Clinical trials fail in recruitment

A few years ago, I hopped in an Uber in Stafford, Virginia headed to Washington Reagan Airport. The driver and I got into conversations about local changes in the area over time, before he asked, “What do you do for a living?” I explained that I worked in clinical research, educating marginalised communities about the now and future benefits of contributing to clinical research. The tone of the conversation changed. He very transparently said, “I don’t know a lot about it, but I’d rather let other people do it.”

That sentence stayed with me. Not because it was unusual, but because it wasn’t. In every community I worked in during my time leading patient engagement, I’ve heard some version of the same, exposing how exclusive education and access still are. The populations that might benefit most are too often the ones no one has tried to engage or identify to start a purposeful conversation to explain that medical treatments react and respond differently to different people. To ensure that everyone, not just certain demographics, can receive safe and effective treatments requires that all populations contribute.

What that driver didn’t see is what most patients have never been told: the medications cleared through trials are only as representative as the people who participate. When a community opts out, the therapies that follow may not have been validated in the very people most likely to one day need them. The cost of “let others participate” gets handed down, generation after generation, in the form of treatments never studied on people like those receiving them.

Clinical trials struggle in many areas, and significantly in recruitment.

The 80% problem

Industry analyses consistently estimate that roughly 80% of clinical trials miss their original enrolment timelines. The Tufts Center for the Study of Drug Development has flagged enrolment as the single largest source of trial delay for years. Each missed week translates into delayed treatments for patients who are waiting and can run into millions of dollars in lost value per day for a sponsor.

The 80% statistic, while striking, is also a little numbing. It abstracts away what is actually happening. Behind every delayed enrolment timeline is a person who could have been contacted, but wasn’t because three days passed before anyone followed up on their interest form, because their physician didn’t know about the protocol, because the screening call came at 2pm on a Tuesday when they were at work. The list goes on.

We have spent the last decade improving almost every part of the trial value chain: protocol design, decentralised monitoring, eConsent, real-world evidence. Yet, the front door to getting the right patient to compatible studies at the right moment has barely budged.

Recruitment is the bottleneck, not a marketing line item

Recruitment, as practiced in most of our industry, is reactive. We post a study on a registry and wait. Interested patients fill out a static form and wait for a callback. We hand a coordinator a 200-page protocol and a list of charts and ask them to work through it. We launch a digital ad campaign and hope the funnel converts. When enrolment slips, we add sites, which dilutes per-site numbers and rarely solves the underlying problem.

Recruitment is a relationship problem layered on top of an awareness challenge. People have to be engaged in a way that respects their time and their context. And you have to be there when they have a question, not three days later when the call-back queue finally clears.

For most coordinators I’ve worked with, that is an impossible expectation. A typical site coordinator juggles multiple protocols, fields hundreds of inquiries a month, and is asked to be a clinical screener, scheduler, trainer, and concierge. It isn’t that they don’t care. It’s that the system asks them to care on a superhuman scale.

Where AI helps and where it doesn’t

This is where the conversation about AI in clinical research has started to mature, and where I think we need to keep pushing.

The early framing of AI in recruitment was lazy: “AI will identify more patients.” That isn’t wrong, but it misses the bigger shift. The real opportunity is moving from a reactive funnel to a proactive process one where eligible patients are identified, engaged, prescreened, and scheduled for time with the coordinator before a coordinator has to dig for them, where outreach is timed to when a patient is actually ready, and where the routine work that fills a coordinator’s day is taken off their plate.

AI is not the recruiter. It is what lets the human recruiter finally show up as a human doing the work that actually requires one.

The most meaningful applications I see are not flashy. They are systems that scan thousands of charts overnight and surface the ones that warrant a coordinator’s attention. They are 24/7 conversational tools that answer a patient’s basic eligibility questions at 9pm, so, when they speak with a coordinator the next morning, they are ready. They are workflows that re-engage a patient who fell out of a screening funnel six weeks ago because life got in the way, instead of letting that person quietly disappear.

On the ground, this looks like a coordinator who finally has time to walk a patient through informed consent, answer their questions, and address scepticism with the empathy and context only a human can bring. Augmentation, not replacement. The technology earns its keep by giving back time to the people who do the irreducibly human work of enrolment.

The path forward

The 80% number won’t move because we built one more dashboard or sent one more mass email. It will move when our industry decides that recruitment is a first-class problem worthy of first-class infrastructure, and when we use AI to support and amplify the people already doing this work.

We finally have the tools to educate and engage people where they are. The question is whether we use those tools to extend the reactive system we already have, or to build the proactive one our communities and patients have always deserved.

My response to the driver in Stafford shared my thoughts as communicated here and connected the choice in front of him to the medicines his own family might already be taking or one day need. There are patients still waiting to be invited into a trial that could change not just their lives, but the standard of care for everyone who follows.

About the author

Niambi Blodgett is a healthcare leader dedicated to advancing equity. Based in Orlando, she has built her career at the intersection of technology, community engagement, and clinical operations. Bldgett currently serves as senior director of customer advocacy at Areti Health, where she leads AI-enabled patient engagement strategies. Before this, she held multiple leadership roles at EmVenio Research and Matrix Medical Network, where she built national recruitment strategies, launched decentralised trial capabilities nationally, built new business units, and led initiatives that achieved 44%+ diversity across research portfolios. Her career began in national security investigations with the US government, where she managed investigative teams across Washington, DC and Central Florida before transitioning into healthcare. A recognised thought leader, Blodgett has shared her insights on platforms like Xtalks, CAN TV, the Health Care Council of Chicago, and Health Equity Congress, focusing on AI solutions, strategic partnerships, and patient engagement.