he widespread availability of data in electronic health records (EHRs), patient registries, and administrative claims databases has ignited an evolutionary shift in life sciences, propelling the field towards a data-centric paradigm across the drug development and post-approval lifecycles. This shift, while touting new possibilities for medical innovation, introduces a complex set of challenges around data integrity, fitness-for-purpose, statistical and epidemiologic methods, patient privacy, data security, and institutional confidentiality.

The growing use of real-world data (RWD) for generation of reliable and robust real-world evidence (RWE) has opened vast avenues for research and clinical application but has also spotlighted the critical need for strong research community and regulatory frameworks. Clinical researchers, regulatory agencies, and other stakeholders, especially those with limited experience or expertise in RWD, find themselves navigating terrain where the abundance of data must be matched with data expertise and methodological rigor, exposing the urgent necessity to address existing knowledge gaps and anticipate future needs.

The goal of this article is to explore the transformative impact of RWD on the life sciences, emphasizing the challenges and opportunities from the shift towards a data-centric paradigm in drug development and post-approval processes. By examining the expanding role of RWD in generating RWE, this article seeks to foster collaboration among researchers, regulatory agencies, and other stakeholders, providing actionable approaches to navigate the complexities of RWD and RWE. It aims to promote methodological rigor, enhance the use of RWD for regulatory decision-making, and address critical gaps in research frameworks, thus ensuring data integrity, protecting patient privacy, and advancing medical knowledge.

Regulatory Use Cases and Challenges in the RWD Era

The integration of RWD into the drug development continuum presents a formidable challenge for regulatory bodies worldwide. Agencies like the US FDA, EMA, and Japan’s PMDA are spearheading efforts to harness the insights RWD offers into treatment outcomes and patient experiences in real-world settings. Despite the promise and multiple guidances from international regulatory agencies to make use of RWD in regulatory decisions, the heterogeneity of RWD sources coupled with a wide range of use cases and inconsistent acceptance of RWD to support decisions poses significant hurdles in gaining acceptance for use in regulatory decisions.

While current regulatory guidance lays the groundwork for integrating RWD into the regulatory framework, the use of RWD and RWE for regulatory decisions, especially medical product approvals and labeling expansion, is still limited. The framework and recommendations outlined in regulatory guidance are often inconsistent with the real-world reality around data completeness, availability, and data collection consistency, and fail to appreciate the complexities of how to use RWD to generate robust evidence. For instance, stringent data anonymization norms across the European Union complicate researchers’ ability to access and utilize patient-level data effectively, illustrating the regional disparities and uncovering the need for an international consensus on RWD management. These data restrictions are starting to appear within the US as state-to-state variation in data privacy laws.

As another example, natural history studies using RWD include data collection intervals based on real-world medical treatment patterns and will not necessarily align with clinical trial visit intervals, thereby requiring methodology and potentially imputation to make direct comparisons to trial data. For example, estimating body mass index (BMI) on a specific date (e.g., 180 days after an index date) to allow for comparison to a trial measurement date may require estimation based on the trend of BMI measurements before and after the target measurement date. Uncertainty in timing of real-world index dates (e.g., the time of treatment failure) may require methods that assign index dates based on an underlying distribution derived from internal or external sources. Further, use of RWD as a comparator for an interventional study raises questions about whether the RWD represent patients who are similar to, sicker than, or healthier than the trial population, questions that can derail a study when posed outside of a clinical framework that allows for uncertainty and the balancing of benefits and risks. These differences between RWD expectations and reality can lead to regulatory hesitancy toward or rejection of the use of RWD for regulatory decisions when the data are likely fit-for-purpose.

Filling the Regulatory Gaps

RWD guidance developed by worldwide regulatory agencies has laid the groundwork for increased use of RWD for regulatory decision-making, but work is needed to realize its full promise. The core recommendations around establishing standard reproducible research protocols, including comprehensive data source evaluations and characterizations, fitness-for-purpose assessments, promoting the pre-registration of RWD studies, appropriate methodological approaches and reporting, and other community and regulatory recommendations are a necessary but not sufficient condition for expanding the use of RWD. Beyond these needs, the expanded acceptance of RWD for regulatory decisions will require investment in training and knowledge transfer between the RWD research community and the regulatory community, and an acknowledgement that RWD has a critical place in regulatory decision-making and ought to be used to promote public health despite the inherent uncertainties as compared to clinical trial data. For example, the FDA Sentinel Program, the Duke-Margolis RWE Collaborative, and the Reagan-Udall Foundation for the FDA regularly hold public workshops and trainings on the use of RWD for regulatory decision-making. Focusing only on the limitations of RWD instead of assessing the balance between benefits and risks within the context of each specific situation will stifle innovation and harm patients.

The Path Ahead

Growing use and acceptance of RWD for the generation of RWE for regulatory decisions will require an all-of-the-above approach. Substantial progress has been made by FDA, EMA, and others through the issuance of multiple RWD guidances focusing on different topics such as data standards, submission expectations, and fitness-for-purpose assessment of data sources, and via publications on expectations and lessons learned. In addition, the research community has developed standards for using RWD, and other stakeholders regularly hold educational seminars on the topic. But more needs to be done to transition from the promise of RWD into action to improve the health of patients. RWD and the generation of RWE exist within an environment of the uncertainty of “known unknowns” paired with “unknown unknowns.”

To address uncertainty around the use of RWD, generators of RWE for regulatory decisions—primarily regulated industry and academic researchers—and the regulatory agencies should have similar levels of expertise when conducting and interpreting RWD studies. While it is largely acknowledged that RWD studies should be conducted by cross-functional research teams with expertise in and understanding of clinical, data, and methodological nuances of the research topic and the conduct of high-quality RWD research, the regulatory agency staff tasked with acceptance and interpretation of the RWE should also have similar expertise. Matching the expertise of the RWE generator and RWE reviewer will help foster an environment of mutual understanding and effective communication, enabling all stakeholders to focus on whether the RWE is robust enough for the intended use. Appropriate use of RWD to generate RWE requires matching fit-for-purpose data to the question, to the method, and to the intended use. The key challenge is to find a path to using RWD to improve public health by balancing the health benefits and potential harms in each situation and for each patient population. A continued effort to improve knowledge and expertise of researchers, regulators, and other stakeholders regarding how to best use RWD to generate robust RWE is necessary for achievement of the promise of RWD to improve health.

Conclusion

The effective integration of RWD into regulatory frameworks demands enhanced collaboration and continuous refinement of methodologies. The growing use of aggregated EHR data presents opportunities and challenges that will require a commitment by stakeholders to foster a robust research community to address issues around data quality, data completeness, methods, and knowledge gaps in the appropriate use of these data for decision-making. These efforts can shape global regulatory practices and drive meaningful advancements in clinical research. Regulatory bodies and stakeholders need to establish standardized approaches and foster an environment conducive to knowledge sharing and transparency. Ultimately, leveraging RWD effectively can transform drug development and regulatory decisions, significantly improving patient outcomes and the health of the public.