Prospective Trial Registration and Publication Rates of Randomized Clinical Trials in Digital Health: A Cross Sectional Analysis of Global Trial Registries

Abstract

AbstractRegistration of clinical trials was introduced to mitigate the risk of publication and selective reporting bias in the realm of clinical research. The prevalence of publication and selective reporting bias in trial results has been evidenced through scientific research. This bias may compromise the ethical and methodological conduct in the design, implementation and dissemination of evidence-based healthcare interventions. Principal investigators of digital health trials may be overwhelmed with challenges that are unique to digital health research, such as the usability of the intervention under test, participant recruitment, and retention challenges that may contribute to non-publication rate and prospective trial registration. Our primary research objective was to examine the prevalence of prospective registration and publication rates in digital health trials. We included 417 trials that enrolled participants in 2012 and were registered in any of the seventeen WHO registries. The prospective registration and publication rates were at (38.4%) and (65.5%) respectively. We identified a statistically significant (P<.001) “Selective Registration Bias” with 95.7% of trials published within a year after registration, were registered retrospectively. We reported a statistically significant relationship (P=.003) between prospective registration and funding sources, with industry-funded trials having the lowest compliance with prospective registration at (14.3%). The lowest non-publication rates were in the Middle East (26.7%) and Europe (28%), and the highest were in Asia (56.5%) and the U.S. (42.5%). We found statistically significant differences (P<.001) between trial location and funding sources with the highest percentage of industry funded trials in Asia (17.3%) and the U.S. (3.3%).