Real-World Evidence – What Are Studies Conducted in Real-World Clinical Practice?
Real-World Evidence (RWE) refers to scientific evidence showing how a medicine, medical device, or therapy performs in everyday clinical practice among real-world patients, rather than only under the carefully controlled conditions of a clinical trial. Just a decade ago, RWE was regarded as a supplement to traditional scientific research. Today, regulatory agencies, healthcare payers, and pharmaceutical companies use it to inform decisions about drug safety, reimbursement, and the design of future studies.
The scale of this transformation is clear. Between February 2025 and February 2026, the European Medicines Agency (EMA) conducted or completed 88 studies based on real-world data, representing a 49% increase compared with the previous year (EMA, 4th RWE Report, 2026). This article explains what RWE studies are, the different types available, who uses them and for what purposes, and what should be considered when planning them.
What Is Real-World Evidence?
One of the most frequently cited definitions comes from the US Food and Drug Administration (FDA). RWE refers to clinical evidence regarding the use and potential benefits or risks of a medical product, derived from the analysis of Real-World Data (RWD) (FDA). The EMA takes a similar approach, describing RWE as information derived from the analysis of routinely collected data on patients' health status or healthcare delivery, obtained outside traditional clinical trials (EMA).
In practice, the key question is: what happens to a medicine once it leaves the controlled environment of a clinical trial? A pivotal clinical trial examines whether a medicine can work. An RWE study investigates whether it works in everyday practice — among older patients, people with multiple conditions, those who occasionally miss a dose, and patients treated in both regional hospitals and university medical centers.
RWE vs. Randomized Controlled Trials (RCTs)
Randomized controlled trials remain the gold standard for assessing treatment efficacy. However, they have limitations that RWE can help address, and vice versa. Rather than competing, the two approaches complement each other.
| Feature | Randomized Controlled Trial (RCT) | Real-World Evidence Study |
|---|---|---|
|
Research question |
Does the treatment work under ideal conditions (efficacy)? | How does the treatment perform in everyday practice (effectiveness)? |
|
Patients |
Highly selected, with numerous exclusion criteria | Broad patient populations, potentially including older adults, children, pregnant women, and patients with multiple conditions |
|
Treatment allocation |
Random assignment | Decisions made by physicians and patients as part of routine care |
|
Sample size |
Usually hundreds to several thousand participants | From hundreds to millions of patients |
|
Follow-up period |
Usually months, sometimes several years | Often many years, including retrospective follow-up |
|
Main advantage |
Strong evidence of causal relationships | Representativeness, scale, and the ability to identify rare or long-term events |
|
Main limitation |
Limited generalizability of results | Risk of bias and confounding factors |
|
Cost and time |
High costs and lengthy preparation | Usually lower costs and faster results when suitable data are available |
This reflects the EMA's position: clinical trials remain central to establishing efficacy and safety, while real-world evidence complements them (The Evidence Base).
What Types of Studies Are Considered RWE?
RWE is not a single study design but a family of approaches. What connects them is their use of data generated during routine healthcare rather than solely through procedures imposed by a clinical trial protocol. The main types include:
-
Prospective observational studies. Patients are enrolled and followed over time, but treatment decisions remain entirely with their physicians. Researchers collect data according to a predefined plan, for example using an electronic case report form (eCRF), without interfering with treatment.
-
Retrospective observational studies. Researchers analyze existing data, such as medical records, hospital databases, and healthcare payer claims. These studies can answer research questions relatively quickly without requiring years of prospective follow-up.
-
Patient and disease registries. Structured, long-term collections of data on individuals with a particular disease or receiving a specific treatment. They are especially valuable in rare diseases, oncology, and cardiology.
-
Post-Authorization Safety Studies (PASS) and Post-Authorization Efficacy Studies (PAES). Conducted after a medicine has received marketing authorization, often at the request of or as a requirement imposed by a regulatory authority. They can help identify rare adverse reactions that may not have been detected in clinical trials involving only a few thousand participants.
-
Drug utilization studies. These examine who uses a particular medicine, how frequently it is prescribed or administered, and for which indications, including off-label use.
-
Pragmatic clinical trials. These are formally randomized clinical trials, but they are conducted under conditions designed to resemble routine clinical practice as closely as possible, with minimal additional procedures. They bridge the gap between conventional RCTs and RWE.
-
External control arms and target trial emulation. Real-world data are used either as a comparator group for a single-arm trial or to emulate a hypothetical randomized trial using observational data. The EMA identifies both approaches among topics increasingly discussed during scientific advice procedures (EMA, 4th RWE Report).
Under Polish and EU law, most RWE studies are classified as non-interventional studies. This generally means they do not require authorization from the President of Poland's Office for Registration of Medicinal Products, Medical Devices and Biocidal Products (URPL) or the procedures applicable to interventional clinical trials of medicinal products.
However, this does not mean that they are free from regulatory obligations. Personal data protection requirements still apply, an ethics committee opinion may be necessary, and PASS studies are subject to specific notification and reporting requirements. Determining the appropriate regulatory classification is one of the first decisions worth discussing with an experienced CRO.
Why Are RWE Studies Conducted, and Who Needs Them?
