Systematically check whether document claims are accurately supported by their cited references.
~20 min with AI, ~2–3 hours without
Expert review required; verification results must be assessed by a qualified reviewer.Document + references → RefCheckr verify-rewrite-recheck loop → Human review of rewrites → Spot-check unflagged → Corrected document
Best for
Pre-checking referencing accuracy before MLR (Medical, Legal, Regulatory) review
Reviewing documents that cite multiple references across many claims
Catching reference drift in documents that have been through several revision rounds
Conducting QC on a deliverable before client or internal review
Inputs
The document to be verified, with clear claim-to-reference mapping
Full text of all cited references (not abstracts alone — partial verification creates a false sense of security)
Any approved messaging framework or key message platform for context on intended claims
Steps
1
Prepare the document
Ensure every verifiable claim has a clear reference citation. Number or label claims if the document does not already do so.
2
Gather all reference materials
Obtain the full text of every cited reference. Use high-quality, text-searchable PDFs. Scanned PDFs with OCR errors cause false flags.
3
Run automated verification
Use RefCheckr to verify each claim against its cited reference, rewrite any that don’t match, and re-check the rewrites for ABPI compliance — RefCheckr closes that loop automatically. If you don’t have RefCheckr, the prompt pattern below produces a verify-only first pass that catches numerical mismatches, language strength discrepancies, and missing qualifiers at scale, leaving the rewrite to you.
4
Review every output
For each rewrite or pass-through that RefCheckr returns, open the cited reference and confirm the AI’s verdict. Do not accept rewrites without checking the source — the closed-loop reduces drift, but a wrong rewrite that passes its own re-check still needs human judgement on context and editorial fit.
5
Spot-check unflagged claims
Automated tools do not catch everything. Manually spot-check 20–30% of unflagged claims, prioritising those with specific numerical data, comparative language, or endpoint results.
6
Correct, annotate, and re-verify
Apply the rewrites you accepted, fix anything the AI missed, update references, and document the verification outcome. If you made further changes after RefCheckr’s closed-loop output, run the loop again to confirm corrections still hold.
Output
A structured verification report covering every verifiable claim, with a clear verdict for each (SUPPORTED, PARTIALLY SUPPORTED, NOT SUPPORTED, CANNOT VERIFY). The report flags numerical discrepancies with exact values from both the document and the reference, highlights language strength mismatches, and notes missing qualifiers — formatted so a reviewer can work through it systematically.
Worked example: verifying a detail aid claim
Document claim:
“Drug Z reduced the risk of disease progression by 45% compared to standard of care (HR 0.55, p=0.002).”¹
Reference 1 (published paper) states:
In the per-protocol population, Drug Z was associated with a 38% reduction in the risk of disease progression compared to standard of care (HR 0.62, 95% CI: 0.44–0.87; p=0.006).
Verification output:
Check
Finding
Verdict
NOT SUPPORTED
Numerical mismatch
Document states HR 0.55 and 45% reduction. Reference states HR 0.62 and 38% reduction.
p-value mismatch
Document states p=0.002. Reference states p=0.006.
Missing qualifier
Reference specifies per-protocol population. Document implies ITT by omitting this qualifier.
Action required
Correct HR to 0.62, risk reduction to 38%, p-value to 0.006. Add “in the per-protocol population” qualifier. Confirm with study team whether ITT results are available.
This type of multi-error claim is common after several revision rounds — each edit drifts the numbers slightly further from the source.
Prompt pattern
You are a medical writing QC assistant. Your task is to verify the following claims against their cited references.For each claim:1. State the claim exactly as it appears in the document2. Identify the cited reference3. Find the relevant section of the reference that should support the claim4. Assess whether the reference supports the claim: SUPPORTED / PARTIALLY SUPPORTED / NOT SUPPORTED / CANNOT VERIFY5. If partially supported or not supported, explain the discrepancy6. Flag any numerical mismatches (different p-values, percentages, sample sizes, etc.)7. Flag any claims where the document uses stronger language than the referenceDocument claims:[INSERT DOCUMENT TEXT OR CLAIM LIST WITH REFERENCE NUMBERS]Reference materials:[INSERT REFERENCE TEXTS — one per reference, clearly labelled]Rules:- Compare claims strictly against the cited reference. Do not use your general knowledge to fill gaps.- If a reference does not contain information relevant to the claim, mark it as NOT SUPPORTED by this reference.- Note any qualifiers in the reference that are missing from the claim (e.g., subgroup, post-hoc, exploratory).
Customisation: For large documents, break claims into batches of 10–15 to keep AI output focused. For promotional materials, add a rule to flag any claim that implies superiority without a head-to-head comparison in the reference.
Why this works
AI excels at the mechanical comparison work — scanning dozens of claim-reference pairs for numerical mismatches, language strength discrepancies, and missing qualifiers far faster than manual review. The human reviewer handles what AI cannot: assessing whether a technically supported claim is used in a misleading context, judging whether partial support is sufficient, and catching important omissions from the references.
Common mistakes
Trusting unflagged claims as verified
A detail aid states “Treatment X reduced hospitalisations by 35%” citing Reference 3. The actual figure is 25%. Automated verification misses it because the reference does mention hospitalisations — just with different numbers. Always spot-check 20–30% of unflagged claims, especially those with specific numerical data.
Acting on false positives without checking
A claim is flagged as unsupported because the tool could not match paraphrased language to the reference text. The claim is actually fine. Review every flag against the source before making changes — false positives create unnecessary rework.
Poor reference PDF quality
A scanned PDF has OCR errors in a results table, and the tool reads “p=0.0O3” instead of “p=0.003” and flags a mismatch. Ensure reference PDFs are high-quality, text-searchable copies. Re-scan or obtain publisher PDFs where possible.
Context-blind verification
A claim about response rates is technically supported by the reference, but the document uses it in a comparative context the reference does not support. The tool sees support; the context is misleading. Assess each claim in the context of the full document narrative, not just as isolated pairs.
Assuming verification covers reference quality
Verification confirms that claims are supported by cited references. It does not check whether the cited reference is the best available evidence or whether newer data exists. Reference selection quality is a separate editorial and scientific judgement.
