Real world evidence (RWE) is gaining recognition as a valued complement to evidence generated through randomized clinical trials (RCT). Today RWE is being used to evaluate the generalizability of trial results to subgroups underrepresented in those trials, to quantify medical product benefits and risks in everyday care settings, and to provide context for single arm trials. Regulators are taking notice.
 
RWE’s Growth in New Drug Approvals and Label Expansions
RWE has long been used for post-authorization safety research and as historical controls for situations when randomization is not feasible, such as for studies of rare diseases. RWE is also being called in when it is difficult to conduct an RCT, such as for studies of a new life-sparing treatment. In these situations, patients and their physicians are understandably reluctant to leave their treatment decisions up to randomization.
 
RWE is also being used to study the long-term benefits and risks of one-time treatments, such as gene therapy and treatments for chronic diseases that may be used for decades.
 
Sometimes it can be as simple as updating a label to address changes in treatment patterns. For example, a label update was given regulatory approval recently in China, based exclusively on RWE. In this situation, a chemotherapeutic agent was first approved in combination with chemotherapies that were no longer in use several years later. RWE was used to show that the treatment was still effective with newer combinations of chemotherapies.
 
RWE for Regulatory Submissions
Researchers who deftly employ RWE may be able to present a more convincing evidence package for regulatory review. For example, RWE can be used to examine subgroups of special interest, both to identify and describe them, and as benchmarks for treatment effectiveness. RWE can also sometimes be used to quantify the safety of treatments used off-label, to the extent such data are available.
 
For example, palbociclib was approved for women with advanced or metastatic breast cancer and specific types of tumors but use by men was widespread and supported by treatment guidelines. Researchers were able to generate RWE from health insurance claims data and EHRs, which documented safety among men using palbociclib, and was then used successfully to support a label expansion to men.
 
RWE can also incorporate treatment randomization. Pragmatic randomized trials are distinguished from traditional RCT in part by their use of endpoints that are meaningful for both patients and clinicians and also by their comparisons to other available treatments (not placebos), which help inform treatment decisions in real-world settings. For example, the Paliperidone Palmitate Research in Demonstrating Effectiveness (PRIDE) was a pragmatic trial that studied the effectiveness of a once-per-month injectable antipsychotic medication compared to daily oral antipsychotics in treating adults with schizophrenia who had recently attempted suicide, had been incarcerated or had a psychiatric hospitalization, among other factors that were used to characterize high risk patients. The trial found that the injectable antipsychotic had a longer time to treatment failure than daily oral treatments, resulting in a modification to the product label.
 
Successful use of RWE requires careful attention to both study design and analytic methods, along with a good understanding of data provenance — how and why the data were created and what factors may influence the likelihood of important information being recorded– and data curation. The bulk of nonrandomized RWE is derived from existing healthcare data, including patient registries, health insurance claims, data contained in electronic health record (EHR) systems, and more recently, smartphone apps and other consumer devices. With growing use of wearable devices and a greater ability to link various other types of RWD,, studies may require little or no direct interaction with clinicians or patients after informed consent, making the process faster and more cost-effective to execute.
 
The Road to a More Holistic Regulatory Submission
U.S. regulators are currently working on guidance around consideration of RWE in submissions. The newly drafted CURES 2.0 Act includes provisions that would help expand the use of RWE. Similarly, the European Medicines Agency and UK regulatory authority are both evaluating how RWE could factor into their own regulatory decisions, as are regulators in Japan, China, and other countries. In particular, the UK has shown recent success in using RWE to aid its response to COVID-19 and the virus’ corresponding treatments and vaccines, and as a steady gauge of infection levels.
 
While RWE will never replace traditional randomized clinical trials, nor should they, RWE offers powerful evidence that can be instrumental in filling knowledge gaps that have limited understanding about the generalizability of research findings to subgroups not studied in those trials. There is still considerable work to be done to establish new standards and foster clarity about the validity, reliability, and utility of RWE. Nonetheless, examples of successful regulatory submissions that use RWE are accumulating rapidly, which builds a favorable outlook for expanded reliance on RWE.