cited a trial registration number (Table ).Moreover, only 283 abstracts (39.3%) specified a primary outcome, and 353 (49.0%) included the word randomized, randomised, or RCT in the title.A total of 442 abstracts (61.4%) specified the number of participants randomized, and 181 (25.1%) specified the number of participants analyzed in each group.In addition, only 101 (14.0%) included dates of recruitment and follow-up.Overall, only 249 abstracts (34.6%) included at least 3 of 6 key CONSORT reporting items.Although trial registration numbers were included in only 97 abstracts, they could be found for an additional 380 by searching registries and publications, thereby increasing the percentage of abstracts that reported registered trials to 66.2% (477 abstracts).Discussion | Our analysis, albeit limited to select major medical conferences, found that clinical trial registration numbers are rarely cited in conference abstracts and adherence to CONSORT guidelines for abstracts is minimal.These findings reflected the abstract submission guidelines for the 8 conferences that were reviewed, none of which mentioned the inclusion of clinical trial registration numbers or CONSORT reporting items.In contrast, almost all abstracts reporting RCT results at the annual meeting of the American Society of Clinical Oncology, held from May 29 to June 2, 2015, at Chicago, Illinois were found to cite trial registration numbers, 5 reflecting the society's abstract guidelines, which required registration numbers to be cited for registered trials. 6The priority given to trial registration and compliance with CONSORT guidelines by that organization arguably reflects their urgency to promote translation of clinical trial results into advances in cancer care.It is unrealistic for most, much less all, CONSORT reporting items to be included in conference abstracts given abstract word-count restrictions.However, such restrictions further emphasize the importance of including trial registration numbers because they may be used to access essentially all CONSORT reporting items in registry data sets. 4oreover, because many trials presented at conferences remain unpublished, detailed and timely information about them may be accessible only through a trial registry.Even when published, access to the publications may not be open, and registration numbers may provide the only means by which clinicians could readily acquire sufficient information about a trial to judge the value and relevance of its results to clinical practice.