突发公共卫生事件的病因假设和病因推断施国庆.ppt

Click to edit Master title style,Click to edit Master text styles,Second level,Third level,Fourth level,Fifth level,Click to edit Master title style,Click to edit Master text styles,Second level,Third level,Fourth level,Fifth level,Click to edit Master title style,Click to edit Master text styles,Second level,Third level,Fourth level,Fifth level,*,突发公共卫生事件的病因假设和病因推断,中国疾病预防控制中心,中国现场流行病学培训项目（,CFETP,）,施国庆,提出假设是现场调查承上启下的关键,准备现场,Prepare for field work,证实暴发,Establish existence of an outbreak,Time,Verify the diagnosis,核实诊断,Define,count,and interview cases,定义，计数，采访病例,描述疾病和人、地点和时间,Characterize person,place and time,Develop a hypothesis,提出假设,Develop analytic study to test hypothesis,检验假设,Reconsider,refine,retest hypothesis,再推敲，修正和再检验假说,实施控制措施,Implement Control,将调查结果进行交流,Communicate Findings,认为一个假设是正确的信念、信服度或经验的多少与假设是否正确无关,The intensity of the belief,conviction or experience that a hypothesis is,true,has no bearing on whether it is true or not.,认为一个假设是错误的信念、信服度或经验的多少与假设是否错误无关,The intensity of the belief,conviction or experience that a hypothesis is false has no bearing on whether it is false or not.,假设的质量取决于数据的宽度和准确性,The quality of the hypothesis depends upon the breadth and accuracy of the data,调查的成功取决于假设的质量,The success of the investigation depends upon the quality of the hypotheses,假设定义,假设是从事实、数据和信息中产生的可,检验,的推断,A hypothesis is a testable proposition developed from facts,data,and information,H,0,和,H,1,需要明确,如，,H0,：病例组暴露于某因素的,Odds,与对照组没有差别,H1：,病例组暴露于某因素的,Odds,高于对照组,如何检验下面的假设？,本次暴发可能是由于水、食物或人传人导致的,例,1：,某病例对照研究结果,饮食因素,集体聚餐史,吃生虾吃生蟹吃生钉螺吃生花甲,生吃瓜菜,吃冰冻八宝粥,吃辣条,接触因素,病例接触史,饭前便后洗手,饮水因素,水源类型,喝生水,喝冬瓜水,喝盐点饮料,OR=3.15,95%CI=1.1-9.3,其他,厕所类型,甲肝疫苗史,外出史,零花钱,去小卖部,完美假设应该包括哪些内容,?,人群,病因和协同因素,预期结果,剂量反应关系,时间关系,孙军玲，等,.,一起伤寒暴发的流行曲线,14 days,3days,Group,No.,Case,AR%,Students,1041,86,8.3,Teachers and staffs,91,00,Total,1132,86,7.6,AR between students and teachers was of statistically difference,X,2,=8.1,，,P=0.004,Group,No.,Case,AR%,Resident students,1000,86,8.6,Non-resident students,41,0,0,Total,1041,86,8.3,8.3%students developed illness,X,2,=2.793,，,P=0.095,Suspected exposure interview,Population,No.,Dining Place:School canteen,Drinking unboiled water,Non-resident students,40,13%,25%,Resident students,74,100%,100%,Teachers lived in school,68,65%,13%,-temporary kitchen worker,Outlier case interview,Onset,8,21,27,Diagnosed,Only one suspected exposure,：,School unboiled water,JUNE,13 days,形成的假设是？