Most IVD trials fail to enroll on schedule. The conventional explanation is that healthy subjects are hard to find, that the eligibility criteria are too strict, that the seasonal windows are unforgiving. None of those explanations are wrong. They are, however, downstream of a more fundamental design error: the trial asks the site to do something other than care for the patient in front of them.
Across more than 300 IVD studies we have run since 2011, the protocols that recruit on schedule share one property. The study activity fits inside the patient visit that was going to happen anyway. The visit was already on the calendar; the consent fits inside the rooming flow; the sample is drawn at the same blood draw the physician ordered for clinical reasons. From the site's perspective, the trial is not an additional burden — it is an overlay on standard work.
We call this standard-of-care alignment. It is the single largest determinant of enrollment velocity we have observed. Patient centricity, as the industry uses the phrase, is the buzzword. Physician fit is the mechanism. The five principles below describe how we apply that mechanism in practice.
Design the protocol around the visit, not the visit around the protocol.
If a patient would not have come in for this visit anyway, the trial will struggle to enroll. The protocol is downstream of the appointment, not upstream of it.
The first question to ask of any IVD protocol is whether the visit it requires already exists in the patient's care pathway. A study that recruits well-women during their annual exam will enroll. A study that requires those same women to come in for a dedicated study visit, untethered from any clinical reason, will not. The patient may consent in principle and forget in practice. The site staff, asked to schedule a non-clinical visit, will deprioritize it against the visits a payer reimburses.
This is not a statement about patient motivation. It is a statement about scheduling friction. The visit calendar at a primary care or OB/GYN practice is built around clinical demand. Anything outside that calendar is an exception, and exceptions are scheduled last.
On a recent multi-cohort healthy-sample study, our adult-female sub-cohorts recruited on cycle phases the practice was already tracking — follicular, luteal, postmenopausal — and the pediatric cohorts recruited at well-child visits the parent had already brought the child in for. Adult women enrolled at roughly twice the rate per active site of the pediatric cohorts, even though pediatric eligibility was wider. The differentiator was not the eligibility envelope. It was that the OB/GYN visit was already on the calendar.
Before writing inclusion criteria, list the existing visit types at your candidate sites. If the protocol cannot be embedded in one of those visit types, redesign the protocol — not the recruitment plan.
A site's standard of care is not a constraint to work around. It is the substrate.
Every divergence from standard practice creates training, monitoring, and consent overhead that compounds across visits.
Most protocols accumulate small deviations from standard of care, each justified individually, each adding to a training burden that scales with the number of sites. By the time the protocol reaches a sponsor's clinical operations review, it requires a half-day site initiation visit, a custom CRF, and an investigator who can recite the protocol from memory. None of these things describe standard medical care.
The fix is not to lower the scientific bar. The fix is to ask, for each protocol-level deviation, whether the divergence from standard of care is doing scientific work — or whether it is a residue of the way the protocol was originally drafted. Most are residue. The fewer divergences, the shorter the IRB review, the faster the SIV, and the lower the protocol deviation rate at monitoring.
On the same study, nine site initiation visits cleared IRB review in January 2025 with a median turnaround of four days from submission to review, including the December–January federal holiday window. Sites engaged because the protocol read as an overlay on existing pediatric and OB/GYN practice, not as a separate operating model. Short, IRB-friendly protocols clear faster than long, custom ones.
Walk through every protocol step with a practicing clinician at the candidate site type. Ask: would you do this anyway? If the answer is no for more than a few steps, the protocol is the problem.
Recruit four cohorts as four studies, not one.
Bucketed recruitment is the single most common point of failure on multi-cohort IVD trials. Each cohort family has its own funnel.
A protocol that enrolls pediatric, adult, and pregnant subjects is not one trial with three cohorts. It is three trials sharing one IRB approval. Pediatric subjects are recruited through pediatric primary care; pregnant subjects through OB/GYN and maternal-fetal medicine practices; adult subjects through mixed primary-care and direct-to-participant outreach. The eligibility scripts are different. The visit cadences are different. The seasonal pressures are different.
