Nearly one in five rheumatology RCTs still unpublished 30 months after completion
Click Here to Manage Email Alerts
Key takeaways:
- Data representing more than 4,000 patients remain unpublished.
- Researchers urged efforts for universal publication of rheumatology trials, as well as analyses of previously unpublished trials.
Publication of trial data in rheumatology appears to correlate with positive results, and after more than 2 years about one-fifth of completed trials remain unpublished, according to data published in Rheumatology.
“We first considered this study after seeing a similar trend in trials of rheumatoid arthritis, where a substantial portion of registered [randomized clinical trials (RCTs)] went unpublished,” Connor Pedersen, a medical student at the Medical College of Wisconsin, told Healio. “Positive outcomes among RA trials were also highly associated with publication, which was concerning for publication bias. Our goal was to expand this kind of study to other rheumatologic diseases that are less represented among trials in order to get a more complete picture of the state of publication in rheumatology.“
To analyze the rate of unpublished studies in rheumatology and investigate the associated factors, Pederson and colleagues searched ClinicalTrials.gov for five rheumatic diseases, including lupus, vasculitis, spondyloarthritis, Sjögren’s syndrome and psoriatic arthritis. The searches spanned from Feb. 20, 2000, through Dec. 31, 2018. Trials were excluded if they were observational, were interventional with non-random subject allocation, extension studies or were phase 1 or 4.
For every included study, the researchers collected data on the identifier, the title, the submission date, the trial start date, the primary completion date, the sponsor, study phase, disease of interest, measured interventions, comparative effectiveness and masking status. The researchers identified which studies were unpublished by conducting searches of databases including PubMed, Google Scholar, Scopus and Web of Science. If the studies could not be matched, the researchers attempted to contact the authors via email.
The analysis included a total of 203 studies that met the criteria, of which 17.2% were unpublished. Data from these trials included information on 4,281 patients. Overall, published trials were more likely to be phase 3 studies — 57.1% published compared with 28.6% unpublished (P < .05) — or have a positive primary outcome measure — 64.9% published vs. 25.7% unpublished (P < .001).
According to the researchers, authors who responded to communication requests cited continued manuscript preparation, funding issues, or negative and unimportant results as the reasons behind trials remaining unpublished.
“The problem of nonpublication among registered RCTs is not limited solely to rheumatoid arthritis, but is present in multiple rheumatologic diseases,” Pederson said. “Trials with positive outcomes were more likely to be published than those with negative outcomes, which is a trend we’ve come to see in other medical fields and is concerning for publication bias. This bias could inflate the treatment’s presumed efficacy or mask potential harms.”
Collapse
Read more about
Click Here to Manage Email Alerts