[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M7FD8D05X7C3WH7BDZNJ2NQP":3,"skill-same:01M7FD8D05X7C3WH7BDZNJ2NQP":68},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":15,"licenseFileUrl":6,"sourceHtmlUrl":16,"categories":17,"keywords":18,"publishedAt":19,"linkOnly":20,"markdownContent":21,"markdownBody":22,"frontMatter":23,"refName":42,"commitSha":43,"skillDirectory":44,"forksCount":14,"repositoryHtmlUrl":45,"repositoryHomepageUrl":45,"ownerHtmlUrl":46,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":47,"install":48,"resources":61},"01M7FD8D05X7C3WH7BDZNJ2NQP","abstracting-clinical-documentation",null,"Extracts codeable diagnoses and procedures from clinical notes with specificity capture. Use when abstracting medical records, identifying codeable conditions, or capturing documentation specificity.","Mr-Nobody-Anonymous\u002FAll-skills","Mr-Nobody-Anonymous","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F184253119?v=4","01M7AEXFJHKTYQW5R0ZCQQTWFV","anonymous","01M7C00M3GQR4QKP9GZA42GFAK",0,"Apache-2.0","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhealthcare\u002Fabstracting-clinical-documentation\u002FSKILL.md",[],[],"2026-11-16 10:05:34",false,"---\nname: abstracting-clinical-documentation\nlanguage: en\ndescription: Extracts codeable diagnoses and procedures from clinical notes with specificity capture. Use when abstracting medical records, identifying codeable conditions, or capturing documentation specificity.\ntags:\n  - medical-coding-and-billing\n  - clinical\nmetadata:\n  author: casemark\n  practice_areas:\n    - Medical Coding\n    - Revenue Cycle\n    - Health Information Management\n  document_types:\n    - Report\n  skill_modes:\n    - Analysis\nsource: \"https:\u002F\u002Fgithub.com\u002FCaseMark\u002Fskills\"\nsource_repository: \"CaseMark\u002Fskills\"\nsource_path: \"skills\u002Fmed\u002Fabstracting-clinical-documentation\u002FSKILL.md\"\nlicense: \"Apache-2.0\"\nimported_at: \"2026-09-20\"\n---\n\n# Abstracting Clinical Documentation\n\nTransforms bulk medical records into structured, chronologically ordered clinical\nabstracts. The output supports downstream use in medical-legal review, insurance\naudit response, CMS quality reporting, peer review, and clinical research.\n\n---\n\n## Why This Skill Exists\n\nClinical documentation abstraction is the bridge between raw medical records and\nevery downstream process that depends on them:\n\n| Consumer | Why They Need a Structured Abstract |\n|---|---|\n| **Billing \u002F Revenue Cycle** | Accurate code assignment (ICD-10-CM\u002FPCS, CPT, HCPCS) requires every codeable condition and procedure surfaced with laterality, severity, and linkage. |\n| **Quality Reporting (CMS\u002FHEDIS)** | Measure numerators and denominators depend on discrete data elements extracted from narrative notes. |\n| **Litigation Support** | Med-legal reviewers need a defensible chronological timeline with source-page citations to assess causation, standard of care, and damages. |\n| **Insurance Audit** | RAC, ZPIC, and MAC auditors compare billed codes against documentation; the abstract is the first line of defense. |\n| **Peer Review \u002F Credentialing** | Case abstracts drive mortality and morbidity conferences, sentinel event reviews, and provider profiling. |\n| **Clinical Research** | Chart abstraction is the primary data-collection method for retrospective studies and registry submissions. |\n\nA poor abstract produces incorrect codes, failed audits, unsupported legal\nopinions, and missed quality measures. Precision here compounds everywhere.\n\n---\n\n## Checkpoint A: Pre-Draft Intake (Mandatory)\n\nBefore touching a single record, confirm the following with the requesting party.\nDo not proceed until every row is resolved.\n\n| Item | Question to Resolve | Why It Matters |\n|---|---|---|\n| **Purpose** | Litigation, audit defense, quality reporting, coding review, research, or peer review? | Determines depth, element focus, and whether de-identification is required. |\n| **Record Scope** | Which facilities, providers, and date ranges are in scope? | Prevents abstracting irrelevant episodes and controls volume. |\n| **Timeframe** | Exact start and end dates, or \"all available\"? | Anchors the chronological spine of the abstract. |\n| **Target Audience** | Coders, attorneys, physicians, quality analysts? | Drives vocabulary level and output format. |\n| **PHI Handling** | Full PHI retained, Limited Data Set, or Safe Harbor de-identified? | Determines whether the 18 HIPAA Safe Harbor identifiers must be stripped. |\n| **Source Inventory** | Has a document inventory or index been provided? | If not, the first deliverable is the source document manifest (see Step 1). |\n| **Known