Treating "healthy samples" as a single recruitment category — the most common framing in the original protocols we inherit on rescue engagements — applies one funnel to four populations and produces one outcome: the protocol stalls at the cohort with the tightest constraint. Built correctly, the four funnels run in parallel and the program enrolls at four times the velocity of a bucketed plan.
The principle is observable across our rescue engagements but particularly clear on cohort-stratified work — for instance, sourcing extreme-value Tg/TgAb subjects, which requires recruitment through endocrinology rather than general primary care. Aspirational feasibility-form responses from sites are systematically inaccurate on populations like these. Claims and EHR analytics have to do the work that feasibility forms cannot.
Build a separate enrollment plan, recruitment script, and visit cadence for each cohort family. If two cohorts share a recruitment plan, you have not actually planned for two cohorts.
Logistics is a protocol-execution problem, not a protocol-amendment problem.
Most logistics gaps — freezer access, courier windows, kit availability — can be solved without amending the protocol. The instinct to amend is usually wrong.
When a site lacks −80 °C freezer capacity, the conventional response is to disqualify the site or amend the protocol's storage requirements. Both are mistakes. Placing a freezer at the site, or arranging same-day-overnight dry-ice shipment to a qualified storage facility, preserves the original protocol and the original endpoints. The amendment was never needed.
This generalizes. Kit shortages can be solved by adjusting sponsor manufacturing cadence, not by relaxing collection criteria. Courier window misalignments can be solved by adjusting the visit schedule, not by extending sample stability claims that lack data. Each protocol amendment costs time at the IRB and creates training delta at every active site. The right default is to ask whether the gap is a logistics problem dressed as a protocol problem.
On a recent multi-cohort study, several qualified clinical sites lacked freezer capacity. We placed equipment where placement was justified by site volume, and arranged dry-ice logistics where it wasn't. No protocol amendment was issued for logistics. The original protocol was made executable through field operations rather than redefined.
When the field flags a logistics gap, the first response should be operational. The second response should be a logistics SOP. The third response, only after the first two have been ruled out, is a protocol amendment.
The IRB queue is not paused during the holiday window.
Sustained follow-up through December and January produces approval timelines that look impossible from the outside. The standard CRO assumption that the window is dead time is empirically wrong.
The IRB calendar is the most undervalued operational lever in a CRO's toolkit. Most CROs treat the December 20 to January 5 window as time off — an assumption that becomes a self-fulfilling prophecy when no one follows up on submitted applications. IRB reviewers, contrary to the assumption, are working through the window, and the queue length is shorter precisely because most CROs are not pushing.
This matters because a substantial fraction of the seasonal cohorts on healthy-volunteer studies — winter respiratory, pregnancy first trimester, pediatric school-year — open their recruitment windows in early January. Filing in late December and securing approval before the window opens is the difference between enrolling that quarter and waiting for the next cycle.
On a recent rescue engagement, we filed a central-IRB application on December 23 and received protocol approval on January 7 — 15 days end-to-end, including the federal holiday window. Subsequent site initiation visits cleared at a median four days each. The fastest single SIV cleared in one day.
If the program hits the IRB in mid-December, do not wait for January to start follow-up. Reviewers are working. The queue is moving. The CROs that win the seasonal window are the ones who treat the holiday weeks as accelerator, not pause.
What this framework rules out.
The five principles above describe what we believe trial design should look like. They also rule out a few things the industry takes for granted.
They rule out patient centricity as a primary design criterion. Patient centricity, as commonly used, conflates two distinct ideas: respecting the patient's time, and recruiting through the patient's enthusiasm. The first is genuinely important and is fully addressed by physician fit, because the patient's time is most respected when the visit was going to happen anyway. The second is operationally weak — patient enthusiasm does not survive the scheduling friction that protocols outside standard care create.
They rule out specialty-only site networks as the right answer to recruitment for most healthy-volunteer work. A network of primary-care and OB/GYN practices, embedded in patient flow, will enroll faster than a specialty research network for any cohort the primary-care practice already sees.
They rule out protocol amendments as a first response to operational friction. Amendments are a regulatory cost paid by every active site. Most operational friction is logistics, and logistics has SOPs, not protocol changes.
The framework is not closed. We refine it as new evidence comes in. When the study outcome matters, you call RDI. The principles above are how we deliver against that promise.