Toma 2006.

Methods	Identification of subsequent full‐length publications Searched electronic database MEDLINE, PubMed, Embase, Cochrane Central Register of Controlled Trials, Google/Google Scholar, SIGLE, CardioSource, Incirculation to December 2005 Search completed by investigator Searched by all authors, keywords, title, and study acronym Matched abstract to full‐length publication by Authors Study name/acronym Title keyword
Data	Included 186 abstracts presented at the 1999 to 2002 American College of Cardiology scientific meetings Included 86 abstracts of late breaking RCTs reporting clinical outcomes, and 100 other randomly chosen RCTs; RCTs with non‐human participants were excluded
Comparisons	Proportion of abstracts published, overall and by meeting Median time to publication by whether 'late‐breaking' or not Survival analysis of publication rate 'Positive' versus not 'positive' Multi‐centered versus single center 'Late‐breaking' versus not 'late‐breaking'
Outcomes	148 of 186 abstracts published 34/43 abstracts presented at the 1999 meeting, 36/48 at the 2000 meeting, 41/49 at the 2001 meeting, and 38/46 at the 2002 meeting published Proportion of abstracts published by time Median time to publication = 11.5 months for 'late‐breaking' abstracts Median time to publication = 22 months for abstracts that were not 'late‐breaking' Survival analysis of proportion published at 60 months = 83.3% Factors related to proportion of abstracts published included 70/84 'positive' (defined as significant results) versus 78/102 not 'positive' abstract results published 74/93 'positive' (defined as experimental better than control) versus 74/93 not 'positive' abstract results published 55/63 abstracts with multiple centers versus 93/123 abstracts with a single center published 79/86 'late‐breaking' abstracts versus 69/100 abstracts that were not 'late‐breaking' published
Notes	Cardiology Reported receiving no funding
Risk of bias
Item	Authors' judgement	Description
Sampling method?	Yes	Included a random selection of abstracts describing RCTs and al late breaking abstracts describing RCTs.
Search for publications?	Yes	Searched 7 databases.
Follow‐up time?	Yes	All meetings before 2002 had at least 48 months follow‐up. The meeting in 2002 only had 36 months follow‐up.
Matching?	Yes	Matched by 3 different criteria.
Adjustment for confounding?	Yes	Examined association of positive results, multi‐center status, and 'late‐breaking' status with publication using Chi2 or tests and time to publication using a multivariable proportional hazards analysis.