Scherer 1994.

Methods	Identification of subsequent full‐length publications Searched electronic database MEDLINE from year of presentation through 1992 Search completed by investigator Searched by all authors Matched abstract to full‐length publication by All authors Contents Contacted abstract authors directly
Data	Included 93 abstracts presented at the 1988 and 1989 American Academy of Ophthalmology and Association for Research in Vision and Ophthalmology meetings Included all abstracts of randomized controlled trials as verified by abstract author.
Comparisons	Proportion of abstracts published Cumulative proportion of abstracts published 'Positive' versus not 'positive' Sample size equal to or above the median versus sample size below the median Multi‐centered versus single center 'High' versus 'low' quality as defined by author
Outcomes	61 of 93 abstracts published Proportion of abstracts published by time Cumulative proportion of abstracts published at 48 months showed proportion published = 63.4% (59/93 abstracts) Factors related to proportion of abstracts published included 33/46 'positive' (defined as significant results) versus 28/47 not 'positive' abstract results published 31/43 abstracts with sample size equal to or above the median versus 24/42 abstracts with sample size below the median 14/19 abstracts with multiple centers versus 45/71 abstracts with a single center published 31/43 abstracts rated by author as 'high' quality versus 29/47 abstracts rated as 'low' quality published
Notes	Ophthalmology/optometry ‐ ophthalmology Government funding
Risk of bias
Item	Authors' judgement	Description
Sampling method?	Yes	Included all abstracts that described specific study design, so low risk of bias.
Search for publications?	Yes	Searched 1 database and contacted abstract authors.
Follow‐up time?	Yes	The meeting in 1988 had 48 months and the meeting in 1989 only 36 months follow‐up.
Matching?	Yes	Matched by 2 different criteria.
Adjustment for confounding?	No	Examined association of positive results, sample size, trial design, multi‐center status, abstract quality, and subspecialty with publication using stratified analysis and Chi2 tests.