Evaluating the Reproducibility of Research in Obstetrics and Gynecology

TL;DR

This study assesses the reproducibility of 300 Obstetrics and Gynecology trials, revealing significant deficiencies in data sharing, protocol availability, and conflict of interest disclosures, which threaten research integrity and clinical decision-making.

Contribution

It provides an objective evaluation of reproducibility indicators in Obstetrics and Gynecology research, highlighting critical gaps and the need for standardized practices.

Findings

Only 10.58% of trials disclosed data availability.

None of the trials provided access to protocols.

Over 54% did not disclose funding sources.

Abstract

Objective: Reproducibility is a core tenet of scientific research. A reproducible study is one where the results can be recreated by different investigators in different circumstances using the same methodology and materials. Unfortunately, reproducibility is not a standard to which the majority of research is currently adherent. Methods: We objectively evaluated 300 trials in the field of Obstetrics and Gynecology for fourteen indicators of reproducibility. These indicators include availability of data, analysis scripts, pre-registration information, study protocols and whether or not the study was available via Open Access. We also assessed the trials for financial conflict of interest statements and source of funding. Results: Of the 300 trials in our sample, 208 contained empirical data that could be assessed for reproducibility. None of the trials in our sample provided a link to their protocols or provided a statement on availability of materials. None were replication studies. Just 10.58% provided a statement regarding their data availability, while only 5.82% provided a statement on preregistration. 25.85% failed to report the presence or absence of conflicts of interest and 54.08% did not state the origin of their funding. Conclusion: Research in the field of Obstetrics and Gynecology is not consistently reproducible and frequently lacks conflict of interest disclosure. Consequences of this could be far-reaching and include increased research waste, widespread acceptance of misleading results and erroneous conclusions guiding clinical decision-making.