Discover how integrating clinical trial data and real-world evidence can strengthen drug development, regulatory decisions, reimbursement, and patient outcomes.
Clinical trials establish whether a medicine can work under controlled conditions. Real-world evidence shows how it performs when patients receive treatment in everyday healthcare settings.
Yet pharmaceutical development still treats these evidence streams as largely separate. Bringing them together could improve Clinical Development, regulatory decisions, reimbursement, and patient outcomes.
The opportunity is not to replace clinical trials, but to make their findings more relevant throughout a medicine’s lifecycle.
Image
Why Must Clinical Trials and Real-World Evidence Work Together?
For decades, randomised controlled trials (RCTs) have occupied the central position in pharmaceutical evidence generation.
Their ability to control variables, reduce bias, and establish causal relationships makes them essential to evaluating the safety and efficacy of investigational medicines.
However, the patients enrolled in clinical trials do not always represent those who eventually receive treatment.
Strict eligibility criteria can exclude older adults, people with multiple chronic conditions, patients taking several medicines, and individuals with complex disease histories.
As a result, trial populations are often more controlled and less diverse than the populations encountered in routine clinical practice.
This creates an important gap betweenwhat clinical trials demonstrate and what healthcare professionals need to understand about treatment performance in everyday practice.
A medicine may demonstrate significant efficacy in a carefully selected Phase III population. Yet questions remain about its performance in patients with different characteristics, longer treatment exposure, varying adherence, or additional health conditions.
RWD refers to information collected during routine healthcare delivery rather than exclusively through traditional clinical trials. Sources include:
Electronic health records and hospital databases.
Disease-specific patient registries.
Insurance claims and healthcare utilisation records.
Medical imaging and laboratory results.
Digital health technologies and remote monitoring.
Patient-reported outcomes and treatment experiences.
When researchers analyse these datasets to generate evidence about a medicine’s use, benefits, and risks, the resulting information is known as real-world evidence (RWE).
The distinction matters. RWD is the underlying information. RWE is the evidence produced through its analysis.
In their March 2026 policy and practice commentary, published in Frontiers in Pharmacology, Alexandros Sagkriotis examines how these two evidence approaches can become complementary components of pharmaceutical development and health technology assessment (HTA).
The central argument is straightforward in that randomised trials and real-world evidence answer different questions, and pharmaceutical decision-making needs both.
The Evidence Gap Between Controlled Trials and Clinical Practice
The strength of an RCT is also one of its limitations.
By controlling the study environment, researchers can isolate treatment effects more confidently. However, this control can make it difficult to predict how findings will translate across broader patient populations.
Consider a patient receiving a new oncology medicine outside a clinical trial.
That patient may be older than the average trial participant, have impaired kidney function, take several concomitant medicines, and receive treatment after multiple previous therapies. These circumstances can influence tolerability, treatment adherence, and clinical outcomes.
A pivotal trialmay provide strong evidence of efficacy, but it may offer limited information about this particular treatment experience.
This is not necessarily a failure of clinical trial design. But it does reflect the different purposes of experimental and observational research.
RCTs are designed primarily to establish whether an intervention causes a particular outcome under defined conditions. RWE helps establish how treatment performs across broader healthcare settings, populations, and periods of observation.
The distinction becomes especially important when evaluating:
Long-term treatment effectiveness and safety.
Rare adverse events that may not emerge during shorter trials.
Outcomes in older patients and people with multiple conditions.
Treatment adherence and discontinuation.
Differences between clinical trial populations and routine patient populations.
Health-related quality of life and functional outcomes.
This means that in some analyses, only approximately one-third of real-world patients would meet the eligibility criteria of pivotal RCTs. Age, comorbidities, disease severity, and previous treatment exposure contribute to this gap.
This has large implications beyond clinical research.
If evidence generated during development does not adequately reflect the population that will eventually receive a medicine, uncertainty can continue into regulatory review, reimbursement negotiations, and clinical adoption.
The objective should therefore be to design evidence generation around the patient population and the decisions that matter throughout the medicine’s lifecycle.
How Can Real-World Evidence Strengthen Clinical Development?
Integrating RWE into clinical development does not mean reducing the importance of randomisation or replacing established trial methodologies.
Instead, it means using observational evidence strategically to address questions that conventional trials cannot fully resolve.
Several approaches are becoming increasingly relevant.
External and Hybrid Control Arms
External control arms use information from patients outside the investigational trial to provide a comparison group.
