Publication rate and characteristics of trials registered in the Australian and New Zealand Clinical Trial Register

Session: 

Oral session: Investigating bias (2)

Date: 

Monday 17 September 2018 - 15:10 to 15:20

Location: 

Carrick 2

All authors in correct order:

Jordan V1, Buchman S1, Showell M1, Berber S2, Willson M2, Ata Allah N3, Farquhar C1
1 University of Auckland, New Zealand
2 University of Sydney, Australia
3 Aleppo University Faculty of Medicine, Syria
Presenting author and contact person

Presenting author:

Vanessa Jordan

Contact person:

Vanessa Jordan
Abstract text
Background: A comprehensive search of the literature is an important step in reducing publication bias in systematic reviews. Since 2007 medical journals have required trial registration as prerequisite to publication. This research was designed to assess whether registered trials are being published and determine if the characteristics of trials that reach publication differ substantially from those that have not.

Objectives: To determine the publication rate of trials initiated in 2007 and 2011 registered in the Australia and New Zealand Clinical trials register (ANZCTR) database and the trial characteristics associated with publication

Methods: The ANZCTR provided a list of randomised controlled trials (RCTs) that began recruitment in the years 2007 and 2011 and their demographics. We contacted investigators to determine publication status and undertook searches on ANZCTR number, author name and/or topic in an attempt to find the corresponding publication on Google Scholar, Cochrane Central Register of Controlled Trials, MEDLINE and Embase. Once found, we recorded the citation and date of publication.

Results: From 779 trials overall the publication rate of trials registered in the ANZCTR is 65%. For trials that began recruitment in 2007 (n = 325) and 2011 (n = 454) the publication rate was 73% and 60%, respectively. For the published trials the mean time to publication was 4.49 years (standard deviation 2.11). Overall 49% of the trials were retrospectively registered with 56% of the published trials being retrospectively registered. The median sample size for published trials was 95 (inter-quartile range (IQR) 45.5 to 220) and unpublished 84 (IQR 42 to 186.5).

Conclusions: There is still a large deficit between trials conducted and trials published. This does not seem to be improving over time. For at least half the trials that are registered, the registration process is being completed retrospectively.

Patient or healthcare consumer involvement: Healthcare consumers altruistically participate in order to improve healthcare long-term. There is an expectation that the result of this research is published. Failure to disseminate clinical research is a betrayal of the relationship between researchers and consumers.

Relevance to patients and consumers: 

Healthcare consumers altruistically participate in order to improve healthcare long-term. There is an expectation that the result of this research is published. Failure to disseminate clinical research is a betrayal of the relationship between researchers and consumers.