Browsing by Author "Pello, Annukka"

Outcome reporting bias (ORB) and publication bias are prevalent problems in medical research and cause a threat to reliability of evidence based medicine. In this work I will 1) present the basic data collection conducted for an ORB and publication bias study, 2) illustrate the usefulness and difficulties in using REC submissions for research, and 3) review the preliminary results of the study by Chan et al. My data collection included all trial protocols in HUS area in 2002 and 2007. There were in total 265 (2002: 139, 2007: 126) trial protocols. Protocols had a lot of variation in their manner and completeness of reporting. The data collection was somewhat cumbersome and included a lot of bureaucracy. However, these kind of studies are important in order to accurately determine ORB and publication bias. Moreover, the forthcoming publication by Chan et al. is, to my knowledge, the first to compare all three sources: protocols, trial registries and publications.