Recruitment in the Age of AI: Q&A with John Worden, Javara

Clinical trial recruitment has long been built around the study, not the patient—a model that creates artificial barriers between research and routine care at exactly the moments when the two should be most connected.

To explore this further, Applied Clinical Trials spoke with John Worden, chief commercial officer at Javara, about why AI alone won't fix recruitment, what responsible patient-facing AI actually requires in a clinical context, and how embedding research access into existing clinical workflows could change what patients are even offered the chance to consider.

ACT: Why has clinical trial recruitment stayed largely one-size-fits-all even as AI is enabling more personalized approaches across the rest of healthcare?

Worden: Clinical trial recruitment has historically been built around the study rather than around the patient. We activate a protocol, identify sites, establish enrollment targets, and then begin looking for patients who meet the criteria. That process has resulted in recruitment strategies that are often broad, one-off, and disconnected from how patients actually receive care.

AI gives us an opportunity to reverse that model. Instead of asking, "How do we find patients for this trial?" we can increasingly understand individual patient populations, their clinical journeys, and which research opportunities may be relevant to them. At Javara, we are already using EHR intelligence to support feasibility, patient identification, and trial matching, while our recruitment technology can simultaneously evaluate patients across multiple studies.

Technology alone will not solve recruitment. The real opportunity is combining that intelligence with the trusted relationship a patient already has with their physician. The future is not just AI-enabled recruitment; it is personalized research access embedded into healthcare.

ACT: Where do the biggest avoidable barriers to trial awareness and participation emerge when research isn't connected to routine care?

Worden: When research operates separately from routine care, we create an artificial divide. The physician may know the patient's history, treatment journey and unmet needs, while the research organization knows about an appropriate clinical trial, but those two worlds don't consistently connect. This creates missed opportunities at exactly the moment when research may be most relevant to the patient.

The reality is that traditional research models operate as an entirely separate destination, adding burden, and often confusion, to both patients and physicians. We’re working to change that.

When research teams are integrated into clinical workflows, patients gain the ability to learn about research from people and organizations they already trust as physicians genuinely evaluate research alongside other care options. Our experience has been that integration reduces friction between clinical research and clinical care and creates more natural opportunities for physician and patient engagement.

Many will be quick to point toward patients’ willingness to participate as the largest barrier today. I would say otherwise. I believe the most powerful barrier facing our industry today (and one that could be avoided) is not about patients’ willingness to participate, but about our collective, historical failure to make these opportunities known, understood, and accessible at the right point in their care journey. It’s time to change that.

ACT: How can AI help identify and support patients for trials without displacing the clinician relationships and human judgment that build trust?

Worden: I think we need to be very clear about AI's role: AI should inform and enable the relationship, not replace it. AI is exceptionally valuable at doing things humans cannot efficiently do at scale, for example interrogating large amounts of clinical data, interpreting unstructured information, identifying potential matches against complex eligibility criteria, and helping research teams prioritize where to focus their attention.

Javara queries EHR data against study inclusion and exclusion criteria, analyze patient populations, support feasibility and identify potential patients for recruitment. We can then surface potential research opportunities within clinical workflows.

But identifying a patient is very different from helping that patient decide whether research is right for them. That is where physicians, research professionals, and ultimately the patients’ perspectives remain essential.

I do not envision a future where an algorithm tells a patient, this trial is right for you; but rather one where technology tells the care team, there may be an opportunity here that otherwise would have been missed. The clinician and research team can then bring context, judgment and empathy to the conversation.

ACT: What does responsible patient-facing AI actually require in a clinical research context?

Worden: I think at its core, technology of this nature would require that AI recognizes that clinical research is fundamentally different from consumer engagement. We are dealing with health information, potentially vulnerable patients and decisions that can carry significant consequences.

Responsible AI therefore starts with privacy, security, transparency and appropriate governance. We also need to understand how algorithms perform across different populations, continuously evaluate accuracy, maintain appropriate human oversight and be very clear with patients about how technology is being used.

It may be bold to stay in today’s AI-obsessed environment, but I strongly believe that AI should never be the final decision-maker in patient care or clinical trial participation. AI can identify, prioritize and personalize. Like I said, it is an amazing tool. But qualified humans should always be involved to validate, communicate, and make (often critical) clinical judgments. If AI, even responsible AI, assumes that level of authority in the future, to me that would represent a collective failure to protect the patients we serve and their ability to make informed decisions about their health and trial participation.

The default should not be: Can AI do this? The default should be: Does using AI in this way improve the patient's experience while protecting their trust?

ACT: What does the next generation of individualized, care-integrated patient engagement need to look like for health systems, sponsors, and sites?

Worden: One of the main sentiments we repeat continuously at Javara is to start with the end in mind. For us, success isn’t just measured by whether we enrolled a trial faster. Of equal importance is whether more patients had an opportunity to consider clinical research through a physician they trust, in a community they love.

From my perspective, next-generation engagement will be built on the AI-insights and clinical data we’re gathering today, as we gain a greater ability to understand patient populations continuously, identify relevant research opportunities earlier, and integrate those opportunities naturally into care. The physician should be able to see that a patient may qualify for research within their existing clinical workflow; the patient should receive information that is relevant and understandable; and a research professional should be available to guide that patient through the process.

Javara's current model is moving in this direction by combining EHR analytics, physician engagement, personalized communications, community outreach and dedicated patient engagement capabilities. We leverage technology to simultaneously prescreen patients across multiple studies rather than treating each protocol as an isolated recruitment exercise.

There is benefit for health systems: making research a more accessible part of care. There is benefit for sponsors: Gaining more predictable and reliable access to qualified patients. And there is benefit for sites: moving away from episodic recruitment campaigns toward an always-on understanding of their patient populations.

I think at its core, recruitment in the Age of AI isn’t about whether technology can solve it all. It is about leveraging AI to fix patient access without losing the human connection.