in patients admitted due to trauma or cerebral disorder and in patients on ventilator therapy or on mannitol. Thirty minutes after tetracosactoid administration p-cortisol response was lower than 200 nmol/l in 56 % of the patients. Conclusions: Several patients had low p-cortisol and attenuated responses to tetracosactoid, indicative of adrenal insufficiency. There seem to be certain risk factors for adrenal hypofunction which may justify more frequent use of physiological doses of corticosteroid in selected patients Received: 5 February 1999 Final revision received: 16 December 1999 Accepted: 21 December 1999