Publication Rates of Heart Failure Clinical Trials Remain Low

Under-reporting of clinical trial results inhibits dissemination of knowledge, limits understanding of therapeutic interventions, and may ultimately harm patients.

This study examined the rates and predictors of heart failure clinical trial publication and how they have changed over time.

This study assessed cross-sectional analysis of all heart failure clinical trials registered on ClinicalTrials.gov with at least 2 years follow-up after trial completion. The content area was chosen for the robust clinical trial activity in the field. The primary outcome was manuscript publication with multivariable proportional hazards adjustment to identify associations with publication.

Of the 1,429 included studies, 806 (56%) were published as manuscripts, 623 were unpublished, and 97 (7%) reported results without manuscript publication. Of the total, 1,243 were completed after 2007, when the mean 1-year publication rate for interventional trials rose from 12.7% to 19.6% (p = 0.049), which was possibly associated with changes in government regulation. However, there was no further sustained improvement over time, and there was no multivariable association between later completion dates and reporting or publication of results. Funding from the National Institutes of Health and use of clinical (death, hospitalization, myocardial infarction, changes in functional classification) rather than nonclinical primary endpoints were associated with earlier publication. Whether the results were consistent with the primary study hypothesis was not associated with likelihood of publication.

The rates of heart failure clinical trial publication or reporting of results remain unacceptably low. Additional efforts by all stakeholders, including investigators, sponsors, regulators, societies, editors, and journals are needed to improve data dissemination.