[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"skill:01M72B5QK12WPY88DPJM2KB0Y4":3,"skill-same:01M72B5QK12WPY88DPJM2KB0Y4":134},{"skillNo":4,"name":5,"nameZh":6,"description":7,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":15,"categories":16,"keywords":17,"publishedAt":18,"linkOnly":19,"markdownContent":20,"markdownBody":21,"frontMatter":22,"refName":28,"commitSha":29,"skillDirectory":30,"forksCount":13,"repositoryHtmlUrl":31,"repositoryHomepageUrl":6,"ownerHtmlUrl":32,"ownerBlogUrl":6,"ownerDescription":6,"updatedAt":33,"install":34,"resources":47},"01M72B5QK12WPY88DPJM2KB0Y4","clinical-decision-support",null,"Generate professional clinical decision support (CDS) documents for pharmaceutical and clinical research settings, including patient cohort analyses (biomarker-stratified with outcomes) and treatment recommendation reports (evidence-based guidelines with decision algorithms). Supports GRADE evidence grading, statistical analysis (hazard ratios, survival curves, waterfall plots), biomarker integration, and regulatory compliance. Outputs publication-ready LaTeX\u002FPDF format optimized for drug development, clinical research, and evidence synthesis.","Sansar35\u002Fclaude-agent","Sansar35","https:\u002F\u002Favatars.githubusercontent.com\u002Fu\u002F145020281?v=4","01M724CQ2SCC86MGNSD87D0D4K","01M724CQ3965G9WJ3QXE3AJ5YY",1,"MIT","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002FSKILL.md",[],[],"2026-11-10 18:39:20",false,"---\nname: clinical-decision-support\ndescription: \"Generate professional clinical decision support (CDS) documents for pharmaceutical and clinical research settings, including patient cohort analyses (biomarker-stratified with outcomes) and treatment recommendation reports (evidence-based guidelines with decision algorithms). Supports GRADE evidence grading, statistical analysis (hazard ratios, survival curves, waterfall plots), biomarker integration, and regulatory compliance. Outputs publication-ready LaTeX\u002FPDF format optimized for drug development, clinical research, and evidence synthesis.\"\nallowed-tools: [Read, Write, Edit, Bash]\n---\n\n# Clinical Decision Support Documents\n\n## Description\n\nGenerate professional clinical decision support (CDS) documents for pharmaceutical companies, clinical researchers, and medical decision-makers. This skill specializes in analytical, evidence-based documents that inform treatment strategies and drug development:\n\n1. **Patient Cohort Analysis** - Biomarker-stratified group analyses with statistical outcome comparisons\n2. **Treatment Recommendation Reports** - Evidence-based clinical guidelines with GRADE grading and decision algorithms\n\nAll documents are generated as publication-ready LaTeX\u002FPDF files optimized for pharmaceutical research, regulatory submissions, and clinical guideline development.\n\n**Note:** For individual patient treatment plans at the bedside, use the `treatment-plans` skill instead. This skill focuses on group-level analyses and evidence synthesis for pharmaceutical\u002Fresearch settings.\n\n## Capabilities\n\n### Document Types\n\n**Patient Cohort Analysis**\n- Biomarker-based patient stratification (molecular subtypes, gene expression, IHC)\n- Molecular subtype classification (e.g., GBM mesenchymal-immune-active vs proneural, breast cancer subtypes)\n- Outcome metrics with statistical analysis (OS, PFS, ORR, DOR, DCR)\n- Statistical comparisons between subgroups (hazard ratios, p-values, 95% CI)\n- Survival analysis with Kaplan-Meier curves and log-rank tests\n- Efficacy tables and waterfall plots\n- Comparative effectiveness analyses\n- Pharmaceutical cohort reporting (trial subgroups, real-world evidence)\n\n**Treatment Recommendation Reports**\n- Evidence-based treatment guidelines for specific disease states\n- Strength of recommendation grading (GRADE system: 1A, 1B, 2A, 2B, 2C)\n- Quality of evidence assessment (high, moderate, low, very low)\n- Treatment algorithm flowcharts with TikZ diagrams\n- Line-of-therapy sequencing based on biomarkers\n- Decision pathways with clinical and molecular criteria\n- Pharmaceutical strategy documents\n- Clinical guideline development for medical societies\n\n### Clinical Features\n\n- **Biomarker Integration**: Genomic alterations (mutations, CNV, fusions), gene expression signatures, IHC markers, PD-L1 scoring\n- **Statistical Analysis**: Hazard ratios, p-values, confidence intervals, survival curves, Cox regression, log-rank tests\n- **Evidence Grading**: GRADE system (1A\u002F1B\u002F2A\u002F2B\u002F2C), Oxford CEBM levels, quality of evidence assessment\n- **Clinical Terminology**: SNOMED-CT, LOINC, proper medical nomenclature, trial nomenclature\n- **Regulatory Compliance**: HIPAA de-identification, confidentiality headers, ICH-GCP alignment\n- **Professional Formatting**: Compact 0.5in margins, color-coded recommendations, publication-ready, suitable for regulatory submissions\n\n## Pharmaceutical and Research Use Cases\n\nThis skill is specifically designed for pharmaceutical and clinical research applications:\n\n**Drug Development**\n- **Phase 2\u002F3 Trial Analyses**: Biomarker-stratified efficacy and safety analyses\n- **Subgroup Analyses**: Forest plots showing treatment effects across patient subgroups\n- **Companion Diagnostic Development**: Linking biomarkers to drug response\n- **Regulatory Submissions**: IND\u002FNDA documentation with evidence summaries\n\n**Medical Affairs**\n- **KOL Education Materials**: Evidence-based treatment algorithms for thought leaders\n- **Medical Strategy Documents**: Competitive landscape and positioning strategies\n- **Advisory Board Materials**: Cohort analyses and treatment recommendation frameworks\n- **Publication Planning**: Manuscript-ready analyses for peer-reviewed journals\n\n**Clinical Guidelines**\n- **Guideline Development**: Evidence synthesis with GRADE methodology for specialty societies\n- **Consensus Recommendations**: Multi-stakeholder treatment algorithm development\n- **Practice Standards**: Biomarker-based treatment selection criteria\n- **Quality Measures**: Evidence-based performance metrics\n\n**Real-World Evidence**\n- **RWE Cohort Studies**: Retrospective analyses of patient cohorts from EMR data\n- **Comparative Effectiveness**: Head-to-head treatment comparisons in real-world settings\n- **Outcomes Research**: Long-term survival and safety in clinical practice\n- **Health Economics**: Cost-effectiveness analyses by biomarker subgroup\n\n## When to Use\n\nUse this skill when you need to:\n\n- **Analyze patient cohorts** stratified by biomarkers, molecular subtypes, or clinical characteristics\n- **Generate treatment recommendation reports** with evidence grading for clinical guidelines or pharmaceutical strategies\n- **Compare outcomes** between patient subgroups with statistical analysis (survival, response rates, hazard ratios)\n- **Produce pharmaceutical research documents** for drug development, clinical trials, or regulatory submissions\n- **Develop clinical practice guidelines** with GRADE evidence grading and decision algorithms\n- **Document biomarker-guided therapy selection** at the population level (not individual patients)\n- **Synthesize evidence** from multiple trials or real-world data sources\n- **Create clinical decision algorithms** with flowcharts for treatment sequencing\n\n**Do NOT use this skill for:**\n- Individual patient treatment plans (use `treatment-plans` skill)\n- Bedside clinical care documentation (use `treatment-plans` skill)\n- Simple patient-specific treatment protocols (use `treatment-plans` skill)\n\n## Visual Enhancement with Scientific Schematics\n\n**⚠️ MANDATORY: Every clinical decision support document MUST include at least 1-2 AI-generated figures using the scientific-schematics skill.