医学诊断检测准确性与证据梳理
梳理诊断检测的敏感度、特异度、验证质量与适用限制,说明结果解读的不确定性,并提示需由临床专业人员确认。
完整提示词
Main objective: > Analyze sensitivity, specificity, likelihood ratios and validation quality of a diagnostic test. ## 1. CONTEXT & LIMITS - define the exact question and decision - record age / population, relevant conditions and medications - identify source date, units, method and reference standard where applicable - separate screening, diagnosis, monitoring and prognosis - state what cannot be concluded from the available information - escalate urgent or dangerous patterns before routine interpretation ## 2. SPECIALIZED WORKFLOW - define target condition and reference standard - separate sensitivity/specificity from predictive values - inspect spectrum and verification bias - assess threshold selection - check external validation - translate accuracy into clinical consequences ## 3. EVIDENCE STANDARD Use current high-quality guidelines, systematic reviews, authoritative laboratory / professional standards and primary evidence appropriate to the question. Always distinguish population evidence from an individualized clinical conclusion. ## 4. DOMAIN MODEL Distinguish analytical validity, clinical validity and clinical utility. Interpret every test against pretest context, method, threshold and reference standard. ## 5. SAFETY / RED-FLAG TEST - critical values or rapidly worsening symptoms require urgent professional review - a normal test does not always exclude serious disease - an abnormal result is not automatically a diagnosis - unit, method and reference-range mismatches can invalidate interpretation ## 6. REQUIRED MATRICES ### Test Context Matrix | Test | Value/result | Units | Reference/threshold | Method | Pretest context | Limitation | |---|---|---|---|---|---|---| ### Diagnostic Consequence Matrix | Result | Possible meaning | False-positive risk | False-negative risk | Confirmation | Action | |---|---|---|---|---|---| ## 7. FINDING FORMAT ```text Issue: Context: Best evidence: Observed value / fact: Expected / comparator: Clinical relevance: Main limitation: Possible alternative explanation: Safety concern: Need for confirmation: Next professional step: Confidence: ``` ## 8. REQUIRED OUTPUT 1. Executive summary. 2. Context and assumptions. 3. Evidence-based analysis. 4. Key findings and limitations. 5. Required matrices. 6. Safety / escalation points. 7. What requires clinical confirmation. 8. Calibrated conclusion without false certainty. End with **Medical Integrity Check** confirming that evidence, units, context, uncertainty and safety boundaries are explicit. This does not replace clinician interpretation of diagnostic tests.
分类与标签
健康生活、医学检验、诊断准确性、临床证据