📚 國考版(醫師國考 / PGY OSCE)
對象:M6 / PGY 國考前。本章 high yield:USPSTF 建議、lead/length bias、cost-effective threshold。
📌 一頁重點整理 (Cram Sheet)
🔥 高 yield 7 條
- WHO 篩檢原則:important + treatable + latent + acceptable test + cost-effective
- Lead time bias:survival 看似長但 mortality 沒變 → 用 mortality 不用 survival
- Length time bias:篩檢偏向 indolent disease
- USPSTF A/B 強推薦;D 不要做;I 不確定
- Cost-effective ≤ $50,000-100,000/QALY
- Overdiagnosis 是真實 harm(mammo 15-40%、PSA 15-37%)
- Lung CA low-dose CT:50-80、≥ 20 PY、目前/15y 戒;mortality ↓ 20%
🔢 必背數字
- WHO 篩檢原則:1968 提出
- Cost-effective threshold:$50,000-100,000 per QALY
- Mammography overdiagnosis:15-40%
- PSA overdiagnosis:15-37%
- Lung CA mortality ↓ via LDCT:20%
- Mammography RR reduction:14-32%
- FOBT colon CA RR reduction:15-30%
- Sigmoidoscopy colon CA RR reduction:40-60%
- Colonoscopy colon CA RR reduction:50-70%
- DEXA NNS:731 women aged 65-69 to prevent 1 hip fx
- BRCA1/2 breast CA risk increase:5-20×
⚠️ 易錯陷阱
- Survival = mortality(錯,差距是 lead time)
- USPSTF C 級 = 不該做(錯,是 individual offer)
- Overdiagnosis 不算 harm(錯,是真 harm)
- 篩檢 always cost-effective(錯,要看 NNS + threshold)
⭐ 高 yield 摘要
USPSTF 等級
- A:強推薦 + 強證據
- B:推薦
- C:individual offer
- D:don’t do
- I:insufficient evidence
4 個 Bias 要避免
- Lead time bias:解法用 mortality
- Length time bias:解法用 mortality
- Selection bias:解法 RCT
- Volunteer bias:解法 RCT + ITT
USPSTF Cancer Screening 必背
| Breast |
Mammography(無 CBE) |
F 50-75(40 個別) |
q2y |
| Cervix |
Pap (21-65) / Pap+HPV (30-65) |
F |
q3y / q5y |
| Colorectal |
FOBT/FIT-DNA/sig/colo |
45-75 |
q1/3/5/10y |
| Lung |
Low-dose CT |
50-80 + ≥20 PY + 目前/15y 戒 |
yearly |
USPSTF Infectious Screening
| HCV |
Anti-HCV + PCR |
18-79(once) |
| HIV |
Immunoassay + confirm |
15-65(at least once) |
| Chlamydia/GC |
NAAT |
F < 25 sexually active |
Chemoprevention 必記
| Aspirin |
40-59 + ASCVD ≥ 10% + 低 bleed risk |
| Folic acid |
Childbearing F |
| Tamoxifen / Raloxifene |
高 breast CA risk F |
| Vitamin D |
> 64 fall risk |
| Statin |
ASCVD ≥ 10% in 40-75 yo |
Lung Cancer Screening 細節(NLST trial)
- 對象:50-80 yo + ≥ 20 pack-year + 目前 OR 15 yr 內戒
- 工具:annual low-dose chest CT
- 結果:lung CA mortality ↓ 20%(NLST 2011;Dutch-Belgian NELSON 2020 confirm)
- False positive:24%(needs follow-up CT)
- Overdiagnosis:~18%
Colon Cancer Screening 比較
| FOBT / FIT |
yearly |
15-30% |
便宜、需 yearly |
| FIT-DNA (Cologuard) |
q1-3y |
中 |
較敏感 |
| Sigmoidoscopy |
q5y |
40-60% |
不需 sedation |
| Colonoscopy |
q10y |
50-70% |
gold standard、incurs sedation/risk |
Breast Cancer Screening 爭議
- 40-49 yo:USPSTF B 級(2024 update 改 → 之前是 C)
- 50-74 yo:A 級
- 75+:證據不足
- USPSTF:每 2 年 vs ACS:每年(30s+)
- 平衡:absolute benefit 1.2/1000 over 12 yr 但 false-positive 50% 累積率
何時停止篩檢
- 一般:age 75 後證據減少
- 個別考量:comorbidity、life expectancy < 5-10 yr
- 病人 preferences
🏆 易混淆對照
Screening vs Case-finding vs Surveillance
| Screening |
完全無症狀人群 |
| Case-finding |
來看其他事的病人主動問 |
| Surveillance |
已知 high-risk 個體(如 BRCA carrier 每年 mammo+MRI) |
Sensitivity/Specificity vs Population
- 同一 test 在不同 prevalence 族群 PPV 大不同
- 篩檢族群通常 prevalence 低 → PPV 低 → 多 false positive
- 解法:高 specificity test + 確診 test 階段
Lead time vs Length time
- Lead = 診斷提早
- Length = 慢病被選
🔢 必背數字總表
- WHO 1968 提出 screening principles
- USPSTF 5 個 grade(A/B/C/D/I)
- Cost-effective threshold:$50K-100K/QALY
- Mammography overdiagnosis:15-40%
- PSA overdiagnosis:15-37%
- Lung CA RR ↓:20%(NLST)
- Mammography RR ↓:14-32%
- DEXA NNS:731(65-69 yo F)
- Breast CA lifetime risk F:10%
- Colon CA lifetime risk:6%
- BRCA1/2 5-20× risk
- Hip fracture lifetime risk:white F 16%
- Dom violence lifetime F:up to 15%
📝 過去考題類型
必考
- USPSTF 篩檢年齡 + 頻率(mammo、Pap、colon、lung CT、AAA、osteoporosis)
- Lead time / Length time bias 概念
- USPSTF A/B/D grade 應用
偶爾考
- Cost-effectiveness threshold
- Number needed to screen 計算
- Overdiagnosis 概念
陷阱題
- Survival 改善 = 篩檢有效(錯,可能 lead time bias)
- USPSTF C = 不該做(錯)
- Pap smear 可預防所有 cervical CA(錯,仍有偽陰性)
- Mammography 每年(錯,USPSTF 是 q2y)
🎯 自我檢測
- Q: USPSTF lung CA screening 條件?
A: 50-80 yo + ≥ 20 PY + 目前 OR 15 yr 內戒
- Q: 男性 65-75 ever smoker 一次性篩檢?
A: AAA US
- Q: USPSTF mammography 推薦?
A: 50-74 yo q2y(40-49 個別考慮,B 級 2024)
- Q: Cervical CA Pap+HPV co-test 對象?
A: 30-65 yo F q5y
- Q: Colonoscopy USPSTF 起始年齡?
A: 45 yo(2021 update 從 50 → 45)
- Q: Lead time bias 解法?
A: 用 mortality 不用 survival
- Q: Length time bias 是什麼?
A: 篩檢偏向偵測 indolent disease
- Q: Cost-effective threshold?
A: $50K-100K per QALY
- Q: Mammography 偵測癌中 overdiagnosis 比例?
A: 15-40%
- Q: BRCA1/2 增加 breast CA risk 多少倍?
A: 5-20×
📚 想深入請看
⚠️ AI 草稿,未經盧醫師驗證。