Clinical protocol authoring that starts from the schedule of assessments and scores every visit for patient burden
A checked first-draft protocol in days, in your approved language.
See one case, screen by screen ↓Before a trial can start, a medical writer turns a nine-page synopsis into a protocol of 90 to 120 pages in the sponsor's template: objectives and estimands, eligibility, intervention, assessments, safety, statistics, oversight, and the schedule of assessments that ties them together. Most of it is copied by hand from the last protocol on the same drug, then changed line by line, while the writer chases the clinical lead, the biostatistician and the medical monitor for the decisions the synopsis leaves open.
What slips through comes back as an amendment. A schedule adapted from a Phase 2 study whose primary endpoint was at Week 12 still has no PASI at Week 16; a contraception period says 20 weeks in Section 5 and 16 weeks in Section 9. And nobody adds up what a visit asks of a patient — the minutes on site, the blood drawn, the needle sticks — until sites say Week 4 is too long, after the first participants are enrolled.
Estimated split, in working days, for a Phase 3 protocol from synopsis to first draft — by hand (about 38 days, the by-hand median on the solution's sample dashboard) and with the solution (the 4-day target).
Six specialist agents read, reuse, write, check and score every visit; a medical writer approves each section.
VLT-302 is a Phase 3 trial of Veltrimab, an anti-IL-23p19 antibody, in moderate-to-severe plaque psoriasis: 660 participants randomised 2:1 at about 120 sites in 14 countries, with co-primary endpoints PASI 90 and sPGA 0/1 at Week 16. Lena Ortiz, senior medical writer, owns the protocol. Here is what happens, screen by screen, in the working solution.
Synopsis v1.0 arrives from clinical development: nine pages. The synopsis reader pulls design, population, objectives, endpoints, dosing, visits and sample size as 31 fields, each with the page it came from, and highlights the passages it used. The closest approved protocols are ranked beside it: VLT-201 at 92% match (same drug and population, Phase 2), VLT-115 at 71%, CRV-201 at 54%.
The synopsis reader quotes endpoint wording, never paraphrases it, and never infers a timepoint the synopsis does not state.
Two minutes later the first draft exists: 14 sections in the sponsor template v3.2 (ICH M11), 58% of it approved language. Lena's screen is Table 1 — 11 visits across, 15 procedures down, a patient burden bar over every visit — with each section drafting live beside it and the share written from approved language. The consistency checker has already found the first problem.
“Week 16 is the co-primary endpoint visit, but PASI, sPGA and BSA are not scheduled there.” — adapted from VLT-201, whose primary visit was Week 12.
Lena clicks “Add them at Week 16” and the scores go in; Section 1.3 follows, saved as v0.2. Then she opens Day 1, the hardest visit: eligibility, PASI, sPGA, BSA, DLQI, vital signs, laboratory tests, pregnancy test, the PK sample, anti-drug antibodies and the first injection, with 15 minutes of check-in and waiting — 1 h 55 min on site, 21 mL of blood, 2 needle sticks.
“Score = 0.25 × 115 min + 0.6 × 21 mL + 9 × 2 sticks = 59. Draws taken together count as one stick. Your limit is 60.”
The medical monitor asks for a liver panel — ALT, AST, ALP, GGT, bilirubin, INR — at Weeks 2, 4 and 8. Lena adds it from “+ Add assessment”: the source is CRV-201 §8.3, approved language with the weeks adapted, and Week 2 is a new visit.
“The patient burden of each visit is re-scored as soon as it is added; Section 8.3 is written from the source shown.”
The new Week 2 visit scores 26. Week 4 now carries a third needle stick and 15 more minutes, and its bar turns red at 70. The scorer proposes the smallest change that brings it back: take the liver panel from the pre-dose PK draw — the same venipuncture, one stick fewer, two more tubes to the central lab. Weeks 2 and 8 stay as their own draw.
“Patient burden scorer · you or Omar Haddad decide.” The scorer proposes; it never applies a suggestion or drops an endpoint or safety assessment.
Lena clicks “Combine with the PK draw”. Week 4 is back to 57, under the limit; Sections 1.3 and 8.3 are rewritten and the draft is saved as v0.4. With the liver panel, blood over the whole trial is 146 mL; the most in any 8 weeks is 68 mL, against the 550 mL limit.
“Week 4 back to 57 · one needle stick fewer · saved as v0.4.”
Section 8, Assessments and Procedures, shows each passage's origin: reused verbatim from VLT-201 Amendment 1 v3.0 pages 41 and 42, adapted from CRV-201 page 49 with every changed word marked, or new from the synopsis. Section 8.3 now reads: “At Week 4 it is taken from the same venipuncture as the pre-dose PK sample.” Each source opens at its passage.
The section writer never invents a number and never hides a change: new text cites the synopsis.
The Week 16 finding is marked fixed — accepted by Lena Ortiz, saved as v0.2. Three remain: the contraception period is 20 weeks in Section 5.1 and 16 weeks in Section 9.4, reused from VLT-201; Section 8.1 allows ±3 days through Week 16 while the schedule footnote allows ±7 days; and sPGA is used in Section 1.1 before it is spelled out. Each comes with one suggested fix to accept or dismiss with a reason.
The consistency checker proposes; it never applies a fix itself.
What the synopsis does not settle becomes a question for a named person: Dr. Maya Chen on whether prior biologic use should be capped, Sam Patel on the testing order for the key secondary endpoints, Dr. Priya Raman on whether a data monitoring committee is needed, and Dr. Maya Chen again on whether the Week 16 switch is blinded. Each answer is written back into its section with name and time.
“VLT-201 excluded participants with more than two prior biologics. The synopsis says nothing. The draft keeps the cap of two.”
