or unpublished on this topic, suggested as a potentially qualifying study by a participating group. The goal of this inclusive approach is to minimize the potential impact of publication bias. The participating collaborative studies are listed in Table 1. All participating studies have approval from their local boards governing ethical conduct of research in human subjects to conduct the agreed upon statistical analyses. A list of the review boards can be found in the acknowledgements. Additional groups that have been invited, but for a variety of reasons are not currently participating, are identified by publications listed in Additional file 1: Table S1.