PURPOSE: We describe steroid use in patients with severe sepsis and assess the relationship between patient characteristics and steroid use.
METHODS: We conducted a single-center, retrospective, observational study evaluating 93 consecutive patients with severe sepsis admitted to the medical intensive care between 1/12/08 and 8/30/09. Baseline and clinical patient characteristics were assessed and the association between these characteristics and the cosyntropin (ACTH) stimulation test with steroid use was evaluated.
RESULTS: Stress dose steroids were administered to 33.3% of the patients. Patients receiving steroids had higher APACHE II scores (24.2±7.4 vs. 32.9±9.2, p< 0.001) and maximum lactate levels (3.3±3.2 vs. 5.2±4.4, p=0.03); were more likely to require multiple antibiotics (87.1% vs. 100%, p=0.04), multiple vasopressors (9.7% vs. 45.2%, p < 0.001), and mechanical ventilation (58.1% vs. 87.1%, p=0.01), and have higher fluid balance in the first 24 hours (3032±2488 vs. 7263±5719, p< 0.001). No significant differences in age, gender or race were found. Among patients receiving stress dose steroids, hydrocortisone was most frequently prescribed (83.9%), followed by methylprednisolone (16.1%), prednisone (9.7%), fludrocortisone (6.5%), and dexamethasone (6.5%). ACTH tests were performed on 26.9% patients. ACTH testing, regardless of result, was associated with steroid administration (p< 0.001). Among patients with ACTH testing who had an increase in cortisol > 9 mcg/dL from baseline, 63.6% did not receive steroids, while 78.6% of patients with an increase ≤ 9 mcg/dL were given steroids. In multivariable logistic regression, ACTH testing (OR 7.72, 95% CI 1.90 to 31.41, p=0.004) was the only significant variable. There was no significant collinearity between ACTH testing and higher-than-median APACHE II scores (p=0.67).
CONCLUSION: Patients receiving stress dose steroids have greater severity of illness. Significant variation exists in steroid regimens and implementation of ACTH testing results. In the multivariable explanatory model, ACTH testing was independently associated with steroid use.
CLINICAL IMPLICATIONS: In the absence of expert consensus, significant variations exist in the use of stress dose steroids in severe septic shock.
DISCLOSURE: Tessy Paul, No Financial Disclosure Information; No Product/Research Disclosure Information