**\n\nThis is not optional. Clinical decision documents require clear visual algorithms. Before finalizing any document:\n1. Generate at minimum ONE schematic or diagram (e.g., clinical decision algorithm, treatment pathway, or biomarker stratification tree)\n2. For cohort analyses: include patient flow diagram\n3. For treatment recommendations: include decision flowchart\n\n**How to generate figures:**\n- Use the **scientific-schematics** skill to generate AI-powered publication-quality diagrams\n- Simply describe your desired diagram in natural language\n- Nano Banana Pro will automatically generate, review, and refine the schematic\n\n**How to generate schematics:**\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o figures\u002Foutput.png\n```\n\nThe AI will automatically:\n- Create publication-quality images with proper formatting\n- Review and refine through multiple iterations\n- Ensure accessibility (colorblind-friendly, high contrast)\n- Save outputs in the figures\u002F directory\n\n**When to add schematics:**\n- Clinical decision algorithm flowcharts\n- Treatment pathway diagrams\n- Biomarker stratification trees\n- Patient cohort flow diagrams (CONSORT-style)\n- Survival curve visualizations\n- Molecular mechanism diagrams\n- Any complex concept that benefits from visualization\n\nFor detailed guidance on creating schematics, refer to the scientific-schematics skill documentation.\n\n---\n\n## Document Structure\n\n**CRITICAL REQUIREMENT: All clinical decision support documents MUST begin with a complete executive summary on page 1 that spans the entire first page before any table of contents or detailed sections.**\n\n### Page 1 Executive Summary Structure\n\nThe first page of every CDS document should contain ONLY the executive summary with the following components:\n\n**Required Elements (all on page 1):**\n1. **Document Title and Type**\n   - Main title (e.g., \"Biomarker-Stratified Cohort Analysis\" or \"Evidence-Based Treatment Recommendations\")\n   - Subtitle with disease state and focus\n   \n2. **Report Information Box** (using colored tcolorbox)\n   - Document type and purpose\n   - Date of analysis\u002Freport\n   - Disease state and patient population\n   - Author\u002Finstitution (if applicable)\n   - Analysis framework or methodology\n   \n3. **Key Findings Boxes** (3-5 colored boxes using tcolorbox)\n   - **Primary Results** (blue box): Main efficacy\u002Foutcome findings\n   - **Biomarker Insights** (green box): Key molecular subtype findings\n   - **Clinical Implications** (yellow\u002Forange box): Actionable treatment implications\n   - **Statistical Summary** (gray box): Hazard ratios, p-values, key statistics\n   - **Safety Highlights** (red box, if applicable): Critical adverse events or warnings\n\n**Visual Requirements:**\n- Use `\\thispagestyle{empty}` to remove page numbers from page 1\n- All content must fit on page 1 (before `\\newpage`)\n- Use colored tcolorbox environments with different colors for visual hierarchy\n- Boxes should be scannable and highlight most critical information\n- Use bullet points, not narrative paragraphs\n- End page 1 with `\\newpage` before table of contents or detailed sections\n\n**Example First Page LaTeX Structure:**\n```latex\n\\maketitle\n\\thispagestyle{empty}\n\n% Report Information Box\n\\begin{tcolorbox}[colback=blue!5!white, colframe=blue!75!black, title=Report Information]\n\\textbf{Document Type:} Patient Cohort Analysis\\\\\n\\textbf{Disease State:} HER2-Positive Metastatic Breast Cancer\\\\\n\\textbf{Analysis Date:} \\today\\\\\n\\textbf{Population:} 60 patients, biomarker-stratified by HR status\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #1: Primary Results\n\\begin{tcolorbox}[colback=blue!5!white, colframe=blue!75!black, title=Primary Efficacy Results]\n\\begin{itemize}\n    \\item Overall ORR: 72\\% (95\\% CI: 59-83\\%)\n    \\item Median PFS: 18.5 months (95\\% CI: 14.2-22.8)\n    \\item Median OS: 35.2 months (95\\% CI: 28.1-NR)\n\\end{itemize}\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #2: Biomarker Insights\n\\begin{tcolorbox}[colback=green!5!white, colframe=green!75!black, title=Biomarker Stratification Findings]\n\\begin{itemize}\n    \\item HR+\u002FHER2+: ORR 68\\%, median PFS 16.2 months\n    \\item HR-\u002FHER2+: ORR 78\\%, median PFS 22.1 months\n    \\item HR status significantly associated with outcomes (p=0.041)\n\\end{itemize}\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #3: Clinical Implications\n\\begin{tcolorbox}[colback=orange!5!white, colframe=orange!75!black, title=Clinical Recommendations]\n\\begin{itemize}\n    \\item Strong efficacy observed regardless of HR status (Grade 1A)\n    \\item HR-\u002FHER2+ patients showed numerically superior outcomes\n    \\item Treatment recommended for all HER2+ MBC patients\n\\end{itemize}\n\\end{tcolorbox}\n\n\\newpage\n\\tableofcontents  % TOC on page 2\n\\newpage  % Detailed content starts page 3\n```\n\n### Patient Cohort Analysis (Detailed Sections - Page 3+)\n- **Cohort Characteristics**: Demographics, baseline features, patient selection criteria\n- **Biomarker Stratification**: Molecular subtypes, genomic alterations, IHC profiles\n- **Treatment Exposure**: Therapies received, dosing, treatment duration by subgroup\n- **Outcome Analysis**: Response rates (ORR, DCR), survival data (OS, PFS), DOR\n- **Statistical Methods**: Kaplan-Meier survival curves, hazard ratios, log-rank tests, Cox regression\n- **Subgroup Comparisons**: Biomarker-stratified efficacy, forest plots, statistical significance\n- **Safety Profile**: Adverse events by subgroup, dose modifications, discontinuations\n- **Clinical Recommendations**: Treatment implications based on biomarker profiles\n- **Figures**: Waterfall plots, swimmer plots, survival curves, forest plots\n- **Tables**: Demographics table, biomarker frequency, outcomes by subgroup\n\n### Treatment Recommendation Reports (Detailed Sections - Page 3+)\n\n**Page 1 Executive Summary for Treatment Recommendations should include:**\n1. **Report Information Box**: Disease state, guideline version\u002Fdate, target population\n2. **Key Recommendations Box** (green): Top 3-5 GRADE-graded recommendations by line of therapy\n3. **Biomarker Decision Criteria Box** (blue): Key molecular markers influencing treatment selection\n4. **Evidence Summary Box** (gray): Major trials supporting recommendations (e.g., KEYNOTE-189, FLAURA)\n5. **Critical Monitoring Box** (orange\u002Fred): Essential safety monitoring requirements\n\n**Detailed Sections (Page 3+):**\n- **Clinical Context**: Disease state, epidemiology, current treatment landscape\n- **Target Population**: Patient characteristics, biomarker criteria, staging\n- **Evidence Review**: Systematic literature synthesis, guideline summary, trial data\n- **Treatment Options**: Available therapies with mechanism of action\n- **Evidence Grading**: GRADE assessment for each recommendation (1A, 1B, 2A, 2B, 2C)\n- **Recommendations by Line**: First-line, second-line, subsequent therapies\n- **Biomarker-Guided Selection**: Decision criteria based on molecular profiles\n- **Treatment Algorithms**: TikZ flowcharts showing decision pathways\n- **Monitoring Protocol**: Safety assessments, efficacy monitoring, dose modifications\n- **Special Populations**: Elderly, renal\u002Fhepatic impairment, comorbidities\n- **References**: Full bibliography with trial names and citations\n\n## Output Format\n\n**MANDATORY FIRST PAGE REQUIREMENT:**\n- **Page 1**: Full-page executive summary with 3-5 colored tcolorbox elements\n- **Page 2**: Table of contents (optional)\n- **Page 3+**: Detailed sections with methods, results, figures, tables\n\n**Document Specifications:**\n- **Primary**: LaTeX\u002FPDF with 0.5in margins for compact, data-dense presentation\n- **Length**: Typically 5-15 pages (1 page executive summary + 4-14 pages detailed content)\n- **Style**: Publication-ready, pharmaceutical-grade, suitable for regulatory submissions\n- **First Page**: Always a complete executive summary spanning entire page 1 (see Document