,帮助形成假设的信息类别,疾病,Disease,时间,Time,地点,Place,人群,Person,类推,Analogy,暴露,Exposures,+,+,+,+,+,病例定义、搜索和调查,合理的分析,认真的访谈,文献的使用,霍乱的一些信息,Facts about cholera,必须有高感染剂量,1,000,000,Requires high infectious dose 1,000,000,在水中自由生存,Free-living in water,耐盐水环境,Tolerates salt,能消化甲壳质,Can digest chitin,对酸高度敏感,Highly sensitive to acid,在食物中繁殖,Multiplies in food,以前有食物源或水源导致的暴发,Previous outbreaks from food or water,诺如病毒Norovirus,可导致胃和或肠道炎症 acute gastroenteritis,症状好转病人粪便可能仍有病毒The virus can stay in your stool for 2 weeks or more after you feel better,拥挤环境可快速人传人Norovirus can spread quickly from person to person in crowded,closed places,Vomiting(95%)was major symptom,65 cases were identified with attack rate of 3.6%,(65/1789),No severe case,No hospitalization,43(66%)cases were treated in outpatient,Sym.&Sig.,%,(N=65),Vomiting,95,（,57,）,Nausea,83,（,50,）,Abdominal pain,83,（,50,）,Diarrhea(,3/d),27,（,16,）,Fever(,37.5),22,（,14,）,Liangliang Cui:Investigation of a gastroenteritis outbreak caused by Norovirus II in a primary school,After index case occurred,the neighboring students developed illness gradually,then affected other buildings,发现非预期表现,：,247,名确诊轮状病毒感染病例的临床特征,Symptom/finding,腹泻,Diarrhea,95%,呕吐,Vomiting,72%,胃口差,Poor feeding,65%,发烧,Fever 37.5 C,21%,流涕,Runny nose,37%,喷嚏,Sneezing,30%,咳嗽,Cough,27%,皮疹,Rash,8%,中位病程,Median Duration,5 days,247,名腹泻门诊就诊的轮状病毒感染病例的临床特征,Characteristics of illness among 247 patients with confirmed rotavirus infection by clinic visit for diarrhea,门诊,Symptom/finding,1,2+,Diarrhea,93%,96%,Vomiting,75%,69%,Poor feeding,63%,69%,Fever 37.5 C,15%,48%,Runny nose,7%,54%,Sneezing,9%,60%,Cough,3%,47%,Rash,0%,21%,Median Duration,4 days,7 days,Co-infection,0%,13%,46,例疑似炭疽临床表现,确诊,n=18,疑似,n=28,No,%,No,%,腹泻,12,67,26,93,发热,8,44,2,7,头痛,7,39,1,4,腹痛,7,39,2,7,咽痛,5,28,3,11,恶心,3,17,1,4,呕吐,3,17,1,4,水肿,3,17,0,0,水疱,2,11,0,0,斑疹,1,6,0,0,皮肤破损,1,1,0,0,临床表现帮助形成假设,247,例病例的临床表现,症状,Symptom/finding,发烧,Fever 37.5C,78%,手部水泡,Hand vesicles,57%,口腔水泡,Mouth vesicles,46%,其他部位水泡,Other vesicles,21%,手足口,HFMD,89%,泡疹性咽峡炎,Herpangina,22%,咳嗽,Cough,10%,喷嚏,Sneezing,8%,病程中位数,Median Duration,5 days,传播途径的假设？,使用科学,/,公共卫生文献,Use the scientific/public health literature,掌握病原和疾病特点,病原和宿主,感染剂量,传播方式,潜隐期,潜伏期,症状和体征,致病机理,病程,治疗效果,时间模式,时间只有一个方向,注意事项：,12,项（暂定）,1、,危险因素一定在疾病发生之前,Time,时间,Onset,发病,Exposure,暴露,Normal,正常状态,Disease,疾病状态,难以判定：尿肌酐与高血压关系？,2、,使用新发病而不是患病病例研究,3、,假设的危险时间应当符合生物学特性,411,例副伤寒的爆发,stop supplying boiled water,停止供应开水,mass antibiotic&screening for fever,抗生素普服和筛查发热病人,Chlorination,水中加氯,Investigation,调查,0,20,40,60,80,1,5,10,15,20,25,30,5,10,15,20,25,November11,月,December12,月,cases,病例,0,30,60,90,120,Rainfall,降雨量,Case,病例,Rainfall,降雨,31,可能暴露时间,已知病原，可根据疾病潜伏期推测可能暴露时间,最长潜伏期,最短潜伏期,平均潜伏期,圆孢子虫病疾病发病时间，休斯顿商业会议，,1996,年,5,月,病原未知可采用,WHO,推荐方法估计可能暴露时间,发病时间的中位值,首末发病时间间隔平均潜伏期,资料来源：,WHO,食品安全事故调查和控制指南,前后各保留至少一个餐次,甲肝人传人模式？