← GuidelineVahti

GuidelineVahti · Delphi pilots

A facilitated consensus round with an auditable record

Run your next Delphi round without spreadsheet chaos.

Most consensus projects can show the final recommendation. Fewer can show exactly how the panel got there. GuidelineVahti gives you that record.

Bring your statements and your panel. We run a modified-Delphi / RAND–UCLA round in GuidelineVahti; you receive the auditable record: consensus, an implementation priority map, chance-corrected agreement, and GRADE/EtD recommendations — with the full round history and a methods appendix. Votes stay anonymous; nothing uploads.

anonymous · aggregates onlyauditable methods recordGRADE / EtDfirst cohort · hands-on
Request a scoped pilot plan → See the tiers ↓ Sample record

How a pilot works

Four steps, one auditable output. You keep the decisions; the tool keeps the record.

  1. Scope. We agree the statements, the panel, the consensus threshold and the stop rule with you.
  2. Vote. Each panelist rates statements 1–9 in the browser, anonymously — no account required, no panel data uploaded.
  3. Analyse. GuidelineVahti computes RAND/UCLA consensus, the importance×feasibility priority map, chance-corrected agreement, and GRADE/EtD recommendations, with between-round feedback if you run a second round.
  4. Deliver. You receive the consensus record, the priority map, the round history, and a methods appendix you can cite — a defensible, reproducible account of what the panel decided.

Pilot tiers

Fixed-scope pilots so you can start small and see the record before committing a full guideline.

Try it

Single pilot round

€1,500

up to 15 panelists · one round

  • One RAND/UCLA appropriateness round
  • Consensus, IQR & distribution per statement
  • Importance×feasibility priority map
  • Anonymous browser voting, set up for you
  • A short summary of the result

Low-risk start: run one statement set. If the record isn't useful, don't proceed to a package.

Request scoped pilot plan

Most chosen

Two-round Delphi + report

€4,500

full two rounds · written report

  • Everything in the single round
  • Two Delphi rounds with between-round feedback
  • Round-to-round trajectory & stability
  • Chance-corrected agreement (Krippendorff's α)
  • A written report of the consensus process
Request scoped pilot plan

Full statement set

Guideline consensus package

€9,500+

multi-statement · versioned archive

  • Everything in the two-round Delphi
  • GRADE / Evidence-to-Decision per recommendation
  • Full statement set across the guideline
  • Versioned archive — the versioned data file is the source record
  • A methods appendix for your manuscript or guideline
Request scoped pilot plan

Prices in EUR, excluding VAT. The first cohort of pilots is small and hands-on — we tailor each one with you, so scope and timing are agreed before anything is booked. Living-guideline programmes that re-vote each cycle: ask about an ongoing arrangement.

What a pilot is — and is not

What you receive. An auditable, reproducible record of your panel's consensus process: the numbers, the round history, the priority ranking, the GRADE/EtD recommendations the panel reached, and a methods appendix documenting how. It's built to satisfy a methods reviewer and to disclose exactly how the consensus was formed. See a sample consensus record →

What it is not. GuidelineVahti records what the panel decided; it does not certify that a guideline is correct, that a recommendation is clinically valid, or that the evidence is sufficient. The panel decides; the AI is never a panelist. Final responsibility for the guideline stays with the panel, the chairs, and the commissioning organisation.

Who it's for

Specialty societies & guideline groupsRun a modified-Delphi/RAND consensus with auditable numbers, not a spreadsheet.
Guideline methodologistsConsensus, priority, agreement and GRADE in one place, in the browser.
Medical-communications teamsA defensible, reproducible consensus record for a client guideline.
Living-guideline programmesRe-vote affected statements each cycle; the versioned data file is the source record.

Common questions

How long does a round take?

Panelists usually vote in a single sitting, and the analysis is immediate once votes are in. Calendar time is mostly scheduling: a single round can run in a week or two; a two-round Delphi depends on how quickly the panel completes each round. We agree the timeline with you at scoping.

How many panelists do we need?

The single-round pilot is set up for up to 15 panelists, which suits most modified-Delphi panels. Larger panels and multi-statement guideline sets fit the group package — tell us your numbers and we'll confirm.

Is voting really anonymous, and does anything upload?

Yes. The tool works with aggregates only — no panelist is ever named — and it runs entirely in the browser with no network calls. Between-round feedback shows each panelist their own prior vote and the group distribution, nothing more.

Do you decide the recommendations?

No. The panel votes; GuidelineVahti applies your agreed consensus threshold and the GRADE logic to what the panel decided. Where the panel is split, it says so rather than manufacturing a recommendation. We facilitate and record — we don't adjudicate the clinical content.

What do we actually receive at the end?

The consensus result per statement, the importance×feasibility priority map, the round history and agreement scores, the GRADE/EtD recommendations, and a methods appendix describing the process — in a form you can put in front of a methods reviewer or into a manuscript.

Start a pilot

Tell us your topic, roughly how many statements and panelists, and your timeline. We'll reply with a scoped plan — no obligation.

Request a scoped pilot plan →

Prefer to see it first? Open a sample consensus record →