Corticosteroid treatment for sepsis: a critical appraisal and meta-analysis of the literature Cronin L, Cook D J, Carlet J, Heyland D K, King D, Lansang M A, Fisher C J Authors' objectives To determine the effect of corticosteroid therapy on morbidity and mortality in patients with sepsis. Searching MEDLINE and EMBASE were searched from 1966 to 1993 using the following MeSH terms: 'sepsis' and 'steroids' or 'corticosteroids', and 'septic shock and 'steroids' or 'corticosteroids'. The Science Citation Index was searched using the terms 'sepsis' and (explode) 'corticosteroid', and 'septic shock' and (explode) 'corticosteroid'. Index Medicus was handsearched from 1951 onwardusing the terms 'corticosteroids', 'steroids', 'sepsis' and 'septic shock'. Personal files, reference lists of relevant primary and review articles, and contact with primary investigators were used to locate additional published and unpublished material. Study selection Study designs of evaluations included in the review Randomised clinical trials were included. Specific interventions included in the review Intravenous corticosteroids (methylprednisolone, dexamethasone, betamethasone and hydrocortisone). Participants included in the review Adults with signs of sepsis or septic shock, including at least 2 of the following: shaking chills and/or temperature greater than 38.9 or less than 35.5 C; heart beat greater than 100 beats/min; systolic blood-pressure less than 90 mmHg; bacteraemia; organ dysfunction. Outcomes assessed in the review Mortality. Complications of corticosteroid therapy (secondary infection, upper gastrointestinal bleeding, progression of organ dysfunction, hyperglycaemia). How were decisions on the relevance of primary studies made? Two investigators independently reviewed the titles and abstracts of all relevant articles. Three investigators (one of whom was blinded to the journal, authors, institution, and the magnitude and direction of results) independently applied the study selection criteria to full manuscripts. The selection criteria were based on study design, population, intervention and outcome. Any discrepancies were resolved by consensus. Assessment of study quality The validity assessment was based on patient selection, patient characteristics at baseline, randomisation, blinding, intervention, contamination, cointervention, explicit description of complications of steroids, withdrawals, use of intention to treat protocol, and explicit definition of septic shock. A methodological quality scoring system was applied independently by two reviewers (one was blinded to the journal, authors, institution, and the magnitude and direction of results). The authors of the primary studies were consulted regarding the accuracy of methodological scores, and additional information was requested if necessary. Any disagreements between reviewers were resolved by discussion and consensus. Data extraction Data were extracted on population, intervention, outcome, and methodological quality by two investigators. Any disagreements were resolved by consensus. Information was requested from primary investigators when data were missing or unclear. Database of Abstracts of Reviews of Effects (DARE) Produced by the Centre for Reviews and Dissemination Copyright © 2016 University of York Page: 1 / 3