Preregistration: A Key to Credible Real‐World Evidence Generation

ABSTRACT

Background

Preregistration of study protocols in a public repository promotes transparency and reproducibility in pharmacoepidemiological research. Despite its clear benefits, preregistration remains underutilized.

Purpose

Here, we discuss the advantages of preregistration, explore barriers to its implementation, and highlight existing repositories for preregistering real‐world evidence study protocols. A relatively new option is the Real‐World Evidence Registry within the Open Science Framework (OSF), which we briefly introduce.

Summary.

• Preregistration of study protocols improves transparency, accountability, and reproducibility in pharmacoepidemiology.

• Despite clear benefits, preregistration is still uncommon in real‐world evidence research.

• Public preregistration helps prevent selective reporting and strengthens trust in research findings.

• Multiple repositories exist for preregistration, including the new Real‐World Evidence Registry (RWE Registry) within the Open Science Framework (OSF).

• Greater adoption of preregistration could enhance the credibility of real‐world data studies.

1. Introduction

A key challenge facing pharmacoepidemiology is ensuring that real‐world evidence (RWE)—evidence derived from real‐world data (RWD) on drug use, safety, and effectiveness—reliably informs clinical and policy decisions. As misinformation spreads and public skepticism toward research grows, the mere presence of rigorous methods no longer guarantees trust. Making the entire research process transparent and reproducible is essential to uphold confidence in RWE and its impact on public health. A powerful approach to strengthening the credibility of RWE is through preregistration of study protocols—a practice that mitigates selective reporting and promotes transparency, accountability, and reproducibility. Yet, despite these recognized benefits, preregistration remains far from standard practice within pharmacoepidemiology.

With this commentary, we aim to advance transparent and reproducible workflows in pharmacoepidemiology and outcomes research by calling attention to the public preregistration of study protocols. Specifically, we discuss the benefits of preregistration in RWE research, explore barriers to its adoption, and highlight how existing tools can help researchers practice transparency effectively. Among these tools, we briefly introduce the Real‐World Evidence Registry (RWE Registry) within the Open Science Framework (OSF), a relatively new platform designed to support protocol preregistering.

1.1. Preregistration in RWE Research

Decades ago, clinical research took a major step forward when preregistering of clinical trial protocols became mandatory. This shift was more than just a procedural formality—it was a decisive move to restore trust and integrity in clinical research. Fueled by regulatory changes and a statement from the International Committee of Medical Journal Editors (ICMJE), preregistration of clinical trials became a prerequisite for publication in major medical journals [1]. Since then, clinical trial registrations have continued to grow, with the number of new registrations on ClinicalTrials.gov rising from 12,825 in 2009 to 22,131 in 2020 [2]. While clinical trial registration is now widely accepted as standard practice, the application of similar registration requirements to RWE studies remains a topic of debate.

RWE studies, which do not involve assigning participants to specific interventions, have historically been exempt from mandatory registration. Some argue that the benefits of preregistering observational studies, including RWE studies, are less clear. Key concerns against preregistration include bureaucratic delays, increased workload for researchers, and restriction on flexibility in exploring new ideas [2]. Additionally, researchers may fear intellectual property theft if hypotheses or methodologies are disclosed early in the research process [2]. Combined, these concerns contribute to a reluctance towards embracing preregistration as standard practice within RWE research.

However, to move our field forward, we must begin to reframe how we think about preregistration in RWE research. Rather than viewing it as an administrative requirement, preregistration must serve as a vital foundation for conducting robust research. For individual research teams, starting with a clear, preregistered protocol promotes team alignment, clarifies roles and responsibilities, and provides a structured foundation from which any necessary midstream changes can be transparently documented and justified. In this way, preregistration is not just a hallmark of good science—it also strengthens day‐to‐day project execution and team coordination. Beyond the individual research team, the benefits of preregistration extend to the wider scientific community. Firstly, preregistering RWE protocols can encourage reporting of null results, thereby reducing publication bias. Secondly, by defining study plans and analysis strategies upfront, preregistration enables comparison between planned and reported outcome measures and analyses, thereby mitigating risks such as selective reporting and data dredging. Collectively, a reduction in these biases bolsters the credibility of RWE research findings. Thirdly, preregistration of protocols fosters a culture of collaborative scrutiny within the RWE community. Specifically, it allows external partners—such as research colleagues, reviewers, and regulatory bodies—to track the research process and assess the study's overall validity once published. Also, when external partners can identify ongoing studies, the risk of duplication is minimized, and research gaps are more easily detected [2]. Finally, preregistration need not come at the expense of scientific flexibility—amendments are not only accepted but expected as part of the research process.

