Clinical Utilization of Stress Dose Hydrocortisone in Adult Patients With Septic Shock: A Retrospective Observational Study at a Large Academic Medical Center

Abstract

Background: The use of stress dose corticosteroids, specifically, hydrocortisone, in septic shock is heterogeneous, and current clinical trials yield conflicting results. Regardless, they are still recommended by guidelines for vasopressor-dependent septic shock. Objectives: This study sought to characterize current practice of hydrocortisone use in patients with septic shock and secondarily to compare clinical outcomes of those who received hydrocortisone to those who did not. Methods: This single center, retrospective cohort study evaluated patients with septic shock admitted to a tertiary care center between 2012 and 2017. Patients receiving hydrocortisone for at least two doses were compared to those without. Results: 3411 septic shock patients were included; 1593 (47%) received hydrocortisone and 1818 (53%) did not. Patients who received hydrocortisone had higher lactate (4.0 vs 3.4 mmol/L; P < .01) and Acute Physiology and Chronic Health Evaluation (APACHE) III scores (104.1 vs 91.0; P < .01). Vasopressor duration was 1.7 days longer in the hydrocortisone group (P < .01), and the hydrocortisone group had higher hospital mortality (52% vs 38%; P < .01). A propensity score–matched population was conducted in patients with APACHE scores >100: vasopressor duration was longer in those who received hydrocortisone (3.9 vs 2.0 days; P < .01), and hospital mortality was higher (59.3% vs 53.1%; P = .036); however, after multivariable adjustment, no association between receipt of hydrocortisone and hospital mortality was detected (OR 1.2 [95% CI .9–1.6]). Conclusions: Patients who received hydrocortisone were more severely ill than those that did not, making retrospective evaluation of this question challenging. These results highlight the wide variability and heterogeneity in hydrocortisone use in clinical practice.