A new study shows just more than 55% of all heart failure-based trials on ClinicalTrials.gov have shared results since 2008.
Mitchell Psotka, MD, PhD
An examination of clinical trial publication rates including the assessment of heart failure (HF) finds the cardiovascular condition is significantly lacking representation in journals.
In an Inova-led study of published clinical trials, plus predictors of future trial publication, a team of investigators reported HF trial publication rates are “unacceptably low” despite regulatory efforts from the US Food and Drug Administration (FDA) to bolster the rates a decade ago.
Led by Mitchell A. Psotka, MD, PhD, of the Inova Heart & Vascular Institute, the team presented their findings at the Heart Failure Society of America (HFSA) 2019 Scientific Sessions in Philadelphia, PA.
In September 2007, the Food and Drug Administration Amendments Act of 2007 (FDAAA) expanded previous guidelines to clinical trials, requiring all published articles to be registered and reported to ClinicalTrials.gov.
As Psotka and colleagues noted, public reporting and trial result publishing is an ethical obligation that “repays societal expenditures on clinical evidence generation.” As a result, under-reported trial results inhibit the dissemination of knowledge, limit the knowledge surrounding therapy interventions, and ultimately promote risk to patients.
Investigators examined the publication rate of clinical trials in HF, predictors of publication, and whether they have changed in response to the FDAAA legislation. They assessed ClinicalTrials.gov-registered studies with the primary condition being HF, reported and completed by June 30, 2017.
Time to publication was defined as time from the trial’s completion on ClinicalTrials.gov, to the time of publication online or in print. Time to reporting was defined as time from trial completion to results being posted.
Clinical endpoints for each trial were defined as either discrete clinical events—including death, hospitalization, myocardial infarction, and changes in functional classification—or non-clinical endpoints.
From 2000-2008, investigators observed 186 completed HF studies—149 interventional and 37 observational/registry. Of those, 117 (62.9%) were published, with a mean time to publication of 30 months.
After January 1, 2008—once the FDAA bill was put into action—investigators observed 1243 HF trials. Of those, the majority were again interventional (n= 918). Mean annual HF trials published online had increased from 16.56 annually, to 138.1 annually.
However, just 689 (55.4%) of trials were published online from January 2008 to June 2017, with a mean time to publication of 18 months.
Psotka and colleagues observed trials funded by the National Institutes of Health (NIH), those having clinical endpoints, and larger patient population sizes were each associated with a greater rate of trial result publication.
Trials funded by the US Veterans Department, or with patient populations from 0-150 were significantly less likely to have trial results published. Whether or not the studies’ results were consistent with the primary trial hypothesis was not shown to be significantly associated with the trial results being published.
In response to the results, investigators called on clinicians, sponsors, steering committees, and even publishing health journals to focus on improvements to the rate of HF clinical trial results publication—keeping in mind the effort is one with an ethical obligation toward patients.
“Future strategies may include facilitating the report of trial results before manuscript publication, standardizing FDAA and journals’ requirements, and encouraging post-hoc trial registration,” Psotka and colleagues wrote.
The study, “Publication Rates for Heart Failure Trials Remain Unacceptably Low Despite Regulation,” was presented at HFSA 2019.