The future of clinical trials is here. So when will the industry embrace it?

In recent years, a deeper understanding of everything from genetics to disease mechanisms has driven a flurry of advances, at an almost dizzying rate, across a multitude of therapy areas. 

Yet clinical trials have, so far, remained immune to this spirit of innovation. And despite the well-documented limitations of traditional randomised controlled trials (RCTs), sponsors have been slow to apply the power of real-world data to interventional studies. 

So, what is stopping them? Because the technology is available, the regulators are on board, and the concept of more inclusive trials that better reflect the way in which drug products are actually used in the real world has already been proved. 

Traditional vs real-world RCT 

RCTs may be the gold standard for evidence generation, but traditional models have their limitations.  

From an operational standpoint, they are expensive, resource hungry, and place a huge burden on sites, investigators, and participants. Recruitment and retention is a stubborn challenge, with actual or predicted low enrolment often delaying, or even halting, the start of studies. 

What’s more, the evidence they generate is not reflective of how the resulting product, if it makes it to market, will be used in the real-world.  

Strict inclusion and exclusion criteria, for example, minimise variation to enable better like-for-like comparisons. But they also mean that drugs are rarely tested in a “true” environment. Participants tend to be white men with few or no co-morbidities: far from the typical patient a clinician may expect to see in clinic.   

There’s also the Hawthorne effect, in which people behave differently when they know they are being observed, and the danger of memory-reliant adverse event (AE) reporting. With study visits often months apart, how can we be sure participants remember to tell investigators about every potential issue? 

The solution is pragmatic decentralised, real-world clinical trials that utilise electronic health records (EHRs) to show how a drug will work in real people, in the real world.  

Such studies enrol more representative, more diverse patient populations, at a faster rate than traditional RCTs. They enable real-time safety reporting, meaning sponsors can detect – and act upon – safety signals early, and place less burden on participants and site staff.  

They are also more agile. By their very nature, they can enable longitudinal follow-up, providing a much deeper, richer dataset over time. And the continuous monitoring model means researchers may be able to spot and investigate emerging questions as they arise. 

Barriers to implementation  

While decentralised trials have become more common since the COVID-19 pandemic, they still tend to be based on the traditional model. The only difference is often an option for remote monitoring at some or all scheduled visits. Of course, this reduces some of the burden on the participant, but all the questions around operational complexity, representation and inefficiencies remain. 

At the same time, we have seen real-world evidence become commonly applied in observational phase IV/post-marketing studies and registries. Yet this model has huge potential earlier in the development pathway, during interventional trials. The sector is leaving much of the real-world data benefit on the table, and the question is why.  

Barriers often cited include the complexity of the technology and the data, and the risk of trying something new. There is the expense of setting up new systems, and concerns around protecting data privacy, while many sponsors worry the model is lacking in regulatory acceptance.  

Yet while all of these may have been true in the past, these barriers have been overcome.  

Industry-leading platforms, such as ConneXon, link directly to EHRs , and protect privacy through de-identification, strict role-based access controls, encryption, audit trails and robust information-governance, legal and ethical safeguards. 

Crucially, global regulators have acknowledged the role of real-world data and real-world evidence in clinical trials. 

The FDA’s 2024 draft guidance on integrating randomised controlled trials into routine clinical practice, for example, says the approach may improve participant convenience and accessibility, allow for more representative populations and produce more generalisable results.i The EMA’s real-world data quality framework, published in March 2026, explains how to assess the quality of data intended for use in medicines regulation, while an MHRA guideline on RCTs using real-world data, from 2025, offers the clearest endorsement of the model. ii,iii “An RCT using a real-world data source is not considered more or less valuable than a traditional RCT, provided the data quality is robust and the trial is well designed,” it says.3 

Proof of concept

NWEH has been at the forefront of real-world evidence generation for more than a decade, since working with GSK on the pioneering Salford Lung Study (SLS) back in 2012. 

Widely cited as the first Phase III pragmatic RCT, the SLS evaluated respiratory treatments in routine NHS clinical practice, using patients’ existing healthcare data.iv Clinically, it found that once-daily fluticasone furoate/vilanterol reduced moderate-to-severe COPD exacerbations by 8.4% compared with usual care, without an increase in SAEs.4 But it also showed that a large, pragmatic RCT could be run using routinely collected healthcare data in everyday NHS practice. 

Since then, we have been developing the technology and the processes to support the model at scale. Throughout, we have maintained an emphasis on human-in-the-loop protocols that ensure the technology revolves around clinical expertise, and not the other way round.  

FARSITE, for example, uses de-identified electronic health records to identify potentially eligible patients. GPs review and approve who is approached, then invite them to participate.  

With this model, we can identify potentially eligible patients in minutes, rather than the weeks or months often associated with traditional feasibility and recruitment approaches. In one trial, NWEH recruited around seven participants per site per month versus 1.6 in a comparable study. In another, two sites using FARSITE accounted for 40% of all global recruitment.v 

Once consented and enrolled, ConneXon takes over. With a direct link to EHRs, ConneXon combines the wealth of routinely collected healthcare data with data captured specifically for the study. Everything from lab results to pharmacy prescriptions, wearable data and patient-reported outcomes can be combined into a single, analysis-ready (SDTM/OMOP) dataset. This allows participants to continue living their normal lives while researchers collect richer, longer-term evidence on how treatments perform in real-world clinical practice. 

This is real-time, continuous monitoring that makes for more efficient trials, reducing manual data entry and transcription errors to improve accuracy and save time and resources.  

It also protects safety. Potential adverse events are detected as soon as they appear on a participant’s EHR, whether that is in primary or secondary care. From there, they are investigated by in-house clinical experts and, if necessary, reported to the sponsor in hours. This is in stark contrast to traditional approaches that rely on investigators asking participants to report an event at periodic site visits. As such, it can take days, weeks, or even months for a sponsor to be made aware of a potential issue – and that’s only if the patient remembers to report it at all. Our real-time safety monitoring detects up to two times more safety events than traditional reporting methods. 

Another benefit of the decentralised, real-world model is its ability to support more responsive trials. As data is available in real time, sponsors are able to identify emerging patterns during the study. That means that, subject to appropriate regulatory and ethical approvals, they can adapt aspects of the trial, or investigate additional questions or endpoints.  

Overall, it is about generating richer, more representative datasets with the ability to impact real-world care. The Phase IIIb VICTORION-Spirit study, for example, not only assessed the efficacy of inclisiran for reducing LDL cholesterol, but also examined the practical barriers and enablers to integrating the treatment into everyday NHS care.vi,vii 

The time is now   

While advances in science and technology are transforming how we discover and develop medicines, the way we test them, in the large part, remains stuck in the past. 

Real-world data-enabled RCTs offer a way to close the gap, reducing the burden on patients and sites, accelerating recruitment, strengthening safety monitoring and generating evidence that better reflects how medicines are used in everyday care. 

The barriers that once made this approach difficult are increasingly being addressed. Regulators are providing clearer frameworks, technology can now connect routinely collected healthcare data with clinical trial systems, and more than a decade of experience has demonstrated that the model can work in practice. 

The question, then, is no longer whether real-world data can transform clinical trials, it is when will the industry take the leap?