Clinical Trial Patient Recruitment: Benchmarks and Strategies

Clinical trial patient recruitment is where most studies lose time. In an analysis of 2,542 randomized trials, 54.1% failed to reach their target enrollment, 79.6% finished late, and only 14.6% managed both on time and fully enrolled. The median delay was 12.2 months, and the median trial enrolled 31% fewer participants than planned [1].

The cost is measurable. Tufts CSDD puts a single delay day at roughly $500,000 in foregone sales plus about $40,000 in direct phase II and III trial costs [2].

This page collects what current research says about why clinical trial recruitment slips and what actually moves it, with every figure sourced. Refero does not run recruitment campaigns. We vet and match the clinical research marketing agencies that do, so what follows is written for the person deciding whether to hire one, and what to ask before they sign.

Clinical trial recruitment statistics at a glance

Every figure here is sourced in full at the foot of the page.

• 54.1% of trials miss their target enrollment, and only 14.6% finish both on time and fully enrolled [1]
• Median enrollment duration is 261 days, with a quarter of trials taking more than 14 months [3]
• A single delay day costs roughly $500,000 in foregone sales plus $40,000 in trial costs [2]
• Accrual failure accounts for 37.4% of all US clinical trial terminations [4]
• Protocols carry a median of 31 eligibility criteria, 9 inclusion and 22 exclusion [3]
• Phase III protocols now collect 5.9 million datapoints, up from 929,203 in 2012 [7]
• 46% of research sites report capacity constraints on taking new studies [8]
• 52% of people who have never participated say they were unaware of a trial right for them [9]
• 51% would prefer to hear about trials from their own doctor. 8% actually did [9]
• Only 6% of pivotal trials matched US population representation across all four largest racial and ethnic groups [11]

What clinical trial patient recruitment covers

Patient recruitment for clinical trials, also called clinical research patient recruitment, is everything between a finished protocol and an enrolled, consented participant: identifying people who might qualify, reaching them, screening them against eligibility criteria, and getting them to a site visit or a remote equivalent.

It is delivered three ways, and the distinction determines who you are actually hiring.

• Site-led. Investigators recruit from their own patient populations and referral networks. Cheapest per patient when the site has the right population, and the default for most trials.
• Sponsor or CRO-led. Central recruitment run across all sites, usually combining media, a call center, and a referral portal.
• Vendor-led. A specialist agency or platform is contracted for media buying, patient identification, prescreening, and referral handoff, either from the start or as rescue recruitment once a trial is behind.

Most trials of any size use a mix. The question is rarely whether to buy recruitment help, but which part of the funnel to outsource.

Clinical trial enrollment timelines: how long recruitment actually takes

Very little published material gives sponsors a benchmark to hold their own timeline against. The best current one comes from 1,286 industry-sponsored phase 2 and 3 trials completed between January 2017 and September 2024: median enrollment duration was 261 days, with an interquartile range of 151 to 424 days [3].

So roughly 8.5 months is the middle of the distribution, and a quarter of trials took more than 14 months to enroll. If a plan assumes materially less than that, the plan is usually what needs revisiting, not the vendor.

Termination data gives the other end of the range. Of 53,263 US interventional trials, 5.59% were terminated specifically because they could not accrue, and accrual failure accounted for 37.4% of all terminations. The authors note this likely understates the real rate, because registry records often do not record a reason [4]. In oncology, 27.1% of discontinued trials ended because of insufficient enrollment [5].

Clinical trial recruitment challenges: why enrollment slips

Eligibility criteria have roughly doubled

The same 2026 analysis found a median of 9 inclusion and 22 exclusion criteria, 31 in total, and identified the number of inclusion criteria as the strongest protocol-design predictor of longer enrollment [3].

The trend is documented. Across NCI-affiliated trials, the median length of combined eligibility criteria rose from 214 unique content words in 2008 to 417 in 2018. Each decile of that growth raised the odds of accrual failure by 9%, and 19.3% of those trials failed to accrue [6].

