Diagnosis of corticosteroid insufficiency in Thai patients with septic shock

The reported incidence of critical illness-related corticosteroid insufficiency (CIRCI) varies widely, depending on the patient population studied and the diagnostic criteria used. Surviving Sepsis Campaign guidelines suggest that corticosteroid therapy should be considered for adult septic shock when hypotension responds poorly to adequate fluid resuscitation and vasopressors, regardless of any results of diagnostic tests. However, steroid treatment may be associated with an increase risk of infection. This study aims to identify the best diagnostic tool for predicting responsiveness to corticosteroid therapy in Thai septic shock patients with poorly responsive to fluid resuscitation and vasopressors.

Twenty-nine septic shock patients who were poorly responsive to fluid therapy and vasopressors were studied. A baseline serum total cortisol was measured in all patients and then 250 mcg corticotropin was injected to patients. Cortisol level was obtained 30 and 60 minutes after injection. All patients were given hydrocortisone (100 mg i.v., then 200 mg i.v. in 24 hrs for at least 5 days). Patients were considered steroid responsive if vasopressor agent could be discontinued within 48 hrs after the first dose of hydrocortisone.

Hospital mortality was 62% in which 45% of the patients were steroid responsive. Baseline serum cortisol was 27.6 +/- 11.4 microg/dl in the steroid-responsive patients compared with 40 +/- 16.9 microg/dl in the steroid-nonresponsive patients (p = 0.03). The area under the ROC curves for predicting steroid responsiveness was 0.72 for baseline cortisol level. Serum cortisol level of 35 microg/dl or less was the most accurate diagnostic threshold to determine hemodynamic response to hydrocortisone treatment (p = 0.04). Using baseline cortisol level of < or = 35 microg/dl to diagnose adrenal insufficiency, the sensitivity was 85%, the specificity was 62% and the accuracy was 72%. A use of (delta cortisol) showed sensitivity of 50%, specificity of 30% and accuracy of 41%.

Baseline cortisol level < or = 35 microg/dl is a useful diagnostic threshold for diagnosis of steroid responsiveness in Thai patients with septic shock and ACTH stimulation test should not be used.