RWE fills evidence gaps that traditional clinical trials cannot fully address. However, different stakeholders within the healthcare system use it for different purposes.
| Stakeholder | How RWE Is Used | Example Research Question |
|---|---|---|
| Regulatory authorities (EMA, URPL, FDA) |
Safety monitoring, signal assessment, updates to product information, and support for marketing authorization decisions |
How frequently does a rare adverse reaction occur, and how soon after treatment initiation? |
| Healthcare payers and HTA agencies (NFZ, AOTMiT) |
Assessment of real-world effectiveness, reimbursement decisions, risk-sharing agreements, and evaluation of drug programs |
Do patients enrolled in a drug program achieve outcomes comparable to those reported in pivotal trials? |
| Pharmaceutical companies and medical device manufacturers |
Post-marketing commitments, evidence of value for payers, new indications, and clinical trial planning |
How many patients meet the eligibility criteria for a planned trial, and where can they be recruited? |
| Physicians and scientific societies |
Updating clinical guidelines and selecting treatments for patients underrepresented in RCTs |
How does a medicine perform in patients over 80 with renal impairment? |
| Patients and patient organizations |
Understanding disease progression, quality of life, and real-world treatment outcomes |
What does the typical treatment pathway look like for someone with my rare disease? |
Several examples demonstrate that these applications are not merely theoretical.
A DARWIN EU network study involving more than 40,000 patients treated with clozapine found that severe hematological complications were very rare and occurred mainly during the first two months of treatment. Based on these findings, the Pharmacovigilance Risk Assessment Committee (PRAC) updated its recommendations, reducing the frequency of blood count monitoring.
Another study, commissioned at the request of HTA agencies and healthcare payers, confirmed a survival benefit associated with pembrolizumab in advanced non-small cell lung cancer using real-world data, consistent with the findings of pivotal clinical trials (EMA, 4th RWE Report).
It is also worth remembering that RWE is not limited to major regulatory decisions. According to an EMA survey of study requesters, approximately 30% of study results directly supported formal regulatory decisions, while another 44% improved clinical or methodological understanding (same source).
RWE in Europe and Poland – Where Do We Stand?
The European Union is developing infrastructure designed to make RWE a permanent part of regulatory decision-making for medicines.
At the center of these efforts is DARWIN EU, a data network coordinated by the EMA. According to the 2026 report, it includes 40 data partners from 18 European countries and covers data from more than 250 million patients. Importantly for the Polish market, Poland and Lithuania joined the network during its most recent expansion (EMA, 4th RWE Report).
The network operates efficiently: the median duration of a DARWIN EU study was 4.8 months from protocol approval to results. The most common study topics included drug utilization, disease epidemiology, drug safety, and the design and feasibility of planned clinical trials (same source).
Further developments are underway. The reform of EU pharmaceutical legislation provides for the use of RWE to complement clinical trial data in regulatory decision-making. At the same time, the European Health Data Space (EHDS) is being developed to facilitate the secure secondary use of health data for research (same source).
Across the Atlantic, the FDA has been developing its RWE program since the introduction of the 21st Century Cures Act in 2016. In December 2025, the agency issued updated guidance on the use of RWE to support regulatory decision-making for medical devices (FDA).
In Poland, real-world data sources include the National Health Fund (NFZ), registries maintained within drug programs, electronic medical records, clinical registries managed by scientific societies, and data collected directly at research sites during prospective studies.
The fragmentation and accessibility of these data remain significant challenges. As a result, many Polish RWE projects rely on their own carefully designed data collection processes.
Limitations and Challenges – Why RWE Requires Rigorous Methodology
Large volumes of data do not automatically lead to reliable conclusions. This is one of the most common misconceptions surrounding RWE.
A poorly designed observational study can produce highly precise but incorrect results. Several issues require particular attention:
-
Confounding factors. Physicians do not assign treatments randomly; their decisions depend on patient characteristics. If patients with more severe conditions are more likely to receive a newer medicine, a direct comparison may produce misleading results. This is why methods such as propensity score matching are needed.
-
Selection and measurement bias. Medical records are primarily created for treatment and reimbursement purposes, not research. Important variables may be missing, diagnoses may be coded inconsistently, and some patients may be lost to follow-up.
-
Data quality and fitness for purpose. The FDA evaluates data according to two main criteria: relevance to the research question and reliability, including completeness, accuracy, and consistency (Knobbe Martens). Not every RWD dataset is suitable for every research question.
-
Heterogeneity of data across countries and healthcare systems. EMA study requesters identify this as a major barrier to comparing results and emphasize the need for better harmonization of coding systems and definitions (EMA, 4th RWE Report).
-
Personal data protection. Health data are classified as a special category of personal data under the General Data Protection Regulation (GDPR). RWE studies require an appropriate legal basis for processing, pseudonymization or anonymization where applicable, data processing agreements, and secure infrastructure.
For this reason, a well-conducted RWE study begins with the same foundations as a clinical trial: a clearly defined research question, a protocol, and a statistical analysis plan prepared before data analysis begins.
Only then can the findings have a strong chance of being accepted by regulators, healthcare payers, or reviewers of scientific journals.
How BioStat Supports Real-World Evidence Studies
The BioStat Research and Development Center has been designing and conducting clinical and observational studies for more than 20 years.
We have completed over 300 studies and have a team of more than 100 experts, including project managers, clinical research associates, statisticians, and data managers (BioStat).
For RWE projects, we can manage the entire process or support selected stages:
-
Study concept and classification – refining the research question, selecting the appropriate study design, and assessing regulatory requirements.
-
Protocol and Statistical Analysis Plan (SAP) – accounting for confounding factors and data limitations.
-
Data collection – using our own eCRF and eTMF systems, in accordance with GCP principles and GDPR requirements.
-
Research site recruitment and monitoring – in Poland and across European markets.
-
Statistical analysis and final report – prepared in a format suitable for submission to regulatory authorities, healthcare payers, or scientific journals.
Are you planning an observational study, patient registry, or post-authorization study? Schedule a free consultation with BioStat experts – we will help you turn your idea into a feasible research project.
Related reading: RWD vs. RWE – What Is the Difference Between Real-World Data and Real-World Evidence?