Protocols collect far more than they used to

Phase III protocols now average 19 endpoints at database lock and 5.9 million datapoints per protocol, up from 929,203 in 2012, a compound growth rate of 15.4% a year. Procedures per participant visit are up 37% over 15 years, and sites per protocol up 63% [7].

Every added procedure is another reason for a participant to decline or drop out, and another minute of site staff time not spent recruiting.

Sites are out of capacity

In a 2024 survey of 852 research sites, 46% reported challenges restricting their capacity to take on new studies and 47% had agreed to fewer studies than the year before. Recruitment and retention was the single biggest challenge for 36% of them [8].

This is the failure sponsors most often misdiagnose as a vendor problem. A site with no bandwidth does not recruit faster because there is more media spend behind it.

Most patients never hear about a trial they could join

The most common reason people give for never having participated is not fear or mistrust. It is that they were not aware of a trial that was right for them, cited by 52% of respondents in a 2025 survey of 12,887 people [9].

Among those who did try to join one and did not, 63% were screened out on eligibility, and 82% of those were still looking for another trial [9]. That is a large, motivated, already-educated population lost at the screening step.

How patients find trials, and how they want to

The gap between those two is the most actionable finding in current recruitment research, and it explains why physician-referral strategies keep underperforming their reputation.

Share of people who would prefer to hear about a trial through each channel, set against the share who actually did [9]:

• Their own doctor: 51% would prefer, 8% actually saw
• Research center doctor or staff: 47% would prefer, 16% actually saw
• Online: 39% would prefer, 58% actually saw
• Patient advocacy group: 30% would prefer, 17% actually saw
• Nurse at their doctor's office: 26% would prefer, 4% actually saw
• A patient recruitment company: 22% would prefer, 15% actually saw

Patients overwhelmingly want to hear about trials from their own physician, and almost never do. Online is the reverse: fewer want it, far more get it [9].

Within that online share, social media accounted for 62% of where people saw a trial, and Facebook for 85% of that. Among people who actually enrolled, 22% first learned about their study online, and of those just 10% came through a government registry. That works out to roughly 2% of all participants arriving via a registry, which is worth knowing before a recruitment plan leans on registry listings [9].

Clinical trial recruitment strategies with evidence behind them

Most recruitment tactics are supported by vendor case studies rather than controlled comparisons. Two things have better evidence than that.

AI-assisted prescreening has been tested in a randomized trial. Screening 4,476 patients for a heart failure trial, an LLM-based tool identified eligible candidates at 20.4% versus 12.7% for manual review, and produced enrollment of 1.6% versus 0.9%. The speed difference was larger than the yield difference: more than 99% of eligible patients were identified within 15 days, against 50 days manually [10]. This is the strongest single piece of recruitment evidence currently published, and it is about screening throughput rather than reach.

Reducing participation burden is what participants themselves rank highest. Asked which services would help most, participants chose home or near-home visits (76%), transportation (72%), caregiver support (70%), compensation (68%), and video visits (67%) [9].

Patient recruitment and retention in clinical trials

Patient retention in clinical trials is usually bought separately from recruitment, and the two fail together. A trial that enrolls on time and then loses participants has not solved anything, and largely the same burden factors drive both.

There is no credible industry-wide dropout rate, which is covered below. What does exist is good data on what makes participants stay. Asked which parts of taking part were most burdensome, participants named biopsies (34%), traveling to the clinic (28%), and diagnostic tests (24%). Travel is the most fixable of those: 35% spent more than an hour getting to a visit each way, and 8% more than two hours [9].

Compensation is the gap nobody discusses. One in three participants received no compensation or reimbursement of any kind, and nearly half of those in Europe [9].

The encouraging half of the picture is that participation itself is not what drives people away. Among participants surveyed, 95% said they would take part in another study and 93% would recommend it to someone else [9]. People who get through the door largely stay, and would come back. The losses happen before enrollment, and in the logistics after it, rather than in the experience of the research.

