"回归均值"为什么会让人误以为某种疗法有效？
How does regression to the mean make treatments look effective?
很多症状会自然波动，人们通常在最难受的时候寻求治疗，而极端值之后往往会回到平均水平——不论有没有干预。于是"用了就好转"的时间关联被误读为因果。感冒、腰痛、血压单次测量偏高都是典型例子。这正是需要对照组的原因：只有比较才能知道好转是否超过自然恢复。
Symptoms fluctuate, and people usually seek help at their worst; extreme values tend to drift back toward average afterwards, with or without intervention. The timing coincidence then reads as cause. Colds, back pain and a single high blood-pressure reading are classic cases. This is exactly why control groups matter: only comparison shows whether improvement exceeds natural recovery.
难度 专业 · 适合 高中(15-18)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#回归均值 #自然恢复
随机分组为什么这么关键？
Why is randomisation so crucial?
随机分配让已知与未知的混杂因素（年龄、病情轻重、生活习惯、依从性倾向）在两组间大致均衡，包括研究者根本没想到的因素。如果由医生挑选谁用新药，病情较轻者更可能被选入，结果会系统性高估疗效。这是随机对照试验在证据等级中地位较高的核心原因。
Random allocation balances known and unknown confounders — age, severity, lifestyle, adherence tendencies — across groups, including factors researchers never thought of. If clinicians chose who receives the new drug, healthier patients would tend to be selected and the effect would be systematically overestimated. This is the core reason randomised controlled trials rank high in evidence hierarchies.
难度 专业 · 适合 高中(15-18)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#随机化 #混杂因素
新药上市前的I、II、III期临床试验分别做什么？
What do phase I, II and III clinical trials each test?
I期在少数健康志愿者或患者中评估安全性、耐受剂量与体内代谢；II期在数十至数百名目标患者中初步看疗效并优化剂量；III期在数百至数千人中与现有标准治疗或安慰剂比较，确认疗效与常见不良反应。上市后的IV期继续监测罕见或长期风险，因为再大的试验也难以发现万分之一的事件。
Phase I tests safety, tolerable dose and metabolism in a small group of healthy volunteers or patients; phase II looks for preliminary efficacy and optimal dose in tens to hundreds of target patients; phase III compares the treatment with standard care or placebo in hundreds to thousands to confirm benefit and common adverse effects. Post-marketing phase IV keeps watching for rare or long-term risks, since no trial is large enough to catch one-in-ten-thousand events.
难度 专业 · 适合 高中(15-18)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#临床试验分期 #药物监管
什么是"反安慰剂效应"？
What is the nocebo effect?
指负面预期导致的真实不适：读过副作用清单的人更容易报告头痛、恶心或乏力，即使服用的是无活性对照。试验中安慰剂组的"副作用"报告率往往不低，这正是反安慰剂的体现。它提醒医患沟通的措辞会影响体验，也说明并非所有服药后的不适都源自药物本身；但真实且严重的反应仍需认真评估。
It is real discomfort generated by negative expectation: people who have read a side-effect list report more headache, nausea or fatigue even on inert controls. Placebo arms in trials often report substantial "side effects" — that is nocebo at work. It shows how wording in clinical communication shapes experience, and that not every symptom after taking a pill comes from the drug; genuine severe reactions still need careful assessment.
难度 专业 · 适合 高中(15-18)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#反安慰剂 #负面预期
证据金字塔里，系统综述为什么排在上层？
Why do systematic reviews sit near the top of the evidence pyramid?
单项研究可能因样本小、人群特殊或偶然波动得出误导性结论。系统综述按预设标准全面检索并评价所有相关研究，Meta分析再把结果定量合并，提高精确度并显示结论是否一致。但它的可靠性受纳入研究质量限制——"垃圾进，垃圾出"，发表偏倚也会让阳性结果被高估。
A single study can mislead through small samples, unusual populations or chance. A systematic review searches and appraises all relevant studies against pre-set criteria, and meta-analysis pools results quantitatively, improving precision and revealing whether findings agree. Its reliability is capped by the included studies — garbage in, garbage out — and publication bias can inflate positive results.
难度 专业 · 适合 高中(15-18)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#系统综述 #Meta分析 #发表偏倚
"知情同意"这个原则是怎么来的？
Where did the principle of informed consent come from?
它源于对研究伦理灾难的反思：二战期间的人体实验催生了1947年《纽伦堡法典》，强调自愿参与；1964年《赫尔辛基宣言》确立了研究伦理规范；美国塔斯基吉梅毒研究长期隐瞒治疗的丑闻推动了1979年《贝尔蒙报告》与伦理审查制度。今天参与研究或接受重要治疗前，都应被充分告知目的、风险与替代选择，并可随时退出。
It grew from reckoning with research disasters: wartime human experiments led to the 1947 Nuremberg Code emphasising voluntary participation; the 1964 Declaration of Helsinki codified research ethics; and the Tuskegee syphilis study, which withheld treatment for decades, prompted the 1979 Belmont Report and formal ethics review. Today, before joining research or major treatment, people must be told the purpose, risks and alternatives, and may withdraw at any time.