Encouragingly, a transition is already underway: RWE research is beginning to follow a trajectory similar to clinical research, driven by guidance from global regulatory bodies [3, 4] and health technology assessment (HTA) agencies [5] recommending preregistration for RWE studies informing policy, access, and reimbursement decisions. Also, a growing number of journals, including the Lancet family, are supporting the preregistration of RWE studies. Leading societies in the field, such as the International Society of Pharmacoepidemiology (ISPE) and the Professional Society for Health Economics and Outcomes Research (ISPOR), support and endorse preregistration of RWE studies. As a clear sign of their commitment, ISPE and ISPOR partnered with Duke‐Margolis in 2021 to establish the RWE Registry [6], a publicly accessible platform for researchers to preregister their RWE study protocols.

Nevertheless, because registration remains largely voluntary, a cultural shift within the research community is still needed. As RWE increasingly influences clinical and policy decisions, we argue that RWE research, like clinical research, must prioritize transparent and reproducible workflows. To facilitate this, several platforms now offer researchers practical ways to preregister their protocols, supporting broader adoption.

1.2. Options for RWE Study Protocol Preregistration

Currently, there are three main options for RWE study protocol preregistration (Table 1), each potentially suitable depending on regulatory obligations or other requirements. ClinicalTrials.gov is the most widely used and recognized platform [2], however, mainly as a platform for clinical trial registration. As it is not optimized for RWE studies, some metadata fields are irrelevant for most RWE registrations. Another option is the HMA‐EMA Catalogue of RWD sources and studies (replacing the previous ENCePP Resource Database and EU‐PAS Register). It focuses on tracking and cataloguing RWD sources and studies to increase transparency, promote collaboration among stakeholders, and support regulatory decision‐making as well as broader healthcare research. Rather than serving as a preregistration platform, it emphasizes metadata collection—such as study protocols, results, and data source details—to facilitate discoverability, enable meta‐studies, and support the use of RWD for safety and effectiveness evaluation and public health purposes. As such, certain metadata fields may be irrelevant if the sole purpose of use is preregistration of study protocols. Notably, the HMA‐EMA Catalogue links studies to their registering institutions and to the data sources, supporting accountability and reproducibility. Further, although the Catalogue did replace the ENCePP Resource Database, the ties with ENCePP are still maintained. ENCePP is coordinated by EMA, and its steering group includes representatives from key organizations like ISPE and ISPOR, alongside elected members from European research centers, ensuring broad scientific representation.

TABLE 1.

Comparative overview of options for RWE study protocol preregistration.

	RWE Registry	HMA‐EMA Catalogue [7]	ClinicalTrials.gov [8]
Purpose	Enhance transparency and reproducibility of RWE through study protocol preregistration.	Help find and use RWD sources studies for research and regulatory use.	Provide public access to clinical study information and ensure transparency.
Focus	RWE studies (e.g., hypothesis‐testing studies, particularly comparative effectiveness studies using RWD).	RWD sources and studies conducted by MAHs, regulators, academia, and research organizations, covering drug use, safety, and effectiveness.	Clinical studies (e.g., studies with prospectively collected data from enrolled participants who have provided informed consent).
Led by	ISPOR, ISPE, and Duke‐Margolis.	HMA and EMA.	United States National Library of Medicine.
Attribution for citable resources	DOI and URL.	URL and PIDs for the study, institution(s), and data source(s), including NCT number and EudraCT number.	DOI, URL, and NCT number.
Protocol requirement?	Yes – upload of study protocol is required, and there is no limit to the number of amendments that can be uploaded.	No – upload of study protocol is not required but encouraged, and there is no limit to the number of versions of the study protocol that can be uploaded.	Yes – upload of study protocol is required, and there is no limit to the number of amendments that can be uploaded.
Embargo feature?	Yes – the public release of the registration can be delayed.	No—a public registration is created upon submission.	No – a public registration is created upon submission.
Attestation regarding data handling prior to registration?	Yes – attestation for data handling before registration is required.	No – attestation for data handling before registration is optional.	No – attestation for data handling before registration is optional.

