Fact Checking
The Fact Check tab in Notis analyzes clinical claims in your note and verifies them against peer-reviewed research. Each claim receives a verdict with confidence rating, explanation, and source citations — helping ensure content accuracy before sharing or publishing.
Accessing Fact Check
Open the AI tools panel by clicking the shield icon in the editor toolbar, or click the Fact Check tab if the panel is already open.
Running a Fact Check
Write or Finalize Your Note
Fact checking works best on complete or near-complete content. Finish drafting before running a check.
Open the Fact Check Tab
Click the shield icon in the editor toolbar, or navigate to the Fact Check tab in the AI panel.
Click Run Fact Check
Click the Run Fact Check button to begin analysis.
Review Results
Claims appear as color-coded cards with verdicts, explanations, and sources.
How Fact Checking Works
The fact-check process uses a multi-step approach to verify claims:
- Claim extraction — The system reads your note and identifies up to 8 factual, verifiable claims (statistics, treatment descriptions, research findings, clinical assertions)
- Classification — Each claim is categorized: well-established clinical knowledge is resolved immediately, while complex or specific claims are routed to academic search
- Verification — Claims needing search are checked against academic literature in parallel, each receiving independent verification
- Results — All claims are merged into a single results view, preserving the order they appear in your note
Understanding Verdicts
Each verified claim receives one of six verdict types. These help quickly identify which statements are well-supported and which need attention.
The claim is backed by credible sources and current research. These statements can be published with confidence.
Example: "Cognitive Behavioral Therapy has demonstrated effectiveness in treating anxiety disorders."
What to do: No action needed. The claim is verified and ready to publish.
The claim contains partially accurate information but oversimplifies, omits important context, misattributes data, or conflates related conditions.
Example: "Postpartum anxiety affects about 1 in 10 new parents."
Why it's misleading: Systematic reviews estimate prevalence around 12–13% for postpartum mothers specifically. The statistic understates prevalence and inaccurately extends to all new parents.
What to do: Review the explanation and sources. Add qualifiers, update figures, or provide more context. Phrases like "research suggests approximately..." help convey appropriate nuance.
The claim was accurate when originally published but has been superseded by more recent research or updated clinical guidelines.
Example: A treatment recommendation from 2015 that has since been revised by newer studies.
What to do: Update the claim with current research findings or note when the information was current.
The claim contradicts established research or contains factually incorrect information.
Example: "Depression is caused solely by chemical imbalances in the brain."
What to do: Remove or substantially revise the claim. Review the provided sources to understand the accurate information.
No credible evidence was found to verify the claim. This doesn't mean the claim is false, but that it cannot be confirmed through available research.
Example: A very specific clinical observation that hasn't been formally studied.
What to do: Consider whether the claim is essential. If so, rephrase as clinical observation rather than established fact, or remove entirely.
The claim is too subjective, vague, or forward-looking to be verified against research sources.
Example: "Many clients find this technique helpful" or "This approach feels more natural for therapists."
What to do: These claims are typically fine to publish but may benefit from qualifiers acknowledging their subjective nature.
Reviewing Results
Claim Cards
Each claim appears as a color-coded card in the results panel:
- Quoted text — The claim wording from your note. This may be edited during fact-checking for clarity but preserves the original meaning
- Verdict badge — Color-coded label (Supported, Misleading, False, etc.)
- Explanation — Why the claim received this verdict, with clinical context
- Sources — Links to research articles or evidence that informed the rating
Counts Summary
At the top of the results, a summary shows counts for each verdict type (e.g., "7 Supported, 2 Misleading, 2 Unsupported"), giving a quick sense of overall content accuracy. Clicking a verdict type filters the claim cards to show only those with that verdict.
Taking Action on Results
Revising Flagged Claims
For claims marked as Misleading, Outdated, False, or Unsupported:
Review the Explanation
Read why the claim was flagged and examine the supporting sources.
Edit in the Note
Click into the editor and locate the claim in your text.
Revise the Claim
Apply appropriate changes:
- Add qualifiers to provide context ("research suggests..." instead of definitive statements)
- Update statistics with current figures
- Include citations from the provided sources
- Remove claims that cannot be adequately supported
Ask Chat for Help
Switch to the Chat tab and ask for help revising specific claims. The assistant has access to your fact-check results.
Using Results in Chat
Fact-check results are automatically available as context in the Chat tab. After running a fact check, switch to Chat and ask:
- "How should I revise the misleading claims?"
- "Can you suggest better wording for the flagged statistics?"
- "What qualifiers should I add to make the unsupported claims more accurate?"
When to Run Fact Checks
- After completing the first draft
- Following major content revisions
- Before sharing or publishing content
- After adding new statistics or research findings
Best Practices
Write complete drafts first: The system works best with substantial content. Running checks on partial notes may miss important context.
Pay attention to statistics: Numbers, percentages, and prevalence figures are the most common source of misleading claims. Always verify these carefully.
Don't ignore misleading verdicts: These often reveal opportunities to add important nuance that makes content more credible and clinically responsible.
Review confidence scores: Lower confidence percentages suggest the system is less certain. Take extra care reviewing these claims manually.
Consider the audience: Some claims that are technically unsupported may be appropriate as clinical observations in certain contexts. Use professional judgment when deciding how to address flags.
Combine with Research: If a claim is flagged, use the Research tab to find current sources that can help you revise the claim accurately.
Limitations
The fact-check tool analyzes claims based on available academic literature:
- Factual statements about research findings
- Clinical statistics and outcome data
- Therapeutic technique descriptions
- Evidence-based treatment approaches
- Very recent research published within the last few months
- Proprietary therapeutic approaches with limited published evidence
- Organizational claims (e.g., specific statistics attributed to PSI or MHA websites)
- Claims requiring deep contextual knowledge of specialized therapeutic modalities