Structure section)\n\n**Visual Elements:**\n- **Colors**: \n  - Page 1 boxes: blue=data\u002Finformation, green=biomarkers\u002Frecommendations, yellow\u002Forange=clinical implications, red=warnings\n  - Recommendation boxes (green=strong recommendation, yellow=conditional, blue=research needed)\n  - Biomarker stratification (color-coded molecular subtypes)\n  - Statistical significance (color-coded p-values, hazard ratios)\n- **Tables**: \n  - Demographics with baseline characteristics\n  - Biomarker frequency by subgroup\n  - Outcomes table (ORR, PFS, OS, DOR by molecular subtype)\n  - Adverse events by cohort\n  - Evidence summary tables with GRADE ratings\n- **Figures**: \n  - Kaplan-Meier survival curves with log-rank p-values and number at risk tables\n  - Waterfall plots showing best response by patient\n  - Forest plots for subgroup analyses with confidence intervals\n  - TikZ decision algorithm flowcharts\n  - Swimmer plots for individual patient timelines\n- **Statistics**: Hazard ratios with 95% CI, p-values, median survival times, landmark survival rates\n- **Compliance**: De-identification per HIPAA Safe Harbor, confidentiality notices for proprietary data\n\n## Integration\n\nThis skill integrates with:\n- **scientific-writing**: Citation management, statistical reporting, evidence synthesis\n- **clinical-reports**: Medical terminology, HIPAA compliance, regulatory documentation\n- **scientific-schematics**: TikZ flowcharts for decision algorithms and treatment pathways\n- **treatment-plans**: Individual patient applications of cohort-derived insights (bidirectional)\n\n## Key Differentiators from Treatment-Plans Skill\n\n**Clinical Decision Support (this skill):**\n- **Audience**: Pharmaceutical companies, clinical researchers, guideline committees, medical affairs\n- **Scope**: Population-level analyses, evidence synthesis, guideline development\n- **Focus**: Biomarker stratification, statistical comparisons, evidence grading\n- **Output**: Multi-page analytical documents (5-15 pages typical) with extensive figures and tables\n- **Use Cases**: Drug development, regulatory submissions, clinical practice guidelines, medical strategy\n- **Example**: \"Analyze 60 HER2+ breast cancer patients by hormone receptor status with survival outcomes\"\n\n**Treatment-Plans Skill:**\n- **Audience**: Clinicians, patients, care teams\n- **Scope**: Individual patient care planning\n- **Focus**: SMART goals, patient-specific interventions, monitoring plans\n- **Output**: Concise 1-4 page actionable care plans\n- **Use Cases**: Bedside clinical care, EMR documentation, patient-centered planning\n- **Example**: \"Create treatment plan for a 55-year-old patient with newly diagnosed type 2 diabetes\"\n\n**When to use each:**\n- Use **clinical-decision-support** for: cohort analyses, biomarker stratification studies, treatment guideline development, pharmaceutical strategy documents\n- Use **treatment-plans** for: individual patient care plans, treatment protocols for specific patients, bedside clinical documentation\n\n## Example Usage\n\n### Patient Cohort Analysis\n\n**Example 1: NSCLC Biomarker Stratification**\n```\n> Analyze a cohort of 45 NSCLC patients stratified by PD-L1 expression (\u003C1%, 1-49%, ≥50%) \n> receiving pembrolizumab. Include outcomes: ORR, median PFS, median OS with hazard ratios \n> comparing PD-L1 ≥50% vs \u003C50%. Generate Kaplan-Meier curves and waterfall plot.\n```\n\n**Example 2: GBM Molecular Subtype Analysis**\n```\n> Generate cohort analysis for 30 GBM patients classified into Cluster 1 (Mesenchymal-Immune-Active) \n> and Cluster 2 (Proneural) molecular subtypes. Compare outcomes including median OS, 6-month PFS rate, \n> and response to TMZ+bevacizumab. Include biomarker profile table and statistical comparison.\n```\n\n**Example 3: Breast Cancer HER2 Cohort**\n```\n> Analyze 60 HER2-positive metastatic breast cancer patients treated with trastuzumab-deruxtecan, \n> stratified by prior trastuzumab exposure (yes\u002Fno). Include ORR, DOR, median PFS with forest plot \n> showing subgroup analyses by hormone receptor status, brain metastases, and number of prior lines.\n```\n\n### Treatment Recommendation Report\n\n**Example 1: HER2+ Metastatic Breast Cancer Guidelines**\n```\n> Create evidence-based treatment recommendations for HER2-positive metastatic breast cancer including \n> biomarker-guided therapy selection. Use GRADE system to grade recommendations for first-line \n> (trastuzumab+pertuzumab+taxane), second-line (trastuzumab-deruxtecan), and third-line options. \n> Include decision algorithm flowchart based on brain metastases, hormone receptor status, and prior therapies.\n```\n\n**Example 2: Advanced NSCLC Treatment Algorithm**\n```\n> Generate treatment recommendation report for advanced NSCLC based on PD-L1 expression, EGFR mutation, \n> ALK rearrangement, and performance status. Include GRADE-graded recommendations for each molecular subtype, \n> TikZ flowchart for biomarker-directed therapy selection, and evidence tables from KEYNOTE-189, FLAURA, \n> and CheckMate-227 trials.\n```\n\n**Example 3: Multiple Myeloma Line-of-Therapy Sequencing**\n```\n> Create treatment algorithm for newly diagnosed multiple myeloma through relapsed\u002Frefractory setting. \n> Include GRADE recommendations for transplant-eligible vs ineligible, high-risk cytogenetics considerations, \n> and sequencing of daratumumab, carfilzomib, and CAR-T therapy. Provide flowchart showing decision points \n> at each line of therapy.\n```\n\n## Key Features\n\n### Biomarker Classification\n- Genomic: Mutations, CNV, gene fusions\n- Expression: RNA-seq, IHC scores\n- Molecular subtypes: Disease-specific classifications\n- Clinical actionability: Therapy selection guidance\n\n### Outcome Metrics\n- Survival: OS (overall survival), PFS (progression-free survival)\n- Response: ORR (objective response rate), DOR (duration of response), DCR (disease control rate)\n- Quality: ECOG performance status, symptom burden\n- Safety: Adverse events, dose modifications\n\n### Statistical Methods\n- Survival analysis: Kaplan-Meier curves, log-rank tests\n- Group comparisons: t-tests, chi-square, Fisher's exact\n- Effect sizes: Hazard ratios, odds ratios with 95% CI\n- Significance: p-values, multiple testing corrections\n\n### Evidence Grading\n\n**GRADE System**\n- **1A**: Strong recommendation, high-quality evidence\n- **1B**: Strong recommendation, moderate-quality evidence  \n- **2A**: Weak recommendation, high-quality evidence\n- **2B**: Weak recommendation, moderate-quality evidence\n- **2C**: Weak recommendation, low-quality evidence\n\n**Recommendation Strength**\n- **Strong**: Benefits clearly outweigh risks\n- **Conditional**: Trade-offs exist, patient values important\n- **Research**: Insufficient evidence, clinical trials needed\n\n## Best Practices\n\n### For Cohort Analyses\n\n1. **Patient Selection Transparency**: Clearly document inclusion\u002Fexclusion criteria, patient flow, and reasons for exclusions\n2. **Biomarker Clarity**: Specify assay methods, platforms (e.g., FoundationOne, Caris), cut-points, and validation status\n3. **Statistical Rigor**: \n   - Report hazard ratios with 95% confidence intervals, not just p-values\n   - Include median follow-up time for survival analyses\n   - Specify statistical tests used (log-rank, Cox regression, Fisher's exact)\n   - Account for multiple comparisons when appropriate\n4. **Outcome Definitions**: Use standard criteria:\n   - Response: RECIST 1.1, iRECIST for immunotherapy\n   - Adverse events: CTCAE version 5.0\n   - Performance status: ECOG or Karnofsky\n5. **Survival Data Presentation**:\n   - Median OS\u002FPFS with 95% CI\n   - Landmark survival rates (6-month, 12-month, 24-month)\n   - Number at risk tables below Kaplan-Meier curves\n   - Censoring clearly indicated\n6. **Subgroup Analyses**: Pre-specify subgroups; clearly label exploratory vs pre-planned