,4、,出现双（多）峰模式提示多种机制,（潜伏期的相对流行曲线）,潜伏期,病例数,提示艾博拉病毒出血热的两种传播方式,Water chlorinated on Monday only,Election banquets,Water chlorinated daily,Water main broken,Water main repaired,Week ending on,July,September,October,November,June,August,Jaundice cases,Data Source:cases by week of onset,June,October 1999.Jaf,r,Jordan,点源二代病例,Secondary cases after a point source,Epi-curve suggesting point source in early outbreak,21 days,Point source,37,5、,病程也能帮助提示和排除假设,聚集死亡点源暴露模式,SUD cases,1984-2004,年,60%,家庭续发病例在,24,小时内死亡,2005,年,92%,家庭续发病例在,24,小时内死亡,days,不支持传染性疾病,6、,感染场所判定的时间合理性,院内感染,在最短潜伏期内住院,潜伏期期间在医院,围产期利斯特菌病有,2,种不同情况,早发型,早发型经胎盘传播,宫内感染可导致死产、早产和新生儿毒血症,绝大多数在出生后,24,小时发病,聚集性早发型是社区食源型爆发表现,迟发型,出生时无发病，出生后数天至数周发病,可能是医院暴露或家庭暴露造成,聚集性迟发型病例通常是医院感染,潜伏期不清楚，很可能超过,2,天,美国正常水平,12/100,，,000,。本次,6,月率是,1700/100,，,000,，是美国基线的,142,倍。,提示：院内感染,7、,事件发生时间的合理性,甲氨蝶呤和阿糖胞苷的问题批号分布在硫酸长春生产之后,生产,MTX,生产,Ara-C,未生产,2007,年,3,月,2007,年,4,月,从,1991,年,12,月,7,日开始的天数,患其他疾病入院,:,出院后感染疟疾,疟疾患者,治疗前感染,8、,发病时间和暴露地点关联性,房间,6,房间,5,房间,4,房间,3,治疗期间感染,9、使用拆分时间与空间、人群联系起来,三个高发村寨发病,高峰及持续时间一致,散童、学生及农民的发病,高峰及持续时间亦基本一致,图,2,三个高发村寨病例发病时间曲线图,图,3,不同职业发病时间曲线图,某医院报告的疟疾本地病例与输入病例流行模式有什么不同,Week,周,Temperature C温度,Cases,Imported,输入,Locally acquired,本地,温度,10、,疾病变化与暴露因素的时间性,一名,18,岁女性重症呼吸道感染的变化,Score,Hours after symptom onset,State,Age,1,Sex,Days,2,Dose,New York,0,2,M,0,3,1,California,0,3,M,0,3,1,Pennsylvania,0,6,M,0,3,1,Pennsylvania,0,2,M,0,4,1,Colorado,0,4,F,0,4,1,California,0,7,M,0,4,2,Kansas,0,2,F,0,5,1,Colorado,0,3,M,0,5,1,New York,0,3,F,0,5,1,N.Carolina,0,4,F,0,5,1,Missouri,11,M,0,5,1,Pennsylvania,0,3,F,0,7,1,California,0,4,F,14,2,Pennsylvania,0,2,M,29,1,California,0,5,M,59,1,7 states,Med,4,60%,M,Med,5,87%,1,美国,1998,1999,年各州四价恒河猴轮状病毒疫苗接种者中肠套叠病例报告,Reported cases of intussusception among,recipients of tetravalent rhesus-based,rotavirus vaccine(RRV-TV)Rotashield),by state United States,1998-1999*,一览表是理解流行病学数据的一个简单组织工具,A line listing is a simple organizational tool for understanding epidemiological data,Najaran,溶血性尿毒症综合症,Hemolytic Uremic Syndrome in Najaran,疾病信息,The disease,毒素由,型痢疾志贺菌引起,Toxin from Shigella dysenteriae type 1,耐氨苄西林,Amplicillin,cotrimoxizole resistance,急性肾衰竭,Acute renal failure,社区内有,800,名痢疾病例,800 cases dysentery in community,40,个痢疾病例住院治疗,40 cases dysentery admitted to hospital,痢疾病人中有,10,例溶血性尿毒症综合症,10 cases of HUS in dysentery