Setting a New Pace: Real-World Evidence is Finding Its Footing in HEOR

By Beth Fand Incollingo

In 2021, something unprecedented occurred in the pharmaceutical arena—the US Food and Drug Administration (FDA) approved a medication based exclusively on retrospective real-world data (RWD), without any evidence from clinical trials.

Vijoice (alpelisib) had been administered on a compassionate-use basis to patients with severe symptoms of PIK3CA-related overgrowth spectrum, and a study of data from their charts led to the FDA’s accelerated approval.¹

That protocol will be less likely to raise eyebrows in the future, as stakeholders are increasingly basing their healthcare decisions on real-world evidence (RWE), which is generated from routinely collected clinical data, including treatments, prescription patterns, patient behaviors, and healthcare outcomes.

Collected from medical claims databases, electronic health records, wearable health monitors, patient-reported outcomes, and product, patient, and disease registries, RWD and the RWE it generates comprise a growing component of health economics and outcomes research (HEOR) worldwide. RWE is being used to help boost prevention efforts, identify patients who are at risk for illness or eligible for clinical trials, shape clinical guidelines, determine the value of medical interventions, establish reimbursement strategies, expand drug safety testing, create public health policy, and support regulatory reviews.

Compared with the data used in clinical trials, RWD can be collected and studied more quickly and affordably, can more realistically demonstrate usage patterns and health outcomes, and can represent larger and more diverse populations. To help pave the way, regulatory authorities in many countries have created RWE guidelines, and ISPOR is also pursuing initiatives—from hosting summits to promoting transparent study design—with the goal of effectively applying RWE to healthcare decision making.² Still, some experts are skeptical about the quality of RWE because nonrandomized studies can leave room for bias or erroneously interpret chance patterns as causal relationships.² Stakeholders also grapple with how best to protect patients’ privacy when using or sharing their health data. To help pave the way, regulatory authorities in many countries have created RWE guidelines, and ISPOR is also pursuing initiatives—from hosting summits to promoting transparent study design—to effectively apply RWE to healthcare decision making.²

“As data become richer and richer, the potential for informing healthcare decisions using real-world evidence will continue to grow,” said Mark McClellan, MD, PhD, director of Duke University’s private, nonprofit Margolis Institute for Health Policy in Washington, DC, which takes a multidisciplinary approach to developing healthcare-related policy solutions. “What’s it’s really enabling is more precise, relevant, timely information on questions that matter to patients but are hard to address using traditional clinical trials.”

“What RWD is really enabling is more precise, relevant, timely information on questions that matter to patients but are hard to address using traditional clinical trials.”— Mark McClellan, MD, PhD

Who’s Compiling RWE?

Internationally, databases that track patients’ health journeys and fuel retrospective and prospective studies are being hosted by a variety of sponsors, including:

• Public entities. The United Kingdom’s Clinical Practice Research Datalink has collected data on about 60 million patients from general practitioners and supported more than 3500 peer-reviewed publications, and the All of Us program run by America’s National Institutes of Health (NIH)—which tracks genomics and health-record information for nearly 850,000 people—has led to 15,000 studies. Meanwhile, the FDA’s Sentinel Initiative, sourced from payer records, is collecting data about millions of patients as a means of monitoring medical product safety.
• Nonprofit efforts. One example is the Margolis Institute’s PCORnet Common Data Model, which is compiling data from US health encounters with more than 47 million people.
• Public/private collaborations, such as Canada’s effort by GEMINI and ICES, focused on gathering information about patients’ health journeys from 35 Ontario hospitals.³
• Private initiatives. Numerous private organizations, including health insurance companies and consulting firms, are actively engaged in RWD collection and research initiatives. A notable example is IQVIA’s real-world database, which encompasses healthcare information on approximately 1.2 billion patients across more than 60 countries. Another prominent example is the Premier Healthcare Database, owned by Premier, a technology-driven healthcare improvement company that collaborates with over two-thirds of US healthcare providers.
  

Premier’s database is a key example of the kinds of practice-changing breakthroughs that can be generated using RWE from deidentified patients, whose health information can legally be studied if they have consented to care under HIPAA, and without approval from institutional review boards.

The Premier Healthcare Database contains electronic medical records and chargemaster billing data from over 1400 US hospitals and 300 healthcare systems. Dating back to 2000, it includes time-stamped service records encompassing diagnoses, treatments, medications, devices, demographics, and provider information from approximately 25% of all US inpatient visits and over 40% of all US hospital-based outpatient visits—more than 1.5 billion encounters in all. To enable the longitudinal tracking of individual health journeys, each patient is assigned a unique identifier.⁴

About one-third of participating hospitals contribute laboratory data, helping to define clinical outcomes, and the database also tracks costs and charges, making it possible to estimate the economic burden of specific diseases and procedures, said Ning Rosenthal, MD, PhD, MPH, assistant vice president of applied research at Premier. To get a broader picture of patients’ treatment journeys across care settings, she added, information from the Premier Healthcare Database can be tokenized and linked to external sources such as medical and pharmacy claim repositories.

