COMMUNICATION STYLE: RESEARCH (College+ reading level, medical literacy assumed) WRITING RULES: - Use precise...

Question asked:

COMMUNICATION STYLE: RESEARCH (College+ reading level, medical literacy assumed) WRITING RULES: - Use precise medical terminology without explanation - Include specific citations: NCT numbers, PMID references, trial names - Provide statistical details: hazard ratios (HR), confidence intervals (95% CI), p-values - Discuss mechanism of action for each treatment option - Include level of evidence (Phase I/II/III, retrospective, meta-analysis) - Note limitations, caveats, and conflicting data - Compare response rates, PFS, OS across options - Reference specific mutations/biomarkers and matching drugs FORMAT: • Use structured paragraphs with clinical detail • Include a "Key Evidence" section with citations • Note "Data Quality" for each recommendation (strong/moderate/limited evidence) • Include "What We Don't Know Yet" section for emerging questions Five oncology perspectives have analyzed a cancer case: 1. Standard of Care (NCCN guidelines) 2. Emerging Evidence (clinical trials, cutting edge) 3. Molecular/Targeted (precision medicine) 4. Watch & Wait (conservative, de-escalation) 5. Whole Person (quality of life, integrative) Standard of Care: Analysis could not be completed. Recommendation: Please try again Emerging Evidence: Analysis could not be completed. Recommendation: Please try again Molecular/Targeted: Analysis could not be completed. Recommendation: Please try again Watch & Wait: Analysis could not be completed. Recommendation: Please try again Whole Person: Analysis could not be completed. Recommendation: Please try again Synthesize ALL FIVE perspectives. Respond in this exact JSON format: { "headline": "One sentence summarizing the clinical bottom line", "synthesis": "2-3 sentences with specific clinical details, citing perspective names", "consensus": ["Specific point of agreement with clinical rationale"], "divergence": ["Specific disagreement noting which perspectives differ and why"], "decision_leverage": { "changes_management": ["Recommendation that directly changes therapy selection or sequencing", "Biomarker-driven treatment decision"], "vs_noise": ["Additional testing without clear treatment implications", "Information that is true but doesn't change management"] }, "questions_for_team": ["1. What objective criteria define treatment failure for this approach?", "2. Is there clinical trial eligibility based on these findings?", "3. What monitoring schedule is appropriate?"], "monitoring_summary": "Specific tests, frequency, and thresholds to track response vs progression" } CRITICAL FRAMING - DECISION LEVERAGE: You MUST distinguish between: - "changes_management": Recommendations that ACTUALLY CHANGE what the patient does (therapy selection, testing that guides therapy, monitoring that triggers decisions) - "vs_noise": Information that is true/interesting but doesn't change management (nice-to-know context, tests without treatment implications, generic advice) This distinction is the most valuable output. Patients need to know what to ACT on vs what to simply understand. QUESTIONS FOR TEAM: Generate 2-5 numbered questions the patient should ask their oncology team. These should be: - Specific to THIS case (not generic) - Actionable (can be answered with a decision or plan) - Prioritized (most important first) Example good questions: - "Given my BRCA2 status, should we discuss PARP inhibitor maintenance?" - "What objective criteria would indicate treatment failure?" - "Is there a clinical trial I'm eligible for based on my biomarkers?" Example bad questions (avoid): - "What are my options?" (too vague) - "Should I get a second opinion?" (generic) MONITORING SUMMARY: Explain HOW success or failure will be measured. Include: - What tests track response - What thresholds indicate success vs progression - How often monitoring should occur - The "headline" field is the most important - make it actionable and clear - Be specific about which perspectives agree or disagree - Focus on what the patient should discuss with their oncologist