Primary tool — closed-loop claim verification, rewrite, and re-check
Complements:Scite.ai for citation context analysis (whether a citation supports, contrasts, or merely mentions a claim). Perplexity for quick fact-checking and finding source material to verify against. Zotero or EndNote for managing and organising the reference library. RefCheckr runs a closed loop — it verifies claims against cited evidence, rewrites any that don’t match, and re-checks the rewrites for ABPI compliance; Scite adds context on how that evidence treats the claim; Perplexity can help locate evidence; Zotero/EndNote store and organise it.
Frequently asked questions
Can AI automatically check whether every claim in a document is supported?
AI can flag likely mismatches at scale, which is valuable for long or multi-reference documents. It cannot confirm that a reference genuinely supports a claim — that still requires a human read of the cited passage. Treat AI verification as a triage layer, not a sign-off.
What's the difference between a reference existing and a reference supporting a claim?
A real reference can still be cited in the wrong context. The paper may discuss the same topic but report a different outcome, a different population, or contradict the claim. Verification means confirming the cited passage says this specific thing, not that the paper exists.
How do I handle a claim supported by multiple references?
Check each reference independently against the specific part of the claim it is cited for. If two references are cited together, confirm that each one supports the claim on its own, or that together they cover it without gaps. Unclear attribution is a common source of drift.
Does AI verification replace a fact-check?
No. AI verification catches a substantial share of mismatches quickly, but it does not replace the human fact-check required before sign-off. Use it to make the human review shorter and more focused, not to remove it.
Can AI detect when a reference is cited in the wrong context?
Sometimes. AI can flag when the cited passage reports a different endpoint, population, or outcome than the claim describes. Subtle context errors — a pooled analysis cited for a single-study claim, a post-hoc finding cited as primary — are harder to catch and benefit from expert review.
Review checklist
Human review checklist
All flagged claims have been manually reviewed against the cited reference
A sample of unflagged claims has been spot-checked for accuracy
Numerical data (p-values, CIs, percentages, sample sizes) has been verified against source
Claims use language consistent with the strength of evidence in the source (no overstatement)
Qualifiers present in the source are preserved in the claims (subgroup, post-hoc, exploratory, etc.)
Safety claims are verified with the same rigour as efficacy claims
Any “NOT SUPPORTED” or “PARTIALLY SUPPORTED” flags have been resolved
Changes made during verification have been documented
The document is ready for formal review with reference accuracy confirmed
{"docsConfig":{"theme":"mint","$schema":"https://mintlify.com/docs.json","name":"Medical Writing AI Playbook","colors":{"primary":"#0F6B5E","light":"#16A085","dark":"#0A4F45"},"logo":{"light":"https://mintcdn.com/pharmatoolsai/gAGdD6wm-6G2DEhI/logo-light.svg?fit=max\u0026auto=format\u0026n=gAGdD6wm-6G2DEhI\u0026q=85\u0026s=94a8f5eebb1eb89326bf8910b46a9864","dark":"https://mintcdn.com/pharmatoolsai/gAGdD6wm-6G2DEhI/logo-dark.svg?fit=max\u0026auto=format\u0026n=gAGdD6wm-6G2DEhI\u0026q=85\u0026s=c18b9685789746bf74dae542e49065d9","href":"/"},"favicon":"https://mintcdn.com/pharmatoolsai/Y6rjNnVLQ1Y-k30w/favicon.png?fit=max\u0026auto=format\u0026n=Y6rjNnVLQ1Y-k30w\u0026q=85\u0026s=6539a69929fecbf123cdd920cdc435e6","appearance":{"default":"light"},"navbar":{"links":[{"href":"https://pharmatools.ai","label":"PharmaTools.AI"},{"icon":"github","href":"https://github.com/nickjlamb/medical-writing-ai-playbook","label":"GitHub"}]},"footer":{"socials":{"website":"https://pharmatools.ai","github":"https://github.com/nickjlamb/medical-writing-ai-playbook"}},"search":{"prompt":"Search workflows, prompts, tools..."},"seo":{"metatags":{"og:site_name":"Medical Writing AI Playbook","og:type":"website","og:image":"https://playbook.pharmatools.ai/og-image.png","og:image:width":"1200","og:image:height":"630","twitter:card":"summary_large_image","twitter:image":"https://playbook.pharmatools.ai/og-image.png"}},"styling":{"latex":false},"navigation":{"pages":[]}},"deployment":{"deploymentMetadataValue":{"subdomain":"pharmatoolsai","hasAudioNarration":false,"actualSubdomain":"pharmatoolsai","customDomains":["playbook.pharmatools.ai"],"gitSource":{"type":"github","owner":"nickjlamb","repo":"medical-writing-ai-playbook","deployBranch":"main","contentDirectory":"","isPrivate":true,"mountPath":"","label":""},"entitlements":{},"buildId":"6afdba35be4a433e33e9dd8c","clientVersion":"0.0.4205","searchFilterCounts":[{"language":"default","tag":"Playbook","count":51},{"language":"default","tag":"AI Workflow","count":18},{"language":"default","tag":"Principles","count":12},{"language":"default","tag":"Tools","count":7},{"language":"default","tag":"Drafting","count":6},{"language":"default","tag":"Overview","count":5},{"language":"default","tag":"Templates","count":5},{"language":"default","tag":"Evidence Workflows","count":5},{"language":"default","tag":"Prompt Patterns","count":4},{"language":"default","tag":"Validation","count":3}]},"assistantConfig":{"enableAskAiButton":false,"enableStarterQuestions":false,"enableContextualStarterQuestions":false,"enableAskAiSearchBar":true},"basePath":"","hasChatPermissions":false,"hasWebMcpPermissions":false,"toggles":[{"name":"kill-isr","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"custom-rbac-roles","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"agent-mintlify-slack","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"agent-integrations-settings","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"rbac-expansion","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"automations-pr-grouping","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"custom-prefix","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"analytics-v7","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"user-flows-analytics","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"no-publish-flow","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"workos-onboarding","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"snapshot-serving-live","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