patients,10,例溶血性尿毒症综合症病例列表,年龄,Age,(yr),性别,Sex,村庄Village,腹泻发病日,Onset,Diarrhea,痢疾发病日,Onset,Dysentery,HUS,发病日,Onset,HUS,住院日,Date in,Hospital,10,F,Bashash,20/03,23/03,24/03,22/03,4,F,Basash,26/03,29/03,31/03,29/03,3,M,Bashash,03/04,03/04,08/04,03/04,3,M,Bashash,11/04,12/04,15/04,12/04,4,F,Bashash,12/04,14/04,17/04,15/04,1,M,Bashash,17/04,18/04,21/0419/04,6,F,Najaran,18/04,20/04,25/04,21/04,2,F,Jerash,22/04,25/04,02/05,28/04,5,M,Bashash,30/04,01/05,07/05,04/05,1,M,Abu Saba,08/05,09/05,14/05,11/05,10,例溶血性尿毒症综合症病例列表,年龄,Age,(yr),性别,Sex,村庄,Village,腹泻发病日,Onset,Diarrhea,痢疾发病日,Onset,Dysentery,住院日,Date in,Hospital,HUS,发病日,Onset,HUS,住院至,HUS,天数,10,F,Bashash,20/03,23/03,22/03,24/03,2,4,F,Basash,26/03,29/03,29/03,31/03,2,3,M,Bashash,03/04,03/04,03/04,08/04,5,3,M,Bashash,11/04,12/04,12/04,15/04,3,4,F,Bashash,12/04,14/04,15/04,17/04,2,1,M,Bashash,17/04,18/04,19/04,21/04,2,6,F,Najaran,18/04,20/04,21/04,25/04,4,2,F,Jerash,22/04,25/04,28/04,02/05,4,5,M,Bashash,30/04,01/05,04/05,07/05,3,1,M,Abu Saba,08/05,09/05,11/05,14/05,3,住院后,2,5,天发生溶血性尿毒症综合症,2-5 day delay of HUS after hospital admission,年龄,Age,(yr),性别,Sex,村庄,Village,由腹泻到,HUS,Diarrhea,To HUS,由痢疾到,HUS,Dysentery to HUS,从住院到,HUS,Hospital to HUS,1,M,Bashash,4,6,2,4,F,Bashash,5,2,2,4,F,Bashash,5,3,2,10,F,Bashash,4,1,2,1,M,Abu Saba,6,5,3,3,M,Bashash,4,3,3,5,M,Bashash,8,7,3,6,F,Najaran,7,5,4,2,F,Jerash,10,3,4,3,M,Bashash,5,5,5,溶血性尿毒症综合症的假设是什么？,What is the hypothesis for HUS,用以治疗痢疾或腹泻的某种抗生素（痢疾志贺菌对其耐药）导致,Najaran,儿童患溶血性尿毒症综合症,Treatment of dysentery or diarrhea with specific antibiotics to which,Shigella dysenteriae,was resistant caused HUS in children of Najaran.,11、,流行期和非流行期拆解,1983-1985,马里兰州各季度,PGH,重症监护病房心跳停止病人,Cardiac arrests,by quarter and year,PGH ICU,Maryland,1983-1985,1983,1984,1985,Number of Cardiac Arrests,季度和年度,Quarter and Year,0,20,30,40,10,心跳停止病例数,1983,1984,1985,0,15,10,5,25,20,季度和年度Quarter and Year,Number of Cardiac Arrests,1983-1985,马里兰州各季度各班次,PGH,重症监护病房心跳停止病人数,Cardiac arrests,by shift,PGH ICU,Maryland,1983-1985,夜班,Night shift,晚班,Evening shift,白班,Day shift,心跳停止病例数,00-03,04-07,08-11,12-15,16-19,20-23,0,20,40,60,80,每日各时段（军队时间）,Hour of Day(Military time,),Number of Cardiac Arrests,1983-1985,马里兰州每日各时段,PGH,重症监护病房心跳停止病人数,Cardiac arrests,by time and period,PGH ICU,Maryland,1983-1985,流行前,Pre-epidemic,流行中,Epidemic,流行后,Post-epidemic,心跳停止病例数,不同护士管理病人的心跳骤停的危险,Relative risk of cardiac arrest by nurse,护士,Nurse,1983-85,Epidemic,Epidemic PM,14,3.9,5.0,11.2,13,1.9,1.6,2.4,196,1.4,0.7,50,106,1.3,1.7,2.8,118,1.0,0.6,5.0,30,0.7,0.4,0.7,129,0.7,0.9,1.7,心跳骤停的重复发生,Recurrent