analyses\n7. **Data Completeness**: Report missing data and how it was handled\n\n### For Treatment Recommendation Reports\n\n1. **Evidence Grading Transparency**: \n   - Use GRADE system consistently (1A, 1B, 2A, 2B, 2C)\n   - Document rationale for each grade\n   - Clearly state quality of evidence (high, moderate, low, very low)\n2. **Comprehensive Evidence Review**: \n   - Include phase 3 randomized trials as primary evidence\n   - Supplement with phase 2 data for emerging therapies\n   - Note real-world evidence and meta-analyses\n   - Cite trial names (e.g., KEYNOTE-189, CheckMate-227)\n3. **Biomarker-Guided Recommendations**:\n   - Link specific biomarkers to therapy recommendations\n   - Specify testing methods and validated assays\n   - Include FDA\u002FEMA approval status for companion diagnostics\n4. **Clinical Actionability**: Every recommendation should have clear implementation guidance\n5. **Decision Algorithm Clarity**: TikZ flowcharts should be unambiguous with clear yes\u002Fno decision points\n6. **Special Populations**: Address elderly, renal\u002Fhepatic impairment, pregnancy, drug interactions\n7. **Monitoring Guidance**: Specify safety labs, imaging, and frequency\n8. **Update Frequency**: Date recommendations and plan for periodic updates\n\n### General Best Practices\n\n1. **First Page Executive Summary (MANDATORY)**: \n   - ALWAYS create a complete executive summary on page 1 that spans the entire first page\n   - Use 3-5 colored tcolorbox elements to highlight key findings\n   - No table of contents or detailed sections on page 1\n   - Use `\\thispagestyle{empty}` and end with `\\newpage`\n   - This is the single most important page - it should be scannable in 60 seconds\n2. **De-identification**: Remove all 18 HIPAA identifiers before document generation (Safe Harbor method)\n3. **Regulatory Compliance**: Include confidentiality notices for proprietary pharmaceutical data\n4. **Publication-Ready Formatting**: Use 0.5in margins, professional fonts, color-coded sections\n5. **Reproducibility**: Document all statistical methods to enable replication\n6. **Conflict of Interest**: Disclose pharmaceutical funding or relationships when applicable\n7. **Visual Hierarchy**: Use colored boxes consistently (blue=data, green=biomarkers, yellow\u002Forange=recommendations, red=warnings)\n\n## References\n\nSee the `references\u002F` directory for detailed guidance on:\n- Patient cohort analysis and stratification methods\n- Treatment recommendation development\n- Clinical decision algorithms\n- Biomarker classification and interpretation\n- Outcome analysis and statistical methods\n- Evidence synthesis and grading systems\n\n## Templates\n\nSee the `assets\u002F` directory for LaTeX templates:\n- `cohort_analysis_template.tex` - Biomarker-stratified patient cohort analysis with statistical comparisons\n- `treatment_recommendation_template.tex` - Evidence-based clinical practice guidelines with GRADE grading\n- `clinical_pathway_template.tex` - TikZ decision algorithm flowcharts for treatment sequencing\n- `biomarker_report_template.tex` - Molecular subtype classification and genomic profile reports\n- `evidence_synthesis_template.tex` - Systematic evidence review and meta-analysis summaries\n\n**Template Features:**\n- 0.5in margins for compact presentation\n- Color-coded recommendation boxes\n- Professional tables for demographics, biomarkers, outcomes\n- Built-in support for Kaplan-Meier curves, waterfall plots, forest plots\n- GRADE evidence grading tables\n- Confidentiality headers for pharmaceutical documents\n\n## Scripts\n\nSee the `scripts\u002F` directory for analysis and visualization tools:\n- `generate_survival_analysis.py` - Kaplan-Meier curve generation with log-rank tests, hazard ratios, 95% CI\n- `create_waterfall_plot.py` - Best response visualization for cohort analyses\n- `create_forest_plot.py` - Subgroup analysis visualization with confidence intervals\n- `create_cohort_tables.py` - Demographics, biomarker frequency, and outcomes tables\n- `build_decision_tree.py` - TikZ flowchart generation for treatment algorithms\n- `biomarker_classifier.py` - Patient stratification algorithms by molecular subtype\n- `calculate_statistics.py` - Hazard ratios, Cox regression, log-rank tests, Fisher's exact\n- `validate_cds_document.py` - Quality and compliance checks (HIPAA, statistical reporting standards)\n- `grade_evidence.py` - Automated GRADE assessment helper for treatment recommendations\n\n","# Clinical Decision Support Documents\n\n## Description\n\nGenerate professional clinical decision support (CDS) documents for pharmaceutical companies, clinical researchers, and medical decision-makers. This skill specializes in analytical, evidence-based documents that inform treatment strategies and drug development:\n\n1. **Patient Cohort Analysis** - Biomarker-stratified group analyses with statistical outcome comparisons\n2. **Treatment Recommendation Reports** - Evidence-based clinical guidelines with GRADE grading and decision algorithms\n\nAll documents are generated as publication-ready LaTeX\u002FPDF files optimized for pharmaceutical research, regulatory submissions, and clinical guideline development.\n\n**Note:** For individual patient treatment plans at the bedside, use the `treatment-plans` skill instead. This skill focuses on group-level analyses and evidence synthesis for pharmaceutical\u002Fresearch settings.\n\n## Capabilities\n\n### Document Types\n\n**Patient Cohort Analysis**\n- Biomarker-based patient stratification (molecular subtypes, gene expression, IHC)\n- Molecular subtype classification (e.g., GBM mesenchymal-immune-active vs proneural, breast cancer subtypes)\n- Outcome metrics with statistical analysis (OS, PFS, ORR, DOR, DCR)\n- Statistical comparisons between subgroups (hazard ratios, p-values, 95% CI)\n- Survival analysis with Kaplan-Meier curves and log-rank tests\n- Efficacy tables and waterfall plots\n- Comparative effectiveness analyses\n- Pharmaceutical cohort reporting (trial subgroups, real-world evidence)\n\n**Treatment Recommendation Reports**\n- Evidence-based treatment guidelines for specific disease states\n- Strength of recommendation grading (GRADE system: 1A, 1B, 2A, 2B, 2C)\n- Quality of evidence assessment (high, moderate, low, very low)\n- Treatment algorithm flowcharts with TikZ diagrams\n- Line-of-therapy sequencing based on biomarkers\n- Decision pathways with clinical and molecular criteria\n- Pharmaceutical strategy documents\n- Clinical guideline development for medical societies\n\n### Clinical Features\n\n- **Biomarker Integration**: Genomic alterations (mutations, CNV, fusions), gene expression signatures, IHC markers, PD-L1 scoring\n- **Statistical Analysis**: Hazard ratios, p-values, confidence intervals, survival curves, Cox regression, log-rank tests\n- **Evidence Grading**: GRADE system (1A\u002F1B\u002F2A\u002F2B\u002F2C), Oxford CEBM levels, quality of evidence assessment\n- **Clinical Terminology**: SNOMED-CT, LOINC, proper medical nomenclature, trial nomenclature\n- **Regulatory Compliance**: HIPAA de-identification, confidentiality headers, ICH-GCP alignment\n- **Professional Formatting**: Compact 0.5in margins, color-coded recommendations, publication-ready, suitable for regulatory submissions\n\n## Pharmaceutical and Research Use Cases\n\nThis skill is specifically designed for pharmaceutical and clinical research applications:\n\n**Drug Development**\n- **Phase 2\u002F3 Trial Analyses**: Biomarker-stratified efficacy and safety analyses\n- **Subgroup Analyses**: Forest plots showing treatment effects across patient subgroups\n- **Companion Diagnostic Development**: Linking biomarkers to drug response\n- **Regulatory Submissions**: IND\u002FNDA documentation with evidence summaries\n\n**Medical Affairs**\n- **KOL Education Materials**: Evidence-based treatment algorithms for thought leaders\n- **Medical Strategy Documents**: Competitive landscape and positioning strategies\n- **Advisory Board Materials**: Cohort analyses and treatment recommendation frameworks\n- **Publication Planning**: Manuscript-ready analyses for peer-reviewed journals\n\n**Clinical Guidelines**\n- **Guideline Development**: Evidence synthesis with GRADE methodology for specialty societies\n- **Consensus Recommendations**: Multi-stakeholder treatment algorithm development\n- **Practice Standards**: Biomarker-based treatment selection criteria\n- **Quality Measures**: Evidence-based performance metrics\n\n**Real-World Evidence**\n- **RWE Cohort Studies**: Retrospective analyses of patient cohorts from EMR data\n- **Comparative Effectiveness**: Head-to-head treatment comparisons in real-world settings\n- **Outcomes Research**: Long-term survival and safety in clinical practice\n- **Health Economics**: Cost-effectiveness analyses by biomarker subgroup\n\n## When to Use\n\nUse this skill when you need to:\n\n- **Analyze patient cohorts** stratified by biomarkers, molecular subtypes, or clinical characteristics\n- **Generate treatment recommendation reports** with evidence grading for clinical guidelines or pharmaceutical strategies\n- **Compare outcomes** between patient subgroups with statistical analysis (survival, response rates, hazard ratios)\n- **Produce pharmaceutical research documents** for drug development, clinical trials, or regulatory submissions\n- **Develop clinical practice guidelines** with GRADE evidence grading and decision algorithms\n- **Document biomarker-guided therapy selection** at the population level (not individual patients)\n- **Synthesize evidence** from multiple trials or real-world data sources\n- **Create clinical decision algorithms** with flowcharts for treatment sequencing\n\n**Do NOT use this skill for:**\n- Individual patient treatment plans (use `treatment-plans` skill)\n- Bedside clinical care documentation (use `treatment-plans` skill)\n- Simple patient-specific treatment protocols (use `treatment-plans` skill)\n\n## Visual Enhancement with Scientific Schematics\n\n**⚠️ MANDATORY: Every clinical decision support document MUST include at least 1-2 AI-generated figures using the scientific-schematics skill.**\n\nThis is not optional. Clinical decision documents require clear visual algorithms. Before finalizing any document:\n1. Generate at minimum ONE schematic or diagram (e.g., clinical decision algorithm, treatment pathway, or biomarker stratification tree)\n2. For cohort analyses: include patient flow diagram\n3. For treatment recommendations: include decision flowchart\n\n**How to generate figures:**\n- Use the **scientific-schematics** skill to generate AI-powered publication-quality diagrams\n- Simply describe your desired diagram in natural language\n- Nano Banana Pro will automatically generate, review, and refine the schematic\n\n**How to generate schematics:**\n```bash\npython scripts\u002Fgenerate_schematic.py \"your diagram description\" -o figures\u002Foutput.png\n```\n\nThe AI will automatically:\n- Create publication-quality images with proper formatting\n- Review and refine through multiple iterations\n- Ensure accessibility (colorblind-friendly, high contrast)\n- Save outputs in the figures\u002F directory\n\n**When to add schematics:**\n- Clinical decision algorithm flowcharts\n- Treatment pathway diagrams\n- Biomarker stratification trees\n- Patient cohort flow diagrams (CONSORT-style)\n- Survival curve visualizations\n- Molecular mechanism diagrams\n- Any complex concept that benefits from visualization\n\nFor detailed guidance on creating schematics, refer to the scientific-schematics skill documentation.\n\n---\n\n## Document Structure\n\n**CRITICAL REQUIREMENT: All clinical decision support documents MUST begin with a complete executive summary on page 1 that spans the entire first page before any table of contents or detailed sections.**\n\n### Page 1 Executive Summary Structure\n\nThe first page of every CDS document should contain ONLY the executive summary with the following components:\n\n**Required Elements (all on page 1):**\n1. **Document Title and Type**\n   - Main title (e.g., \"Biomarker-Stratified Cohort Analysis\" or \"Evidence-Based Treatment Recommendations\")\n   - Subtitle with disease state and focus\n   \n2. **Report Information Box** (using colored tcolorbox)\n   - Document type and purpose\n   - Date of analysis\u002Freport\n   - Disease state and patient population\n   - Author\u002Finstitution (if applicable)\n   - Analysis framework or methodology\n   \n3. **Key Findings Boxes** (3-5 colored boxes using tcolorbox)\n   - **Primary Results** (blue box): Main efficacy\u002Foutcome findings\n   - **Biomarker Insights** (green box): Key molecular subtype findings\n   - **Clinical Implications** (yellow\u002Forange box): Actionable treatment implications\n   - **Statistical Summary** (gray box): Hazard ratios, p-values, key statistics\n   - **Safety Highlights** (red box, if applicable): Critical adverse events or warnings\n\n**Visual Requirements:**\n- Use `\\thispagestyle{empty}` to remove page numbers from page 1\n- All content must fit on page 1 (before `\\newpage`)\n- Use colored tcolorbox environments with different colors for visual hierarchy\n- Boxes should be scannable and highlight most critical information\n- Use bullet points, not narrative paragraphs\n- End page 1 with `\\newpage` before table of contents or detailed sections\n\n**Example First Page LaTeX Structure:**\n```latex\n\\maketitle\n\\thispagestyle{empty}\n\n% Report Information Box\n\\begin{tcolorbox}[colback=blue!5!white, colframe=blue!75!black, title=Report Information]\n\\textbf{Document Type:} Patient Cohort Analysis\\\\\n\\textbf{Disease State:} HER2-Positive Metastatic Breast Cancer\\\\\n\\textbf{Analysis Date:} \\today\\\\\n\\textbf{Population:} 60 patients, biomarker-stratified by HR status\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #1: Primary Results\n\\begin{tcolorbox}[colback=blue!5!white, colframe=blue!75!black, title=Primary Efficacy Results]\n\\begin{itemize}\n    \\item Overall ORR: 72\\% (95\\% CI: 59-83\\%)\n    \\item Median PFS: 18.5 months (95\\% CI: 14.2-22.8)\n    \\item Median OS: 35.2 months (95\\% CI: 28.1-NR)\n\\end{itemize}\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #2: Biomarker Insights\n\\begin{tcolorbox}[colback=green!5!white, colframe=green!75!black, title=Biomarker Stratification Findings]\n\\begin{itemize}\n    \\item HR+\u002FHER2+: ORR 68\\%, median PFS 16.2 months\n    \\item HR-\u002FHER2+: ORR 78\\%, median PFS 22.1 months\n    \\item HR status significantly associated with outcomes (p=0.041)\n\\end{itemize}\n\\end{tcolorbox}\n\n\\vspace{0.3cm}\n\n% Key Finding #3: Clinical Implications\n\\begin{tcolorbox}[colback=orange!5!white, colframe=orange!75!black, title=Clinical Recommendations]\n\\begin{itemize}\n    \\item Strong efficacy observed regardless of HR status (Grade 1A)\n    \\item HR-\u002FHER2+ patients showed numerically superior outcomes\n    \\item Treatment recommended for all HER2+ MBC patients\n\\end{itemize}\n\\end{tcolorbox}\n\n\\newpage\n\\tableofcontents  % TOC on page 2\n\\newpage  % Detailed content starts page 3\n```\n\n### Patient Cohort Analysis (Detailed Sections - Page 3+)\n- **Cohort Characteristics**: Demographics, baseline features, patient selection criteria\n- **Biomarker Stratification**: Molecular subtypes, genomic alterations, IHC profiles\n- **Treatment Exposure**: Therapies received, dosing, treatment duration by subgroup\n- **Outcome Analysis**: Response rates (ORR, DCR), survival data (OS, PFS), DOR\n- **Statistical Methods**: Kaplan-Meier survival curves, hazard ratios, log-rank tests, Cox regression\n- **Subgroup Comparisons**: Biomarker-stratified efficacy, forest plots, statistical significance\n- **Safety Profile**: Adverse events by subgroup, dose modifications, discontinuations\n- **Clinical Recommendations**: Treatment implications based on biomarker profiles\n- **Figures**: Waterfall plots, swimmer plots, survival curves, forest plots\n- **Tables**: Demographics table, biomarker frequency, outcomes by subgroup\n\n### Treatment Recommendation Reports (Detailed Sections - Page 3+)\n\n**Page 1 Executive Summary for Treatment Recommendations should include:**\n1. **Report Information Box**: Disease state, guideline version\u002Fdate, target population\n2. **Key Recommendations Box** (green): Top 3-5 GRADE-graded recommendations by line of therapy\n3. **Biomarker Decision Criteria Box** (blue): Key molecular markers influencing treatment selection\n4. **Evidence Summary Box** (gray): Major trials supporting recommendations (e.g., KEYNOTE-189, FLAURA)\n5. **Critical Monitoring Box** (orange\u002Fred): Essential safety monitoring requirements\n\n**Detailed Sections (Page 3+):**\n- **Clinical Context**: Disease state, epidemiology, current treatment landscape\n- **Target Population**: Patient characteristics, biomarker criteria, staging\n- **Evidence Review**: Systematic literature synthesis, guideline summary, trial data\n- **Treatment Options**: Available therapies with mechanism of action\n- **Evidence Grading**: GRADE assessment for each recommendation (1A, 1B, 2A, 2B, 2C)\n- **Recommendations by Line**: First-line, second-line, subsequent therapies\n- **Biomarker-Guided Selection**: Decision criteria based on molecular profiles\n- **Treatment Algorithms**: TikZ flowcharts showing decision pathways\n- **Monitoring Protocol**: Safety assessments, efficacy monitoring, dose modifications\n- **Special Populations**: Elderly, renal\u002Fhepatic impairment, comorbidities\n- **References**: Full bibliography with trial names and citations\n\n## Output Format\n\n**MANDATORY FIRST PAGE REQUIREMENT:**\n- **Page 1**: Full-page executive summary with 3-5 colored tcolorbox elements\n- **Page 2**: Table of contents (optional)\n- **Page 3+**: Detailed sections with methods, results, figures, tables\n\n**Document Specifications:**\n- **Primary**: LaTeX\u002FPDF with 0.5in margins for compact, data-dense presentation\n- **Length**: Typically 5-15 pages (1 page executive summary + 4-14 pages detailed content)\n- **Style**: Publication-ready, pharmaceutical-grade, suitable for regulatory submissions\n- **First Page**: Always a complete executive summary spanning entire page 1 (see Document Structure section)\n\n**Visual Elements:**\n- **Colors**: \n  - Page 1 boxes: blue=data\u002Finformation, green=biomarkers\u002Frecommendations, yellow\u002Forange=clinical implications, red=warnings\n  - Recommendation boxes (green=strong recommendation, yellow=conditional, blue=research needed)\n  - Biomarker stratification (color-coded molecular subtypes)\n  - Statistical significance (color-coded p-values, hazard ratios)\n- **Tables**: \n  - Demographics with baseline characteristics\n  - Biomarker frequency by subgroup\n  - Outcomes table (ORR, PFS, OS, DOR by molecular subtype)\n  - Adverse events by cohort\n  - Evidence summary tables with GRADE ratings\n- **Figures**: \n  - Kaplan-Meier survival curves with log-rank p-values and number at risk tables\n  - Waterfall plots showing best response by patient\n  - Forest plots for subgroup analyses with confidence intervals\n  - TikZ decision algorithm flowcharts\n  - Swimmer plots for individual patient timelines\n- **Statistics**: Hazard ratios with 95% CI, p-values, median survival times, landmark survival rates\n- **Compliance**: De-identification per HIPAA Safe Harbor, confidentiality notices for proprietary data\n\n## Integration\n\nThis skill integrates with:\n- **scientific-writing**: Citation management, statistical reporting, evidence synthesis\n- **clinical-reports**: Medical terminology, HIPAA compliance, regulatory documentation\n- **scientific-schematics**: TikZ flowcharts for decision algorithms and treatment pathways\n- **treatment-plans**: Individual patient applications of cohort-derived insights (bidirectional)\n\n## Key Differentiators from Treatment-Plans Skill\n\n**Clinical Decision Support (this skill):**\n- **Audience**: Pharmaceutical companies, clinical researchers, guideline committees, medical affairs\n- **Scope**: Population-level analyses, evidence synthesis, guideline development\n- **Focus**: Biomarker stratification, statistical comparisons, evidence grading\n- **Output**: Multi-page analytical documents (5-15 pages typical) with extensive figures and tables\n- **Use Cases**: Drug development, regulatory submissions, clinical practice guidelines, medical strategy\n- **Example**: \"Analyze 60 HER2+ breast cancer patients by hormone receptor status with survival outcomes\"\n\n**Treatment-Plans Skill:**\n- **Audience**: Clinicians, patients, care teams\n- **Scope**: Individual patient care planning\n- **Focus**: SMART goals, patient-specific interventions, monitoring plans\n- **Output**: Concise 1-4 page actionable care plans\n- **Use Cases**: Bedside clinical care, EMR documentation, patient-centered planning\n- **Example**: \"Create treatment plan for a 55-year-old patient with newly diagnosed type 2 diabetes\"\n\n**When to use each:**\n- Use **clinical-decision-support** for: cohort analyses, biomarker stratification studies, treatment guideline development, pharmaceutical strategy documents\n- Use **treatment-plans** for: individual patient care plans, treatment protocols for specific patients, bedside clinical documentation\n\n## Example Usage\n\n### Patient Cohort Analysis\n\n**Example 1: NSCLC Biomarker Stratification**\n```\n> Analyze a cohort of 45 NSCLC patients stratified by PD-L1 expression (\u003C1%, 1-49%, ≥50%) \n> receiving pembrolizumab. Include outcomes: ORR, median PFS, median OS with hazard ratios \n> comparing PD-L1 ≥50% vs \u003C50%. Generate Kaplan-Meier curves and waterfall plot.\n```\n\n**Example 2: GBM Molecular Subtype Analysis**\n```\n> Generate cohort analysis for 30 GBM patients classified into Cluster 1 (Mesenchymal-Immune-Active) \n> and Cluster 2 (Proneural) molecular subtypes. Compare outcomes including median OS, 6-month PFS rate, \n> and response to TMZ+bevacizumab. Include biomarker profile table and statistical comparison.\n```\n\n**Example 3: Breast Cancer HER2 Cohort**\n```\n> Analyze 60 HER2-positive metastatic breast cancer patients treated with trastuzumab-deruxtecan, \n> stratified by prior trastuzumab exposure (yes\u002Fno). Include ORR, DOR, median PFS with forest plot \n> showing subgroup analyses by hormone receptor status, brain metastases, and number of prior lines.\n```\n\n### Treatment Recommendation Report\n\n**Example 1: HER2+ Metastatic Breast Cancer Guidelines**\n```\n> Create evidence-based treatment recommendations for HER2-positive metastatic breast cancer including \n> biomarker-guided therapy selection. Use GRADE system to grade recommendations for first-line \n> (trastuzumab+pertuzumab+taxane), second-line (trastuzumab-deruxtecan), and third-line options. \n> Include decision algorithm flowchart based on brain metastases, hormone receptor status, and prior therapies.\n```\n\n**Example 2: Advanced NSCLC Treatment Algorithm**\n```\n> Generate treatment recommendation report for advanced NSCLC based on PD-L1 expression, EGFR mutation, \n> ALK rearrangement, and performance status. Include GRADE-graded recommendations for each molecular subtype, \n> TikZ flowchart for biomarker-directed therapy selection, and evidence tables from KEYNOTE-189, FLAURA, \n> and CheckMate-227 trials.\n```\n\n**Example 3: Multiple Myeloma Line-of-Therapy Sequencing**\n```\n> Create treatment algorithm for newly diagnosed multiple myeloma through relapsed\u002Frefractory setting. \n> Include GRADE recommendations for transplant-eligible vs ineligible, high-risk cytogenetics considerations, \n> and sequencing of daratumumab, carfilzomib, and CAR-T therapy. Provide flowchart showing decision points \n> at each line of therapy.