Objective: Controversy remains as to whether low-dose corticosteroids can reduce the mortality and morbidity of acute lung injury (ALI) or the acute respiratory distress syndrome (ARDS) without increasing the risk of adverse reactions. We aimed to evaluate all studies investigating prolonged corticosteroids in low-to-moderate dose in ALI or ARDS.
Data sources: MEDLINE, EMBASE, Current Content, and Cochrane Central Register of Controlled Trials, and bibliographies of retrieved articles.
Study selection: Randomized controlled trials (RCTs) and observational studies reported in any language that used 0.5-2.5 mg.kg.d of methylprednisolone or equivalent to treat ALI/ARDS.
Data extraction: Data were extracted independently by two reviewers and included study design, patient characteristics, interventions, and mortality and morbidity outcomes.
Data synthesis: Both cohort studies (five studies, n = 307) and RCTs (four trials, n = 341) showed a similar trend toward mortality reduction (RCTs relative risk 0.51, 95% CI 0.24-1.09; p = 0.08; cohort studies relative risk 0.66, 95% CI 0.43-1.02; p = 0.06). The overall relative risk was 0.62 (95% CI 0.43-0.91; p = 0.01). There was also improvement in length of ventilation-free days, length of intensive care unit stay, Multiple Organ Dysfunction Syndrome Score, Lung Injury Scores, and improvement in Pao2/Fio2. There was no increase in infection, neuromyopathy, or any major complications. There was significant heterogeneity in the pooled studies. Subgroup and meta-regression analyses showed that heterogeneity had minimal effect on treatment efficacy; however, these findings were limited by the small number of studies used in the analyses.
Conclusion: The use of low-dose corticosteroids was associated with improved mortality and morbidity outcomes without increased adverse reactions. The consistency of results in both study designs and all outcomes suggests that they are an effective treatment for ALI or ARDS. The mortality benefits in early ARDS should be confirmed by an adequately powered randomized trial.