These patients may come from disease registries, electronic health records, or previously conducted studies.
This approach can be particularly valuable in rare diseases, oncology, and situations where recruiting a conventional control group is difficult or ethically challenging.
A hybrid control arm combines participants from a traditional trial with appropriately selected external data.
However, the approach requires careful design.
Researchers must establish comparable eligibility criteria, treatment histories, outcome definitions, and follow-up periods. They must also address differences between the investigational population and external patients.
Without these safeguards, apparent treatment differences may reflect population characteristics rather than genuine therapeutic effects.
Pragmatic and Registry-Based Trials
Pragmatic trials bring randomisation closer to routine healthcare settings.
Rather than testing a medicine exclusively under tightly controlled research conditions, these studies examine treatment within more representative clinical environments.
Registry-based randomised trials can use established patient registries to support recruitment, follow-up, and outcome collection.
These approaches can preserve the benefits of randomisation while improving relevance to routine clinical practice.
They are particularly useful for chronic diseases, where treatment patterns and outcomes may evolve over several years.
Target Trial Emulation
Target trial emulation uses observational data to recreate the structure of a hypothetical randomised trial.
Researchers specify the eligibility criteria, treatment strategies, follow-up period, outcomes, and analytical approach that would apply to the proposed trial.
They then use real-world datasets to approximate that design.
This framework can improve the transparency of observational research and help reduce common analytical problems, including inappropriate patient selection and timing-related bias.
It does not make observational research equivalent to randomisation. Unmeasured confounding can remain, and conclusions must reflect that uncertainty.
Nevertheless, target trial emulation offers a more structured way to investigate treatment questions when conventional trials are unavailable or impractical.
How AI Is Opening New Opportunities in Fertility Care for Pharma
"If I had to get great minds in a room, I'd be looking at it from a tech perspective. How can we take all this knowledge, apply it to policies that can be easily distributed?” Clara Toombs, Fertilia AI.
Image
Regulators Are Opening the Door, but Clinical Evidence Standards Still Matter
Regulatory acceptance of RWE has developed considerably.
Authorities increasingly recognise that traditional clinical trials cannot answer every question about medicines, particularly in rare diseases, personalised medicine, and conditions requiring long-term monitoring.
However, greater acceptance does not mean that all real-world datasets are suitable for regulatory decisions.
There are also many approaches across several major regulatory jurisdictions that make statistical consistency complex.
The FDA and real-world evidence
The U.S. Food and Drug Administration (FDA) has developed its Real-World Evidence Program to support the appropriate use of RWD in regulatory decision-making.
Its applications include safety surveillance, label expansions, and, under specific circumstances, evidence supporting effectiveness.
The emphasis is on clearly defined research questions, appropriate data sources, prespecified study designs, and transparent analytical methods.
This is an important distinction for sponsors. The availability of a large healthcare dataset does not automatically make it suitable for a regulatory submission.
Relevance depends on the research question, the quality of the underlying information, and the reliability of the analytical approach.
Europe’s expanding evidence infrastructure
The European Medicines Agency (EMA) has also expanded its approach to RWE.
One significant development is DARWIN EU (Data Analysis and Real-World Interrogation Network).
This federated network connects healthcare data sources to support regulatory research and analysis across Europe.
Its importance extends beyond individual studies. It offers a framework for accessing information from different healthcare systems while respecting local data governance arrangements.
For pharmaceutical companies conducting multinational development programmes, these developments could support more consistent evidence generation across European markets.
However, differences in national data access procedures, privacy interpretations, and governance requirements continue to create operational difficulties.
Its work reflects growing interest in using real-world information to support innovative licensing pathways, post-authorisation evaluation, and ongoing assessment of medicines.
Japan’s Pharmaceuticals and Medical Devices Agency (PMDA), meanwhile, has taken a more cautious but increasingly structured approach, particularly around registry information and post-marketing studies.
Other jurisdictions, including Canada, Australia, South Korea, China, Saudi Arabia, and Brazil, have also developed guidance or initiatives addressing RWD and RWE.
Despite these developments, regulatory acceptance is not yet consistent across markets.
A dataset or study design considered acceptable by one authority may not meet the requirements of another.
This creates additional complexity for sponsors planning global clinical development and regulatory submissions.
Image
Regulatory Approval and Reimbursement Require Different Evidence
One of the most important issues around RWE concerns the relationship between regulatory approval and HTA.