"stripTemperatureMiddleware","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"support-platform-pylon-zendesk","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"snapshot-serving","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"instant-rollback","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"granular-notifications-server","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"granular-notifications","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"automations-fable","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"fixed-ai-action-pricing","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"automation-resolve-merge-conflict","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"scoped-publishing","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"mintql","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"page-level-analytics","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"editor-version-history-v2","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"billing-page-v2","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"full-page-search","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"workspace-projects","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"assistant-v5","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"fork-pr-previews","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"revalidation-jwt","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"workos-sso-scim","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"hobby-abuse-scan","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"dashboard-inbox","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"workos-signup","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"turbopuffer-dual-write","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"workos-enabled","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"approve-review-flow","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"editor-metadata","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"sso-setup-deadline-banner","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"turbopuffer-serve","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"live-agents","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"starter-pricing","enabled":true,"variant":{"name":"control","enabled":true,"payload":{"type":"string","value":"control"},"feature_enabled":true},"impressionData":false},{"name":"llms-txt-v2","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false},{"name":"wildcard-email-auth","enabled":true,"variant":{"name":"disabled","enabled":false,"feature_enabled":true},"impressionData":false}],"googleFontsInlined":false},"customCss":[{"filePath":"style.css","content":"body:has(.playbook-hero) header:has(#page-title){display:none!important}"}],"page":{"pagePath":"workflows/verify-claims-against-references","slug":"/workflows/verify-claims-against-references","pageMetadata":{"title":"Verify Claims Against References","sidebarTitle":"Verify Claims","description":"Systematically check whether document claims are accurately supported by their cited references.","icon":"search","href":"/workflows/verify-claims-against-references"},"description":{"compiledSource":"\"use strict\";\nconst {jsx: _jsx} = arguments[0];\nconst {useMDXComponents: _provideComponents} = arguments[0];\nfunction _createMdxContent(props) {\n const _components = {\n p: \"p\",\n ..._provideComponents(),\n ...props.components\n };\n return _jsx(_components.p, {\n children: \"Systematically check whether document claims are accurately supported by their cited references.\"\n });\n}\nfunction MDXContent(props = {}) {\n const {wrapper: MDXLayout} = {\n ..._provideComponents(),\n ...props.components\n };\n return MDXLayout ? _jsx(MDXLayout, {\n ...props,\n children: _jsx(_createMdxContent, {\n ...props\n })\n }) : _createMdxContent(props);\n}\nreturn {\n default: MDXContent\n};\n","frontmatter":{},"scope":{}},"pageType":"default","mdx":{"mdxSource":{"compiledSource":"\"use strict\";const{Fragment:_Fragment,jsx:_jsx,jsxs:_jsxs}=arguments[0];const{useMDXComponents:_provideComponents}=arguments[0];const RiskBadge=({level=`low`})=>{const styles={\"low\":{bg:`#D1FAE5`,fg:`#065F46`,label:`Low`},\"low-medium\":{bg:`#ECFCCB`,fg:`#365314`,label:`Low–Medium`},\"medium\":{bg:`#FEF3C7`,fg:`#854D0E`,label:`Medium`},\"medium-high\":{bg:`#FFEDD5`,fg:`#9A3412`,label:`Medium–High`},\"high\":{bg:`#FECACA`,fg:`#991B1B`,label:`High`},\"critical\":{bg:`#DC2626`,fg:`#FFFFFF`,label:`Critical`}};const s=styles[level]||styles.low;return _jsxs(`span`,{style:{display:`inline-block`,backgroundColor:s.bg,color:s.fg,padding:`2px 12px`,borderRadius:`999px`,fontSize:`0.75rem`,fontWeight:700,letterSpacing:`0.03em`,textTransform:`uppercase`,verticalAlign:`middle`,whiteSpace:`nowrap`},children:[`Risk tier · `,s.label]})};function _createMdxContent(props){const _components={a:`a`,blockquote:`blockquote`,code:`code`,em:`em`,hr:`hr`,li:`li`,p:`p`,pre:`pre`,span:`span`,strong:`strong`,tbody:`tbody`,td:`td`,th:`th`,thead:`thead`,tr:`tr`,ul:`ul`,..._provideComponents(),...props.components},{Accordion,AccordionGroup,CodeBlock,Heading,Step,Steps,Table,Tip,Warning,_MdxComponentBoundary}=_components;if(!Accordion)_missingMdxReference(`Accordion`,true);if(!AccordionGroup)_missingMdxReference(`AccordionGroup`,true);if(!CodeBlock)_missingMdxReference(`CodeBlock`,true);if(!Heading)_missingMdxReference(`Heading`,true);if(!Step)_missingMdxReference(`Step`,true);if(!Steps)_missingMdxReference(`Steps`,true);if(!Table)_missingMdxReference(`Table`,true);if(!Tip)_missingMdxReference(`Tip`,true);if(!Warning)_missingMdxReference(`Warning`,true);if(!_MdxComponentBoundary)_missingMdxReference(`_MdxComponentBoundary`,true);return _jsxs(_Fragment,{children:[_jsxs(Warning,{children:[_jsx(_MdxComponentBoundary,{name:`RiskBadge`,children:_jsx(RiskBadge,{level:`high`})}),_jsx(_components.p,{children:`~20 min with AI, ~2–3 hours without\nExpert review required; verification results must be assessed by a qualified reviewer.