“After the 21st Century Cures Act was passed, the FDA was asked by Congress to speed up the drug-approval process by leveraging real-world evidence.”— Ning Rosenthal, MD, PhD, MPH

What Can RWE Accomplish?

For providers and payers, Rosenthal said, this database is “a gold mine.” Information from Premier’s database has fueled more than 1200 peer-reviewed publications, including studies on the cost, effectiveness, safety, and outcomes of interventions across a host of medical specialties.⁵

Stakeholders licensing the Premier Healthcare Database have included US regulatory agencies such as the FDA, Centers for Disease Control and Prevention, and Department of Health and Human Services, as well as private healthcare, pharmaceutical, and medical device companies. And, of course, Premier’s experts conduct their own studies.

“Because our database is both large and nationally representative,” Rosenthal said, “it can be leveraged to develop predictive models that support earlier diagnosis and treatment decisions by clinicians—ultimately improving patient outcomes.”

Rosenthal described a recent real-world study in which investigators developed an algorithm using laboratory data from the database to find over 400,000 patients who had presented to participating health systems with symptoms of chronic kidney disease but never received a diagnosis. Now, Premier is collaborating with hospitals to improve chronic kidney disease diagnosis and follow-up care—a move expected to save not only lives but dollars, as preventable dialysis and hospitalizations tend to be costly in this population.⁶

With its healthcare database, Premier is also boosting clinical study efficiency by enabling the selection of trial sites based on their access to eligible patients, Rosenthal said. Another key use for the healthcare database is in studies supporting drug approval or label expansion.

“After the 21st Century Cures Act⁷ was passed in the United States in 2016, the FDA was asked by Congress to speed up the drug-approval process by leveraging real-world evidence,” Rosenthal said. “Since then, the FDA has released guidelines for conducting studies using real-world data to support regulatory submissions.”

That trend made headlines in 2021, due not only to Vijoice’s approval but to the use of RWE as primary evidence to support an indication expansion for the drug tacrolimus. In that case, data from an observational study arm—compared against historical controls—supported the drug’s indication for the prevention of organ rejection in lung transplants.⁸

Wearable health monitors are another way to collect RWD—a practice common in China, where information about sleep quality, glucose levels, heart rate, and blood pressure can be gathered by the public health system and sometimes used in studies.

But in America, wearable monitor tracking and research are lagging—as is the collection and application of health information pulled from social media. While All of Us⁹ has fueled studies¹⁰ by collecting Fitbit data from 60,000 patients, Premier has not gathered much information from wearables because a lot of it is proprietary to device vendors, Rosenthal said.

To help the United States catch up, Margolis convened a working group that made protocol recommendations, including creating a research community to publish standards and employing wearables to enroll patients in real-world studies and secure their consent,¹¹ said Rachele Hendricks-Sturrup, DHSc, MSc, MA, who leads the Institute’s RWE Collaborative.

Real-world studies have proven to be more streamlined, reducing the path to drug approval from as long as 5 years to just 1 year.— Xin Sun, PhD

Leading the Way in China

Another key test of RWE is happening in the Boao Lecheng International Medical Tourism Pilot Zone, where initiatives include treating seriously ill patients with interventions that have been approved in other countries, but not yet in China. This enables researchers to track the real-world safety and health outcomes associated with novel products as an alternative path to Chinese regulatory approval.

“For most of these products, no trials will be conducted in China,” said Xin Sun, PhD, a professor and director of clinical epidemiology at West China Hospital’s Evidence-Based Medicine Center. “This process will be like a bridge, enabling our regulatory authorities to consider trial evidence from other countries along with real-world evidence from the Chinese population.”

Compared with clinical trials, Sun said, real-world studies have proven to be more streamlined, reducing the path to drug approval from as long as 5 years to just 1 year.

As of December 2024, 40 products approved in other countries had undergone real-world study in Boao Lecheng, with 17 receiving regulatory approval.¹² However, many of those have not yet been deemed reimbursable, Sun said, as they’ll need to accrue more cost-effectiveness data first.

Involved in much of that research, Sun’s team is also using RWE both inside and outside the pilot zone to conduct post-market studies of treatments already approved in China. The team is also generating RWE to support regulatory applications for herbal medicines, which are heavily used in Chinese clinical practice.