These processes are connected, but they do not assess medicines for precisely the same purpose.
Regulators primarily evaluate whether a medicine meets the required standards for quality, safety, and efficacy, while HTA organisations assess its relative clinical benefit, economic value, and implications for healthcare resource allocation.
Consequently, the evidence required to support marketing authorisation may not be sufficient to secure reimbursement.
This distinction is particularly visible in oncology.
A medicine may receive accelerated or conditional approval based on promising results from a single-arm trial, supported by an external control group. However, a reimbursement authority may subsequently request stronger comparative evidence, longer survival data, or more reliable information about treatment outcomes.
The same RWD can therefore contribute to regulatory decision-making but remain insufficient for an HTA assessment.
For sponsors, this creates a strategic problem. Evidence planning that focuses exclusively on obtaining marketing authorisation may leave significant questions unanswered when a product reaches the market.
Therefore, regulatory and reimbursement strategies should be developed together, rather than treated as separate stages of the development process.
Early discussions with regulators, HTA bodies, and payers can help sponsors understand which outcomes, comparators, and patient populations their evidence programmes need to address.
This is particularly important for advanced therapies, rare diseases, and treatments targeting narrowly defined patient groups.
Find out how Quality-by-Design improves reliability and compliance in pharma and how you can maximise innovation through quality-based pharma development systems.
Image
What Do Oncology and Ophthalmology Reveal About Integrated Evidence?
The therapeutic areas of oncology and ophthalmology illustrate how clinical trial evidence and RWE can complement one another.
These therapeutic areas present different challenges, but both demonstrate the value of extending evidence generation beyond conventional trial environments.
Oncology: Addressing uncertainty beyond approval
Oncology is a particularly relevant setting for hybrid evidence generation.
Clinical trials frequently involve small populations, biomarker-defined subgroups, and rapidly changing standards of care.
Researchers may also rely on surrogate endpoints, such as progression-free survival (PFS), because waiting for overall survival (OS) results can delay development.
However, improvements in PFS do not always translate consistently into longer survival across different cancers and treatment classes.
RWE can provide additional information about treatment sequencing, survival outcomes, and longer-term effectiveness after authorisation.
External control arms can also help contextualise single-arm studies where conventional randomisation presents substantial challenges.
However, oncology datasets have limitations. Disease progression may not be recorded consistently, patients may receive different treatment combinations, and confounding by indication can distort comparisons.
These problems reinforce the need for prespecified study protocols and robust sensitivity analyses.
Ophthalmology: Combining patient registries, imaging, and patient outcomes
Ophthalmology offers a different example.
Many eye diseases progress over extended periods, making longitudinal registries valuable sources of clinical information.
These datasets can connect treatment records with imaging results, visual function, and patient-reported outcomes.
Artificial intelligence (AI) is adding another dimension.
Image-analysis systems can extract quantitative measurements from routinely collected retinal images, allowing researchers to examine disease progression and treatment response across larger patient populations.
For example, registry and imaging data can help assess the durability of anti-vascular endothelial growth factor (anti-VEGF) treatment regimens and investigate whether treatment intervals can be optimised.
These insights can extend the evidence generated in pivotal trials and inform post-authorisation assessment.
However, imaging devices, acquisition protocols, and annotation practices must be sufficiently consistent to support reliable comparisons.
The broader lesson is that the most useful evidence strategy depends on the disease, the treatment, the patient population, and the clinical question.
There is no single integration model that will work equally well across every therapeutic area.
The Methodological Challenges Sponsors Cannot Ignore
Combining clinical trial evidence with RWE introduces opportunities, but it also creates scientific and operational risks.
The credibility of integrated evidence depends on more than the volume of available patient information. Researchers must demonstrate that the data are appropriate, the analytical methods are defensible, and the findings can withstand independent scrutiny.
Several issues require particular attention.
Confounding and selection bias
Patients receiving different treatments may have fundamentally different characteristics.
Statistical approaches such as propensity score matching, inverse probability weighting, and doubly robust estimation can address measured differences. However, they cannot automatically eliminate unmeasured confounding.
Incomplete and inconsistent data
Electronic health records and registries may contain missing information, inconsistent coding, or gaps in follow-up.
These problems can affect outcome measurement and undermine comparisons between patient populations.
Data provenance and reproducibility
Sponsors need to demonstrate where information originated, how it was processed, and whether analytical results can be independently reproduced.