`}),_jsx(_components.p,{children:`Document + references → RefCheckr verify-rewrite-recheck loop → Human review of rewrites → Spot-check unflagged → Corrected document`})]}),`\n`,_jsx(Heading,{level:`2`,id:`best-for`,children:`Best for`}),`\n`,_jsxs(_components.ul,{children:[`\n`,_jsx(_components.li,{children:`Pre-checking referencing accuracy before MLR (Medical, Legal, Regulatory) review`}),`\n`,_jsx(_components.li,{children:`Reviewing documents that cite multiple references across many claims`}),`\n`,_jsx(_components.li,{children:`Catching reference drift in documents that have been through several revision rounds`}),`\n`,_jsx(_components.li,{children:`Conducting QC on a deliverable before client or internal review`}),`\n`]}),`\n`,_jsx(Heading,{level:`2`,id:`inputs`,children:`Inputs`}),`\n`,_jsxs(_components.ul,{children:[`\n`,_jsx(_components.li,{children:`The document to be verified, with clear claim-to-reference mapping`}),`\n`,_jsx(_components.li,{children:`Full text of all cited references (not abstracts alone — partial verification creates a false sense of security)`}),`\n`,_jsx(_components.li,{children:`Any approved messaging framework or key message platform for context on intended claims`}),`\n`]}),`\n`,_jsx(Heading,{level:`2`,id:`steps`,children:`Steps`}),`\n`,_jsxs(Steps,{children:[_jsx(Step,{title:`Prepare the document`,children:_jsx(_components.p,{children:`Ensure every verifiable claim has a clear reference citation. Number or label claims if the document does not already do so.`})}),_jsx(Step,{title:`Gather all reference materials`,children:_jsx(_components.p,{children:`Obtain the full text of every cited reference. Use high-quality, text-searchable PDFs. Scanned PDFs with OCR errors cause false flags.`})}),_jsx(Step,{title:`Run automated verification`,children:_jsx(_components.p,{children:`Use RefCheckr to verify each claim against its cited reference, rewrite any that don’t match, and re-check the rewrites for ABPI compliance — RefCheckr closes that loop automatically. If you don’t have RefCheckr, the prompt pattern below produces a verify-only first pass that catches numerical mismatches, language strength discrepancies, and missing qualifiers at scale, leaving the rewrite to you.`})}),_jsx(Step,{title:`Review every output`,children:_jsx(_components.p,{children:`For each rewrite or pass-through that RefCheckr returns, open the cited reference and confirm the AI’s verdict. Do not accept rewrites without checking the source — the closed-loop reduces drift, but a wrong rewrite that passes its own re-check still needs human judgement on context and editorial fit.`})}),_jsx(Step,{title:`Spot-check unflagged claims`,children:_jsx(_components.p,{children:`Automated tools do not catch everything. Manually spot-check 20–30% of unflagged claims, prioritising those with specific numerical data, comparative language, or endpoint results.`})}),_jsx(Step,{title:`Correct, annotate, and re-verify`,children:_jsx(_components.p,{children:`Apply the rewrites you accepted, fix anything the AI missed, update references, and document the verification outcome. If you made further changes after RefCheckr’s closed-loop output, run the loop again to confirm corrections still hold.`})})]}),`\n`,_jsx(Heading,{level:`2`,id:`output`,children:`Output`}),`\n`,_jsx(_components.p,{children:`A structured verification report covering every verifiable claim, with a clear verdict for each (SUPPORTED, PARTIALLY SUPPORTED, NOT SUPPORTED, CANNOT VERIFY). The report flags numerical discrepancies with exact values from both the document and the reference, highlights language strength mismatches, and notes missing qualifiers — formatted so a reviewer can work through it systematically.`}),`\n`,_jsxs(Accordion,{title:`Worked example: verifying a detail aid claim`,id:`worked-example-verifying-a-detail-aid-claim`,children:[_jsx(_components.p,{children:_jsx(_components.strong,{children:`Document claim:`})}),_jsxs(_components.blockquote,{children:[`\n`,_jsx(_components.p,{children:`“Drug Z reduced the risk of disease progression by 45% compared to standard of care (HR 0.55, p=0.002).”¹`}),`\n`]}),_jsx(_components.p,{children:_jsx(_components.strong,{children:`Reference 1 (published paper) states:`})}),_jsxs(_components.blockquote,{children:[`\n`,_jsx(_components.p,{children:`In the per-protocol population, Drug Z was associated with a 38% reduction in the risk of disease progression compared to standard of care (HR 0.62, 95% CI: 0.44–0.87; p=0.006).`}),`\n`]}),_jsx(_components.p,{children:_jsx(_components.strong,{children:`Verification output:`})}),_jsxs(Table,{children:[_jsx(_components.thead,{children:_jsxs(_components.tr,{children:[_jsx(_components.th,{children:`Check`}),_jsx(_components.th,{children:`Finding`})]})}),_jsxs(_components.tbody,{children:[_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.strong,{children:`Verdict`})}),_jsx(_components.td,{children:`NOT SUPPORTED`})]}),_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.strong,{children:`Numerical mismatch`})}),_jsx(_components.td,{children:`Document states HR 0.55 and 45% reduction. Reference states HR 0.62 and 38% reduction.`})]}),_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.strong,{children:`p-value mismatch`})}),_jsx(_components.td,{children:`Document states p=0.002. Reference states p=0.006.`})]}),_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.strong,{children:`Missing qualifier`})}),_jsx(_components.td,{children:`Reference specifies per-protocol population. Document implies ITT by omitting this qualifier.`})]}),_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.strong,{children:`Action required`})}),_jsx(_components.td,{children:`Correct HR to 0.62, risk reduction to 38%, p-value to 0.006. Add “in the per-protocol population” qualifier. Confirm with study team whether ITT results are available.`})]})]})]}),_jsx(_components.p,{children:`This type of multi-error claim is common after several revision rounds — each edit drifts the numbers slightly further from the source.`})]}),`\n`,_jsx(Heading,{level:`2`,id:`prompt-pattern`,children:`Prompt pattern`}),`\n`,_jsx(CodeBlock,{numberOfLines:`22`,language:`text`,children:_jsx(_components.pre,{className:`shiki shiki-themes github-light-default dark-plus`,style:{backgroundColor:`#ffffff`,\"--shiki-dark-bg\":`#0B0C0E`,color:`#1f2328`,\"--shiki-dark\":`#D4D4D4`},language:`text`,children:_jsxs(_components.code,{language:`text`,numberOfLines:`22`,children:[_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`You are a medical writing QC assistant. Your task is to verify the following claims against their cited references.`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`For each claim:`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`1. State the claim exactly as it appears in the document`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`2. Identify the cited reference`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`3. Find the relevant section of the reference that should support the claim`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`4. Assess whether the reference supports the claim: SUPPORTED / PARTIALLY SUPPORTED / NOT SUPPORTED / CANNOT VERIFY`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`5. If partially supported or not supported, explain the discrepancy`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`6. Flag any numerical mismatches (different p-values, percentages, sample sizes, etc.)