Landry 1996.

Methods	Identification of subsequent full‐length publications Searched electronic database MEDLINE from January 1990 through November 1994 Person completing the search not reported Searched by 'principal' author and keywords Matched abstract to full‐length publication by 'Principal author' Keywords Contents
Data	Included 168 abstracts presented at the 1990 American Burn Association meeting Included all abstracts
Comparisons	Proportion of abstracts published 'Positive' versus not 'positive' Clinical research versus basic science research
Outcomes	44 of 168 abstracts published Proportion of abstracts published by time not reported Factors related to proportion of abstracts published included 24/58 'positive' (defined as significant results) versus 20/110 not 'positive' abstract results published 16/65 abstracts describing clinical research versus 15/63 abstracts describing basic science research published
Notes	Emergency medicine ‐ trauma, burns Funding not reported
Risk of bias
Item	Authors' judgement	Description
Sampling method?	Yes	Included all abstracts with reasonable exceptions.
Search for publications?	No	Searched only 1 database.
Follow‐up time?	Yes	The meeting had 48 months follow‐up.
Matching?	Yes	Matched by 3 different criteria.
Adjustment for confounding?	No	Examined association of positive results, type of science, university affiliation, doctoral degree, use of statistics, and clarity of statistics with publication using stratified analysis and Chi2 tests.