Gaps** | Are any records known to be missing or outstanding? | Gaps must be flagged, not silently skipped. |\n\n---\n\n## Workflow\n\n### Step 1: Source Document Manifest\n\nCatalog every record before abstracting. Use the following classification:\n\n| Document Type | Typical Content | Abstraction Priority |\n|---|---|---|\n| History & Physical (H&P) | Chief complaint, HPI, ROS, exam, assessment\u002Fplan | **High** -- anchors encounter context |\n| Progress Notes | Daily\u002Finterval updates, reassessments | **High** -- tracks clinical trajectory |\n| Operative Reports | Pre\u002Fpost-op dx, procedure narrative, findings, specimens | **High** -- procedure detail source |\n| Discharge Summary | Hospital course, discharge dx, follow-up plan | **High** -- episode closure |\n| Consultation Notes | Specialist assessment and recommendations | **Medium-High** |\n| Lab \u002F Pathology Reports | Results, reference ranges, interpretations | **Medium** -- supports dx specificity |\n| Imaging Reports | Findings, impressions, recommendations | **Medium** |\n| Medication Administration Records (MAR) | Drug, dose, route, frequency, duration | **Medium** |\n| Nursing Notes \u002F Flowsheets | Vitals, I&O, wound care, fall risk | **Low-Medium** -- ancillary but relevant for quality\u002Flitigation |\n| Therapy Notes (PT\u002FOT\u002FST) | Functional status, goals, progress | **Low-Medium** |\n| Referral \u002F Authorization Forms | Requested services, clinical justification | **Low** -- administrative context |\n| Unsigned \u002F Addended Notes | Late entries, amendments, corrections | **Critical to flag** -- often missed |\n\n**Deliverable:** A completed Source Document Manifest table listing every\ndocument with: document type, date, author\u002Fprovider, page range or file\nreference, and a completeness flag (Complete \u002F Partial \u002F Missing).\n\n### Step 2: Chronological Data Extraction\n\nWalk the records in strict date order. For each encounter or clinical event,\nextract the following data elements:\n\n| Data Element | Capture Requirements |\n|---|---|\n| **Encounter Date** | Exact date (and time if relevant, e.g., ED visits). |\n| **Setting** | Inpatient, outpatient, ED, ambulatory surgery, telehealth, SNF, home health. |\n| **Provider(s)** | Name, credentials, specialty. Note attending vs. consulting vs. referring. |\n| **Chief Complaint \u002F Reason** | Patient-stated reason or referral indication. |\n| **Diagnoses** | Each diagnosis with: ICD-10-CM-level specificity, laterality, acuity (acute\u002Fchronic), onset context (new vs. established), and clinical status (active, resolved, rule-out). |\n| **Procedures** | Each procedure with: CPT\u002FHCPCS-level description, approach, laterality, qualifier, and outcome. |\n| **Medications** | Name, dose, route, frequency, start\u002Fstop dates, indication if documented. |\n| **Lab Results** | Test name, result value, units, reference range, abnormal flag. |\n| **Imaging Findings** | Modality, body region, key findings, impression. |\n| **Referrals \u002F Orders** | To whom, for what, urgency. |\n| **Hospitalizations** | Admit\u002Fdischarge dates, principal and secondary dx, procedures, disposition. |\n| **Functional Status** | ADLs, mobility, cognitive status (especially for quality\u002Flitigation). |\n\nUse the chronological abstract table format in\n[`references\u002FABSTRACTION-TEMPLATE.md`](references\u002FABSTRACTION-TEMPLATE.md).\n\n### Step 3: Terminology Normalization\n\n- Map all diagnoses to ICD-10-CM descriptors. Do NOT assign codes; capture the\n  natural-language equivalent at code-level specificity (e.g., \"acute\n  ST-elevation myocardial infarction of the left anterior descending artery\"\n  not just \"heart attack\").\n- Map procedures to CPT\u002FHCPCS descriptors where identifiable.\n- Standardize medication names to generic (brand in parentheses).\n- Normalize lab units where source uses non-standard reporting.\n\n### Step 4: Distinguish Fact from Inference\n\nThis is the most common source of abstraction error. Apply these rules without\nexception:\n\n| Category | Rule | Example |\n|---|---|---|\n| **Documented Fact** | Appears in the clinical narrative authored by the treating provider. | \"Assessment: Type 2 diabetes mellitus with diabetic nephropathy, stage 3 CKD.\" |\n| **Inference \u002F Implication** | Suggested by data but not explicitly stated by a provider. | Lab shows HbA1c 9.2% but no diabetes diagnosis documented. |\n| **Assessment vs. Diagnosis** | An assessment is a clinical impression; a confirmed diagnosis requires explicit provider documentation or supporting evidence. Do not conflate. | \"R\u002FO PE\" is an assessment, not a diagnosis of pulmonary embolism. |\n| **Primary Source** | The treating provider's own documentation. | Attending physician's progress note. |\n| **Secondary Reference** | Another provider quoting or summarizing. | Consultant note stating \"per Dr. Smith, patient has CHF.