Low-Dose Corticosteroid Treatment in Septic Shock: A Propensity-Matching Study*

Objective: Given conflicting data and current guidelines, low-dose corticosteroids are often used in the treatment of septic shock. To evaluate the therapeutic benefit of early low-dose corticosteroid in patients with septic shock. Design: Retrospective, multicenter, propensity-matched cohort study. Setting: ICUs of 28 academic and community hospitals in three countries between 1996 and 2007. Subjects: Six thousand six hundred sixty-three eligible patients with septic shock of whom 1,838 received IV low-dose corticosteroid treatment within 48 hours of the diagnosis of septic shock and were matched to a comparable group who did not receive low-dose corticosteroid. Measurements and Main Results: The primary outcome was 30-day mortality. Mortality analyses were stratified by severity of illness (Acute Physiology and Chronic Health Evaluation II quartile). Using a Cox proportional hazards model, corticosteroid therapy was associated with similar 30-day mortality when compared with the matched control cohort (652/1,838 [35.5%] vs 641/1,838 [34.9%]; hazard ratio, 0.98; 95% CI, 0.88–1.10; p = 0.77). In the subgroup of patients with the Acute Physiology and Chronic Health Evaluation II score quartile more than or equal to 30, low-dose corticosteroid was associated with lower mortality (232/461 [50.6%] vs 251/450 [55.8%]; hazard ratio, 0.81; 95% CI, 0.68–0.97; p = 0.02). In logistic regression models, corticosteroid therapy was not associated with reductions in ICU (556/1,838 [30.3%] vs 558/1,838 [30.4%]; odds ratio, 0.99; 95% CI, 0.86–1.15; p = 0.94) or hospital mortality (797/1,838 [43.4%] vs 773/1,838 [42.1%]; odds ratio, 1.05; 95% CI, 0.93–1.20; p = 0.42). Similarly, there were no significant differences in ventilator- (median and interquartile range, 13 [0–25] vs 15 [0–25]; p = 0.8) and pressor/inotrope-free days (median and interquartile range, 25 [3–27] vs 24 [2–28]; p = 0.63) up to 30 days between groups. Conclusion: Early administration of low-dose corticosteroid is not associated with decreased mortality when it is administered to unselected patients with septic shock. A beneficial effect of low-dose corticosteroid on mortality may exist in patients with the highest severity of illness. Future trials of low-dose corticosteroid in septic shock should consider restricting the study population to this cohort.