cardiac arrests,病人,Pt,日期,Date,护士,Nur,病人,Pt,日期,Date,护士,Nur,A,1983/07/10,26,D,1984/05/17,14,A,1983/08/04,118,D,1084/06/05,14,B,1984/01/02,14,E,09/01,113,B,1984/01/02,14,E,09/03,14,B,1984/01/05,14,F,10/02,14,C,1995/03/15,154,F,10/11,14,C,1995/03/19,33,F,10/12,14,12,、发病时间可与其他地理等要素结合分析,霍乱与海拔关联分析,家庭聚死猝死发病时间与村庄海拔有关联,Number of cases,Rainfall in the week (mm),Week of hospitalization,Mar|Apr|May|Jun|Jul|Aug|Sep|Oct|Nov,环境因素传播模式,Environmentally spread disease,Data Source:Cases of leptospirosis and rainfall index by weekSalvador,Brazil,March 10,November 2,1996.,环境因素传播模式,Environmentally spread disease,20,天,25,天,25,天,25,天,25,天,海岸贝类养殖整顿,数据来源：,CFETP5,期,2006,年某沿海城市副伤寒疫情调查,地区模式,为什么在这个地方而不是其他地方发生？,在地区上，有什么因素接近侵袭人群？,与未侵袭人群相比，没有或不密切的因素？,地区分布应注意与危险因素结合起来分析,水源名称,供应范围,供应人数,病例数,罹患率,(%),镇自来水,A,村,10,个组,2976,93,3.1,镇自来水,B,村,8,个组,3282,102,3.1,镇自来水,C,村,4,个组,1152,52,4.5,使用镇自来水小计,7410,247,3.3,山泉水,B,村,2,个组,340,6,1.8,自备井水,C,村,6,个组,1168,3,0.26,不使用镇自来水小计,1508,9,0.60,中国现场流行病学培训项目现场报告,2011,年第,21,期）,2011,年,4-5,月贵州省,X,镇菌痢暴发高发村寨水源分布及发病情况,这三个村位于镇自来水供应范围内,A,B,C,地图含有多种可能与疫情有关的要素时，如供水网，水井位置，厕所位置，河流等,示意图同样也可显示疾病主要传播特点,非洲一起出血热疫情的不同村庄发病风险图,病因的分布与疾病的分布相似,:,线状分布,线状分布,N,风向,Wind direction,+,卫生所,Health Center,3,按病例的发病时间排序,Cases ordered by onset,?,无明确发病时间的病例,Cases with unknown onset,1,首发病例,First case,*,J Heukelbach et.al.Trop Med Internat,Health,6:635-42,2001,2005,年四川农民猪链球菌感染,Streptococcus suis,infection in Sichuan farmers,2005,6.24 7.4,7.5 7.15,100,公里,100,公里,2005,年四川农民猪链球菌感染,Streptococcus suis,infection in Sichuan farmers,2005,7.16 7.26,7.27 8.8,100,公里,100,公里,实例：炭疽,11.42 km,7.21 km,动物病例家庭,早期,3,例死亡动物地点,无动物病例家庭,人炭疽病例,簇状分布,What factors are close in space to the affected persons,住处,Residence,商业,Business,某地不明原因猝死,散状分布：疾病已知，但分布广泛的爆发,寻找可能的共同暴露一起涉及多州的沙门菌爆发,疫情发现,2006,年,11,月至,12,月,A,州报告分离,2,例,Tennessee,型沙门菌株,B,州报告分离,5,例,CDC,通过,PulseNet,发现同,PFGE,模式：,JNXX01.0011,病例定义,美国居民，,2006,年,8,月,1,日及以后发病，感染同样,PFGE,型的田纳西沙门菌,食物史访谈与比较,暴露,例数,%,香蕉,24,78,蛋,22,71,医疗机构,5,16,奶,22,72,花生酱,26,85,番茄,26,83,火鸡,15,48,其他,略,略,31,例病人食物访谈结果（,7,天）,获得全国人群食用史比例进行比较？,食物,食用人数,香蕉,6365,70,蛋,6820,75,奶,7093,78,花生酱,5365,59,番茄,6911,76,火鸡,2000,22,全国已有的调查数据：,2006,年,10,州（,5-7,月）,9093,人调查结果（,7,天）,病例对照研究结果,暴露病例数,(a),未暴露病例数,(c),病例暴露,%,暴露对照数,(b),未暴露对照数,(c),对照暴露,%,OR,mOR(95%CI),食用花生酱,50,12,80.6,80,44,64.5,2.29,-,食用两次及以上花生酱,39,20,66.1,49,75,39.5,2.98,3.5(I 1.4-9.9),食用,A,牌花生酱,37,24,606,15,104,12.6,11.2,12.1(3.6-63.3),食用,B,牌花生酱,5,56,8.2,1,118,0.8,10.5,9.1(1.0-433),食用火鸡,15,46,24.6,18,104,14.6,1.89,-,65,个病例和,124,个对照（电话抽话按地区匹配，病例病前,1,周食物史）,散状分布：,未知病因，且分布广泛,海湾战争综合症,Gulf War Syndrome,慢性疲乏综合症,Chronic Fatigue Syndrome,群体性臆症，如恐艾症,Mass Hysteria,格林巴例综合征,Guillian Barre Syndrome,云南不明原因猝死,关键是不是一种疾病？