Realizing the promise of RWE: A need for greater regulatory clarity

An Expert View from Mike D’Ambrosio (pictured above, left, with Tala Fakhouri), Senior Vice President and Global Head, Real World Research, Parexel, and Tala Fakhouri, Ph.D., M.P.H., Chief AI and Regulatory Strategy Officer, Parexel.

As the quality of real-world data (RWD) improves and its availability expands, we are increasingly seeing the use of real-world evidence (RWE) to help demonstrate the safety, effectiveness and value of therapies. Eight years into the US Food and Drug Administration's (FDA) framework for RWE evaluation, however, ambiguity still exists.

Sponsors either risk investing in studies that may be ultimately deemed unsuitable for regulatory decision-making or avoid pursuing RWE opportunities altogether, potentially delaying evidence generation and patient access to new therapies.

In both scenarios, patients lose out. Yet neither outcome is inevitable. With industry colleagues, we recently published an article in Therapeutic Innovation & Regulatory Science ¹ examining several operational and policy barriers that continue to limit RWE adoption in regulatory decision-making and outlining ways the FDA could help bridge those gaps. While methodological questions surrounding RWE have received significant attention, the industry's next challenge may be more practical: ensuring that regulatory frameworks keep pace with the realities of modern healthcare data. Without greater regulatory clarity, valuable evidence may remain underutilized, limiting opportunities to bring new insights and therapies to patients more efficiently.

Aligning data submission requirements with RWD realities

One of the most significant barriers facing sponsors today is that regulatory submission standards were largely developed for protocol-driven randomized controlled trials (RCTs), while RWD originates from diverse healthcare environments and may include structured and unstructured data as well as derived variables generated through complex transformations. Converting these datasets into FDA-accepted formats often requires extensive manual effort that can obscure provenance, introduce bias and lead to the loss of clinically relevant information.

Recognizing this challenge, in April 2025 the FDA's Drug and Biologics Centers opened a public docket to explore Health Level Seven International Fast Healthcare Interoperability Resources (HL7 FHIR) to support submission of structured clinical study data created from RWD sources. Feedback from respondents identified several potential benefits, including improved interoperability and reduced data processing burden. At the same time, substantial challenges were identified, including the absence of established analysis-ready data structures, variability in FHIR implementations and limited workflow metadata to support provenance assessment.

While FHIR can reduce the burden associated with data extraction and transfer, it was not designed to produce analysis-ready regulatory datasets. The FDA could therefore evaluate FHIR as one component of a broader modernization effort rather than a standalone solution. As part of that effort, the FDA could consider adding HL7 FHIR to its Data Standards Catalog and building on The Clinical Data Interchange Standards Consortium's (CDISC) existing FHIR-to-SDTM mapping work to support more efficient transformation of healthcare data into regulatory submission formats. Emerging technologies, including transparent and auditable AI-enabled solutions, may also help harmonize heterogeneous RWD sources and map them into FDA-accepted standards while preserving data integrity and traceability.

In parallel, the FDA could make greater use of its existing waiver authority to exempt certain RWD submissions from standard format conversion requirements and continue modernizing its own review infrastructure to support data that remain closer to their native formats. Such changes could reduce unnecessary operational burden while preserving rigorous scientific review.

Clarifying expectations for data access, verification and regulatory review

In 2023 RWE guidance, the FDA states that reviewers expect access to patient-level data sufficient to replicate study analyses and evaluate the reliability of submitted evidence. At the same time, many of today's most valuable RWD assets are maintained within secure environments operated by registries, health systems and third-party data providers. Privacy requirements, contractual obligations and institutional policies can make traditional approaches to data submission difficult or, in some cases, impractical. As a result, sponsors may be unable to incorporate otherwise valuable datasets into regulatory evidence packages.

In December 2025, the FDA announced that identifiable patient-level data would not always be required for certain medical device submissions and indicated that similar approaches could be considered for drugs and biologics. This presents an important opportunity to provide greater clarity regarding when aggregated data may be acceptable, what forms of source verification are appropriate, which records should be considered acceptable source documentation and what information reviewers need to understand how RWD were collected, transformed and analyzed. It also raises the open question of whether the FDA may ultimately extend similar least-burdensome principles beyond medical devices to drug and biologic submissions.

More broadly, regulators and industry should explore approaches that allow patient-level information to remain within secure environments while still enabling rigorous regulatory review. Federated data networks such as the FDA's Sentinel System and Data Analysis and Real World Interrogation Network (DARWIN EU) demonstrate that privacy protection, traceability and robust analyses can coexist. These models may offer useful lessons as the industry works to expand the role of RWE in regulatory decision-making.

Finally, additional clarity may be needed beyond the FDA. Questions regarding how privacy laws affect the sharing and review of patient-level information often extend into broader Health and Human Services policy considerations. Greater alignment across regulators and policymakers would help sponsors, data providers and health systems navigate these requirements with greater confidence.

Specifying how consent and IRB oversight apply to non-interventional studies

Under current FDA regulations, clinical investigations intended to support safety or effectiveness determinations must comply with 21 CFR Parts 50 and 56, which require informed consent and oversight by an institutional review board (IRB). While the FDA has provided guidance on the use of RWE, questions remain regarding how these requirements should be applied across different types of non-interventional research.

Non-interventional studies exist along a continuum. Some involve retrospective analyses of de-identified data with no additional data collection. Others rely on data generated prospectively through routine clinical care without protocol-driven interaction with patients. Still others may include questionnaires, laboratory testing or other protocol-specified activities. These designs present different levels of participant interaction and risk, yet sponsors often face uncertainty regarding the regulatory expectations that apply to them.

Many RWE studies are conducted using data collected under broad consent frameworks, Health Insurance Portability and Accountability Act (HIPAA) authorizations or approved waivers that differ from traditional clinical trial consent processes. As the use of RWE expands, additional clarity regarding whether retrospective secondary analyses of de-identified data require application of Parts 50 and 56, while providing explicit expectations regarding IRB review, informed consent, waivers and modifications for studies using prospectively generated RWD and for studies involving protocol-specified activities may further support implementation. Such guidance would help sponsors design studies more efficiently while ensuring that participant protections remain aligned with the level of risk involved.

Turning regulatory clarity into patient benefit

Underutilization of RWE in drug approvals is detrimental but addressable. By encouraging adoption of fit-for-purpose data standards, clarifying expectations for patient-level data access and seeking explicit guidance on patient protections, sponsors and their partners can help increase confidence in the use of RWE across clinical development.

The result would be more efficient evidence generation, greater use of data from rare diseases and underserved populations and a clearer pathway for translating real-world insights into regulatory decision-making. Ultimately, patients stand to benefit when valuable evidence can be generated, evaluated and applied with greater confidence and consistency.

References

1. Fakhouri, T.H., Kalamegham, R., Mahoney, N. et al. Real-World Evidence for Drug Approvals: Moving Toward Greater Regulatory Clarity. Ther Innov Regul Sci (2026). https://doi.org/10.1007/s43441-026-01000-3