难度 进阶 · 适合 初中(12-15)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
世界医学协会（WMA）《赫尔辛基宣言》；美国《贝尔蒙特报告》（1979）
World Medical Association (WMA). Declaration of Helsinki; The Belmont Report (1979).
独立核验依据 · Independent verification: 核验依据 1 OPENLIBRARY 标识 /works/OL19667349W：The Belmont report（首版
#知情同意 #研究伦理
健康新闻里"相关不等于因果"能举个例子吗？
Can you give an example of "correlation isn't causation" in health news?
早年研究发现喝咖啡的人肺癌较多，后来发现原因是当时咖啡爱好者更常吸烟，吸烟是真正的原因（混杂因素）。类似地，"吃早餐的人更瘦"也可能因为规律吃早餐的人整体生活方式更健康。判断因果需要控制混杂、时间顺序、剂量反应关系，最好有随机试验支持。
Early studies found more lung cancer among coffee drinkers; it later turned out coffee lovers smoked more, and smoking was the real cause — a confounder. Likewise "breakfast eaters are slimmer" may reflect that regular breakfasters have healthier lifestyles overall. Establishing causation needs control of confounders, correct time order, dose–response, and ideally randomised evidence.
难度 进阶 · 适合 初中(12-15)
CC BY 4.0 · 来源待核验 · Source pending review
来源 · Source
出处备注：著录仅给书名，无版次与年份；同一书名常对应多个版次，不臆测具体版本。
《Gordis流行病学》
Gordis Epidemiology
#相关与因果 #混杂
安慰剂效应到底是什么？
What exactly is the placebo effect?
指没有活性成分的治疗因期待、仪式感和医患互动而带来的真实主观改善，甚至伴随内源性阿片肽、多巴胺释放等可测的生理变化。它对疼痛、恶心、疲劳等主观症状影响明显，但不能缩小肿瘤或清除细菌。这也是为什么新疗法必须与安慰剂对照比较，才能分清哪部分是药物本身的作用。
It is the genuine subjective improvement produced by an inert treatment through expectation, ritual and the clinician–patient interaction, sometimes with measurable physiology such as endogenous opioid and dopamine release. It notably affects subjective symptoms like pain, nausea and fatigue, but it does not shrink tumours or clear bacteria. That is why new treatments must be compared against placebo to separate the drug's own effect.
难度 进阶 · 适合 大众
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#安慰剂 #期待效应
临床试验为什么要"双盲"？
Why do clinical trials use double blinding?
如果患者知道自己用的是新药，期待会放大主观改善；如果评估者知道分组，也可能在测量与记录时无意偏向。双盲让受试者和研究者都不知道分组，使两组的心理预期与评估标准尽量一致，从而把药物的真实效果分离出来。涉及手术或明显外观差异的干预难以双盲，这时可对结局评估者设盲。
If patients know they got the new drug, expectation inflates subjective improvement; if assessors know the allocation, they may unconsciously skew measurement and recording. Double blinding keeps both participants and researchers unaware, equalising expectation and assessment standards so the drug's real effect stands out. Surgery or visibly different interventions are hard to double-blind, in which case outcome assessors are blinded instead.
难度 进阶 · 适合 初中(12-15)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#双盲 #偏倚控制
为什么医生有时说"再观察几天"？
Why do doctors sometimes say "let's watch it for a few days"?
因为许多常见病（如病毒性感冒、轻度扭伤）会自行好转，而症状在早期往往还不足以区分不同病因。观察等待可以避免不必要的检查与用药风险，同时医生会告诉你哪些"警示信号"出现时需立即回诊。这是一种有依据的策略，不等于不管。
Many common problems — viral colds, mild sprains — resolve on their own, and early symptoms often cannot yet distinguish between causes. Watchful waiting avoids unnecessary tests and medication risks, and the clinician should tell you which warning signs mean coming back immediately. It is an evidence-based strategy, not neglect.
难度 入门 · 适合 小学高段(9-12)
CC BY 4.0 · 已核验 · Verified 2026-11-03
来源 · Source
葛均波 等主编,《内科学》第9版,人民卫生出版社,2018
Ge Junbo et al. eds., Internal Medicine, 9th ed., PMPH, 2018
独立核验依据 · Independent verification: 核验依据 1 ISBN 9787117265416（豆瓣书目条目: 内科学）
#观察等待 #临床决策