The moment a program starts to drift
One study is manageable in any tool. The trouble starts with the second one. The second protocol is copied from the first, so now two documents describe the same product, the same dose rationale, mostly the same safety language. The brochure describes it too, and so do two consent forms, two sets of visit worksheets, two screeners. The program now has one set of facts living in a dozen places, and every amendment, every new edition and every review comment moves one copy and not the others.
Teams compensate with discipline: comparison reads, tracked-changes archaeology, spreadsheets of which document cites which version. Discipline does not scale. It is exactly the work that gets cut when timelines compress, and a missed inconsistency surfaces at the worst possible readers, an IRB, an inspector, a site coordinator mid-visit. The answer is not more careful copying. It is to stop copying facts and start linking them.
The three moves of a growing program
Programs grow in a small number of recognizable moves, and Trelice has a specific answer for each.
A new indication for the same product
The oncology compound gets its dermatology study. The protocol is ninety percent reusable: product, formulation, dose rationale, safety monitoring, most procedures. In Trelice you derive the new protocol from the approved one, picking what carries over, and the copied fields stay linked. When safety language tightens in either protocol, the other one raises a drift alert, with both values side by side and a person deciding whether the two indications should still agree. The product facts themselves, mechanism, formulation, adverse events of special interest, come from the Investigator’s Brochure, which both protocols share, so a new brochure edition reaches both studies the day it is released.
A new application of the same drug
Same molecule, new formulation, new route, a combination, a different population. These studies share less protocol text than a new indication but exactly as much product history: the nonclinical package, the clinical experience to date, the safety profile. That is brochure territory. One brochure per product carries those facts; every protocol that uses the product is prefilled from the current edition, each filled field linked to the section it came from. Approved protocols send their study facts back to the brochure as proposed updates, so the clinical experience table of the next edition is assembled from the program’s own records rather than from memory.
A parallel protocol version in the same study
Sometimes one study needs two protocols running side by side: a sub-study at a subset of sites, a regional version with different local requirements, a next version prepared while the current one still runs. These are the most dangerous copies, because the overlap is nearly total and the differences are deliberate and few. A derivation keeps the overlap linked and makes the differences explicit: a field where the two versions should differ is unlinked once, on purpose, with the decision recorded, and every other difference that ever appears raises an alert. The map of what is shared and what diverges is in the system, not in the lead writer’s head.
Drift detection, across protocols and downstream
All of this rests on one mechanism applied at every joint: a link between two places that state the same fact, and an alert when they stop agreeing. Within a protocol, golden threads keep the sample size, the criteria and the design consistent across sections, and the schedule is the one place a timepoint is entered, with procedures and endpoints reading from it. Between protocols, derivation links watch every copied field in both directions. Between the brochure and its protocols, releasing an edition checks every linked field of every protocol of every study that uses the product, and flags the consent forms whose risk language the edition touched.
Downstream, every generated document flags for regeneration when its protocol changes, so the amendment cascade that starts with one field ends with a worklist naming each affected consent form, worksheet and screener. The alerts share a grammar everywhere: both values side by side, the choice to take the other value, keep your own, or unlink deliberately, and an audit entry for whichever you choose. A reviewer can trust an approval because unresolved drift blocks it: a section with an open alert cannot be approved until someone has decided.
Nothing resolves silently. Trelice never auto-corrects a document to match another; it makes disagreement visible, cheap to review and impossible to approve past. That distinction is what lets the same machinery satisfy both the writer who wants speed and the quality lead who wants control.
Scaling the paperwork, not just the protocols
Growth also multiplies the formatting work: every new study means new documents that must come out in the house style, under the house template. In Trelice both are configuration learned from your own files. A document format template, learned from one sample, styles every document of every study: fonts, headers, cover pages, logo. A house protocol layout puts every new protocol in your heading tree and numbering, with your own title-page fields, as the default for the study type. The twentieth study’s documents come out as right as the first, with no per-study formatting effort at all.
What stays true at any scale
The guarantees do not loosen as the program grows. Every section of every document, in every study, goes through the same role-based review workflows, and every review decision is recorded in a permanent, append-only audit trail. Released brochure editions and approved protocol versions are frozen snapshots; later exports print exactly what was approved. Each organization’s data is strictly isolated. Dashboards and pending tasks aggregate across studies, so one program lead can see every open review, every drift alert and every stale document in the portfolio from one screen.
Where this propels a team forward
A few situations where the linking pays for itself fastest. A biotech taking its lead compound into a second indication derives the protocol in an afternoon and spends its writing time on what is genuinely new. A sponsor running a sub-study keeps it provably aligned with the parent through two amendment cycles, with the differences on a reviewable list instead of in a comparison read. A program head preparing the annual brochure edition accepts the queued study facts from the year’s approvals instead of chasing five departments for them, and the release itself produces the worklist of protocols and consent forms to look at. A CRO authoring for several sponsors keeps a format and layout per client and lets every document come out in the right house shape without a formatting pass.
The common thread: the work that used to grow with the square of the document count, checking everything against everything, becomes a stream of specific, pre-localized questions, each answered once and recorded. That is what makes the next study cheaper than the last one instead of more expensive.
Getting started
The pieces are covered in depth on their own pages: derived protocols and drift alerts, the Investigator’s Brochure, document formats and layouts, and the downstream documents every protocol generates. Pricing follows the program, not the seat count, so growing the team around a growing program costs nothing extra per reviewer.