\" |\n\nMark inferences with `[INFERRED -- not explicitly documented]`. Mark secondary\nreferences with `[PER: source provider\u002Fdocument]`.\n\n### Step 5: CDI Query Generation\n\nWhen documentation is ambiguous or incomplete, generate Clinical Documentation\nImprovement (CDI) queries rather than guessing. Format:\n\n```\nCDI QUERY #[n]\nRecord:    [Document type, date, provider]\nIssue:     [What is ambiguous or missing]\nClinical Indicators: [Supporting data that suggests the undocumented condition]\nSuggested Clarification: [Specific question for the provider]\nImpact:    [Coding, quality measure, or legal relevance]\n```\n\nCommon CDI triggers:\n- Lab\u002Fimaging findings inconsistent with documented diagnoses\n- Medications prescribed without a corresponding documented indication\n- Severity or acuity not specified (e.g., \"heart failure\" without systolic\u002Fdiastolic, EF, or NYHA class)\n- Cause-and-effect linkage absent (e.g., \"CKD\" and \"diabetes\" documented but no causal relationship stated)\n- Unsigned or contradicted documentation\n\n### Step 6: PHI Handling\n\nIf de-identification is required, apply HIPAA Safe Harbor method. The 18\nidentifiers to strip or generalize:\n\n1. Names\n2. Geographic data smaller than state\n3. Dates (except year) for dates related to an individual\n4. Phone numbers\n5. Fax numbers\n6. Email addresses\n7. SSN\n8. Medical record numbers\n9. Health plan beneficiary numbers\n10. Account numbers\n11. Certificate\u002Flicense numbers\n12. Vehicle identifiers and serial numbers\n13. Device identifiers and serial numbers\n14. Web URLs\n15. IP addresses\n16. Biometric identifiers\n17. Full-face photographs\n18. Any other unique identifying number or code\n\nFor Limited Data Sets, retain dates and geographic data at zip-code level.\n\n---\n\n## Output Structure\n\nThe final deliverable includes these sections in order:\n\n1. **Abstract Header** -- Patient identifier (or de-identified placeholder), DOB\u002Fage, sex, date range, purpose, abstractor, date of abstraction.\n2. **Source Document Manifest** -- Table from Step 1.\n3. **Chronological Case Abstract** -- Table from Step 2 (see template).\n4. **Problem List Summary** -- Consolidated active\u002Fresolved problem list with onset dates and current status.\n5. **Medication Summary** -- Current medications with start dates and indications.\n6. **CDI Queries** -- Any generated queries from Step 5 (or \"None identified\").\n7. **Abstraction Notes** -- Gaps, limitations, assumptions, and flags.\n8. **Attestation** -- Statement that the abstract reflects only documented facts, with inferences and gaps explicitly marked.\n\n---\n\n## Checkpoint B: Post-Draft Alignment (Mandatory)\n\nBefore delivering the final abstract, verify each item:\n\n| Check | Pass Criteria |\n|---|---|\n| **Chronological integrity** | Every entry is in strict date order. No undated entries unless flagged. |\n| **Source traceability** | Every data element cites its source document (type, date, provider, page\u002Fsection). |\n| **Fact vs. inference marking** | All inferred data elements carry the `[INFERRED]` tag. All secondary references carry `[PER:]`. |\n| **Completeness against manifest** | Every document in the manifest has been abstracted or explicitly marked as \"Not abstracted -- [reason]\". |\n| **Terminology consistency** | Diagnoses use ICD-10-CM-level language throughout. Medications use generic names. |\n| **PHI compliance** | If de-identification was required, all 18 identifiers have been scrubbed. Spot-check 5 random entries. |\n| **CDI queries generated** | Ambiguous documentation has been flagged, not silently resolved. |\n| **No silent gaps** | Missing records, illegible sections, and unsigned notes are explicitly called out. |\n| **Purpose alignment** | Output depth and element focus match the stated purpose from Checkpoint A. |\n\n---\n\n## Quality Audit\n\n### Common Abstraction Errors (Ranked by Impact)\n\n| # | Error | Impact | Prevention |\n|---|---|---|---|\n| 1 | Conflating assessment with confirmed diagnosis | Overcoding, audit liability, legal exposure | Apply fact-vs-inference rules strictly |\n| 2 | Missing unsigned or late-addended notes | Incomplete clinical picture, missed dx\u002Fprocedures | Check for addenda on every note; include document status in manifest |\n| 3 | Skipping ancillary records (nursing, therapy, MAR) | Missing functional status, medication errors, quality measure gaps | Manifest must include all document types, not just physician notes |\n| 4 | Losing chronological order | Causal relationships become unclear; legal timelines break | Sort by encounter date before writing a single line |\n| 5 | Copying secondary references as primary facts | Attribution errors; circular documentation | Always trace to the originating provider's note |\n| 6 | Omitting laterality, severity, or acuity | Codes default to \"unspecified,\" triggering audits | Extract every specificity qualifier the documentation supports |\n| 7 | Inconsistent terminology | Same condition described differently across entries | Normalize at Step 3; use a single descriptor per condition throughout |\n| 8 | PHI leakage in de-identified output | HIPAA violation | Run Safe Harbor checklist against final output |\n\n### Audit Sampling\n\nFor abstracts exceeding 50 entries, perform a minimum 10% random sample audit\nagainst source documents. For litigation-purpose abstracts, audit 100% of\nentries related to the contested clinical issues.\n\n---\n\n## Key Rules\n\n- Never present inferred clinical data as documented fact.\n- Every data element must trace to a specific source document, date, and provider.\n- Chronological order is non-negotiable. If a date is uncertain, flag it; do not omit the entry.\n- When documentation is ambiguous, generate a CDI query -- do not resolve ambiguity silently.\n- Mark with `[VERIFY]` any data point that cannot be confirmed from available records.\n- Include appropriate disclaimers: the abstract is a summary tool, not a substitute for the complete medical record.\n- Escalate to human review when clinical complexity exceeds abstraction confidence.\n\n\n---\n\n## Guidelines\n\n- Clinical documentation abstraction is a high-stakes activity. Errors propagate to coding, billing, quality reporting, litigation support, and clinical research. Treat abstraction with the same rigor as clinical documentation itself\n- Always prioritize the treating provider's own documentation over secondary references. A consultant quoting another provider is not equivalent to the original provider's own note for abstraction purposes\n- Chronological integrity is the backbone of a defensible abstract. Every data point must be anchored to a specific date, source document, and provider. Undated entries or entries without source attribution undermine the entire abstract\n- CDI queries are a feature, not a flaw. Generating queries when documentation is ambiguous demonstrates rigor and prevents the silent propagation of assumptions through the coding and billing pipeline\n- For litigation-purpose abstracts, apply a higher standard: 100% audit of entries related to contested clinical issues, explicit notation of documentation gaps, and clear distinction between documented facts and clinical inferences\n- PHI handling is not optional for any abstraction activity. Verify de-identification requirements before beginning work and validate compliance before delivering the final product\n- When abstracting records from multiple facilities or providers, maintain consistent terminology normalization across all sources. The same condition described differently by different providers must be reconciled to a single standardized descriptor throughout the abstract\n- Escalate to a clinical subject matter expert when the abstraction requires interpretation of complex clinical scenarios (e.g., distinguishing between related diagnoses, interpreting ambiguous operative findings, or determining whether a documented condition meets coding criteria)","# Abstracting Clinical Documentation\n\nTransforms bulk medical records into structured, chronologically ordered clinical\nabstracts. The output supports downstream use in medical-legal review, insurance\naudit response, CMS quality reporting, peer review, and clinical research.\n\n---\n\n## Why This Skill Exists\n\nClinical documentation abstraction is the bridge between raw medical records and\nevery downstream process that depends on them:\n\n| Consumer | Why They Need a Structured Abstract |\n|---|---|\n| **Billing \u002F Revenue Cycle** | Accurate code assignment (ICD-10-CM\u002FPCS, CPT, HCPCS) requires every codeable condition and procedure surfaced with laterality, severity, and linkage. |\n| **Quality Reporting (CMS\u002FHEDIS)** | Measure numerators and denominators depend on discrete data elements extracted from narrative notes. |\n| **Litigation Support** | Med-legal reviewers need a defensible chronological timeline with source-page citations to assess causation, standard of care, and damages. |\n| **Insurance Audit** | RAC, ZPIC, and MAC auditors compare billed codes against documentation; the abstract is the first line of defense. |\n| **Peer Review \u002F Credentialing** | Case abstracts drive mortality and morbidity conferences, sentinel event reviews, and provider profiling. |\n| **Clinical Research** | Chart abstraction is the primary data-collection method for retrospective studies and registry submissions. |\n\nA poor abstract produces incorrect codes, failed audits, unsupported legal\nopinions, and missed quality measures. Precision here compounds everywhere.