What 611 sites told WCG about recruitment in 2025

Who answered the WCG 2025 site survey?

Of the 611 responding sites, 60% were academic medical centers, health systems or community hospitals, and 35% were independent sites, physician practices or site networks. Eighty percent were in the United States. Most respondents were operational staff: 28% in research administration and 27% coordinators or research nurses. Only 7% were principal investigators.^⁠[1]

Three caveats. The report does not say how sites were invited, so it describes the sites that answered, not all sites. Top-three and multiple-answer questions limit how percentages can be combined. And WCG sells IRB review and site services, including site resource augmentation and study start-up support, per the report's closing section.^⁠[1]

Where does recruitment rank among site challenges?

Sites picked their top three issues from a fixed list. Recruitment and retention was named by 28%, down from 36% in 2024.^⁠[1]

Read the drop narrowly: fewer sites ranked recruitment above other problems, which is not the same as trials enrolling faster. It also varies by site type: 32% of independent sites and physician practices named it, against 26% of larger sites, a gap WCG links in part to smaller patient databases and less brand recognition.^⁠[1]

What did sites do about their challenges?

WCG asked which solutions sites had put in place for their challenges overall, not for recruitment alone. Sites could select all that apply.

• Hired additional staff: 34%
• Trained staff or PIs: 33%
• Implemented technology solutions: 26%
• Took on more industry-sponsored trials: 23%
• Partnered with clinical services companies: 19%
• Limited studies to specific therapeutic areas, or joined fewer studies: 16% each^⁠[1]

Why do sites reach for people before technology?

The survey does not ask why, but its other answers suggest a reason: sites already carry a technology load. Asked for the single biggest driver of trial complexity, 22% chose the number of technologies and vendors a trial requires, second only to complex protocols. Sponsor-provided systems such as CTMS, eReg and eISF were a top-three issue for 20% of all sites and 25% of independent sites and physician practices. Only 9% named their own site technology.^⁠[1]

The complaint is mostly about systems each sponsor requires, not tools a site picks. But any new system means IT and security review, training and new coordinator routines, while a new hire follows a process every site already knows.

What are the limits of solving recruitment with headcount?

Hiring is itself one of the problems. Site staffing was a top-three issue for 30% of sites. Over the past year, 45% of sites held staffing levels steady, 28% grew and 27% shrank.^⁠[1] In ACRP's 2025 workforce survey of more than 735 clinical research professionals, 52% said hiring and retaining staff is worse than it was 5 to 10 years ago.^⁠[2]

Headcount also scales in a straight line with the work. In a 2012 study at one cancer center, finding, screening and enrolling a single patient took 3.4 to 8.8 staff hours, depending on the type of study.^⁠[3] Protocol design adds to that load: 59% of WCG's respondents ranked complex protocols, including eligibility criteria, as the biggest driver of trial complexity.^⁠[1] A new coordinator adds hours, but the hours each enrollment costs stay the same. Our comparison of Bond and manual chart review breaks down where those hours go.

Does recruitment technology actually save coordinator time?

Sometimes. In a pilot at a Yale Cancer Center GI oncology clinic, reported in 2023, a matching algorithm excluded 146 of 161 patients in one week, leaving 15 charts for staff, and missed none of the six eligible patients. Review time fell from 3.11 to 1.82 minutes per patient.^⁠[5] In a randomized evaluation on 355 oncology charts, AI assistance raised coordinators' chart-level accuracy from 71.1% to 76.5% but saved no time.^⁠[4]

The difference is what happens after the tool runs. In the randomized study, coordinators read every document for all 12 criteria, and the authors note it was not powered to detect a time difference.^⁠[4] Time is more likely to drop when the tool narrows who gets reviewed and shows where each criterion was met, so the coordinator checks evidence instead of hunting for it. Bond's Identify stage works this way: it ranks candidates and links each criterion to its evidence. Bond's website cites 50%+ less chart review, a figure to confirm on your own study.^⁠[6]

What should a site do with these findings?

Treat hiring, training and technology as one plan, not rival options. WCG's recommendations to sites cover both staff training and technology, along with tracking key metrics against industry benchmarks and launching pilot projects.^⁠[1] For recruitment, that looks like this.

1. 1

   Measure one study

   Log coordinator hours on chart review, outreach and scheduling for one active study over two weeks, with referrals and enrollments.

2. 2

   Find the bottleneck

   Fix the step with the most hours per enrolled patient first. The coordinator chart review checklist standardizes the manual baseline.

3. 3

   Pilot one tool on that step

   Same study, same staff. Agree on measures in advance: hours per referral, referrals per week, and screen-fail rate at the first visit, which the screen failure guide covers.

4. 4

   Train the people who will use it

   Agree on what a coordinator must verify for each flagged patient. If every chart still gets a full read, as in the randomized study above, time may not drop.

5. 5

   Decide on a set date

   Set the review date before the pilot starts, and keep the tool only if the numbers moved. Either way, you have data for sponsor feasibility calls.