`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`7. Flag any claims where the document uses stronger language than the reference`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`Document claims:`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`[INSERT DOCUMENT TEXT OR CLAIM LIST WITH REFERENCE NUMBERS]`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`Reference materials:`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`[INSERT REFERENCE TEXTS — one per reference, clearly labelled]`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`Rules:`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`- Compare claims strictly against the cited reference. Do not use your general knowledge to fill gaps.`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`- If a reference does not contain information relevant to the claim, mark it as NOT SUPPORTED by this reference.`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{children:`- Note any qualifiers in the reference that are missing from the claim (e.g., subgroup, post-hoc, exploratory).`})}),`\n`,_jsx(_components.span,{className:`line`,children:_jsx(_components.span,{})})]})})}),`\n`,_jsx(Tip,{children:_jsxs(_components.p,{children:[_jsx(_components.strong,{children:`Customisation:`}),` For large documents, break claims into batches of 10–15 to keep AI output focused. For promotional materials, add a rule to flag any claim that implies superiority without a head-to-head comparison in the reference.`]})}),`\n`,_jsx(Heading,{level:`2`,id:`why-this-works`,children:`Why this works`}),`\n`,_jsx(_components.p,{children:`AI excels at the mechanical comparison work — scanning dozens of claim-reference pairs for numerical mismatches, language strength discrepancies, and missing qualifiers far faster than manual review. The human reviewer handles what AI cannot: assessing whether a technically supported claim is used in a misleading context, judging whether partial support is sufficient, and catching important omissions from the references.`}),`\n`,_jsx(Heading,{level:`2`,id:`common-mistakes`,children:`Common mistakes`}),`\n`,_jsxs(AccordionGroup,{children:[_jsx(Accordion,{title:`Trusting unflagged claims as verified`,id:`trusting-unflagged-claims-as-verified`,children:_jsx(_components.p,{children:`A detail aid states “Treatment X reduced hospitalisations by 35%” citing Reference 3. The actual figure is 25%. Automated verification misses it because the reference does mention hospitalisations — just with different numbers. Always spot-check 20–30% of unflagged claims, especially those with specific numerical data.`})}),_jsx(Accordion,{title:`Acting on false positives without checking`,id:`acting-on-false-positives-without-checking`,children:_jsx(_components.p,{children:`A claim is flagged as unsupported because the tool could not match paraphrased language to the reference text. The claim is actually fine. Review every flag against the source before making changes — false positives create unnecessary rework.`})}),_jsx(Accordion,{title:`Poor reference PDF quality`,id:`poor-reference-pdf-quality`,children:_jsx(_components.p,{children:`A scanned PDF has OCR errors in a results table, and the tool reads “p=0.0O3” instead of “p=0.003” and flags a mismatch. Ensure reference PDFs are high-quality, text-searchable copies. Re-scan or obtain publisher PDFs where possible.`})}),_jsx(Accordion,{title:`Context-blind verification`,id:`context-blind-verification`,children:_jsx(_components.p,{children:`A claim about response rates is technically supported by the reference, but the document uses it in a comparative context the reference does not support. The tool sees support; the context is misleading. Assess each claim in the context of the full document narrative, not just as isolated pairs.`})}),_jsx(Accordion,{title:`Assuming verification covers reference quality`,id:`assuming-verification-covers-reference-quality`,children:_jsx(_components.p,{children:`Verification confirms that claims are supported by cited references. It does not check whether the cited reference is the best available evidence or whether newer data exists. Reference selection quality is a separate editorial and scientific judgement.`})})]}),`\n`,_jsx(Heading,{level:`2`,id:`tool-stack`,children:`Tool stack`}),`\n`,_jsxs(Table,{children:[_jsx(_components.thead,{children:_jsxs(_components.tr,{children:[_jsx(_components.th,{children:`Tool`}),_jsx(_components.th,{children:`Role`})]})}),_jsx(_components.tbody,{children:_jsxs(_components.tr,{children:[_jsx(_components.td,{children:_jsx(_components.a,{href:`/tools/refcheckr`,children:`RefCheckr`})}),_jsx(_components.td,{children:`Primary tool — closed-loop claim verification, rewrite, and re-check`})]})})]}),`\n`,_jsxs(_components.p,{children:[_jsx(_components.strong,{children:`Complements:`}),` `,_jsx(_components.a,{href:`https://scite.ai`,children:`Scite.ai`}),` for citation context analysis (whether a citation supports, contrasts, or merely mentions a claim). `,_jsx(_components.a,{href:`https://www.perplexity.ai`,children:`Perplexity`}),` for quick fact-checking and finding source material to verify against. `,_jsx(_components.a,{href:`https://www.zotero.org`,children:`Zotero`}),` or `,_jsx(_components.a,{href:`https://endnote.com`,children:`EndNote`}),` for managing and organising the reference library. RefCheckr runs a closed loop — it verifies claims against cited evidence, rewrites any that don’t match, and re-checks the rewrites for ABPI compliance; Scite adds context on how that evidence treats the claim; Perplexity can help locate evidence; Zotero/EndNote store and organise it.`]}),`\n`,_jsx(Heading,{level:`2`,id:`frequently-asked-questions`,children:`Frequently asked questions`}),`\n`,_jsxs(AccordionGroup,{children:[_jsx(Accordion,{title:`Can AI automatically check whether every claim in a document is supported?`,id:`can-ai-automatically-check-whether-every-claim-in-a-document-is-supported`,children:_jsx(_components.p,{children:`AI can flag likely mismatches at scale, which is valuable for long or multi-reference documents. It cannot confirm that a reference genuinely supports a claim — that still requires a human read of the cited passage. Treat AI verification as a triage layer, not a sign-off.