Talk to us

Why interested patients never reach randomization

Enrollment shortfalls in clinical trials are usually diagnosed as a shortage of willing patients, yet many interested and eligible people are lost after they first come forward. Leading sponsors and sites are beginning to manage every referral, prescreen and screening visit as an obligation with a named owner and a daily cost.

Clinical researchSegeta Insights
Clinical research
The number$$840K

a day of enrollment delay for a sponsor

What is at stake

A 30-day enrollment slip on a registration trial can put about $$25 million at stake.

Where days are lost
  1. Referral or inquiry to first contact
  2. Prescreen to screening visit
  3. Screening to randomization

Clinical trial patient recruitment is usually managed as a demand problem. When a study falls behind its enrollment curve, sponsors add sites, widen advertising and commission outreach to communities the protocol has not yet reached. That response assumes the binding constraint is the number of patients who want to take part, and a growing body of evidence suggests that the constraint sits elsewhere.

Patient interest is rarely the scarce input in a modern trial. In a 2023 survey by the Center for Information and Study on Clinical Research Participation, 91 percent of people who had never joined a trial said they were interested in doing so, yet a 2024 analysis by Tufts CSDD and Ancora.ai researchers reports that 69 percent of patients identified as interested are never prescreened.1 The loss happens in the interval between a referral, a prescreen and a screening visit, where the referring physician, the recruitment vendor, the call center and the site coordinator each hold part of the patient's record and none owns the next call. The sections below trace where that interval costs a study its days, how an engagement model that keeps site staff accountable can close it and what each day of delay is worth.

Enrollment pressure is rising as protocols grow more demanding

Most trials eventually reach their enrollment target, and they pay for it in time. A Tufts CSDD analysis of more than 150 trials and nearly 16,000 sites found that 89 percent of trials met their enrollment goals, yet most took nearly twice as long as planned to get there.2 The same research found that 11 percent of activated sites never enrolled a patient and nearly half under-enrolled or enrolled none, while proprietary benchmarks placed roughly one third of sites at zero or one patient; these foundational figures remain the reference points the industry uses.3 More recent Tufts research, cited in 2024, reports that only 60 percent of activated sites in a trial enroll a patient.1

The path from referral to randomization has also narrowed. Tufts CSDD found that the average screen failure rate in Phase II and III trials completed between 2016 and 2019 reached 36.3 percent, and 57 percent in central nervous system studies, while the mean number of procedures in a Phase III protocol has risen from 187 to 301 over the past decade.45 The Food and Drug Omnibus Reform Act of 2022 adds a further obligation, because it requires sponsors of Phase III and other late-stage registration studies to submit diversity action plans with enrollment goals by race, ethnicity, sex and age.6 The requirement applies only to studies that begin enrollment 180 days after FDA finalizes its guidance, which was withdrawn from FDA's website in January 2025, yet the statute remains in force and compliant sponsors will need to show who was referred, prescreened and screened.7

Clinical trial patient recruitment loses most patients at three handoffs

A recruitment funnel typically crosses three organizations before a patient is randomized: a referring physician or advocacy group, a vendor or call center, and the site itself. In a pilot by Tufts CSDD and Ancora.ai covering 133 cancer patients seeking trials, only 62 inquiries to trial contacts drew any response and only 7 patients were successfully referred.1

Of 133 patients helped to seek a cancer trial, only 7 were successfully referred.

Exhibit 1Of 133 patients seeking a cancer trial, only 7 were successfully referredPatients remaining at each stage of a trial inquiry pilot, number of patients
Patients supported133
Inquiry drew a response62
Actionable information53
Accurate information26
Successfully referred7

Source: Ralic, Getz et al., DIA Global Forum, Last Mile Leak pilot, 2024

01Referral or inquiry to first contact

A referral arrives by fax, portal, email or web form, and the site coordinator who receives it is often managing several studies and a full visit schedule. The pilot found that contact details listed on ClinicalTrials.gov produced a response for only 35 percent of patients, and about half of the responses that did arrive carried inaccurate information.1 Each day without first contact is likely to lower the chance that the patient, or the referring physician, will try again.