Lena approves the protocol section by section; after her, each section goes to its reviewer — Sam Patel for objectives and estimands, Dr. Maya Chen for population and intervention, Omar Haddad for the schedule and procedures, Dr. Priya Raman for safety. When all 14 are approved, a full consistency re-check runs, the team reviews, and the draft goes to Dana Okafor, head of medical writing, for e-signature with a reason; the version is then locked. On the Protocols list, VLT-212 is at that point: 14 of 14 sections approved, in team review.
Every version is kept; nothing is overwritten.
The same protocol, seen by the five people who write, review and sign it — what their weeks looked like, and what they look like now.
Six specialist agents read, reuse, write, check and score every visit; a medical writer approves each section.
Reads a trial synopsis and pulls out design, population, objectives, endpoints, dosing, visits and sample size as 31 structured fields, each with the exact wording and the page it came from.
Finds approved passages in final protocols that fit each section, ranked by population, phase and therapeutic area, and returns each with its document, version, page and the differences the writer must see.
Writes each template section from the synopsis fields and the retrieved passages — 14 writers in parallel, in the sponsor template and house style. Reuses a passage verbatim when it fits and marks every changed word when it does not.
Checks the draft against the synopsis and itself: endpoint timepoints against the schedule of activities, objectives against endpoints and estimands, eligibility, dose and regimen, visit windows and abbreviations. For each mismatch it shows both passages and proposes one fix.
Scores every visit in the schedule of assessments for minutes on site, blood drawn and needle sticks or injections, on the weights in Settings, and proposes the smallest change that brings a visit back under the limit: combine draws, move a procedure to an adjacent visit, or use a local lab.
Turns gaps in the synopsis into short questions for the right person and writes the answers back into the section.
Approves every section. The agents propose; a named person decides.
13 screens from the working solution, on its sample data. Pick one to see it large.
How visits are scored, which language may be reused and who decides are settings — every change applies to new scores and drafts.
| Setting | Default | Choose from |
|---|---|---|
| Burden limit per visit | 60 | Up or down in steps of 5, from 30 to 90 — every visit is re-scored |
| Burden weights | 0.25 / min · 0.6 / mL · 9 / stick | 0.25 / min · 0.6 / mL · 9 / stick · 0.2 / min · 0.8 / mL · 12 / stick |
| Blood drawn in any 8 weeks | 550 mL | 550 mL · 400 mL · 10.5 mL/kg |
| Who decides on a burden suggestion | Omar Haddad · Clinical operations lead | A named person — the suggestion is never applied on its own |
| Protocol template | Sponsor template v3.2 (ICH M11) | Sponsor template v3.2 (ICH M11) · Sponsor template v3.1 (legacy) |
| Word export | Sources as Word comments | Sources as Word comments · Clean copy, sources in an appendix |
| Oldest language allowed | 5 years | 1 to 15 years since the source protocol was finalized |
| Consistency checks | 6 of 7 on | Endpoint timepoints in the schedule · objectives ↔ endpoints ↔ estimands · eligibility · dose and regimen · visit windows · abbreviations · references in Section 14 |
Every figure is labelled: a target the solution is built to, an estimate, a typical published result, or a proven one.
“demo” = seen in the working solution, on its sample protocol data · “target” = the design goal, measured in the live solution · “estimated” = our estimate · “typical” = published figures (Tufts CSDD protocol amendment studies, 2016 and 2024 — 45% of substantial amendments avoidable, a median $535,000 direct cost per Phase 3 amendment, 76% of protocols amended; Tufts CSDD / ZS patient burden studies; ICH M11). People, studies, drugs and sponsors named on this page are fictional — characters and sample data in the working solution.
Software that helps medical writers turn a trial synopsis into a full protocol in the sponsor's template. Clinical Protocol Authoring adds six agents that read the synopsis, reuse approved language from final protocols with its source, write the sections, build the schedule of assessments, check the draft against the synopsis and score every visit for patient burden. The medical writer approves every section.
Every draft follows your protocol template — its sections, headings and standard text. In the working solution that is a sponsor template v3.2 built on ICH M11, with 14 sections, and the template is a setting.
The approved-language retriever searches final protocols only, preferring the same drug and population. A passage is reused verbatim when it fits; when it is adapted, every changed word is marked. Each passage keeps its document, version and page, and language from protocols later amended is flagged.
For each visit: 0.25 points per minute on site, 0.6 per mL of blood and 9 per needle stick or injection, with draws taken together counted once. A visit above the limit — 60 by default — is flagged with the smallest change that brings it back, and the blood drawn in any 8 weeks is held against 550 mL. The weights and limit are settings.
Endpoint timepoints missing from the schedule of activities, objectives that do not match endpoints and estimands, eligibility that differs from the synopsis population, dose and regimen that differ between Section 6 and the schedule, visit windows that disagree, and abbreviations used before they are spelled out. Each finding shows both passages and one proposed fix.
Yes. The agents draft, check and propose; none of them applies a fix or a burden suggestion on its own. The medical writer approves every section, named reviewers follow, and the head of medical writing signs the final version with a reason before it is locked.
Yes. The draft exports to Word with its sources as Word comments, or as a clean copy with the sources in an appendix.
The Agentic Solution Engine builds and deploys it from your requirements and documents — your protocol template, style guide, procedure burden table and a set of your final protocols — and it goes live once every quality gate has passed. We will walk you through it on one of your own synopses first.
An inspection-ready trial master file every day, each site scored on what is due today.
1-min demoClinical Study Report AuthoringResults drafted within days of database lock, linked to tables.
1-min demoProtocol Design OptimizationAvoidable amendments fixed before the committee meets, each flag tied to a past amendment.
1-min demoWe'll run Clinical Protocol Authoring on one of your synopses, your template and a sample of your final protocols.