Privacy and governance
Cross-border research introduces additional challenges involving consent, data protection, access permissions, and national interpretations of privacy legislation.
AI introduces further considerations.
Machine-learning systems can extract useful information from imaging, genomic datasets, and unstructured clinical records. However, algorithmic bias, limited explainability, and inconsistent performance across patient populations require careful management.
Synthetic data and digital twins also offer interesting opportunities for methodological development and simulation.
However, they should not be confused with established sources of decision-critical clinical evidence. Their current applications do not remove the need for reliable underlying patient data.
These considerations reinforce an important principle: Integrating evidence should strengthen scientific scrutiny, not lower the standard of proof.
Image
How to Build Evidence Generation Into the Entire Drug Lifecycle?
Perhaps the most significant implication for sponsors is the need to change when and how RWE planning takes place.
Historically, organisations have often generated substantial real-world datasets after a medicine reaches the market.
This information can support pharmacovigilance, effectiveness assessments, and post-authorisation commitments.
However, retrospective evidence generation can leave important gaps. Data sources may not capture the outcomes required by regulators or payers, and researchers may discover too late that essential information was never collected.
There can be a more proactive approach, with RWD generation incorporated into development planning from an earlier stage but this suggests several practical priorities for sponsors who should:
1. Plan evidence requirements early. Identify which clinical questions cannot be answered adequately through conventional trials and determine how RWD could address them.
2. Engage decision-makers before submission. Discuss evidence expectations with regulators and HTA organisations to reduce the risk of generating studies that do not meet their requirements.
3. Invest in reliable data infrastructure. Establish partnerships with registries, healthcare providers, academic institutions, and federated data networks.
4. Include patient-generated outcomes. Capture treatment burden, quality of life, adherence, and functional outcomes alongside conventional clinical endpoints.
5. Maintain evidence generation after launch. Use prospectively designed studies to monitor safety, effectiveness, and treatment patterns as medicines enter broader patient populations.
Sponsors can also use predefined post-launch evidence networks operating for an initial period, potentially three to five years depending on the medicine and its remaining evidence uncertainties.
Such arrangements could provide a more structured approach to post-authorisation evaluation, with the intensity of evidence generation adjusted as confidence in a medicine’s safety and effectiveness increases.
This would require cooperation between sponsors, healthcare systems, regulators, and independent research organisations.
It would also require clearer governance and stronger incentives for long-term investment in RWE infrastructure.
How to Move From Separate Evidence Streams to Better Development Decisions?
The integration of clinical trials and RWE represents more than another methodological development in pharmaceutical research.
It challenges the traditional approach to evidence generation, in which controlled clinical studies establish a medicine’s value and observational research provides supplementary information afterwards.
Modern drug development requires a more connected model.
Clinical trials remain essential for establishing causal treatment effects, while RWE adds information about broader patient populations, longer-term outcomes, treatment patterns, and the realities of healthcare delivery.
Neither approach is sufficient on its own.
For sponsors, the commercial and strategic implications are significant. Better integrated evidence planning could reduce uncertainty during regulatory assessment, strengthen reimbursement submissions, improve post-launch monitoring, and provide a clearer understanding of how treatments perform across different patient populations.
However, this will depend on early planning, transparent methodologies, reliable data infrastructure, and closer collaboration between stakeholders.
The challenge is no longer simply generating more evidence. It is generating the right evidence, at the right time, for the decisions that determine whether medicines reach patients and deliver meaningful clinical value.
What is the difference between clinical trial evidence and real-world evidence?
Clinical trial evidence is generated through structured research studies, often using randomisation and controlled conditions to establish treatment safety and efficacy. Real-world evidence comes from analysing patient information collected during routine healthcare delivery. It provides additional insight into treatment effectiveness, safety, and outcomes across broader populations.
How can real-world evidence improve clinical development?
RWE can help pharmaceutical companies understand treatment outcomes in patients who may not meet conventional trial eligibility criteria. It can also support external control arms, pragmatic trials, long-term safety monitoring, treatment adherence assessments, and post-authorisation effectiveness studies.
Do regulators accept real-world evidence for drug approval?
Yes. Regulators, including the FDA, EMA, and MHRA, increasingly recognise RWE as a potential source of evidence for regulatory decisions. However, acceptance depends on the research question, data quality, study design, analytical methods, and transparency. RWE does not automatically replace randomised clinical trial evidence.