Diversity and representation

Representation is both a regulatory expectation and a recruitment constraint, and the current numbers are poor. Across 341 phase III pivotal trials supporting FDA approvals from 2017 to 2023, only 6% achieved enrollment aligned with the US population distribution of all four largest racial and ethnic groups. In 2023 trials, Black representation was appropriate in 23%, Hispanic in 30%, and Asian in 81% [11].

Where the gap comes from matters for anyone buying recruitment. A study of cancer trials found they enrolled 45% fewer Black patients than proportional representation would predict, and that the choice of enrolling region and practice accounted for more than half of that shortfall [12]. Much of the disparity is therefore set by site selection, before any recruitment media runs.

One note on sourcing, because it is widely misreported: the FDA publishes representation in its Drug Trials Snapshots as ranges by therapeutic area, not as single aggregate percentages [15]. Any claim of the form "the FDA says X% of trial participants were Black" is misreading the source.

Three numbers to stop repeating

Recruitment content recycles a handful of figures that do not survive a look at their sources. If a vendor's deck leans on these, it tells you something about the rest of the deck.

"$19,533 to replace a patient who drops out." Follow the citation chain and it ends at a sponsored trade post and an opinion column, neither containing original research. The companion claim that 15 to 40% of patients drop out comes from the same chain. There is no primary source for either.

"20% of research sites enroll zero patients." This traces to an unsourced conference slide from 2000. The measured figure is 11%, from a Tufts CSDD study of 16,000 sites across 151 global phase II and III trials, which exists partly to correct the 20% claim. That study also found 48% of sites under-enrolled and 59% met or exceeded target [14].

A single industry-wide dropout rate. There is not one. An umbrella review of 88 systematic reviews across 33 conditions declined to report a pooled figure, noting that publicly funded trials typically lose up to 12% but rates as high as 70% have been reported [13]. Anyone quoting one number for all trials is inventing it.

How to choose a clinical trial recruitment agency or vendor

Drawn from the failure modes above rather than from a generic vendor checklist.

1. What is your median time from campaign launch to first randomized patient in my therapeutic area? Referrals are easy to produce and randomizations are not. Insist the metric be randomization.
2. What share of your referrals historically clear site prescreening? A high referral count with a low qualification rate transfers work to sites that are already short on capacity.
3. Have you worked on a protocol with criteria like mine? Share the actual inclusion and exclusion list. At a median of 31 criteria, general experience in the indication is not the same as experience with your protocol.
4. How do you handle screen failures? Since 82% of people who fail screening are still looking, a partner with no plan for them is discarding your most motivated population.
5. What are you assuming about site capacity? If the plan does not account for sites already turning down studies, media spend will outrun the ability to consent people.
6. How do you reach patients whose doctors are not raising trials with them? With 51% wanting to hear from their physician and 8% doing so, this is the actual gap being bought.
7. Is your representation strategy site selection or media? More than half the disparity in one large study came from where trials were placed, not how they were advertised.
8. Which of your numbers are sourced? Ask for the provenance of any percentage in the pitch, then check one.

If you would rather not run that evaluation alone, Refero screens healthcare marketing and patient recruitment agencies against five published criteria and introduces up to three that fit your brief. It is free for buyers, and there is no obligation to hire anyone we introduce. Tell us what you need.

Frequently asked questions

What is patient recruitment in clinical trials?

Everything between a finished protocol and an enrolled, consented participant: finding people who may qualify, reaching them, screening them against the protocol's eligibility criteria, and getting them enrolled. It is delivered by sites, centrally by the sponsor or CRO, by specialist vendors, or by a combination of all three.

How long does patient recruitment take?

Across 1,286 industry-sponsored phase 2 and 3 trials completed between 2017 and 2024, median enrollment duration was 261 days, roughly 8.5 months, with an interquartile range of 151 to 424 days. A quarter of trials took more than 14 months.

What are the biggest clinical trial recruitment challenges?