\n\n---\n\n## Checkpoint A: Pre-Draft Intake (Mandatory)\n\nBefore touching a single record, confirm the following with the requesting party.\nDo not proceed until every row is resolved.\n\n| Item | Question to Resolve | Why It Matters |\n|---|---|---|\n| **Purpose** | Litigation, audit defense, quality reporting, coding review, research, or peer review? | Determines depth, element focus, and whether de-identification is required. |\n| **Record Scope** | Which facilities, providers, and date ranges are in scope? | Prevents abstracting irrelevant episodes and controls volume. |\n| **Timeframe** | Exact start and end dates, or \"all available\"? | Anchors the chronological spine of the abstract. |\n| **Target Audience** | Coders, attorneys, physicians, quality analysts? | Drives vocabulary level and output format. |\n| **PHI Handling** | Full PHI retained, Limited Data Set, or Safe Harbor de-identified? | Determines whether the 18 HIPAA Safe Harbor identifiers must be stripped. |\n| **Source Inventory** | Has a document inventory or index been provided? | If not, the first deliverable is the source document manifest (see Step 1). |\n| **Known Gaps** | Are any records known to be missing or outstanding? | Gaps must be flagged, not silently skipped. |\n\n---\n\n## Workflow\n\n### Step 1: Source Document Manifest\n\nCatalog every record before abstracting. Use the following classification:\n\n| Document Type | Typical Content | Abstraction Priority |\n|---|---|---|\n| History & Physical (H&P) | Chief complaint, HPI, ROS, exam, assessment\u002Fplan | **High** -- anchors encounter context |\n| Progress Notes | Daily\u002Finterval updates, reassessments | **High** -- tracks clinical trajectory |\n| Operative Reports | Pre\u002Fpost-op dx, procedure narrative, findings, specimens | **High** -- procedure detail source |\n| Discharge Summary | Hospital course, discharge dx, follow-up plan | **High** -- episode closure |\n| Consultation Notes | Specialist assessment and recommendations | **Medium-High** |\n| Lab \u002F Pathology Reports | Results, reference ranges, interpretations | **Medium** -- supports dx specificity |\n| Imaging Reports | Findings, impressions, recommendations | **Medium** |\n| Medication Administration Records (MAR) | Drug, dose, route, frequency, duration | **Medium** |\n| Nursing Notes \u002F Flowsheets | Vitals, I&O, wound care, fall risk | **Low-Medium** -- ancillary but relevant for quality\u002Flitigation |\n| Therapy Notes (PT\u002FOT\u002FST) | Functional status, goals, progress | **Low-Medium** |\n| Referral \u002F Authorization Forms | Requested services, clinical justification | **Low** -- administrative context |\n| Unsigned \u002F Addended Notes | Late entries, amendments, corrections | **Critical to flag** -- often missed |\n\n**Deliverable:** A completed Source Document Manifest table listing every\ndocument with: document type, date, author\u002Fprovider, page range or file\nreference, and a completeness flag (Complete \u002F Partial \u002F Missing).\n\n### Step 2: Chronological Data Extraction\n\nWalk the records in strict date order. For each encounter or clinical event,\nextract the following data elements:\n\n| Data Element | Capture Requirements |\n|---|---|\n| **Encounter Date** | Exact date (and time if relevant, e.g., ED visits). |\n| **Setting** | Inpatient, outpatient, ED, ambulatory surgery, telehealth, SNF, home health. |\n| **Provider(s)** | Name, credentials, specialty. Note attending vs. consulting vs. referring. |\n| **Chief Complaint \u002F Reason** | Patient-stated reason or referral indication. |\n| **Diagnoses** | Each diagnosis with: ICD-10-CM-level specificity, laterality, acuity (acute\u002Fchronic), onset context (new vs. established), and clinical status (active, resolved, rule-out). |\n| **Procedures** | Each procedure with: CPT\u002FHCPCS-level description, approach, laterality, qualifier, and outcome. |\n| **Medications** | Name, dose, route, frequency, start\u002Fstop dates, indication if documented. |\n| **Lab Results** | Test name, result value, units, reference range, abnormal flag. |\n| **Imaging Findings** | Modality, body region, key findings, impression. |\n| **Referrals \u002F Orders** | To whom, for what, urgency. |\n| **Hospitalizations** | Admit\u002Fdischarge dates, principal and secondary dx, procedures, disposition. |\n| **Functional Status** | ADLs, mobility, cognitive status (especially for quality\u002Flitigation). |\n\nUse the chronological abstract table format in\n[`references\u002FABSTRACTION-TEMPLATE.md`](references\u002FABSTRACTION-TEMPLATE.md).