`})}),_jsx(Accordion,{title:`What's the difference between a reference existing and a reference supporting a claim?`,id:`whats-the-difference-between-a-reference-existing-and-a-reference-supporting-a-claim`,children:_jsxs(_components.p,{children:[`A real reference can still be cited in the wrong context. The paper may discuss the same topic but report a different outcome, a different population, or contradict the claim. Verification means confirming the cited passage says `,_jsx(_components.em,{children:`this specific thing`}),`, not that the paper exists.`]})}),_jsx(Accordion,{title:`How do I handle a claim supported by multiple references?`,id:`how-do-i-handle-a-claim-supported-by-multiple-references`,children:_jsx(_components.p,{children:`Check each reference independently against the specific part of the claim it is cited for. If two references are cited together, confirm that each one supports the claim on its own, or that together they cover it without gaps. Unclear attribution is a common source of drift.`})}),_jsx(Accordion,{title:`Does AI verification replace a fact-check?`,id:`does-ai-verification-replace-a-fact-check`,children:_jsx(_components.p,{children:`No. AI verification catches a substantial share of mismatches quickly, but it does not replace the human fact-check required before sign-off. Use it to make the human review shorter and more focused, not to remove it.`})}),_jsx(Accordion,{title:`Can AI detect when a reference is cited in the wrong context?`,id:`can-ai-detect-when-a-reference-is-cited-in-the-wrong-context`,children:_jsx(_components.p,{children:`Sometimes. AI can flag when the cited passage reports a different endpoint, population, or outcome than the claim describes. Subtle context errors — a pooled analysis cited for a single-study claim, a post-hoc finding cited as primary — are harder to catch and benefit from expert review.`})})]}),`\n`,_jsx(Heading,{level:`2`,id:`review-checklist`,children:`Review checklist`}),`\n`,_jsx(Accordion,{title:`Human review checklist`,id:`human-review-checklist`,children:_jsxs(_components.ul,{children:[`\n`,_jsx(_components.li,{children:`All flagged claims have been manually reviewed against the cited reference`}),`\n`,_jsx(_components.li,{children:`A sample of unflagged claims has been spot-checked for accuracy`}),`\n`,_jsx(_components.li,{children:`Numerical data (p-values, CIs, percentages, sample sizes) has been verified against source`}),`\n`,_jsx(_components.li,{children:`Claims use language consistent with the strength of evidence in the source (no overstatement)`}),`\n`,_jsx(_components.li,{children:`Qualifiers present in the source are preserved in the claims (subgroup, post-hoc, exploratory, etc.)`}),`\n`,_jsx(_components.li,{children:`Safety claims are verified with the same rigour as efficacy claims`}),`\n`,_jsx(_components.li,{children:`Any “NOT SUPPORTED” or “PARTIALLY SUPPORTED” flags have been resolved`}),`\n`,_jsx(_components.li,{children:`Changes made during verification have been documented`}),`\n`,_jsx(_components.li,{children:`The document is ready for formal review with reference accuracy confirmed`}),`\n`]})}),`\n`,_jsx(_components.hr,{}),`\n`,_jsxs(_components.p,{children:[_jsx(_components.strong,{children:`Next steps:`}),` After verifying references, run `,_jsx(_components.a,{href:`/workflows/check-promotional-compliance`,children:`Check Promotional Compliance`}),` for promotional content or `,_jsx(_components.a,{href:`/workflows/check-document-consistency`,children:`Check Document Consistency`}),` for long documents, then complete `,_jsx(_components.a,{href:`/workflows/final-human-review`,children:`Final Human Review`}),` before submission.`]}),`\n`,_jsx(_components.hr,{}),`\n`,_jsx(_components.p,{children:_jsx(_components.em,{children:`Last reviewed: 15 April 2026`})})]})}function MDXContent(props={}){const{wrapper:MDXLayout}={..._provideComponents(),...props.components};return MDXLayout?_jsx(MDXLayout,{...props,children:_jsx(_createMdxContent,{...props})}):_createMdxContent(props)}return{RiskBadge,default:MDXContent};function _missingMdxReference(id,component){throw new Error(`Expected `+(component?`component`:`object`)+\" `\"+id+\"` to be defined: you likely forgot to import, pass, or provide it.\")}","frontmatter":{},"scope":{"config":{},"pageMetadata":{"title":"Verify Claims Against References","sidebarTitle":"Verify Claims","description":"Systematically check whether document claims are accurately supported by their cited references.","icon":"search","href":"/workflows/verify-claims-against-references"}}},"mdxSourceWithNoJs":null,"noJsSplices":[[145,834,")=>null;function _cr"]]},"mdxExtracts":{"tableOfContents":[{"title":"Best for","slug":"best-for","depth":2,"children":[]},{"title":"Inputs","slug":"inputs","depth":2,"children":[]},{"title":"Steps","slug":"steps","depth":2,"children":[]},{"title":"Output","slug":"output","depth":2,"children":[]},{"title":"Prompt pattern","slug":"prompt-pattern","depth":2,"children":[]},{"title":"Why this works","slug":"why-this-works","depth":2,"children":[]},{"title":"Common mistakes","slug":"common-mistakes","depth":2,"children":[]},{"title":"Tool stack","slug":"tool-stack","depth":2,"children":[]},{"title":"Frequently asked questions","slug":"frequently-asked-questions","depth":2,"children":[]},{"title":"Review checklist","slug":"review-checklist","depth":2,"children":[]}],"codeExamples":{}},"jsLinks":["/_mintlify/custom/pharmatoolsai/25cbb83bee4cdc5c/counter-worker/src/index.js"],"subdomain":"pharmatoolsai","siteName":"Medical Writing AI Playbook","title":"Verify Claims Against References - Medical Writing AI Playbook","navMaps":{"firstHrefInVersion":{},"firstHrefInLanguage":{}},"lastModified":"2026-05-04T11:16:46.643Z","originalFileLocation":"workflows/verify-claims-against-references.mdx"},"headHtml":"\u003cscript data-mint-head>(function(){var d=document.documentElement,p=null;if(!p){try{p=localStorage.getItem(\"isDarkMode\")}catch(e){}if(p==='true')p='dark';else if(p==='false')p='light';if(p!=='light'\u0026\u0026p!=='dark'\u0026\u0026p!=='system')p=\"light\"}var r=p==='system'?(matchMedia(\"(prefers-color-scheme: dark)\").matches?'dark':'light'):p;d.classList.remove(r==='dark'?'light':'dark');d.classList.add(r);d.setAttribute(\"data-theme-preference\",p)})();\u003c/script>\u003cmeta data-mint-head name=\"description\" content=\"Systematically check whether document claims are accurately supported by their cited references.