Preston 2005.

Methods	Identification of subsequent full‐length publications Searched electronic database PubMed to January 2004 Search completed by investigator Searched by all authors and keywords Matched abstract to full‐length publication by All authors Title Contents Study institution (for posters)
Data	Included 815 abstracts presented at the 1994 to 1998 Orthopaedic Trauma Association meetings Included all abstracts
Comparisons	Proportion of abstracts published, overall and by meeting Mean time to publication by meeting and type of presentation Oral versus poster presentation
Outcomes	476 of 815 abstracts published 96/153 abstracts presented at the 1994 meeting, 98/168 at the 1995 meeting, 79/156 at the 1996 meeting, 91/157 at the 1997 meeting, and 114/181 at the 1998 meeting published Proportion of abstracts published by time Mean time to publication = 24.8 months (SD = 17.9) for oral presentations and 21.6 months (SD = 19.4) for poster presentations Mean time to publication = 21.9 months (SD = 18.1) for oral presentations and 28.2 months (SD = 27.8) for poster presentations for 1994 Orthopaedic Trauma Association meeting Mean time to publication = 26.1 months (SD = 20.3) for oral presentations and 19.7 months (SD = 18.8) for poster presentations for 1995 Orthopaedic Trauma Association meeting Mean time to publication = 31.1 months (SD = 19.3) for oral presentations and 22.1 months (SD = 17.3) for poster presentations for 1996 Orthopaedic Trauma Association meeting Mean time to publication = 20.9 months (SD = 13.8) for oral presentations and 18.6 months (SD = 15.9) for poster presentations for 1997 Orthopaedic Trauma Association meeting Mean time to publication = 25.4 months (SD = 16.8) for oral presentations and 21.3 months (SD = 17.3) for poster presentations for 1998 Orthopaedic Trauma Association meeting Factors related to proportion of abstracts published included 222/329 abstracts presented orally versus 254/486 abstracts presented as posters published
Notes	Emergency medicine ‐ orthopedic trauma Funding not reported
Risk of bias
Item	Authors' judgement	Description
Sampling method?	Yes	Included all abstracts.
Search for publications?	No	Searched only 1 database.
Follow‐up time?	Yes	All meetings had at least 6 years follow‐up.
Matching?	Yes	Matched by 4 different criteria.
Adjustment for confounding?	No	Examined association of meeting year and type of presentation with publication using stratified analysis and Fisher's Exact test.