While more companies are using real-world data (RWD) and real-world evidence (RWE) in research, there are still barriers that continue to plague its uptake.
One of the main barriers for companies using RWD and RWE is the quality of data and whether it can be interpreted properly for use in research.
Go deeper with GlobalData
Discover B2B Marketing That Performs
Combine business intelligence and editorial excellence to reach engaged professionals across 36 leading media platforms.
The primary concerns sponsors have with the quality of RWD/RWE stem from how it is collected and logged, which impacts its interpretability. However, advancements in technology are helping address these challenges.
While RWD and RWE are sometimes used interchangeably, they have different definitions by the US Food and Drug Administration (FDA) and European Medicines Agency (EMA).
The FDA defines RWD as the data relating to patient health status and/or the delivery of healthcare routinely collected from a variety of sources, while RWE is the clinical evidence regarding a medical product’s use and potential benefits or risks derived from analysis of RWD.
The EMA has a similar definition, classifying RWD as routinely collected data relating to a patient’s health status or the delivery of healthcare, sourced from a variety of settings beyond traditional clinical trials, while RWE is clinical evidence on the use and potential benefits or risks of a medical product, derived from the analysis of RWD.
US Tariffs are shifting - will you react or anticipate?
Don’t let policy changes catch you off guard. Stay proactive with real-time data and expert analysis.
By GlobalDataRWD and RWE use in trials on the rise
While barriers exist, sponsors are trusting RWD more, with a larger number of trials utilising RWD and RWE in 2024, according to GlobalData analysis.
The number of initiated studies using RWD and RWE elements peaked in 2024, with 16% of the total studies using a RWD and RWE element launched that year. This is compared with 13% in 2023 and 15% for 2021 and 2022.
Of those trials initiated in 2024, 47% are ongoing and recruiting, 39% are planned, 5% are ongoing, not recruiting, 5% are completed, with the rest either ongoing recruiting by invitation, withdrawn or terminated.
Oncology is the main therapy area to use RWD and RWE elements, with 34% of studies that are using RWD and RWE being in this area. Central nervous system trials account for 12% of studies using RWD/RWE, with cardiovascular the next most used indication at 10%.
The studies included in this analysis were initiated between 2017 and 2025.
GlobalData is the parent company of Clinical Trials Arena.
Quality of RWD needs to improve at point of collection
Experts agree that RWD and RWE can only be used effectively if it is of high quality.
As clinicians are not collecting the data for use in research, this can cause some issues, says Anne Donovan, Vice President and General Manager of Health Language at Wolters Kluwer, Health.
“When you think about the full potential of RWD, it obviously starts at the point of collection. Fundamentally, healthcare is built around predominantly managing reimbursement as opposed to managing clinical richness. As a result, the systems are not as well oriented towards capturing the richest and relevant clinical data,” Donovan explains.
Hospital data also tends to be a single moment in time rather than continued analysis or data collected at set timepoints, which also creates issues in its use in research, says Dr. Gen Li, President of AI-data company Phesi. “When you want to look at the longevity of the data, it becomes more difficult,” Li explains.
For clinical trials, there are certain endpoints measured which may not be as closely monitored by clinicians, adds Vish Srivastava, Co-Founder and CEO of Century Health – making it difficult to use this kind of data for research purposes.
“The harder problem is, do you have the right raw material to generate the clinical variables that are necessary for the downstream analysis?” Srivastava asks. “Drawing on respiratory, for example, we need to know exacerbations, we need to know what therapies the patient is on, and we need to know long-term what disease progression looks like in those cases. For this, we are reliant on what is recorded as a part of the standard of care.”
Li agrees that there could be elements of data that are critical to research that are missing. “As they are being collected in the care setting, it is not necessarily designed in the same way as in a clinical development setting. This means that most of the things that are critically important for clinical development are missing from the data,” Li says.
AI and ML helping to interpret data
However, recent advances in artificial intelligence (AI) and machine learning (ML) have made it easier to process unstructured data, Srivastava adds.
“Historically, it required armies of clinicians to manually look through these records, and we can now automate that. Moving the needle on data quality that I think is really what’s going to unlock use cases for RWD,” Srivastava explains.
However, this does not solve every issue, as companies are also managing this data on a global scale, meaning they are contending with different languages, dialects and abbreviations in medical notes. Different countries may also use different systems to collect data, meaning it may be presented differently to research teams.
“It’s not impossible, but it is definitely challenging,” Donovan says. “One of the reasons that healthcare is so challenging is that there are so many different coding systems and data systems. You have LOINC [Logical Observation Identifiers Names and Codes] for labs, SNOMED [Systematised Nomenclature of Medicine] for clinical data, and ICD-10 [International Classification of Diseases] for diagnosis data. There’s not just one foundational terminology, so how do you make sure all that data is properly captured and translated? What you call a condition in one area of the world might be very different from what you call a condition in another area of the world.”
Donovan says this challenge can be overcome and given the vast amount of data needed to create such a system, it would make it a “very powerful system that would provide powerful datasets”.
While quality and interpretability issues of RWD/RWE are well known, the US Food and Drug Administration (FDA) currently still holds this data to the same standard as data collected for the sole purpose of research, highlights Srivastava.
“Today, there is no separate standard for how you measure data quality and auditability for RWD versus trial data, which means that the absolute highest bar for data that’s collected is also applied to RWD,” Srivastava says. As a result, data is not always categorised into standard clinical endpoints, which may make it unsuitable for research.
RWD won’t replace RCTs
While RWD controls are being used commonly in rare diseases due to recruitment difficulties, it is agreed amongst experts that this comparator will never replace the gold standard – randomised controlled trials (RCTs). However, it can be used as a tool to speed up the drug development process.
“I don’t necessarily see the ideal end state as being we no longer do RCTs, and we kind of evaluate new treatments in the real world exclusively,” Srivastava notes. “I think that we’re building towards a hybrid world where we can shorten the time a drug takes to get to market. Currently, it takes 12 years, so how can we cut that in half?”
Srivastava emphasises that RWD will be crucial, as it will help identify which patients are most likely to respond to therapies ahead of running a study, which will allow a sponsor to ensure they are setting a suitable trial design.
Li agrees that RWD will become a part of the research toolkit, assisting most in rare disease studies where traditional RCTs are challenging, indications with well-established standard of care, and providing supplementary insights to traditional trials.