\n```\n\n## Key Features\n\n### Biomarker Classification\n- Genomic: Mutations, CNV, gene fusions\n- Expression: RNA-seq, IHC scores\n- Molecular subtypes: Disease-specific classifications\n- Clinical actionability: Therapy selection guidance\n\n### Outcome Metrics\n- Survival: OS (overall survival), PFS (progression-free survival)\n- Response: ORR (objective response rate), DOR (duration of response), DCR (disease control rate)\n- Quality: ECOG performance status, symptom burden\n- Safety: Adverse events, dose modifications\n\n### Statistical Methods\n- Survival analysis: Kaplan-Meier curves, log-rank tests\n- Group comparisons: t-tests, chi-square, Fisher's exact\n- Effect sizes: Hazard ratios, odds ratios with 95% CI\n- Significance: p-values, multiple testing corrections\n\n### Evidence Grading\n\n**GRADE System**\n- **1A**: Strong recommendation, high-quality evidence\n- **1B**: Strong recommendation, moderate-quality evidence  \n- **2A**: Weak recommendation, high-quality evidence\n- **2B**: Weak recommendation, moderate-quality evidence\n- **2C**: Weak recommendation, low-quality evidence\n\n**Recommendation Strength**\n- **Strong**: Benefits clearly outweigh risks\n- **Conditional**: Trade-offs exist, patient values important\n- **Research**: Insufficient evidence, clinical trials needed\n\n## Best Practices\n\n### For Cohort Analyses\n\n1. **Patient Selection Transparency**: Clearly document inclusion\u002Fexclusion criteria, patient flow, and reasons for exclusions\n2. **Biomarker Clarity**: Specify assay methods, platforms (e.g., FoundationOne, Caris), cut-points, and validation status\n3. **Statistical Rigor**: \n   - Report hazard ratios with 95% confidence intervals, not just p-values\n   - Include median follow-up time for survival analyses\n   - Specify statistical tests used (log-rank, Cox regression, Fisher's exact)\n   - Account for multiple comparisons when appropriate\n4. **Outcome Definitions**: Use standard criteria:\n   - Response: RECIST 1.1, iRECIST for immunotherapy\n   - Adverse events: CTCAE version 5.0\n   - Performance status: ECOG or Karnofsky\n5. **Survival Data Presentation**:\n   - Median OS\u002FPFS with 95% CI\n   - Landmark survival rates (6-month, 12-month, 24-month)\n   - Number at risk tables below Kaplan-Meier curves\n   - Censoring clearly indicated\n6. **Subgroup Analyses**: Pre-specify subgroups; clearly label exploratory vs pre-planned analyses\n7. **Data Completeness**: Report missing data and how it was handled\n\n### For Treatment Recommendation Reports\n\n1. **Evidence Grading Transparency**: \n   - Use GRADE system consistently (1A, 1B, 2A, 2B, 2C)\n   - Document rationale for each grade\n   - Clearly state quality of evidence (high, moderate, low, very low)\n2. **Comprehensive Evidence Review**: \n   - Include phase 3 randomized trials as primary evidence\n   - Supplement with phase 2 data for emerging therapies\n   - Note real-world evidence and meta-analyses\n   - Cite trial names (e.g., KEYNOTE-189, CheckMate-227)\n3. **Biomarker-Guided Recommendations**:\n   - Link specific biomarkers to therapy recommendations\n   - Specify testing methods and validated assays\n   - Include FDA\u002FEMA approval status for companion diagnostics\n4. **Clinical Actionability**: Every recommendation should have clear implementation guidance\n5. **Decision Algorithm Clarity**: TikZ flowcharts should be unambiguous with clear yes\u002Fno decision points\n6. **Special Populations**: Address elderly, renal\u002Fhepatic impairment, pregnancy, drug interactions\n7. **Monitoring Guidance**: Specify safety labs, imaging, and frequency\n8. **Update Frequency**: Date recommendations and plan for periodic updates\n\n### General Best Practices\n\n1. **First Page Executive Summary (MANDATORY)**: \n   - ALWAYS create a complete executive summary on page 1 that spans the entire first page\n   - Use 3-5 colored tcolorbox elements to highlight key findings\n   - No table of contents or detailed sections on page 1\n   - Use `\\thispagestyle{empty}` and end with `\\newpage`\n   - This is the single most important page - it should be scannable in 60 seconds\n2. **De-identification**: Remove all 18 HIPAA identifiers before document generation (Safe Harbor method)\n3. **Regulatory Compliance**: Include confidentiality notices for proprietary pharmaceutical data\n4. **Publication-Ready Formatting**: Use 0.5in margins, professional fonts, color-coded sections\n5. **Reproducibility**: Document all statistical methods to enable replication\n6. **Conflict of Interest**: Disclose pharmaceutical funding or relationships when applicable\n7. **Visual Hierarchy**: Use colored boxes consistently (blue=data, green=biomarkers, yellow\u002Forange=recommendations, red=warnings)\n\n## References\n\nSee the `references\u002F` directory for detailed guidance on:\n- Patient cohort analysis and stratification methods\n- Treatment recommendation development\n- Clinical decision algorithms\n- Biomarker classification and interpretation\n- Outcome analysis and statistical methods\n- Evidence synthesis and grading systems\n\n## Templates\n\nSee the `assets\u002F` directory for LaTeX templates:\n- `cohort_analysis_template.tex` - Biomarker-stratified patient cohort analysis with statistical comparisons\n- `treatment_recommendation_template.tex` - Evidence-based clinical practice guidelines with GRADE grading\n- `clinical_pathway_template.tex` - TikZ decision algorithm flowcharts for treatment sequencing\n- `biomarker_report_template.tex` - Molecular subtype classification and genomic profile reports\n- `evidence_synthesis_template.tex` - Systematic evidence review and meta-analysis summaries\n\n**Template Features:**\n- 0.5in margins for compact presentation\n- Color-coded recommendation boxes\n- Professional tables for demographics, biomarkers, outcomes\n- Built-in support for Kaplan-Meier curves, waterfall plots, forest plots\n- GRADE evidence grading tables\n- Confidentiality headers for pharmaceutical documents\n\n## Scripts\n\nSee the `scripts\u002F` directory for analysis and visualization tools:\n- `generate_survival_analysis.py` - Kaplan-Meier curve generation with log-rank tests, hazard ratios, 95% CI\n- `create_waterfall_plot.py` - Best response visualization for cohort analyses\n- `create_forest_plot.py` - Subgroup analysis visualization with confidence intervals\n- `create_cohort_tables.py` - Demographics, biomarker frequency, and outcomes tables\n- `build_decision_tree.py` - TikZ flowchart generation for treatment algorithms\n- `biomarker_classifier.py` - Patient stratification algorithms by molecular subtype\n- `calculate_statistics.py` - Hazard ratios, Cox regression, log-rank tests, Fisher's exact\n- `validate_cds_document.py` - Quality and compliance checks (HIPAA, statistical reporting standards)\n- `grade_evidence.py` - Automated GRADE assessment helper for treatment recommendations\n\n",{"name":5,"description":7,"allowed-tools":23},[24,25,26,27],"Read","Write","Edit","Bash","main","d046d088ce5a79a083f0abc0fe1b7ded73bc5b1f","claude-config\u002Fskills\u002Fclinical-decision-support","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent","https:\u002F\u002Fgithub.com\u002FSansar35","2026-05-10 16:04:53",{"source":8,"skillName":5,"commands":35,"prompt":45,"zipAvailable":46},[36,39,42],{"tool":37,"command":38},"npx","npx skills add Sansar35\u002Fclaude-agent --skill clinical-decision-support",{"tool":40,"command":41},"bunx","bunx skills add Sansar35\u002Fclaude-agent --skill clinical-decision-support",{"tool":43,"command":44},"pnpm","pnpm dlx skills add Sansar35\u002Fclaude-agent --skill