The measured ones, in rough order of impact: eligibility criteria that have roughly doubled in length since 2008, protocol complexity that now runs to 5.9 million datapoints per phase III study, research sites that are out of capacity with 46% reporting constraints on new studies, and a public awareness gap in which 52% of people who have never participated say they simply did not know of a trial that fit them.

How do you improve patient recruitment in clinical trials?

Current evidence points at three things: simplify eligibility criteria, since inclusion criteria count is the strongest protocol predictor of slow enrollment; reduce participation burden through home visits, transportation, and remote options, which participants rank highest themselves; and use structured prescreening, where AI-assisted review beat manual review in a randomized trial on both yield and speed.

How do you recruit patients for clinical trials?

In practice through a mix of site referrals from investigators' own patient populations, digital media (where social accounts for the majority of what patients actually see), patient advocacy partnerships, registry listings, and EHR-based identification at the site level. Most trials of any size combine several channels.

How do you choose a clinical trial recruitment vendor?

Judge on randomizations rather than referrals, on referral-to-prescreen qualification rate, on protocol-specific rather than indication-general experience, and on whether their published performance figures have sources you can check. The eight questions above cover it.

How does AI improve patient recruitment?

The one randomized test published so far compared LLM-based prescreening with manual review across 4,476 patients. It identified eligible candidates at 20.4% versus 12.7%, enrolled 1.6% versus 0.9%, and found more than 99% of eligible patients within 15 days rather than 50. The gain is in screening throughput, not in reaching more people.

Why is patient recruitment important in clinical trials?

Because it is the most common way trials fail on time and on budget. In one large analysis 54.1% of trials missed their enrollment target, accrual failure accounts for 37.4% of all US trial terminations, and Tufts CSDD estimates a delay day costs roughly $500,000 in foregone sales plus $40,000 in trial costs.

Sources

1. Shadbolt C, et al. Analysis of Rates of Completion, Delays, and Participant Recruitment in Randomized Clinical Trials in Surgery. JAMA Network Open, 2023. Link
2. Smith ZP, DiMasi JA, Getz KA. New Estimates on the Cost of a Delay Day in Drug Development. Therapeutic Innovation & Regulatory Science (Tufts CSDD), 2024. Link
3. Tsuchiya H, et al. Association of Protocol Design-Related Factors With Enrollment Duration in Clinical Trials. Clinical and Translational Science, 2026. Link
4. Ma S, et al. Predicting accrual success for better clinical trial resource allocation. Scientific Reports, 2025. Link
5. Wu B, et al. Analysis of discontinued radiotherapy trials registered on ClinicalTrials.gov. Radiotherapy and Oncology, 2025. Link
6. Peterson JS, et al. Growth in eligibility criteria content and failure to accrue among NCI-affiliated clinical trials. Cancer Medicine, 2023. Link
7. Getz K, et al. Insights Informing Strategies for Optimizing the Collection of Clinical Trial Data. Therapeutic Innovation & Regulatory Science (Tufts CSDD and TransCelerate), 2025. Link
8. WCG Clinical. 2024 Clinical Research Site Challenges Report. Link
9. CISCRP. 2025 Perceptions and Insights Study. Link
10. Unlu O, et al. Manual vs AI-Assisted Prescreening for Trial Eligibility Using Large Language Models: A Randomized Clinical Trial. JAMA, 2025. Link
11. Longitudinal clinical trial enrollment trends across 341 US FDA-approved drugs. Communications Medicine, 2025. Link
12. Shah SJ, et al. The role of regional and practice trial sites in distorted randomized cancer trial enrollment. Journal of the National Cancer Institute, 2025. Link
13. McChrystal R, et al. Participant and trial characteristics reported in predictive analyses of trial attrition. Trials, 2025. Link
14. Tufts CSDD. Enrollment Performance: Weighing the Facts. Applied Clinical Trials, 2012. Link
15. US Food and Drug Administration. Drug Trials Snapshots. Link

Refero publishes no performance figure it cannot attribute. If you find one on this page without a source, that is an error and we want to hear about it: hello@referohq.com.