,寻找共同暴露,人群模式,平时不在一起人群的疾病点源爆发,共同暴露易于确定和检验,假设产生容易，很快出结果,平时在一起人群,学校、部队等集中人群,同吃、同住，为什么现在爆发？,1.,发病风险不等,In distribution C,3 groups have excess risk,Cases,People,Ob,Ex,O/E,3,0,1,0,3,3,1,3,4,0,1.3,0,4,4,1.3,3,5,0,1.7,0,5,0,1.7,0,6,1,2,0.5,7,5,2.3,2.1,8,2,2.7,0.8,45,15,15,非洲一起出血热疫情的年龄和性别罹患率,年龄（岁）,男性女性,合计,50,7.7,5.8,6.5,合计,5.9,6.7,6.4,通过副伤寒人群的分布我们可以有哪些推论？,What can we learn from the distribution of paratyphoid by person?,群体,Group,人口,Population,罹患率,AR(%),学生,Students,1005,40,一年级,Grade One,394,56,二年级,Grade Two,308,33,三年级,Grade Three,303,27,厨房工人,Kitchen workers,10,60,教师,Teachers,72,1,总数,Total,1087,38,Step 5,另一个简单的例子,A simple example,有,258,人参加婚礼,Among 258 persons attending a wedding,来自,9,个村庄的,23,名妇女在就餐时出现头痛、恶心、头晕和晕厥症状,23 women from 9 villages develop headache,nausea,dizziness,syncope while eating dinner at a wedding party.,140,名妇女在帐篷中吃晚餐,140 women ate dinner in a tent,118,名男士在外面吃晚餐,118 men ate dinner outside,婚礼于,11,月的乡村举办，人们通过燃烧木炭取暖,Held in the countryside in November charcoal fires used to keep warm,我们可以得出什么假设,参加婚宴的妇女由于暴露密闭帐篷内燃烧的炭火所产生的一氧化炭致病。,Illness among women attending the wedding party resulted from exposure to carbon monoxide generated from charcoal fires in the enclosed tent.,2.,各组有同等的发病风险,In distribution A,all groups have about the same risk,Cases,People,Ob,Ex,O/E,3,0,1,0,3,1,1,1,4,1,1.3,1.3,4,1,1.3,1.3,5,2,1.7,1.3,5,2,1.7,1.3,6,2,2.0,1,7,3,2.3,1.3,8,3,2.7,1.1,45,15,15,继续寻找不同风险！,在“假设形成的访谈”中，什么暴露最能解释？,询问病例潜伏期内的暴露,Ask individuals about exposures during the incubation period,询问病例的日常行为和习惯,Ask individuals about their general behaviors and habits,找出社区中特别的环境、社会和行政方面的因素,Find out about specific environmental conditions,social and administrative situation of the community,有什么异常或特殊事件发生？,What unusual or special occurances happened?,学校井,Schools well,0.2M,1.5M,村里的井,Village well,现场观察非常重要,注意：水从管道连接处渗出后滴下来,Note the water dripping from a leak in the pipe joint,例外病例的可能性,录入错误,背景病例，与本次疫情无关,本次疫情的传染来源,最早暴露的病例,暴露剂量大的病例,最晚暴露的病例,续发病例,特殊暴露,假设并不一定要和所有观测到的现象一致,Not necessary to be consistent with all observations,一次暴发或一种疾病可能有多种原因,Multiple causes for an outbreak or disease,不完善的疾病分类,Imperfect disease classification,观测到的不一致的现象可能会通过更深入的调查得到解释,Observations that appear inconsistent may actually explain on deeper investigation.,病因推论,Hill,关于病因的思考,在决定采取针对性干预前应考虑关联的哪些方面？,Sir Austin Bradford Hill.(1965)The environment and disease:Association or Causation?,关联的强度（Strength）,生态学关联和个体关联,关联强度,关联的时间性（Te