02Prescreen to screening visit

A prescreen in the site's CRM or a vendor's call center tool establishes likely eligibility, yet the screening visit is booked in a separate scheduling system, often after medical records arrive from another provider. In the same pilot, 9 of the 26 patients who received accurate information decided not to proceed, a decision that slow scheduling and repeated paperwork may make more likely.1 Planning benchmarks suggest that a typical trial needs about 12 available subjects for every patient randomized, so losses at this step compound quickly.8

03Screening to randomization

Screening procedures, laboratory results and investigator review sit in the EDC, the lab portal and the coordinator's inbox, and a patient can wait days for the result that decides eligibility. With average screen failure above one third, sites that cannot rescreen or redirect patients promptly lose both the patient and the coordinator hours spent on the visit.4

A connected engagement model gives every referral an accountable owner

Leading sponsors and sites are beginning to manage each referral as an obligation with an owner, a deadline and a cost for every day it stays open. A live layer reads the systems already in use, including the CTMS, EDC, site CRM, scheduling, email and patient messaging channels, and runs a four-step loop:

  1. Detect

    a referral that has gone without first contact past the agreed window, such as a web inquiry still unanswered after two business days.

  2. Explain

    why it stalled and what it affects, such as records still awaited from the referring clinic, and the site's gap to its enrollment and diversity goals.

  3. Recommend

    the next step and its owner, such as a callback by a backup coordinator, a records request drafted for the referring physician or a match to another open study.

  4. Execute and track

    each action to a booked screening visit or a documented outcome, measuring days from inquiry to randomization by site and by population.

An AI-native engagement surface can answer a patient's questions about the study in plain language, collect prescreening answers, book visits and send reminders across text, voice and messaging apps, drawing only on IRB-approved materials. Artificial intelligence does the reading, linking and drafting, while the site coordinator and the investigator remain accountable for every eligibility judgment, consent conversation and clinical decision. Sponsors and CROs can see the same record, so a study lead can tell which sites are losing patients after referral and which simply lack them.

A day of enrollment delay can cost a sponsor about $$840,000

Cost of delay in clinical development is the daily value lost when a trial milestone slips, combining the direct cost of running the trial for another day with the prescription sales a product forgoes for each day its approval moves out.

A 30-day enrollment slip on a registration trial can put about $$25 million at stake.

Exhibit 2. Estimated daily cost of an enrollment slip on a Phase II or III trial

ComponentBasisEstimate per day
Direct trial operating costTufts CSDD mean direct daily cost for Phase II and III trials across therapeutic areas, in 2023 dollars 9About $$40,000
Prescription sales lost or deferredTufts CSDD estimate of sales value per day of delay for a product that reaches the market; assumes enrollment sits on the critical path to submission 9About $$800,000
Total cost per day of enrollment delaySum of the two componentsAbout $$840,000

Source: Segeta estimate based on Tufts Center for the Study of Drug Development data; assumptions are illustrative.

On these assumptions, a 30-day enrollment slip costs about $$25 million and a 60-day slip about $$50 million, of which $$1.2 million and $$2.4 million are direct trial costs. Phase III trials run higher, at a mean of about $$55,700 a day in direct costs, while the sales figure applies in full only to products that are ultimately approved, so a risk-adjusted value for an asset still in development would be lower.9 Adding sites late in a study can recover part of the slip, yet each new site carries its own activation time, contract and budget.

What study leads and site directors can do now

  1. 01Measure the time from referral to first contact

    Report the days between each inquiry and the first two-way conversation, by site and by referral source, and agree a target window with every site.

  2. 02Name an owner for every open referral

    Assign each inquiry to a coordinator with a named backup, so leave, turnover or a full visit schedule does not leave a patient waiting for a response.

