Thyroid reference ranges: enrolling more people doesn't help
Thyroid immunoassay panel · CLSI EP28 reference-interval study · 564 healthy adults, 9 sites, 18 months
The challenge the client gave us
Before a thyroid test can report a result as "normal" or "abnormal," someone has to establish what normal looks like in healthy people. The client needed RDI to run that reference-interval study for a thyroid panel, following the CLSI EP28 guideline that governs these studies.
The catch: EP28 dictates the demographic mix of the healthy reference group — the right spread of ages, sexes, and other characteristics. So the binding constraint isn't how many people you enroll; it's who each site can actually reach. Once a demographic cell is full, more volunteers from that cell are useless.
At a glance
| Healthy adults enrolled | 564 |
| Sites | 9 |
| Timeline | 18 months |
| Peak month | +155 (March 2022) |
| Front-loaded | 446 of 564 (81%) by June 2022 |
| Governing guideline | CLSI EP28-A3 |
Timeline
- Jan 2022 Program opens
- Mar 2022 Peak month (+155)
- Jun 2022 459 enrolled (81% of final)
- Jul 2022 Demographic "top-up" phase begins
- 2023 Slow top-up at ~9/month to fill hard cells
- Jun 2023 Program closes at 564
What we did as a CRO
| CRO responsibility | What RDI did on this study |
|---|---|
| Regulatory & study start-up | Full-service CRO execution against the EP28 guideline |
| Site selection & feasibility | Selected 9 sites weighted for demographic-mix delivery, not just raw throughput |
| Site activation & training | Activated the network and trained sites on the eligibility and demographic-strata requirements |
| Recruitment & enrollment | Verified healthy-adult eligibility at screening; managed enrollment slots at the program level — closing over-represented demographics at one site while keeping under-represented ones open at another |
| Kits, samples & lab | Kit supply and specimen handling for the thyroid panel |
| Monitoring & data | Central monitoring; tracked each demographic stratum against its EP28 target |
| Project management & close-out | Ran a front-loaded bulk phase, then a long, deliberate top-up phase to fill the hardest demographic cells |
What made it hard
- The target is a mix, not a number. Hitting 564 healthy adults is easy; hitting the EP28-specified spread of ages and sexes is not.
- Full cells go dead. Once a demographic group is filled, additional volunteers in it can't be used — so the last subjects are the slowest and most specific to find.
- Every subject had to be verified healthy against the exclusion criteria at screening.
How we solved it
- We managed enrollment as a portfolio of demographic slots across 9 sites, closing over-represented cells and steering under-represented ones — instead of chasing total volume.
- We front-loaded the bulk (81% delivered in the first ~5 months) so the schedule had room for the slow part.
- We ran a patient top-up phase at roughly 9 subjects/month specifically to fill the hard demographic cells that raw recruiting would never reach.
The result
564 healthy adults across the full EP28 demographic spread — 81% enrolled in the first five months, and the remainder delivered through a targeted top-up that treated the demographic mix, not the headcount, as the real finish line.
Source: Diagnostica™ at program close, June 2023. Sponsor identity withheld; site labels anonymized. Operations report, not a clinical-results publication.