Tiered framework proposed to match clinical trial data sharing to scientific purpose

Clinical trial data sharing is increasingly endorsed by journals, funders, regulators and research organisations, yet access to individual participant data remains inconsistent and is often unavailable when trial results are published. A new debate article in BMC Medicine argues that transparency efforts should be matched to the scientific purpose a given data object can actually serve, and proposes a three-tier framework for data that trialists or data custodians intentionally release.

The authors, led by Ulrich Mansmann with Ka Hin Tai, Giulia Varvarà and Florian Naudet, note that delays in data access reflect several distinct components, including scientific and governance review, contractual and legal procedures, data preparation, de-identification, transfer and subsequent harmonisation. When original data are unavailable, researchers sometimes reconstruct approximate participant-level event-time data from published Kaplan-Meier curves, an approach the article treats separately because such pseudo-data are generated externally from publications rather than shared from the trial database itself.

The framework’s first tier comprises original publication-time reduced datasets. According to the authors, these support endpoint-level numerical reproducibility and analyses that are functions of the released observations. The second tier consists of reproducibility packages containing analysis-ready data, code, derivation logic and documentation sufficient to reproduce the reported analytic process. The third tier provides governed access to broader de-identified individual participant data, enabling substantively new analyses and reassessment of operational and analytic decisions.

A key point of the argument is that these tiers should not be confused with one another. Endpoint-level numerical reproducibility, the authors write, is not the same as reproduction of the complete analytic process, and neither equals broader scientific reuse. Privacy protection, governance and the processing environment are described as cross-cutting dimensions that do not map uniquely to any single tier and apply differently to each.

The proposal responds to a recent suggestion by Trinquart and Stockler that trialists release minimal analysable time-to-event data at publication. The authors frame that proposal as raising a broader question: what kinds of data and documentation are required for different forms of reproducibility and scientific reuse? Their answer is a graduated model in which each tier supports a distinct class of scientific activity, from checking published numbers to re-running full analysis pipelines to conducting entirely new research on the underlying data.

In their conclusions, the authors argue that publication-time reduced datasets can provide meaningful and timely transparency when broader individual participant data are not yet accessible. They propose that such datasets should form a practical floor for timely transparency, while reproducibility packages and governed access to broader data offer progressively richer opportunities for reproducing, interrogating and extending clinical trial analyses. The value of reduced datasets, they caution, should be recognised without overstating what they can demonstrate.

Subject of Research: A proposed three-tier framework for clinical trial data transparency matching data access to scientific purpose

Article Title: Clinical trial data transparency should match data access to scientific purpose: a tiered framework for reproducibility and reuse

Article References: Mansmann, U., Tai, K. H., Varvarà, G., & Naudet, F. (2026). Clinical trial data transparency should match data access to scientific purpose: a tiered framework for reproducibility and reuse. BMC Medicine, 24(1), Article 588. https://doi.org/10.1186/s12916-026-05349-y

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05349-y

Keywords: clinical trials, data sharing, individual participant data, reproducibility, data transparency, Kaplan-Meier reconstruction, governance, de-identification, BMC Medicine, research methodology, Clinical, trial