“For investigational herbal treatments,” Sun said, “real-world research could eventually take the place of phase II clinical trials.”

Overcoming Barriers

In a 5-year plan, Margolis’s Collaborative has outlined 4 goals for real-world study: to incorporate reliable and relevant RWE into regulatory decisions, payer evaluations, and routine care while helping to make it transferable across regions and countries.¹³

Yet, accomplishing that will mean overcoming significant barriers, Sun said, including the varying comprehensiveness of RWD across medical specialties and the inability for researchers to follow up with patients when working with deidentified data.

Real-world study design can also raise concerns, Rosenthal said. “In head-to-head studies, the populations treated with drug A and those with drug B can be quite different, since patient characteristics may affect doctors’ prescribing patterns,” she said. “The good news is that we can use a propensity score method to create a matched sample, along with multivariable regression methods to control known confounders, making real-world studies robust and, in some cases, not inferior to clinical trials.”

Margolis aims to support the practice of leveraging observational RWD in causal inference studies in their published white paper,¹⁴ and has conveyed tools that can be used in practice to vet the relevance, reliability, and quality of real-world data,¹⁵ Hendricks-Sturrup said.

Still, other obstacles remain. “Fragmented healthcare and fee-for-service payments, common in the United States, don’t lend themselves to reliable longitudinal tracking,” McClellan said. “Fortunately, that is changing because of healthcare reforms that focus payments more on patient outcomes.”

Yet another challenge is that healthcare providers—from hospitals to national healthcare systems—don’t often work together to gather RWD, and they typically don’t share it with each other. Usually, that’s due to differing rules governing ethics and privacy.

“Given that we’re dealing with data and technology that can range from being highly regulated to not regulated at all, we have to identify best practices, frameworks, tools, and principles that can guide our decision making.”— Rachele Hendricks-Sturrup, DHSc, MSc, MA

To guide the playing field, regulatory authorities including those in the United States, Europe, the United Kingdom, Canada, and China have published guidelines on data quality and relevance, registries, and external control arms. Margolis’s Collaborative compiles those guidelines online so it’s easier for stakeholders to compare them,¹⁶ Hendricks-Sturrup said.

ISPOR has been supporting such efforts by publishing best-practices documents and hosting summits designed to help standardize RWE methodologies, including one planned for September 2025 that will highlight the Asia-Pacific region and feature Sun as a panelist. ISPOR has also created a tool that helps decision makers evaluate RWE-driven comparative-effectiveness research.

In a combined effort, ISPOR, Margolis, and other partners have created a Real-World Transparency Initiative and a related registry through which scientists can share their study designs to ensure that their tests will follow a prespecified analytic protocol.

“Given that we’re dealing with data and technology that can range from being highly regulated to not regulated at all, we have to identify best practices, frameworks, tools, and principles that can guide our decision making on a case-by-case basis,” Hendricks-Sturrup said.

Preparing for Success

Despite challenges to its implementation, RWE is bringing expediency and affordability to healthcare decisions, and that could be especially helpful as new challenges arise—for example, slashed funding for scientific research and healthcare regulation in the United States.¹⁷

McClellan, who served as former commissioner of both the FDA and the Centers for Medicare & Medicaid Services (CMS), noted that “there’s a lot going on at CMS around advancing real-world data interoperability to support better patient-care decisions, such as protocols to slow the progression of diabetes or coronary artery disease.”

He added that the FDA can turn to RWE¹⁸ to resolve critical epidemiology questions, such as which subpopulations should be eligible to receive specific vaccines.

For the research community to fully embrace RWE, education about the process will be needed, Sun said.

“In China, we have polarized opinions about real-world evidence, with some people treating it very seriously while others do not,” he said. “It all depends on how the evidence is interpreted, and that means that training efforts will be really important.”

Sun added that RWE can only have a global impact if communication improves between researchers in high-income countries, who are using the strategy, and those in the developing world, where its applications remain limited.

That’s my dream,” Sun said. “It will be difficult, but filling those gaps is worth doing.”

References:

1. Drugs.com. FDA Approves Vijoice. Published April 6, 2022. Accessed July 14, 2025. https://www.drugs.com/newdrugs/fda-approves-vijoice-alpelisib-pik3ca-related-overgrowth-spectrum-pros-5817.html.
2. ISPOR. Real World Evidence. Accessed July 14, 2025. https://www.ispor.org/strategic-initiatives/real-world-evidence.
3. IC/ES. New data partnership to expand insights on hospital care in Ontario. Published June 26, 2025. Accessed July 14, 2025. https://www.ices.on.ca/announcements-and-events/new-data-partnership-to-expand-insights-on-hospital-care-in-ontario.
4. Premier Applied Sciences. Premier Healthcare Database: Data That Informs and Performs. Published December 2024. Accessed July 14, 2025. https://offers.premierinc.com/rs/381-NBB-525/images/PremierHealthcareDatabaseWhitepaper.pdf.
5. Premier. Peer-reviewed Publications. Published June 2025. Accessed July 14, 2025. C:/Users/16093/Downloads/Peer-Reviewed-Journal-Publications_Premier-Authors_2000-June2025.pdf.
6. Premier. Premier, Inc. Announces Cross-Industry Initiative to Improve Diagnosis and Care for Chronic Kidney Disease. Published May 8, 2024. Accessed July 14, 2025. https://premierinc.com/newsroom/press-releases/premier-inc-announces-cross-industry-initiative-to-improve-diagnosis-and-care-for-chronic-kidney-disease.
7. National Institutes of Health. The 21st Century Cures Act. Accessed July 14, 2025. https://www.nih.gov/research-training/medical-research-initiatives/21st-century-cures-act.
8. Drugs.com. US Food and Drug Administration Expands Indication for Prograf for Prevention of Organ Rejection in Adult and Pediatric Lung Transplant Recipients. Published July 20, 2021. Accessed July 14, 2025. https://www.drugs.com/newdrugs/u-s-food-administration-expands-indication-prograf-prevention-organ-rejection-adult-pediatric-lung-5603.html.
9. NIH. All of Us Adds Data from 50% More Participants in Largest Data Expansion to Date. Published February 24, 2025. Accessed July 14, 2025. https://allofus.nih.gov/article/announcement-all-of-us-adds-data-from-50-more-participants-in-largest-data-expansion-to-date.
10. Feliciano KB, Aminorroaya A, Dhingra L, Thangaraj P, Adejumo P, Khera R. Wearable device use patterns among those with cardiovascular risk in All of Us Research Program: an evaluation against a nationally representative database. JACC. 2024;83(13_Supplement):2045. https://www.jacc.org/doi/10.1016/S0735-1097%2824%2904035-X.
11. Atreja A, Bates D, Clancy S, et al. Mobilizing mHealth Innovation for Real-World Evidence Generation. Duke Margolis Center for Health Policy. Accessed July 14, 2025. Published March 2020. https://healthpolicy.duke.edu/sites/default/files/2020-03/duke-margolis_mhealth_action_plan.pdf?utm.
12. Kyodo News PR Wire. The Third Boao International Conference on Real World Studies of Medical Products was held in Boao, Hainan Province. Published December 2, 2024. Accessed July 14, 2025. https://kyodonewsprwire.jp/release/202412020913.
13. Duke Margolis Center for Health Policy. Duke-Margolis Center for Health Policy 2023-2027 Strategy to Advance Real-World Evidence Policy. Accessed July 14, 2025. https://healthpolicy.duke.edu/sites/default/files/2024-04/Margolis%20RWE%20Collab%20Advisory%20Strat%20Plan_new%20design.pdf.
14. Emmott N, Nafie M, Kim D, Hendricks-Sturrup R. Real-World Evidence to Support Causal Inference: Methodological Considerations for Non-Interventional Studies. Duke Margolis Institute for Health Policy. Published June 18, 2024. Accessed July 14, 2025. https://healthpolicy.duke.edu/sites/default/files/2024-06/RWE%20Support%20Causal%20Inference.pdf.
15. D’Ambrosio M, Rodriguez P, Emmott N, Hendricks-Sturrup R. Operational Tools and Best Practices to Support Electronic Health Record-Sourced Data Quality, Relevance, and Reliability at the Data Accrual Phase. Duke Margolis Institute for Health Policy. Published February 14, 2025. Accessed July 14, 2025. https://tinyurl.com/38ys3eem.
16. Duke Margolis Institute for Health Policy. International Harmonization of Real World Evidence Standards Dashboard. Accessed July 14, 2025. https://healthpolicy.duke/edu/projects/international-harmonization-real-world-evidence-standards-dashboard.
17. Chadwick D. The Breaking Point: Funding Uncertainty Threatens Scientific Innovation. American Physiological Society. Published July 2025. Accessed July 14, 2025. https://www.physiology.org/publications/news/the-physiologist-magazine/2025/july/the-breaking-point--funding-uncertainty-threatens-scientific-innovation?SSO=Y.
18. US Food and Drug Administration. CBER Biologics Effectiveness and Safety (BEST) System. Updated March 14, 2022. Accessed July 14, 2025. https://www.fda.gov/vaccines-blood-biologics/safety-availability-biologics/cber-biologics-effectiveness-and-safety-best-system.

Beth Fand Incollingo is a freelance writer who reports on scientific, medical, and university issues.