  3. 03Price enrollment slippage in dollars per day

    Attach the daily cost of delay to the enrollment curve, so a stalled referral queue is discussed with the same urgency as a late site activation.

  4. 04Track the funnel by population and site

    Record referral, prescreen, screening and randomization outcomes by demographic group, which builds the evidence a diversity action plan will require.

  5. 05Give patients one channel to the study

    Offer plain-language answers, prescreening, booking and reminders through text, voice and messaging, with site staff reviewing every eligibility decision.

Sponsors cannot create eligible patients or shorten a protocol's screening procedures on their own, yet they can control how quickly an interested patient reaches a person who can act. Studies that manage every referral as an owned, priced obligation are likely to enroll sooner with fewer sites, and they may find that the same records make diversity commitments easier to evidence.

Questions leaders ask

Why do clinical trials struggle to enroll patients?

Interest is rarely the main constraint, since surveys find most people who have never joined a trial would consider it. Many interested patients are lost between referral, prescreen and screening, where the referring physician, the recruitment vendor and the site each hold part of the record. Tufts CSDD research also finds high screen failure rates and many sites enrolling few or no patients.

What is a typical screen failure rate in clinical trials?

Tufts CSDD found an average screen failure rate of 36.3 percent across Phase II and III trials completed between 2016 and 2019, up from 34.7 percent in 2012. Rates vary widely by therapeutic area, reaching about 57 percent in central nervous system studies, and growing protocol complexity is likely to keep pushing them higher.

How many clinical trial sites fail to enroll a patient?

Foundational Tufts CSDD research found that 11 percent of activated sites never enrolled a patient and nearly half under-enrolled or enrolled none, while other benchmarks put about one third of sites at zero or one patient. More recent Tufts research cited in 2024 reports that only 60 percent of activated sites in a trial enroll a patient.

How much does a day of delay cost in a clinical trial?

A 2024 Tufts CSDD white paper estimates that a day of delay is worth about $$800,000 in lost prescription drug or biologic sales, and that Phase II and III trials carry about $$40,000 a day in direct costs. Together, a 30-day enrollment slip on a registration trial can put about $$25 million at stake for an approved product.

Notes and sources
  1. Danielle Ralic, Beatriz Monreal, Katie Vieyra, Ruby Madison Ford and Kenneth Getz, "Documenting the 'Last Mile' Leak in the Patient Recruitment Pipeline," DIA Global Forum, September 2024, globalforum.diaglobal.org.
  2. Regulatory Affairs Professionals Society, "Study: Clinical trial site activation and enrollment," reporting Tufts CSDD Impact Report, "89% of trials meet enrollment, but timelines slip, half of sites under-enroll," January 2013, raps.org. Cited as a foundational study.
  3. Kenneth Getz, "Enrollment Performance: Weighing the Facts," Applied Clinical Trials, May 2012, appliedclinicaltrialsonline.com. Cited as a foundational study.
  4. Kenneth Getz, "Can Recruitment and Retention Get Any Worse?," Applied Clinical Trials, December 2019, appliedclinicaltrialsonline.com.
  5. Abby Proch, "Endpoints, Procedures, And Deviations Are On The Rise. Why?," Clinical Leader, January 2026, clinicalleader.com.
  6. Foley & Lardner, "Clinical Trials: FDA Publishes Draft Guidance on Diversity Action Plans," July 2024, foley.com.
  7. AgencyIQ, "FDA pulls diversity action plan guidance, other diversity programs after executive order," January 2025, agencyiq.com.
  8. Beth Harper, "Rethinking Recruitment Plans During The Pandemic," Clinical Leader, April 2020, clinicalleader.com.
  9. Zachary Smith, Joseph DiMasi and Kenneth Getz, "Quantifying the Value of a Day of Delay in Drug Development," Tufts Center for the Study of Drug Development white paper, August 2024, csdd.tufts.edu.