Content area
Full text
About the Authors:
Vojtech Huser
* E-mail: [email protected]
Affiliation: Laboratory for Informatics Development, NIH Clinical Center, Bethesda, Maryland, United States of America
James J. Cimino
Affiliation: Laboratory for Informatics Development, NIH Clinical Center, Bethesda, Maryland, United States of America
Introduction
The purposes of trial registries, such as ClinicalTrials.gov, include serving as a repository for shared clinical trials results and providing the ability to track journal articles reporting trial results in medical literature. ClinicalTrials.gov registry has been documenting trials since 2000 and is the world’s largest registry. Of particular note, it provides electronic access to its database. It is an important tool for many constituents, including researchers (14%) and potential research volunteers (23%), with 95 million page views per month and 60,000 unique visitors daily [1,2].
We are interested in exploring mechanisms for automatically identifying results of completed trials that have either shared data sets within the registry or linked the trial records to published, peer-reviewed journal articles. In this study, we combined ClinicalTrials.gov data with data from PubMed, the popular citation database from the National Library of Medicine (NLM), to look at a large set of clinical trials and investigate how results of trials are made publicly available. We include in our study, an analysis of change in the number of trials reporting their results in some form since the establishment of policies and laws that encourage such reporting. Previous reports show that 32% to 53% of trials do not report their outcome [3,4]. To our knowledge, our study is the first study that combines publication data with deposited basic summary results. Our study also looks at the largest set of registered trials compared with previous publication rate analyses.
Policies for Reporting Trial Results
From an evidence-based perspective, clinicians and researchers need to know about all human clinical trials in a given domain and consider results of all of them. Worldwide, there are policies of varying strengths that encourage publishing trial results in journal articles and legal mandates to report data about trials to clinical trial registries. In the US, principal investigators of trials supported by National Institute of Health (NIH) grants are encouraged to publish journal articles reporting trial results [5]. There are generally two types of mandates related to the use of trial...