\">\u003cmeta data-mint-head name=\"application-name\" content=\"Medical Writing AI Playbook\">\u003cmeta data-mint-head name=\"generator\" content=\"Mintlify\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"192x192\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/android-chrome-192x192.png\" media=\"(prefers-color-scheme: light)\">\u003clink data-mint-head rel=\"apple-touch-icon\" type=\"image/png\" sizes=\"180x180\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/apple-touch-icon.png\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"16x16\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/favicon-16x16.png\" media=\"(prefers-color-scheme: light)\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"32x32\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/favicon-32x32.png\" media=\"(prefers-color-scheme: light)\">\u003clink data-mint-head rel=\"shortcut icon\" type=\"image/x-icon\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/favicon.ico\" media=\"(prefers-color-scheme: light)\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"192x192\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon-dark/android-chrome-192x192.png\" media=\"(prefers-color-scheme: dark)\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"16x16\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon-dark/favicon-16x16.png\" media=\"(prefers-color-scheme: dark)\">\u003clink data-mint-head rel=\"icon\" type=\"image/png\" sizes=\"32x32\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon-dark/favicon-32x32.png\" media=\"(prefers-color-scheme: dark)\">\u003clink data-mint-head rel=\"shortcut icon\" type=\"image/x-icon\" href=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon-dark/favicon.ico\" media=\"(prefers-color-scheme: dark)\">\u003cmeta data-mint-head property=\"og:title\" content=\"Verify Claims Against References - Medical Writing AI Playbook\">\u003cmeta data-mint-head property=\"og:description\" content=\"Systematically check whether document claims are accurately supported by their cited references.\">\u003cmeta data-mint-head property=\"og:site_name\" content=\"Medical Writing AI Playbook\">\u003cmeta data-mint-head property=\"og:url\" content=\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references\">\u003cmeta data-mint-head property=\"og:type\" content=\"website\">\u003cmeta data-mint-head property=\"og:image\" content=\"https://playbook.pharmatools.ai/og-image.png\">\u003cmeta data-mint-head property=\"og:image:width\" content=\"1200\">\u003cmeta data-mint-head property=\"og:image:height\" content=\"630\">\u003cmeta data-mint-head name=\"twitter:card\" content=\"summary_large_image\">\u003cmeta data-mint-head name=\"twitter:title\" content=\"Verify Claims Against References - Medical Writing AI Playbook\">\u003cmeta data-mint-head name=\"twitter:description\" content=\"Systematically check whether document claims are accurately supported by their cited references.\">\u003cmeta data-mint-head name=\"twitter:image\" content=\"https://playbook.pharmatools.ai/og-image.png\">\u003cmeta data-mint-head name=\"twitter:image:width\" content=\"1200\">\u003cmeta data-mint-head name=\"twitter:image:height\" content=\"630\">\u003cmeta data-mint-head name=\"msapplication-config\" content=\"/mintlify-assets/_mintlify/favicons/pharmatoolsai/wLzuLM2o8yqKIevZ/_generated/favicon/browserconfig.xml\">\u003cmeta data-mint-head name=\"apple-mobile-web-app-title\" content=\"Medical Writing AI Playbook\">\u003cmeta data-mint-head name=\"msapplication-TileColor\" content=\"#0F6B5E\">\u003clink data-mint-head rel=\"canonical\" href=\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references\">\u003clink data-mint-head rel=\"alternate\" type=\"application/xml\" href=\"/sitemap.xml\">\u003clink data-mint-head rel=\"alternate\" type=\"text/markdown\" href=\"/workflows/verify-claims-against-references.md\">\u003cscript data-mint-head type=\"application/ld+json\">{\"@context\":\"https://schema.org\",\"@graph\":[{\"@type\":\"Organization\",\"@id\":\"https://playbook.pharmatools.ai/#organization\",\"name\":\"Medical Writing AI Playbook\",\"url\":\"https://playbook.pharmatools.ai\",\"logo\":{\"@type\":\"ImageObject\",\"url\":\"https://mintcdn.com/pharmatoolsai/gAGdD6wm-6G2DEhI/logo-light.svg?fit=max\u0026auto=format\u0026n=gAGdD6wm-6G2DEhI\u0026q=85\u0026s=94a8f5eebb1eb89326bf8910b46a9864\"}},{\"@type\":\"WebSite\",\"@id\":\"https://playbook.pharmatools.ai#website\",\"name\":\"Medical Writing AI Playbook\",\"url\":\"https://playbook.pharmatools.ai\",\"publisher\":{\"@id\":\"https://playbook.pharmatools.ai/#organization\"}},{\"@type\":\"WebPage\",\"@id\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references#webpage\",\"url\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references\",\"name\":\"Verify Claims Against References\",\"description\":\"Systematically check whether document claims are accurately supported by their cited references.\",\"dateModified\":\"2026-05-04T11:16:46.643Z\",\"isPartOf\":{\"@id\":\"https://playbook.pharmatools.ai#website\"},\"breadcrumb\":{\"@id\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references#breadcrumb\"}},{\"@type\":\"BreadcrumbList\",\"@id\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"AI Workflow\",\"item\":\"https://playbook.pharmatools.ai/workflows/find-evidence\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Verify Claims Against References\",\"item\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references\"}]},{\"@type\":[\"Article\",\"TechArticle\"],\"@id\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references#article\",\"headline\":\"Verify Claims Against References\",\"name\":\"Verify Claims Against References\",\"description\":\"Systematically check whether document claims are accurately supported by their cited references.