Background: Eight clinical trials involving the administration of preoperative i.v. methylprednisolone have been undertaken in order to decrease the considerable inflammatory response to esophageal resection, in an effort to decrease the supposedly associated morbidity and mortality
Method: A meta-analysis was performed for eight clinical end-points. Due to quality problems in seven of the eight included studies, a Bayesian meta-analysis using a skeptical prior derived from the results of the classical analysis was also performed.
Results: The end-points including any organ dysfunction (OR = 0.30), respiratory complication (OR = 0.41), sepsis (OR = 0.37), liver dysfunction (OR = 18), cardiovascular dysfunction (OR = 0.28), and surgical anastomotic leak (OR = 0.42) were significantly decreased by methylprednisolone pretreatment. Following the Bayesian analysis, despite the use of skeptical priors, there is a 95% probability to obtain a relative risk reduction of at least 23% to 54%, depending of the end-point, by methylprednisolone pretreatment.
Conclusion: We are in the presence of a potential benefit that cannot be accepted at face value due to the quality problems of the included studies. But in the presence of a remaining potential benefit after a Bayesian analysis starting from a skeptical prior, the best option would be the planning of a large multicenter prospective randomized study.