Skip to main content
Pull structured, evidence-grounded key messages from clinical or scientific sources — ready for use in messaging frameworks, briefing documents, content planning, or stakeholder communications.
Risk tier: Medium  ·  Review requirement: Enhanced review with source cross-checkKey messages extracted here are starting points for development, not approved promotional claims. Messages that will inform promotional content must go through MLR review before use.

​
What AI does

  • Extracts candidate messages from a dense paper in minutes, giving you a structured set to evaluate rather than a blank page
  • Drafts each message in a consistent format: message + supporting data point + evidence strength + qualifiers
  • Generates 2–3 alternative framings of the same finding for comparison
  • Organises messages by theme (efficacy, safety, PROs, practical considerations) against a messaging framework

​
What AI cannot do

AI extracts everything — it does not know which messages matter for your project, whether the evidence is robust enough to support a given message, or how to frame findings for a specific audience. Strategic and editorial judgement is yours.

​
Before you start

  • Read the source paper or data package yourself. Understand the study, its strengths, and its limitations before asking AI to extract messages.
  • Have the full source text available, not just an abstract.
  • Know the intended use for the messages (HCP communications, internal briefing, payer value story) — this shapes what AI should prioritise.
  • If an approved messaging framework exists, have it to hand for cross-reference.

​
Steps

1

Review the source evidence

Read the paper or data source yourself. Understanding the study’s strengths and limitations is essential for evaluating whether AI-extracted messages are appropriate.
2

Define context and intended use

Decide who will use these messages and for what purpose. Generic message extraction produces generic output. Be specific about audience and use case in your prompt.
3

Generate candidate key messages

Provide the full source text and intended use context to your AI tool using the prompt pattern below. Run the extraction.
4

Verify each message against the source

For every message, find the specific data point or finding in the source that supports it. Check for overstatement, selective emphasis, or unsupported interpretation.
5

Check evidence strength labels

Confirm that messages derived from primary endpoints, secondary endpoints, and exploratory/post-hoc analyses are correctly labelled. Non-inferiority results must not be framed as superiority.
6

Verify safety messages are included

Require at least one safety or tolerability message and one limitations message. An efficacy-only set is not balanced and will not reflect the evidence fairly.
7

Refine and prioritise

Edit messages for accuracy, clarity, and relevance to your project objectives. Remove or flag any messages that go beyond what the evidence supports.
8

Cross-reference with existing messaging

If an approved messaging framework exists, check alignment and identify any genuinely new messages that need to be flagged for strategic discussion.
9

Verify with RefCheckr

Run the final messages through RefCheckr to confirm source support before using them in deliverables.

​
Prompt pattern

You are a medical writing assistant specialising in evidence-based messaging. Your task is to extract key messages from the following source document.

For each key message:
1. State the message clearly in one sentence
2. Cite the specific data point or finding that supports it (include numbers, endpoints, and statistical results)
3. Note the strength of the supporting evidence (e.g., primary endpoint, secondary endpoint, subgroup analysis, post-hoc)
4. Flag any qualifiers or limitations that should accompany the message

Organise messages into these categories where applicable:
- Efficacy
- Safety and tolerability
- Patient-reported outcomes / quality of life
- Mechanism / pharmacology
- Practical considerations (dosing, administration, etc.)

Rules:
- Base messages only on the provided source. Do not include information from outside the source material.
- Do not overstate findings. If the data shows non-inferiority, do not frame it as superiority. If a result is from a subgroup, say so.
- Include relevant safety messages, not just efficacy highlights.
- Flag any message where the supporting evidence is exploratory or hypothesis-generating.

Source document:
[INSERT FULL TEXT]

Intended use for these messages: [SPECIFY — e.g., HCP slide deck, internal briefing, messaging framework development]

​
Human review checklist

  • Every key message is directly supported by a specific, cited finding in the source
  • No message overstates the evidence (e.g., non-inferiority framed as superiority)
  • Subgroup and post-hoc findings are clearly identified as such
  • Safety and tolerability messages are included and fairly represent the data
  • Limitations and qualifiers are noted for each message
  • Messages are appropriate for the stated intended use
  • No messages introduce information not present in the source
  • Messages do not use promotional language unless the content is subject to MLR review
  • If an existing messaging framework exists, new messages are consistent or flagged as additions

​
Common failure modes

RiskWhat to look for
Overstated messages”Superior efficacy” claimed from a non-inferiority trial — verify message against the specific trial design and endpoint
Cherry-picked findingsStriking subgroup result foregrounded without the modest overall population result
Conflated endpointsPrimary and secondary endpoints merged into a single message that makes both sound like primary results
Missing contextPopulation, subgroup, comparator, or analysis-type qualifiers absent from a message
Promotional framingLanguage like “best-in-class” or “transformative” — review for promotional signals before sharing with brand or strategy teams

​
Relevant tools

RefCheckr

Verify that extracted key messages are supported by the cited source.

MedCheckr

Pre-screen messages for promotional compliance signals before use in promotional content.

​
Next steps

Build a Content Outline

Structure a deliverable around your key messages.

Verify Claims

Confirm every message is supported by the cited source.

Adapt for Audiences

Reframe messages for different stakeholder groups.