Clinical trial site burden is the work a study puts on a research site beyond the participant in the chair: the logins, the training, the budget and contract back-and-forth, the data entered twice. It matters this autumn because demand and capacity are moving apart. Bain & Company's 2026 survey of 120 sponsors, CROs and sites, reported by BioXconomy on September 30, 2026, found that "twice as many sites expect their capacity to worsen over the next three years," and that "slow contracting and managing progress across multiple software platforms top the list of site frustrations." Sponsors cannot fix what a site cannot show them. The useful response is to measure your own burden, study by study, before the next feasibility questionnaire arrives.
In short
- 54% of trials now run at least 20% late, up from 44% before 2020 (Bain, via BioXconomy, 2026).
- Nearly 70% of sites have six or more logins per study, and 55% call technology the most burdensome start-up task (Advarra, 2023).
- 17.5 hours per study per month go to training, on the Society for Clinical Research Sites' site surveys (SCRS, 2024).
- 55% of sites took longer than five months to activate a trial in 2025, and contract and budget delays frequently held up 66% of them (WCG, via Clinical Trial Vanguard, 2026).
Where the hours go
Bain named two frustrations, and the older site surveys fill in their size.
Systems. Advarra's 2023 survey of more than 500 North American respondents found nearly 70% had six or more logins per study, and 67% said sponsor technology setup and training was more burdensome than five years earlier. Writing in DIA's Global Forum in January 2025, Christine Senn cited Advarra's finding that a typical oncology study needs the site's own systems "plus as many as 17 additional systems," that sites copy or transcribe data between systems about 60% of the time, and that only 32% can use their own credentials.
Training. In a December 2024 call to action, SCRS said its earlier site landscape surveys put study training at an average of 17.5 hours per study per month, around six hours per role, and asked the industry for a 25% cut by the end of 2025.
Start-up. In Advarra's survey 41% rated budgeting and 36% rated contracting extremely or very burdensome, and only 36% said sponsors often reuse an existing contract. WCG's 2025 report, summarised by ACRP, ranked study start-up second among site challenges at 31%, behind trial complexity at 35%.
The protocol itself. TransCelerate and Tufts CSDD found in 2025 that up to 32.5% of Phase III per-patient data is non-core, and that this data accounts for 25% to 30% of the burden on participants and sites.
None of that shows up as a line in the study budget. It shows up as a coordinator who cannot take the next study.
Why the manual way of noticing it fails
Most sites know a study is heavy only after it has been heavy for months: the coordinator is late on queries, a monitoring visit finds gaps, someone leaves. By then the cost is paid. ACRP's 2025 workforce piece put the cost of replacing a coordinator at an estimated $50,000 to $60,000, and WCG's 2025 report had site staffing as a top challenge for 30% of sites. A site that waits for the symptoms is measuring burden in resignations.
The other failure is the feasibility form. A sponsor asks how many participants you can enroll. Almost nobody asks how many hours the study will cost you, so the site answers the first question and absorbs the second.
Five numbers to count on every active study
- Systems and logins. List every sponsor and CRO system the study needs, with who holds a login. It takes an hour per study and gives you the figure Advarra's respondents reported: six or more is common, and an oncology study can need many more.
- Training hours per month. Ask each role to log study training for one month. Set it beside SCRS's 17.5 hours per study per month. A study far above it is either new or badly designed, and the sponsor should hear which.
- Days from budget receipt to signed contract. Date both ends for every study started this year. Note whether the sponsor used your existing master agreement or budget terms; Advarra found most do not.
- Days to activation. From site selection to first participant screened. WCG reported more than five months at 55% of sites in 2025. Your own median is the number to quote in the next feasibility response.
- Coordinator hours per enrolled participant. A rough weekly estimate is enough. It is the number that tells you whether a study is worth its budget, and it is the one that rises first when a protocol amendment adds procedures.
The mistake most sites make at step 5
They measure enrollment and not effort. A study that enrolls well and eats twice the coordinator hours of its neighbour looks like a success on the enrollment report and is the reason the next good study gets turned away. Count both, side by side, every month. Our post on protocol amendment implementation shows how much work a single amendment adds, and the regulatory binder guide covers the documents whose dates add to it.
What sponsors say they will change
Bain's authors, as BioXconomy reported, put both top frustrations partly on the sponsor's side: sponsors can improve contracting, and system proliferation follows from sponsor and CRO technology decisions. The same survey found as many as 65% of sponsors and CROs piloting AI tools. That may or may not reach the site. What a site can do now is arrive at the conversation with its own numbers, so that a request to reuse contract terms or accept site credentials comes with the hours it would save.
What we build for this
A performance dashboard pulled from the systems a site already runs, with per-coordinator workload beside screening and enrollment, and a study start-up tracker that dates budget, contract, IRB, training and first screen with owners and blockers, are two of the systems we build for research sites. They count the burden. They do not reduce a protocol's procedures; that is the sponsor's decision, and your numbers are the case for it. For the step that comes after activation, see the first hour after a referral.
Capacity is falling at the sites that cannot say where their hours went. Start counting this month.