\n\n### Step 3: Terminology Normalization\n\n- Map all diagnoses to ICD-10-CM descriptors. Do NOT assign codes; capture the\n  natural-language equivalent at code-level specificity (e.g., \"acute\n  ST-elevation myocardial infarction of the left anterior descending artery\"\n  not just \"heart attack\").\n- Map procedures to CPT\u002FHCPCS descriptors where identifiable.\n- Standardize medication names to generic (brand in parentheses).\n- Normalize lab units where source uses non-standard reporting.\n\n### Step 4: Distinguish Fact from Inference\n\nThis is the most common source of abstraction error. Apply these rules without\nexception:\n\n| Category | Rule | Example |\n|---|---|---|\n| **Documented Fact** | Appears in the clinical narrative authored by the treating provider. | \"Assessment: Type 2 diabetes mellitus with diabetic nephropathy, stage 3 CKD.\" |\n| **Inference \u002F Implication** | Suggested by data but not explicitly stated by a provider. | Lab shows HbA1c 9.2% but no diabetes diagnosis documented. |\n| **Assessment vs. Diagnosis** | An assessment is a clinical impression; a confirmed diagnosis requires explicit provider documentation or supporting evidence. Do not conflate. | \"R\u002FO PE\" is an assessment, not a diagnosis of pulmonary embolism. |\n| **Primary Source** | The treating provider's own documentation. | Attending physician's progress note. |\n| **Secondary Reference** | Another provider quoting or summarizing. | Consultant note stating \"per Dr. Smith, patient has CHF.\" |\n\nMark inferences with `[INFERRED -- not explicitly documented]`. Mark secondary\nreferences with `[PER: source provider\u002Fdocument]`.\n\n### Step 5: CDI Query Generation\n\nWhen documentation is ambiguous or incomplete, generate Clinical Documentation\nImprovement (CDI) queries rather than guessing. Format:\n\n```\nCDI QUERY #[n]\nRecord:    [Document type, date, provider]\nIssue:     [What is ambiguous or missing]\nClinical Indicators: [Supporting data that suggests the undocumented condition]\nSuggested Clarification: [Specific question for the provider]\nImpact:    [Coding, quality measure, or legal relevance]\n```\n\nCommon CDI triggers:\n- Lab\u002Fimaging findings inconsistent with documented diagnoses\n- Medications prescribed without a corresponding documented indication\n- Severity or acuity not specified (e.g., \"heart failure\" without systolic\u002Fdiastolic, EF, or NYHA class)\n- Cause-and-effect linkage absent (e.g., \"CKD\" and \"diabetes\" documented but no causal relationship stated)\n- Unsigned or contradicted documentation\n\n### Step 6: PHI Handling\n\nIf de-identification is required, apply HIPAA Safe Harbor method. The 18\nidentifiers to strip or generalize:\n\n1. Names\n2. Geographic data smaller than state\n3. Dates (except year) for dates related to an individual\n4. Phone numbers\n5. Fax numbers\n6. Email addresses\n7. SSN\n8. Medical record numbers\n9. Health plan beneficiary numbers\n10. Account numbers\n11. Certificate\u002Flicense numbers\n12. Vehicle identifiers and serial numbers\n13. Device identifiers and serial numbers\n14. Web URLs\n15. IP addresses\n16. Biometric identifiers\n17. Full-face photographs\n18. Any other unique identifying number or code\n\nFor Limited Data Sets, retain dates and geographic data at zip-code level.\n\n---\n\n## Output Structure\n\nThe final deliverable includes these sections in order:\n\n1. **Abstract Header** -- Patient identifier (or de-identified placeholder), DOB\u002Fage, sex, date range, purpose, abstractor, date of abstraction.\n2. **Source Document Manifest** -- Table from Step 1.\n3. **Chronological Case Abstract** -- Table from Step 2 (see template).\n4. **Problem List Summary** -- Consolidated active\u002Fresolved problem list with onset dates and current status.\n5. **Medication Summary** -- Current medications with start dates and indications.\n6. **CDI Queries** -- Any generated queries from Step 5 (or \"None identified\").\n7. **Abstraction Notes** -- Gaps, limitations, assumptions, and flags.\n8. **Attestation** -- Statement that the abstract reflects only documented facts, with inferences and gaps explicitly marked.