18-19 Physician-driven, AI-supported: addressing the clinical trial access gap

With more than 500,000 clinical trials1 currently registered worldwide and over 8,0002 drugs in development, the clinical research ecosystem has never been more complex and healthcare professionals (HCPs) are feeling the strain.

Advances in science and technology have deepened the medical community’s understanding of many diseases and their subtypes, helping to pave the way for new trials and new therapies. But this progress has also come with added complexity: trial designs are increasingly sophisticated, eligibility criteria are more intricate and more restrictive, and access to trial information has become fragmented across registries and regions. Navigating this evolving landscape is a growing challenge, both for patients and the HCPs at the frontlines of their care.

Identifying and pre-screening patients for relevant clinical trials today is still dominated by manual, time-intensive processes. This makes access to trials more challenging for many patients and can slow the pace of medical innovation. In fact, a survey of US HCPs conducted recently by myTomorrows found that 72% of respondents believe searching for relevant clinical trials takes too much time3 – time that is in short supply due to the heavy workloads of an already thinly-stretched clinical workforce.

In other words, the current state of the clinical trial ecosystem is one in which HCPs, despite their commitment to patient care, lack the appropriate tools and support needed to efficiently match patients to trial opportunities.

These complexities should never limit access to clinical trials for patients and HCPs or the advancement of emerging therapies. To reduce friction in the search for investigational treatments, it is critical to understand and address the systemic challenges HCPs face.

This article outlines four of the primary barriers that hinder HCPs from connecting patients to clinical trials: fragmented trial registries; inadequate pre-screening tools; delays in patient referrals, and limited familiarity with AI tools for identifying trials.

It also explores actionable solutions that may help streamline how HCPs find and connect with relevant trials and examines how AI is positioned to supercharge the trial discovery and access process for patients, HCPs and biopharma.

HCPs trying to discover and connect patients with clinical trials are tasked with navigating a slew of complex, siloed registries and platforms, each with their own unique interface and search criteria.

Of the HCPs who participated in the recent survey, 56% indicate that they are required to consult too many different systems to identify potential trial matches3. Nearly a quarter said that the information in these registries was outdated or inaccurate.

Some registries, for example, only include trials conducted in specific countries, meaning information on cross-border trials isn’t surfaced. Other registries – such as ClinicalTrials.gov. the EU Clinical Trials Register, WHO ICTRP and more – have no precedent for sharing data and sources, so HCPs are forced to embark on broad, manual searches. Disconnected data sources can result in isolated systems and inefficiencies that lead to real-world delays. For HCPs, this can limit their ability to identify additional care options – for their patients, it can significantly influence the opportunities those patients have to participate in research.

Fortunately, emerging AI-powered trial-matching platforms are making their way into the hands of clinical professionals, allowing them to combine data from various registries into a single search process. By integrating this information directly with electronic health record systems, these AI tools can also help HCPs identify relevant trials more easily during their routine patient care. Paired with ongoing efforts to standardise data fields across global registries and encourage timely updates from trial sponsors, the accuracy and usability of these platforms is continually improving.

Even when HCPs are able to locate relevant trials for their patients, assessing eligibility can be just as time-consuming and difficult to navigate. It often requires comparing multiple studies with distinct and complicated eligibility criteria.
Determining eligibility for clinical trials is based on a patient’s unique medical profile, factoring in a sophisticated combination of elements, such as genetic biomarkers, lab values, comorbidities and treatment history. In addition, many pre-screening tools require structured data that is not always available in electronic health records, especially when HCPs are working with PDFs, scanned files or handwritten notes. The result is a mountain of manual work that can slow both trial referrals and enrolment.

It’s unsurprising then, that the aforementioned survey found that 60% of HCPs see current clinical trial pre-screening tools as too complicated or insufficient3.

Moreover, pre-screening and eligibility testing tools must be updated to reflect the latest FDA guidelines. By keeping these tools current and adaptable to evolving guidelines and trial designs, all stakeholders can potentially gain improved access to trials.

After identifying a relevant clinical trial match, referring a patient to a recruiting trial site is the next step. However, often the referral process is a significant hurdle that delays a patient’s potential enrolment. In fact, over half of the HCPs surveyed (56%) view the referral processes as disjointed and slow and 41% say they need to use multiple tools to make a single referral3.

The consequences can be detrimental. A recent analysis of over 2,500 randomised surgical trials found that just one in five were completed on time, with median delays exceeding 12 months. Nearly half failed to reach their recruitment targets and over 20% ended early, often after enrolling less than 10% of their intended participants4. These delays reflect deep structural gaps in how patients are identified, referred and enrolled and indicate the need for referral process improvements.

One possible solution is to combine comprehensive pre-screening tools directly with HCPs’ workflows, for example through integration with preexisting EHR systems. This implementation would allow HCPs to proactively flag trial eligibility based on clinical and demographic criteria, simplifying the referral process and reducing the administrative burden for HCPs.

AI-powered trial-matching tools are quickly emerging as a valuable support for HCPs in their search for pre-approval options. While 61% of HCPs say they find existing GenAI tools to support their clinical trial search to be trustworthy, these tools are still new and often misunderstood. More than one-third of HCPs remain hesitant to use them, citing limited familiarity, uncertainty about how the tools work and concerns around data privacy and compliance3.

Building confidence in these powerful tools will require thoughtful investment in HCP training and education. HCPs need hands-on guidance to understand how these tools work, how to interpret their outputs and how to apply them responsibly in patient care. Without this foundation, even the most advanced AI will remain a black box that many HCPs are reluctant to use.

However, it’s not just HCPs who must readjust their workflows to adapt. The tools themselves must be developed with HCP workflows in mind, offering clear clinical utility rather than added complexity. Healthcare systems need to implement clear internal policies for responsible AI use, and regulatory bodies must establish consistent oversight and guidelines so that AI solutions meet the highest standards of patient privacy and data protection.

Physician-driven, AI-supported: addressing the clinical
trial access gap

To open up more clinical paths for patients, HCPs need simpler search tools, better access to information and integrated referral systems that fit into their clinical workflows

Disconnected data and fragmented registries

Pre-screening pitfalls

Delayed referrals, missed opportunities

AI tools for precision trial matching