Cheng 1998.

Methods	Identification of subsequent full‐length publications Searched electronic database Cochrane Cystic Fibrosis Group register, Embase, MEDLINE to 1995 Search completed by investigator Search criteria not reported Matched abstract to full‐length publication by All authors Titles Content Methodology
Data	Included 178 abstracts found in Cochrane Cystic Fibrosis group's register of trials Included all abstracts of RCTs except duplicates and those published before presentation
Comparisons	Proportion of abstracts published Median time to publication Survival analysis of publication rate 'Positive' versus not 'positive' Sample size equal to or above the median versus sample size below the median
Outcomes	57 of 178 abstracts published Proportion of abstracts published by time Median time to publication = 18 months Survival analysis of proportion published at 60 months = 40.1% Factors related to proportion of abstracts published included 43/113 'positive' (defined as experimental better than control) versus 14/42 not 'positive' abstract results published 36/92 abstracts with sample size equal to or above the median versus 25/78 abstracts with sample size below the median published
Notes	Other clinical specialties ‐ cystic fibrosis Funding by National Health Service
Risk of bias
Item	Authors' judgement	Description
Sampling method?	Yes	Included all RCTs with reasonable exceptions.
Search for publications?	Yes	Searched 3 databases.
Follow‐up time?	Unclear	Follow‐up of abstracts as short as 3 months to multiple years. Unclear how many abstracts were followed up for < 48 months.
Matching?	Yes	Matched by 4 different criteria.
Adjustment for confounding?	No	Examined association of positive results and sample size with publication using stratified analysis and time to publication with univariate Cox regression models.