The review concluded that steroid prophylaxis may reduce morbidity after cardiac surgery and did not increase the risk of postoperative infections. The review was generally well conducted. The authors’ conclusions reflect the results, but do not fully reflect the uncertainties with the evidence base; hence, caution is warranted when interpreting the conclusions.

To assess the effects of steroid treatment on mortality and morbidity after cardiac surgery.

PubMed and the Cochrane Library were searched from 1966 to January 2009. Search terms were reported. Reference lists of selected studies and relevant reviews were searched.

Randomised controlled trials (RCTs) in adult patients who underwent elective cardiac surgery and received pre-operative or intra-operative steroid prophylaxis were eligible for inclusion. Trials had to be placebo-controlled and have a randomised double-blind design. The relevant clinical endpoints were postoperative mortality, new onset atrial fibrillation, postoperative infections, postoperative chest tube drainage, length of intubation, length of intensive care unit stay and overall hospital stay. Trials were excluded if they included confounding treatments or included cohorts of patients that had been studied in another trial.

The included trials studied methylprednisolone, dexamethasone, hydrocortisone, prednisolone and methylprednisolone plus dexamethasone, primarily pre- and intra-operatively (some trials continued steroids postoperatively). Most trials included diabetic patients; some trials excluded insulin dependent diabetic patients. Most trials were published after 1999; some trials were published between 1979 and 1998.

Three reviewers independently screened titles and abstracts. It was unclear how many reviewers applied the inclusion criteria to the full papers.

Two authors independently assessed trial quality using the Jadad scale of randomisation, blinding, allocation concealment and withdrawals and drop-outs to give a maximum score of 5. Disagreements between reviewers were resolved by discussion with a third reviewer.

Data were extracted on clinical, biological and physiological outcomes and used to calculate odds ratios (ORs) or mean differences (MDs), each with 95% confidence intervals (CIs).

Two reviewers independently extracted data. Disagreements were resolved by a third reviewer.

A random-effects meta-analysis was used to calculate pooled odds ratios and mean differences, each with 95% confidence intervals. Statistical heterogeneity was assessed using the I² statistic. A continuity correction was applied in the instance of zero events in one arm; in the case of zero events in both arms, the trial was excluded from analysis. Subgroup analysis was conducted on the basis of steroid drug and diabetic status.

Thirty-one RCTs were included in the review (n=1,974 patients). Trial sample sizes were generally small: four trials had a sample size of more than 100, 15 trials had a sample size between 31 and 100 patients and 12 trials had a sample size of 30 patients or less. Trial quality was variable: five trials scored 5 out of 5, 10 scored 4, 12 scored 3 and four scored 2.

Compared with placebo, there was no statistically significant difference with steroids in terms of mortality, re-exploration, postoperative infections, sternal wound infections and ventilation length. Steroids offered a statistically significantly reduced risk of atrial fibrillation (OR 0.56, 95% CI 0.44 to 0.72; I²=0%; 15 RCTs) and a statistically significantly reduced blood loss (MD -204.2mL, 95% CI -287.4 to -121, I²=0%; seven RCTs). Steroids also offered a statistically significantly shorter length of intensive care unit stay (MD -6.7 hours, 95% CI -10.5 to -2.8, I²=89%; 17 RCTs) and a shorter overall stay (-0.8 days, 95% CI -1.4 to -0.2, I²=46%; 16 RCTs).

Subgroup analysis indicated differences in certain outcomes between methotrexate and dexamethasone and between diabetic and non-diabetic patients.

Steroid prophylaxis may reduce morbidity after cardiac surgery and does not increase the risk of postoperative infections.

Inclusion criteria for the review were clearly defined. Two relevant data sources were searched. It was unclear whether any language restrictions were imposed, so the risk of language bias was uncertain. Publication bias was not assessed and could not be ruled out. Attempts were made to reduce reviewer error and bias during data extraction, quality assessment and screening of titles and abstracts. Assessment of trial quality using a standard checklist indicated variable quality evidence, which the authors acknowledged. The sample size was less than 30 patients in more than one third of the included trials, which the authors noted. Trials were combined using random-effects meta-analysis, although significant statistical heterogeneity remained in some of the analyses (likely due to variation in type and dose of steroid across trials). The authors noted that trials were conducted over a 30-year span, during which time treatments and techniques changed.

The review was generally well conducted. The authors’ conclusions reflect the results, but do not fully reflect the uncertainties with the evidence base; hence, caution is warranted when interpreting the conclusions.

Practice: The authors did not state any implications for practice.

Research: The authors stated that a forthcoming large RCT (NCT00427388) may give useful insights. Future studies should determine the best steroid for cardiac surgery. A large RCT would be needed to determine the effects of steroids on postoperative mortality. Outcomes such as neurologic or gastrointestinal complications needed to be examined using standard definitions.

Not stated.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.