What is a hybrid control arm in a clinical trial?
A hybrid control arm combines patients enrolled in a conventional clinical trial with appropriately selected external patients identified through real-world datasets. It can provide additional comparative information when traditional randomised control groups are difficult to establish. Researchers must address differences in patient characteristics, eligibility criteria, outcomes, and follow-up.
Why is integrating real-world evidence important for sponsors?
Integrating RWE into clinical development can help companies address evidence gaps, understand treatment performance beyond controlled trials, strengthen regulatory and reimbursement strategies, and monitor medicines after launch. It also supports more patient-centred development by incorporating outcomes that matter in everyday clinical practice.
Nicole (BSc Molecular Medicine, Honours Medical Biochemistry) has many years of pharmaceutical experience, having worked for top CROs and biopharma companies for more than a decade.
Did you enjoy the content?
2
Why not support
Nicole Dale
by giving this content a like
Anthropic's new drug discovery programme marks a strategic shift from sole AI software provider to active R&D participant. Here's what it means for pharmaceutical innovation.
Anthropic’s Gates Foundation Partnership and Healthcare AI
See how Anthropic's Gates Foundation partnership and Long-Term Benefit Trust could lead the way in shaping responsible AI governance across healthcare innovation.
AI in Healthcare: OpenAI and Epic Integration Connects EHRs
OpenAI and Epic integration brings AI healthcare into EHR workflows, connecting patient records with ChatGPT for Healthcare to advance research and patient care.
<p>Founded in 2004, Ametris (formerly ActiGraph) is a global digital health solutions provider transforming real-world patient data into validated clinical evidence. Our end-to-end platform combines advanced wearables, regulatory-aligned analytics, and expert scientific support across every phase of a clinical trial. We serve leading pharmaceutical, biotech, and public health research organizations worldwide, backed by two decades of experience and a collaborative, science-first approach.</p><p>In 2026, Ametris joined Signant Health, uniting wearable-derived digital outcome measures with industry-leading eCOA solutions to give sponsors a single partner for integrated evidence generation, combining continuous sensor data with patient-reported outcomes.</p>
Image
Clinical Development
<p><a href="https://www.medidata.com/en/">Medidata </a>has been powering smarter treatments and healthier people by supporting clinical trials through groundbreaking technological innovations for over 25 years. Medidata is a leader in AI-powered solutions across the entire trial lifecycle to accelerate clinical breakthroughs, reduce patient and site burden, and bring new therapies to market faster. Supported by one of the largest clinical datasets in the industry of more than 38,000 trials and 12 million patients, Medidata's <a href="https://www.medidata.com/en/experiences/study-experience/">Study</a>, <a href="https://www.medidata.com/en/experiences/patient-experience/">Patient</a>, and <a href="https://www.medidata.com/en/experiences/data-experience/">Data</a> Experiences are trusted by more than 1 million registered users across approximately 2,300 customers. A Dassault Systèmes brand (Euronext Paris: FR0014003TT8, DSY.PA), Medidata is headquartered in New York City and is recognized as a leader by Everest Group and IDC. Discover more at <a href="https://www.medidata.com/">www.medidata.com</a><strong>. </strong>Listen to our latest podcast, <a href="https://www.medidata.com/en/from-dreamers-to-disruptors-podcast/"><em>from Dreamers to Disruptors</em></a>,<strong> </strong>and follow us at @medidata-solutions on LinkedIn.</p>
Image
Clinical Development
<p>Caidya is a global, therapeutically focused drug development partner helping biotech and biopharma sponsors make smarter decisions, move faster, and advance together. Our connected teams combine scientific insight, operational expertise, and local knowledge across critical markets to generate actionable evidence and keep complex development programs moving forward.</p>
Image
HealthTech & AI
<p>We are rebuilding biotech for the AI era. When a breakthrough is delayed, the world waits.</p><p>Getting a molecule from discovery to patients, or a crop from lab to field, involves thousands of slow, manual, disconnected steps. AI has the potential to change this, compressing decades of R&D work into years. But that only happens when clean, structured scientific data and AI are built into how science gets done. </p><p>Benchling is the AI platform for biotech R&D. Scientists use Benchling to design experiments, capture structured data, and run AI agents and models directly in their workflows. Over 200,000 scientists around the world trust Benchling to power their most important work, from academic labs to Sanofi, Moderna, and more than half of the world's top 50 biopharma.</p>
Comments (0)