Publication bias is present in blood and marrow transplantation: an analysis of abstracts at an international meeting.

Publication bias is the preferential publication of research with positive results, and is a threat to the validity of medical literature. Preliminary evidence suggests that research in blood and marrow transplantation (BMT) lacks publication bias. We evaluated publication bias at an international conference, the 2006 Center for International Blood and Marrow Transplant Research (CIBMTR)/American Society for Blood and Marrow Transplantation (ASBMT) "tandem" meeting. All abstracts were categorized by type of research, funding status, number of centers, sample size, and direction of the results. Publication status was then determined for the abstracts by searching PubMed. Of 501 abstracts, 217 (43%) were later published as complete manuscripts. Abstracts with positive results were more likely to be published than those with negative or unstated results (P = .001). Furthermore, positive studies were published in journals with a mean impact factor of 6.92, whereas journals in which negative/unstated studies were published had an impact factor of only 4.30 (P = .02). We conclude that publication bias exists in the BMT literature. Full publication of research, regardless of direction of results, should be encouraged and the BMT community should be aware of the existence of publication bias.