\",\"url\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references\",\"mainEntityOfPage\":{\"@id\":\"https://playbook.pharmatools.ai/workflows/verify-claims-against-references#webpage\"},\"image\":\"https://playbook.pharmatools.ai/og-image.png\",\"dateModified\":\"2026-05-04T11:16:46.643Z\",\"publisher\":{\"@id\":\"https://playbook.pharmatools.ai/#organization\"},\"isPartOf\":{\"@id\":\"https://playbook.pharmatools.ai#website\"}}]}\u003c/script>\u003cscript data-mint-head type=\"application/ld+json\">{\"@context\":\"https://schema.org\",\"@type\":\"WebSite\",\"name\":\"Medical Writing AI Playbook\",\"publisher\":{\"@type\":\"Organization\",\"name\":\"Medical Writing AI Playbook\"},\"provider\":{\"@type\":\"Organization\",\"name\":\"Mintlify\",\"url\":\"https://mintlify.com\"}}\u003c/script>","agentPointerHref":"/workflows/verify-claims-against-references.md","siteName":"Medical Writing AI Playbook","nav":{"global":null,"tabs":[{"tab":"Playbook","groups":[{"group":"Overview","pages":[{"title":"Home","sidebarTitle":"Home","href":""},{"title":"Start Here","icon":"arrow-right","href":"/start"},{"title":"Glossary","sidebarTitle":"Glossary","icon":"book-open","href":"/glossary"},{"title":"Changelog","sidebarTitle":"Changelog","icon":"list-check","href":"/changelog"},{"title":"About","sidebarTitle":"About","icon":"user-pen","href":"/about"}]},{"group":"Principles","pages":[{"title":"Human-in-the-Loop Decision Making","icon":"user-check","href":"/principles/human-in-the-loop"},{"title":"Source Grounding","icon":"book-open","href":"/principles/source-grounding"},{"title":"The Description–Discernment Loop","sidebarTitle":"Description \u0026 Discernment","icon":"arrows-rotate","href":"/principles/description-discernment"},{"title":"Understanding AI Risk in Medical Writing","sidebarTitle":"Understanding AI Risk","icon":"gauge","href":"/principles/risk-levels"},{"title":"AI Risk Framework for Medical Writing","icon":"shield-check","href":"/principles/ai-risk-framework"},{"title":"AI Failure Modes in Medical Writing","sidebarTitle":"AI Failure Modes","icon":"triangle-exclamation","href":"/principles/ai-failure-modes"},{"title":"Review and Accountability","icon":"clipboard-check","href":"/principles/review-and-accountability"},{"title":"Declaring AI Use","icon":"file-pen","href":"/principles/declaring-ai-use"},{"title":"AI Regulation in Pharma","icon":"scale-balanced","href":"/principles/ai-regulation"},{"title":"Choosing Your Model","icon":"compass","href":"/principles/choosing-your-model"},{"title":"Agentic Workflows","icon":"circle-nodes","href":"/principles/agentic-workflows"},{"title":"AI in Peer Review","icon":"magnifying-glass","href":"/principles/ai-in-peer-review"}]},{"group":"AI Workflow","pages":[{"group":"Evidence Workflows","pages":[{"title":"Find Evidence","sidebarTitle":"Find Evidence","icon":"microscope","href":"/workflows/find-evidence"},{"title":"Summarise a Source Paper","sidebarTitle":"Summarise a Paper","icon":"file-text","href":"/workflows/summarise-source-paper"},{"title":"Prepare a Congress or Poster Summary","sidebarTitle":"Congress Summary","icon":"presentation","href":"/workflows/prepare-congress-or-poster-summary"},{"title":"Extract Study Data","sidebarTitle":"Extract Study Data","icon":"file-text","href":"/workflows/extract-study-data"},{"title":"Extract Key Messages from Evidence","sidebarTitle":"Extract Key Messages","icon":"key","href":"/workflows/extract-key-messages"}]},{"group":"Drafting","pages":[{"title":"Build a First Content Outline","sidebarTitle":"Build an Outline","icon":"list-check","href":"/workflows/build-content-outline"},{"title":"Write a Manuscript","sidebarTitle":"Write a Manuscript","icon":"pen-line","href":"/workflows/write-a-manuscript"},{"title":"Draft a Regulatory Document","sidebarTitle":"Regulatory Document","icon":"file-text","href":"/workflows/draft-regulatory-document"},{"title":"Convert Statistical Outputs to Narrative","sidebarTitle":"Convert Stats to Narrative","icon":"file-text","href":"/workflows/convert-stats-to-narrative"},{"title":"Create a Medical Slide Deck","sidebarTitle":"Create a Slide Deck","icon":"presentation","href":"/workflows/create-medical-slide-deck"},{"title":"Generate Concept Visuals with AI","sidebarTitle":"Generate Concept Visuals","icon":"image","href":"/workflows/generate-concept-visuals"}]},{"group":"Adaptation","pages":[{"title":"Adapt Content for Different Audiences","sidebarTitle":"Adapt for Audiences","icon":"users","href":"/workflows/adapt-for-different-audiences"},{"title":"Create a Plain Language Summary","sidebarTitle":"Plain Language Summary","description":"Draft accessible patient-facing summaries from clinical study reports and published research.","icon":"heart-pulse","href":"/workflows/create-plain-language-summary"}]},{"group":"Validation","pages":[{"title":"Verify Claims Against References","sidebarTitle":"Verify Claims","icon":"search","href":"/workflows/verify-claims-against-references"},{"title":"Check Promotional Compliance","sidebarTitle":"Compliance Check","description":"Pre-screen content for promotional compliance signals before formal MLR review.","icon":"shield","href":"/workflows/check-promotional-compliance"},{"title":"Check Document Consistency","sidebarTitle":"Check Document Consistency","icon":"search","href":"/workflows/check-document-consistency"}]},{"group":"Delivery","pages":[{"title":"Repurpose Content Across Channels","sidebarTitle":"Repurpose Content","icon":"repeat","href":"/workflows/repurpose-content-across-channels"},{"title":"Final Human Review Before Use","sidebarTitle":"Final Review","icon":"clipboard-check","href":"/workflows/final-human-review"}]}]},{"group":"Prompt Patterns","pages":[{"title":"Source Analysis Prompts","icon":"file-search","href":"/prompts/source-analysis-prompts"},{"title":"Outlining Prompts","icon":"list-tree","href":"/prompts/outlining-prompts"},{"title":"Adaptation Prompts","sidebarTitle":"Adaptation Prompts","icon":"file-text","href":"/prompts/adaptation-prompts"},{"title":"Review Prompts","icon":"check-circle","href":"/prompts/review-prompts"}]},{"group":"Tools","pages":[{"title":"PubCrawl","icon":"microscope","href":"/tools/pubcrawl"},{"title":"RefCheckr","icon":"search","href":"/tools/refcheckr"},{"title":"MedCheckr","icon":"shield","href":"/tools/medcheckr"},{"title":"Patiently AI","icon":"heart","href":"/tools/patiently-ai"},{"title":"PosterLens","icon":"image","href":"/tools/posterlens"},{"title":"Tool ecosystem","sidebarTitle":"AI ecosystem","icon":"compass","href":"/tools/ecosystem"},{"title":"Which tool when","sidebarTitle":"What I use","icon":"signs-post","href":"/tools/decision-tree"}]},{"group":"Templates","pages":[{"title":"Disclosure Language Template","icon":"rectangle-list","href":"/templates/disclosure-language"},{"title":"Freelancer AI-Use Statement","icon":"id-card","href":"/templates/freelancer-ai-use-statement"},{"title":"AI Audit-Trail Log Template","icon":"clipboard-list","href":"/templates/ai-audit-trail-log"},{"title":"MLR-with-AI Review Checklist","icon":"list-check","href":"/templates/mlr-ai-review-checklist"},{"title":"Pre-submission QC Checklist","icon":"check-double","href":"/templates/pre-submission-qc-checklist"}]}]}]},"basePath":"","title":"Verify Claims Against References - Medical Writing AI Playbook","deploymentId":"dpl_FZMKUpaaKiVETbfqhoUe4sPFDTmY","buildId":"dpl_FZMKUpaaKiVETbfqhoUe4sPFDTmY","dynamics":[],"chunks":[]}