Hydrocortisone infusion for severe community-acquired pneumonia

Am J Respir Crit Care Med 2005;171:242-248

Introduction

Severe community acquired pneumonia is associated with a high mortality rate due to sepsis related complications. Non-survivors show high circulating levels of cytokines especially interleukin (IL)-6. The investigators hypothesized that the administration of hydrocortisone early in the course of the disease would reduce the incidence of sepsis-related complications and death by modulating the inflammatory response.

Study

A total of 46 patients were enrolled in this prospective, randomized, double blind placebo controlled multicentre trial. Patients with a clinical and radiographic suspicion for pneumonia were eligible if they met two minor or one major 1993 ATS criterion for severe pneumonia. The patients assigned to the hydrocortisone group (n = 23) were treated with a 200 mg loading dose followed by 10 mg/hr for 7 days. Primary endpoints were improvement of the PaO₂/FiO₂ ratio and the multiple organ dysfunction syndrome score by day 8 and development of delayed septic shock. Secondary endpoints were the duration of mechanical ventilation, length of ICU and hospital stay, survival to hospital discharge and survival to 60 days. Patients were analyzed according to the intention-to-treat principle. All patients who entered the trial were properly accounted for and follow-up was complete. Unfortunately, baseline characteristics of the two groups were not completely comparable. Patients in the placebo group were a little older, had a higher incidence of comorbidities but had a lower chest radiograph score, a lower C-reactive protein level and a higher baseline PaO₂/FiO₂ ratio. The study was stopped when an interim analysis after 46 patients showed a significant difference between the two groups in the improvement of the PaO₂/FiO₂ ratio and hospital mortality. On study day 8 there was an improvement of the PaO₂/FiO₂ ratio in 9 patients randomized to placebo versus 20 patients randomized to hydrocortisone (p= 0.0018). The MODS score was also significantly different as was the incidence of delayed septic shock (9 versus 0 respectively) Mortality at hospital discharge was 30% in the placebo group and 0% in the hydrocortisone group. At 60 days this was 38% and 0% respectively.

Discussion

Although the study appears to be properly conducted some concerns may be raised. Due to the small number of patients, randomization resulted in important imbalances among prognostic factors. Only one out of 11 patients in the placebo group developing septic shock received rescue hydrocortison therapy. Nowadays hydrocortison therapy in these patients is considered the standard of care. Another concern is the fact that the study was stopped after (the second?) interim analysis. Explicit stopping rules are not provided and the authors give no evidence that statistical techniques compensated for these repeated analyses. Under these circumstances it is important to consider the "regression to the mean" principle. As the authors already state a larger randomized clinical trial with consideration of all prognostic factors and adequate treatment of septic shock is necessary to confirm the hypothesis that hydrocortisone infusion result in an improved outcome in patients with severe community acquired pneumonia. Until then, hydrocortison should not be given for this specific indication.

M. Hilkens, Fellow Intensive Care
J.G. van der Hoeven, Internist-Intensivist