Abbreviations: DOI, digital object identifier; EMA, European Medicines Agency; EudraCT, European Union Drug Regulating Authorities Clinical Trials; HMA, Heads of Medicines Agencies; ISPE, International Society for Pharmacoepidemiology; ISPOR, Professional Society for Health Economics and Outcomes Research; MAHs, marketing authorization holders; NCT, National Clinical Trial; PIDs, persistent identifiers; RWD, real‐world data; RWE, real‐world evidence; URL, uniform resource locator.

By contrast, the RWE Registry is a preregistration platform, designed to require minimal metadata beyond protocol upload. Additionally, the platform includes a project wiki that can store associated project materials and automatically assigns citable DOIs for any code, data, paper, and supplemental material that the investigator chooses to link to the registration. This setup enhances research transparency and ensures proper attribution for associated resources.

Ultimately, the choice of platform for RWE study protocol preregistration depends on the specific goals of the study and regulatory requirements. As the RWE Registry may be less familiar to researchers, we highlight it here as a practical example of preregistration of RWE studies.

1.3. Protocol Registration the RWE Registry

In brief, the RWE Registry is hosted on the OSF platform and provides a structured way to preregister RWE study protocols using the RWE Recommended Minimum Study Registration Template. To preregister a study protocol, researchers log in, complete a brief metadata form, upload their protocol, and submit it for review. Each registration receives a citable digital object identifier (DOI).

To register amendments to the protocol, a revised protocol or an amendment can be uploaded to the associated OSF project, with the changes linked in an updated registration note.

An illustrative example is the preregistration for T he BMJ study “Risk of adverse events after covid‐19 in Danish children and adolescents and effectiveness of BNT162b2 in adolescents: cohort study,” [9] which includes both the original protocol and updated protocol versions (see osf.io/7ejh5 and osf.io/xv8pt).

For more technical guidance on registering a protocol, please refer to the resources available on the RWE Registry website.

2. Conclusion

With the growing focus on transparency and reproducibility in RWE research [10, 11, 12], it is worth considering what has driven the shift towards preregistration in other fields. For clinical trials, it took legislative changes and publication restrictions to adopt preregistration. The social sciences embraced preregistration in response to a reproducibility “crisis.” [13] The field of pharmacoepidemiology does not need to wait for either legislation or crisis. By adopting preregistration as a standard practice, we can build a culture of transparency, accountability, and reproducibility. With preregistration tools readily available and clear benefits such as improved study planning and reduced biases, there is an opportunity to enhance confidence in research findings, benefiting patients, policymakers, and the broader public health community.

2.1. Plain Language Summary

Research using real‐world data, such as information on how medicines are used and how well they work, is important for guiding healthcare decisions. However, trust in this type of research depends on making the process open and easy to verify. One way to do this is by “preregistering” a study plan—that is, publicly sharing what the researchers will do before the study begins. This helps prevent biased reporting and ensures that results are more reliable. Although preregistration has clear benefits, it is not yet widely used in research using real‐world data. This commentary explains why preregistration matters, the challenges researchers face in using it, and the tools available to make it easier. It also introduces a new online resource, the Real‐World Evidence Registry on the Open Science Framework, which allows researchers to preregister their studies and share them with others. By using preregistration more often, researchers can build greater public trust and reliably inform healthcare policy and practice.

Author Contributions

Emma Simonsen: conceptualization, writing – original draft. Shirley V. Wang: conceptualization, writing – original draft, writing – review and editing. Helene Kildegaard: conceptualization, writing – review and editing. Anton Pottegård: conceptualization, writing – review and editing, supervision.

Conflicts of Interest

Shirley V. Wang has been an ad hoc consultant to Cytec Inc., MITRE, a FFRDC for the Centers for Medicare and Medicaid Services, and the ACI Group for unrelated work. The remaining authors have nothing to report.

Simonsen E., Wang S. V., Kildegaard H., and Pottegård A., “Preregistration: A Key to Credible Real‐World Evidence Generation,” Pharmacoepidemiology and Drug Safety 34, no. 9 (2025): e70215, 10.1002/pds.70215.

Data Availability Statement

No data was used in the research described in the article.