Discontinuation and non-publication of randomized controlled trials on cervical cancer or precancer

Abstract

Background Research waste is a considerable problem in clinical trials, with nonpublication being a significant contributor. We aimed to determine the prevalence of discontinuation and nonpublication of randomized controlled trials (RCTs) on cervical cancer or precancer. Methods We searched ClinicalTrials.gov for registered RCTs investigating cervical cancer or precancer that started between January 2000 and December 2020. The primary and secondary outcomes were trial nonpublication and premature discontinuation, respectively. Publication status was determined by systematic searches of peer-reviewed journals using the PubMed and Scopus databases. Results A total of 113 RCTs met the inclusion criteria. Among the 85 trials completed before December 2020, 44 (51.8%) were prematurely discontinued and 40 (47.1%) were unpublished. A single-center design (61.4% vs. 34.1%, P = .012) and lack of external funding (59.1% vs. 36.6%, P = .038) were significantly associated with trial discontinuation. Large-scale (target sample size >400; 46.7% vs. 17.5%, P = .004) and externally funded trials (66.7% vs. 35.0%, P = .004) were more likely to be published. Multivariate logistic analysis revealed that a large sample size [odd ratio (OR): 4.125, 95% confidence interval (CI): 1.511–11.259, P = .006] and presence of external funding (OR: 3.714, 95% CI: 1.513–9.117, P = .004) were independent positive factors for trial publication. Conclusion A significant proportion of RCTs related to cervical cancer or precancer were discontinued early or remain unpublished, resulting in a waste of research resources.