Reply to Yang

We have to thank Yang for his well-taken comments on our manuscript [1] based on his dedicated and careful description and analysis of studies included in our meta-analysis [2] about pre-emptive aortic counterpulsation in cardiac surgery. In the following, we would like to comment on his remarks.

First of all, we would like to emphasize that we tried to strictly avoid the inclusion of studies that did not meet the predefined criteria. Therefore, we excluded studies that were pooled in previously published analyses, e.g. because of a missing control group or a second treatment group (e.g. levosimendan) as a comparator [3]. We agree that the randomized controlled trials (RCTs) from Christenson have to be interpreted with caution due to some major limitations that, however, at least partly apply to all RCTs on this topic as extensively discussed in our manuscript [4, 5]. However, we cannot agree to exclude all or some of Christenson's publications from our analysis due to a suspected, albeit unproven double publication. In personal communication with the Cochrane Heart team, Christenson assured the independency of his publications published in 1997 despite overlapping inclusion periods [6]. Against this background, we included all three Christenson trials from 1997 in our work. In contrast to all other meta-analyses, we performed a subgroup analysis with restriction to studies published since 2009 with exclusion of all Christenson studies still showing a survival benefit for the counterpulsation group.

On the basis of these clarifications, we appreciate the concerns of the author about the study selection process and the over-representation of the Christenson population, but conclude that methodology as well as conclusions of our manuscript still prove to be correct.

REFERENCES