\n\n---\n\n## Checkpoint B: Post-Draft Alignment (Mandatory)\n\nBefore delivering the final abstract, verify each item:\n\n| Check | Pass Criteria |\n|---|---|\n| **Chronological integrity** | Every entry is in strict date order. No undated entries unless flagged. |\n| **Source traceability** | Every data element cites its source document (type, date, provider, page\u002Fsection). |\n| **Fact vs. inference marking** | All inferred data elements carry the `[INFERRED]` tag. All secondary references carry `[PER:]`. |\n| **Completeness against manifest** | Every document in the manifest has been abstracted or explicitly marked as \"Not abstracted -- [reason]\". |\n| **Terminology consistency** | Diagnoses use ICD-10-CM-level language throughout. Medications use generic names. |\n| **PHI compliance** | If de-identification was required, all 18 identifiers have been scrubbed. Spot-check 5 random entries. |\n| **CDI queries generated** | Ambiguous documentation has been flagged, not silently resolved. |\n| **No silent gaps** | Missing records, illegible sections, and unsigned notes are explicitly called out. |\n| **Purpose alignment** | Output depth and element focus match the stated purpose from Checkpoint A. |\n\n---\n\n## Quality Audit\n\n### Common Abstraction Errors (Ranked by Impact)\n\n| # | Error | Impact | Prevention |\n|---|---|---|---|\n| 1 | Conflating assessment with confirmed diagnosis | Overcoding, audit liability, legal exposure | Apply fact-vs-inference rules strictly |\n| 2 | Missing unsigned or late-addended notes | Incomplete clinical picture, missed dx\u002Fprocedures | Check for addenda on every note; include document status in manifest |\n| 3 | Skipping ancillary records (nursing, therapy, MAR) | Missing functional status, medication errors, quality measure gaps | Manifest must include all document types, not just physician notes |\n| 4 | Losing chronological order | Causal relationships become unclear; legal timelines break | Sort by encounter date before writing a single line |\n| 5 | Copying secondary references as primary facts | Attribution errors; circular documentation | Always trace to the originating provider's note |\n| 6 | Omitting laterality, severity, or acuity | Codes default to \"unspecified,\" triggering audits | Extract every specificity qualifier the documentation supports |\n| 7 | Inconsistent terminology | Same condition described differently across entries | Normalize at Step 3; use a single descriptor per condition throughout |\n| 8 | PHI leakage in de-identified output | HIPAA violation | Run Safe Harbor checklist against final output |\n\n### Audit Sampling\n\nFor abstracts exceeding 50 entries, perform a minimum 10% random sample audit\nagainst source documents. For litigation-purpose abstracts, audit 100% of\nentries related to the contested clinical issues.\n\n---\n\n## Key Rules\n\n- Never present inferred clinical data as documented fact.\n- Every data element must trace to a specific source document, date, and provider.\n- Chronological order is non-negotiable. If a date is uncertain, flag it; do not omit the entry.\n- When documentation is ambiguous, generate a CDI query -- do not resolve ambiguity silently.\n- Mark with `[VERIFY]` any data point that cannot be confirmed from available records.\n- Include appropriate disclaimers: the abstract is a summary tool, not a substitute for the complete medical record.\n- Escalate to human review when clinical complexity exceeds abstraction confidence.\n\n\n---\n\n## Guidelines\n\n- Clinical documentation abstraction is a high-stakes activity. Errors propagate to coding, billing, quality reporting, litigation support, and clinical research. Treat abstraction with the same rigor as clinical documentation itself\n- Always prioritize the treating provider's own documentation over secondary references. A consultant quoting another provider is not equivalent to the original provider's own note for abstraction purposes\n- Chronological integrity is the backbone of a defensible abstract. Every data point must be anchored to a specific date, source document, and provider. Undated entries or entries without source attribution undermine the entire abstract\n- CDI queries are a feature, not a flaw. Generating queries when documentation is ambiguous demonstrates rigor and prevents the silent propagation of assumptions through the coding and billing pipeline\n- For litigation-purpose abstracts, apply a higher standard: 100% audit of entries related to contested clinical issues, explicit notation of documentation gaps, and clear distinction between documented facts and clinical inferences\n- PHI handling is not optional for any abstraction activity. Verify de-identification requirements before beginning work and validate compliance before delivering the final product\n- When abstracting records from multiple facilities or providers, maintain consistent terminology normalization across all sources. The same condition described differently by different providers must be reconciled to a single standardized descriptor throughout the abstract\n- Escalate to