CONTEXT(CLONE)
Sign in

Watch guideline and label changes against local clinical protocols

For: Medication safety pharmacist or clinical governance lead at a hospital group

Pattern: WatchtowerNeeds a connectorDesigned for 20 to 300 agents
Get early accessSign in to use

The pain today

National guidelines, product labels and safety notices change all year. Local protocols and order set texts that depend on them are reviewed on a calendar, so a changed dose limit or contraindication can sit unnoticed until the next cycle.

The ask

“Here are our local clinical protocols and order set texts, and the list of guidelines, labels and safety notices they are based on. When one of those sources changes, tell me exactly which of our documents depended on the changed passage and what now differs.”

Plain words, as you would say it to a colleague. Edit it to fit your case before you send it.

What you attach or connect

  • Local protocols and order set texts
  • List of source guidelines, labels and notices
  • Document owner register

The unit of work

One worker task per one source passage and the local text built on it.

Why a swarm fits

Hundreds of source passages feed hundreds of local documents. Each link is small and independent, and when one source changes only the documents that cite it need to be re-read.

Not for

Deciding what the new local dose or rule should be, or alerting clinicians at the point of care.

The decision tree

5 typed decisions, each with an action for every answer

At fixed moments in a run, the engine puts one narrow question to a decision model. The decision model never writes text: it answers yes or no with a probability, picks from listed options, or gives a score, about a small slice of the material. The engine then does exactly what this tree says, which is what makes the run auditable. The thresholds are the template's design values, not measured results.

  1. Planner, while planning

    Scope checkYes or no, with a probability

    Before work starts on a unit

    Does the changed text fall inside a passage that one of our local documents cites as its basis?

    Sees only: The changed source passage and the dependency list for that source

    Why: Skips source updates that no local document depends on.

    • Yes: 0.60 or higherthenAccept
    • Unsure: 0.30 up to 0.60thenEscalate to a strong model
    • No: below 0.30thenSkip this unit

  2. After workers, the judge checks

    Evidence checkA choice among options

    After a worker answers

    Does the new source text change a dose, a limit, a contraindication, a monitoring step or a patient group, compared with the old text?

    Sees only: Old and new versions of one source passage

    Why: Separates editorial revisions from changes a committee needs to see.

    • Changes one of those itemsthenAccept
    • Wording or layout onlythenSkip this unit
    • Cannot tell from the two passagesthenEscalate to a strong model

  3. Evidence checkYes or no, with a probability

    After a worker answers

    Does the local protocol passage state the same dose, limit or restriction as the new source passage?

    Sees only: One local passage and the new source passage it depends on

    Why: Confirms a real difference exists before a document owner is alerted.

    • Yes: 0.85 or higherthenAccept
    • Unsure: 0.50 up to 0.85thenEscalate to a strong model
    • No: below 0.50thenReject and retry

  4. Run control, between rounds

    Another round?Yes or no, with a probability

    Between rounds

    Have all local passages that cite the changed source passage been re-checked in this round?

    Sees only: The dependency list for the changed passage with check status

    Why: Ends the round when dependents are exhausted instead of re-reading the whole library.

    • Yes: 0.70 or higherthenStop
    • Unsure: 0.40 up to 0.70thenContinue
    • No: below 0.40thenContinue

  5. Accountable person, before anything is settled

    Person decidesYes or no, with a probability

    Before anything is reported as settled

    Does the difference concern prescribing, administration, monitoring or a contraindication in a local clinical document?

    Sees only: One alert with old source, new source and local text

    Why: Ensures the committee and the clinical owner see every care-relevant difference; nothing changes locally without them.

    Accountable: The drug and therapeutics committee and the named clinical owner decide whether and how a local protocol changes.

    • Yes: 0.40 or higherthenAsk a person
    • Unsure: 0.15 up to 0.40thenAsk a person
    • No: below 0.15thenAccept

The fleet: who does what

Model tiers by role, not brands: you choose the models. Strong reasoning models plan and reconcile, small fast models do the wide work, and the judge is a decision model from a different family, so it does not share the workers' blind spots.

  1. Planner

    A strong reasoning model maps which local passage depends on which source passage and schedules the re-checks.

    Decisions here:1. Scope check

  2. Workers

    Small fast workers from an open-weight family compare one changed source passage with one dependent local passage.

    Designed for 20 to 300 agents, one worker task per one source passage and the local text built on it. Each worker receives only its own unit.

  3. Judge, from a different model family

    A decision model from a different family decides whether the change is material to the local text or editorial only.

    Decisions here:2. Evidence check3. Evidence check

  4. Reconciler

    A strong reasoning model bundles material differences per document owner and keeps unclear cases visible.

    Decisions here:4. Another round?

  5. Accountable person

    The drug and therapeutics committee and the named clinical owner decide whether and how a local protocol changes.

    Decisions here:5. Person decides

Checked before anything is accepted

  • Each alert quotes the old source text, the new source text and the local text
  • Only documents that depend on the changed passage are re-checked
  • Editorial changes are logged separately from material ones

What comes back

  • Change alerts per local document with three-way quotes
  • Review queue per document owner
  • Dependency map of local documents to sources
  • Log of changes judged editorial

What to measure

  • Time from source change to alert
  • Alerts the committee judged relevant
  • Material changes found later that were missed
  • Cost per watched source

Names of measures only. No result is claimed for this template.

Templates open in the workspace chat with the ask filled in. Nothing runs until you send it.

Get early accessSign in to use

More in Healthcare and life sciences

Grade protocol deviations across trial sites, with evidence

For: Clinical operations lead or trial quality manager at a sponsor or contract research organisation

Deviation logs and monitoring reports pile up across sites.

Pattern: Map, verify, reduceNeeds scale6 decisionsDesigned for 40 to 400 agents

Check adverse-event narratives against their case data

For: Pharmacovigilance lead or safety writer preparing case narratives for a study report or aggregate report

Hundreds of case narratives are written from line listings.

Pattern: Cross-examinationNeeds scale5 decisionsDesigned for 30 to 300 agents