clinical-decision-support","安装技能，技能地址：https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002FSKILL.md",true,[48,54,60,64,68,72,76,80,84,88,92,96,100,104,108,112,118,122,126,130],{"path":49,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":52,"sourceHtmlUrl":53},".skill-source.json","REFERENCE","参考文档",800,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002F.skill-source.json",{"path":55,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":58,"sourceHtmlUrl":59},"assets\u002Fbiomarker_report_template.tex","ASSET","资产文件",13326,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Fbiomarker_report_template.tex",{"path":61,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":62,"sourceHtmlUrl":63},"assets\u002Fclinical_pathway_template.tex",8325,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Fclinical_pathway_template.tex",{"path":65,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":66,"sourceHtmlUrl":67},"assets\u002Fcohort_analysis_template.tex",12601,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Fcohort_analysis_template.tex",{"path":69,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":70,"sourceHtmlUrl":71},"assets\u002Fcolor_schemes.tex",7029,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Fcolor_schemes.tex",{"path":73,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":74,"sourceHtmlUrl":75},"assets\u002Fexample_gbm_cohort.md",9316,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Fexample_gbm_cohort.md",{"path":77,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":78,"sourceHtmlUrl":79},"assets\u002Frecommendation_strength_guide.md",11322,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Frecommendation_strength_guide.md",{"path":81,"resourceType":56,"resourceTypeDesc":57,"sizeBytes":82,"sourceHtmlUrl":83},"assets\u002Ftreatment_recommendation_template.tex",18843,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fassets\u002Ftreatment_recommendation_template.tex",{"path":85,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":86,"sourceHtmlUrl":87},"README.md",5021,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002FREADME.md",{"path":89,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":90,"sourceHtmlUrl":91},"references\u002Fbiomarker_classification.md",29270,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Fbiomarker_classification.md",{"path":93,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":94,"sourceHtmlUrl":95},"references\u002Fclinical_decision_algorithms.md",21393,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Fclinical_decision_algorithms.md",{"path":97,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":98,"sourceHtmlUrl":99},"references\u002Fevidence_synthesis.md",33342,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Fevidence_synthesis.md",{"path":101,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":102,"sourceHtmlUrl":103},"references\u002Foutcome_analysis.md",24628,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Foutcome_analysis.md",{"path":105,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":106,"sourceHtmlUrl":107},"references\u002Fpatient_cohort_analysis.md",16551,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Fpatient_cohort_analysis.md",{"path":109,"resourceType":50,"resourceTypeDesc":51,"sizeBytes":110,"sourceHtmlUrl":111},"references\u002Ftreatment_recommendations.md",19647,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Freferences\u002Ftreatment_recommendations.md",{"path":113,"resourceType":114,"resourceTypeDesc":115,"sizeBytes":116,"sourceHtmlUrl":117},"scripts\u002Fbiomarker_classifier.py","SCRIPT","脚本",13315,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fscripts\u002Fbiomarker_classifier.py",{"path":119,"resourceType":114,"resourceTypeDesc":115,"sizeBytes":120,"sourceHtmlUrl":121},"scripts\u002Fbuild_decision_tree.py",15891,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fscripts\u002Fbuild_decision_tree.py",{"path":123,"resourceType":114,"resourceTypeDesc":115,"sizeBytes":124,"sourceHtmlUrl":125},"scripts\u002Fcreate_cohort_tables.py",18023,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fscripts\u002Fcreate_cohort_tables.py",{"path":127,"resourceType":114,"resourceTypeDesc":115,"sizeBytes":128,"sourceHtmlUrl":129},"scripts\u002Fgenerate_survival_analysis.py",15258,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fscripts\u002Fgenerate_survival_analysis.py",{"path":131,"resourceType":114,"resourceTypeDesc":115,"sizeBytes":132,"sourceHtmlUrl":133},"scripts\u002Fvalidate_cds_document.py",12735,"https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fclinical-decision-support\u002Fscripts\u002Fvalidate_cds_document.py",[135,142,149,156,163,171],{"skillNo":136,"name":137,"nameZh":6,"description":138,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":139,"categories":140,"keywords":141,"publishedAt":18},"01M72DVAN1JYEA302MT0C5MGD6","zustand-store-ts","Create Zustand stores following established patterns with proper TypeScript types and middleware.","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzustand-store-ts\u002FSKILL.md",[],[],{"skillNo":143,"name":144,"nameZh":6,"description":145,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":146,"categories":147,"keywords":148,"publishedAt":18},"01M72DV8J5NBRRM4RPDFRV95N7","zod-validation-expert","Expert in Zod — TypeScript-first schema validation. Covers parsing, custom errors, refinements, type inference, and integration with React Hook Form, Next.js, and tRPC.","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzod-validation-expert\u002FSKILL.md",[],[],{"skillNo":150,"name":151,"nameZh":6,"description":152,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":153,"categories":154,"keywords":155,"publishedAt":18},"01M72DV5EMHW0BHP9BZ167MDCB","zinc-database","Access ZINC (230M+ purchasable compounds). Search by ZINC ID\u002FSMILES, similarity searches, 3D-ready structures for docking, analog discovery, for virtual screening and drug discovery.","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzinc-database\u002FSKILL.md",[],[],{"skillNo":157,"name":158,"nameZh":6,"description":159,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":160,"categories":161,"keywords":162,"publishedAt":18},"01M72DV216QY0E693ZPEVQ0DBS","zarr-python","Chunked N-D arrays for cloud storage. Compressed arrays, parallel I\u002FO, S3\u002FGCS integration, NumPy\u002FDask\u002FXarray compatible, for large-scale scientific computing pipelines.","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzarr-python\u002FSKILL.md",[],[],{"skillNo":164,"name":165,"nameZh":6,"description":166,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":167,"sourceHtmlUrl":168,"categories":169,"keywords":170,"publishedAt":18},"01M72DTXQREMCRKA8NHCVKHN4W","zapier-workflows","Manage and trigger pre-built Zapier workflows and MCP tool orchestration. Use when user mentions workflows, Zaps, automations, daily digest, research, search, lead tracking, expenses, or asks to \"run\" any process. Also handles Perplexity-based research and Google Sheets data tracking.","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzapier-workflows\u002FLICENSE.txt","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzapier-workflows\u002FSKILL.md",[],[],{"skillNo":172,"name":173,"nameZh":6,"description":174,"descriptionZh":6,"repositoryFullName":8,"ownerLogin":9,"ownerAvatarUrl":10,"ownerAccountNo":11,"ownerDisplayName":6,"repositoryNo":12,"starsCount":13,"licenseSpdxId":14,"licenseFileUrl":6,"sourceHtmlUrl":175,"categories":176,"keywords":177,"publishedAt":18},"01M72DTT17TBSJ0Y8J2XZ4C9TZ","zapier-make-patterns","No-code automation democratizes workflow building. Zapier and Make (formerly Integromat) let non-developers automate business processes without writing code. But no-code doesn't mean no-complexity - these platforms have their own patterns, pitfalls, and breaking points.  This skill covers when to use which platform, how to build reliable automations, and when to graduate to code-based solutions. Key insight: Zapier optimizes for simplicity and integrations (7000+ apps), Make optimizes for power","https:\u002F\u002Fgithub.com\u002FSansar35\u002Fclaude-agent\u002Fblob\u002Fmain\u002Fclaude-config\u002Fskills\u002Fzapier-make-patterns\u002FSKILL.md",[],[]]