a clinical subject matter expert when the abstraction requires interpretation of complex clinical scenarios (e.g., distinguishing between related diagnoses, interpreting ambiguous operative findings, or determining whether a documented condition meets coding criteria)",{"name":5,"tags":24,"source":27,"license":15,"language":28,"metadata":29,"description":7,"imported_at":39,"source_path":40,"source_repository":41},[25,26],"medical-coding-and-billing","clinical","https:\u002F\u002Fgithub.com\u002FCaseMark\u002Fskills","en",{"author":30,"skill_modes":31,"document_types":33,"practice_areas":35},"casemark",[32],"Analysis",[34],"Report",[36,37,38],"Medical Coding","Revenue Cycle","Health Information Management","2026-09-20","skills\u002Fmed\u002Fabstracting-clinical-documentation\u002FSKILL.md","CaseMark\u002Fskills","main","2d1750717747097b602acf30bd09a7cbd4b4c6ea","awesome_skills\u002Fhealthcare\u002Fabstracting-clinical-documentation","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous","2026-11-09 11:45:36",{"source":8,"skillName":5,"commands":49,"prompt":59,"zipAvailable":60},[50,53,56],{"tool":51,"command":52},"npx","npx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill abstracting-clinical-documentation",{"tool":54,"command":55},"bunx","bunx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill abstracting-clinical-documentation",{"tool":57,"command":58},"pnpm","pnpm dlx skills add Mr-Nobody-Anonymous\u002FAll-skills --skill abstracting-clinical-documentation","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhealthcare\u002Fabstracting-clinical-documentation\u002FSKILL.md",true,[62],{"path":63,"resourceType":64,"resourceTypeDesc":65,"sizeBytes":66,"sourceHtmlUrl":67},"references\u002FABSTRACTION-TEMPLATE.md","REFERENCE","参考文档",7342,"https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhealthcare\u002Fabstracting-clinical-documentation\u002Freferences\u002FABSTRACTION-TEMPLATE.md",[69,77,84,91,98,105],{"skillNo":70,"name":71,"nameZh":6,"description":72,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":74,"categories":75,"keywords":76,"publishedAt":19},"01M7FFFZPC2025JQQHPHDX154J","dei-assessment","Assess diversity, equity, and inclusion with metrics. TRIGGERS - Use when user needs help with dei-assessment related tasks.","MIT","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fdei-assessment\u002FSKILL.md",[],[],{"skillNo":78,"name":79,"nameZh":6,"description":80,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":81,"categories":82,"keywords":83,"publishedAt":19},"01M7FFFWN8H4WA7A6Y398QKB6J","dei-action-plan","Create DEI action plans with goals and initiatives. TRIGGERS - Use when user needs help with dei-action-plan related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fdei-action-plan\u002FSKILL.md",[],[],{"skillNo":85,"name":86,"nameZh":6,"description":87,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":88,"categories":89,"keywords":90,"publishedAt":19},"01M7FFFSHTM0QZRVZ2NRJTJ0PY","culture-fit-assessment","Design culture add assessment processes. TRIGGERS - Use when user needs help with culture-fit-assessment related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fculture-fit-assessment\u002FSKILL.md",[],[],{"skillNo":92,"name":93,"nameZh":6,"description":94,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":95,"categories":96,"keywords":97,"publishedAt":19},"01M7FFFMA12RKEQA38KHJSXXPS","cowork-to-code-bridge","Use an already-installed, independently verified cowork-to-code bridge to run narrowly approved actions on the user's own macOS, Linux, or WSL2 machine through a local file queue.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcowork-to-code-bridge\u002FSKILL.md",[],[],{"skillNo":99,"name":100,"nameZh":6,"description":101,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":102,"categories":103,"keywords":104,"publishedAt":19},"01M7FFFH1F6M8YRGW5V487GH1K","counter-offer-response","Draft counter-offer responses for retention. TRIGGERS - Use when user needs help with counter-offer-response related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcounter-offer-response\u002FSKILL.md",[],[],{"skillNo":106,"name":107,"nameZh":6,"description":108,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":12,"repositoryNo":13,"starsCount":14,"licenseSpdxId":73,"licenseFileUrl":6,"sourceHtmlUrl":109,"categories":110,"keywords":111,"publishedAt":19},"01M7FFFDGEHV8Y7J9N40HZXW52","continuous-feedback-system","Design continuous feedback systems. TRIGGERS - Use when user needs help with continuous-feedback-system related tasks.","https:\u002F\u002Fgithub.com\u002FMr-Nobody-Anonymous\u002FAll-skills\u002Fblob\u002Fmain\u002Fawesome_skills\u002Fhr-people\u002Fcontinuous-feedback-system\u002FSKILL.md",[],[]]