Too Many HF Trial Results Underreported or Never Published

The clinical trial community has not made it a priority to report and publish results of a significant percentage of heart failure studies conducted in recent years, an analysis suggests.

Lead investigator Mitchell A. Psotka, MD, PhD (Inova Heart and Vascular Institute, Falls Church, VA), said the problem of unpublished and underreported trial results in cardiology has been well recognized in the scientific literature for some time, and unfortunately nothing seems to be propelling reform.

“There has been no change over time despite attempts at regulations and interventions by journal publishers and others to try to improve this,” he told TCTMD. “Our study shows there really has been no growth in this area over the last decade.”

The analysis, published June 22, 2020, in the Journal of the American College of Cardiology, examined randomized interventional trials and observational trials that began before and after passage of the US Food and Drug Administration Amendments Act of 2007 (FDAAA), which mandated trials involving FDA-regulated interventions to register and report their results to ClinicalTrials.gov, an online global database under the direction of the US National Library of Medicine.

Modest Effect of Reporting Requirement

Psotka and colleagues found that while there was a bump in registration of heart failure trials to ClinicalTrials.gov after passage of the FDAAA, only 56% of the more than 1,400 studies were ever published in a journal. Of those that were terminated or incomplete, only 20% ever reported their results, and of those that were completed but never published, only 12% ever reported their results.

Nearly 20% of all registered interventional trials and 17% of all observational studies were published within 1 year of completion through 2017. However, the implementation of the FDAAA had only a modest effect on the 1-year publication rate over the 10-year period, increasing it from 12.7% to 19.6% (P = 0.049).

Among the funding sources, institutional or local academic sources accounted for approximately half of interventional heart failure clinical trials and observational studies. Industry sponsorship accounted for 40%, followed by the National Institutes of Health (NIH) and the US Department of Veterans Affairs. Compared with the others, NIH-funded trials had a higher likelihood of being published (71% vs 55%; P < 0.001). Interventional trials that were funded by the NIH had a publication rate at 1 year of 30%, compared with 19% for those with other sources of funding and 20% for industry-funded trials.

A significant difference in the clinical information provided was seen between studies that were published versus unpublished. While the majority of published studies reported their primary hypothesis and 61% reported data in support of that hypothesis, only 14% of unpublished studies reported their primary hypothesis to ClinicalTrials.gov and just 10% reported data to make their case.

After multivariable adjustment, NIH funding and use of a clinical primary endpoint, rather than a nonclinical primary endpoint, were associated with more rapid publication of interventional trials.

By not releasing results, particularly for trials that do not meet their primary endpoint, we are depriving society of that investment that patients have made. Mitchell Psotka

Among the 55 terminated trials that were never published, 27% were reported as a conference abstract or on ClinicalTrials.gov alone and only eight were reported within 1 year of trial termination. In multivariable modeling, use of a clinical endpoint and longer trial duration were associated with earlier publication of terminated trials.

A Disservice to Patients

To TCTMD, Psotka said failure to publish results of completed or incomplete trials is not only a disservice to science and the research community—it also shows disregard for the patients who agreed to participate.

“Patients enter into a contract with the researchers, the sponsors, and everyone else performing a clinical trial when they sign up. The contract is that they are participating in research not only for the benefit of themselves, to attempt to get a new therapeutic before it may be available, but also altruistically for the benefit of everyone,” he said. “By not releasing results, particularly for trials that do not meet their primary endpoint, we are depriving society of that investment that patients have made. They are putting their own health on the line to provide benefit for society, and we are not repaying that debt to them when we keep these results hidden from the public eye and from scientific and public understanding.”

In an editorial, Mandeep R. Mehra, MD (Brigham and Women’s Hospital and Harvard Medical School, Boston, MA), and Luanda P. Grazette, MD, MPH (University of Southern California, Los Angeles), say the underreporting of clinical trial data has persisted into the contemporary era as “a sin of omission” despite the ability today to compute data with high efficiency, rapidly disseminate information via the internet, and have simultaneous publications released in conjunction with meeting presentations.

Reasons for delaying or avoiding reporting trial results may include publication bias “that steers more positive results toward publications,” Mehra and Grazette note. In some cases, sponsors may seek less timely dissemination of negative information due to perceived financial consequences, they write. As for what we needs to be done, they advocate for a kind of “coercive paternalism,” such as making it mandatory to publish results at the time of FDA filing for regulatory approval, and suggest that underreporting results should possibly be considered scientific misconduct as well as “a public health matter because it is an impediment to medical discovery and poses plausible threats to patient safety.”

Psotka said the answers to improving trial reporting and transparency are likely multifactorial, “but we think everyone needs to do a better job of getting together and making publication and the reporting of clinical trial results a priority.” He added that his study also shows that “from top to bottom there are many issues within the sharing of information and publication.”