医学统计学10.Experimentdesign；SurveydesigChapter 13 Design of medical research医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig The aim of research design: l To get a reliable result in the lowest cost of manpower, time and money;l To estimate the random error within the observed data;l To promote the efficiency of the research There are two kinds of research: l Laboratory experiment and Clinical triall Medical survey医学统计学10.Experimentdesign；Surveydesig13.1 Principles of research design1. Control 2. Balance 3. Randomization 4. Replication 医学统计学10.Experimentdesign；Surveydesig 1. ControllIf there is no control, then there is no comparison. OthersEffect of othersEffect of treatmentTreatmentSubjectOthersControlEffect of controlSubjectEffect of others医学统计学10.Experimentdesign；SurveydesiglSome forms of control:(1) Empty control(2) Placebo control (for animal, Experimental control)(3) Mutual control(4) Self control(5) Standard control(6) Historical control医学统计学10.Experimentdesign；Surveydesig(1) Empty controlOthersEffect of othersEffect of treatmentTreatmentSubject AOthersSubject BEffect of others医学统计学10.Experimentdesign；Surveydesig(2) Placebo controlOthersEffect of othersEffect of treatmentTreatmentSubject AOthers PlaceboEffect of placeboSubject BEffect of others医学统计学10.Experimentdesign；Surveydesig(3) Mutual controlOthersEffect of othersEffect of Treatment A Treatment I Subject AOthersTreatment II Effect ofTreatment B Subject BEffect of others医学统计学10.Experimentdesign；Surveydesig(4) Self controlOthersEffect of othersEffect of treatmentTreatmentSubjectOthersControlEffect of controlSubjectEffect of others医学统计学10.Experimentdesign；Surveydesig(5) Standard controlOthersEffect of othersEffect of treatmentTreatmentSubjectThere is no control group, but compare with certain “standard”医学统计学10.Experimentdesign；Surveydesig(6) Historical controlThere is no control group, but compare with “historical result”OthersEffect of othersEffect of treatmentTreatmentSubject医学统计学10.Experimentdesign；Surveydesig 2. Balance: The experimental group and control group are almost the same in all aspects except the treatment. OthersEffect of othersEffect of treatmentTreatmentSubjectOthersControlEffect of controlSubjectEffect of others医学统计学10.Experimentdesign；Surveydesig 3. Randomization Many factors, we know that they may influence the results, but they are very difficult to deal with Randomization is the best choice! Example To improve the homogeneity of subjects, collect a number of students with same age and gender; randomly arrange them into two groups to make them balance in height and weight. 医学统计学10.Experimentdesign；SurveydesiglRandomization is the prerequisite of statistical inference.lRandomization CasuallRandomization means that all subjects in population have same probability to be sampled out for research. 医学统计学10.Experimentdesign；Surveydesig 4. ReplicationlOne meaning of replication : The results can be reproduced in different labs and by different researchers.lAnother meaning of replication : The study should be performed in a big enough sample. Altman & Dore checked 90 papers: l39% mentioned their sample size and why.lSample sizes of 27% papers were too small to make a conclusion.医学统计学10.Experimentdesign；Surveydesig How to estimate sample size in the design stage? Four parameters are needed: (1) : the maximal probability of type error is allowed(2) : the maximal probability of type I error is allowed(3) : the minimal difference between two means is allowed(4) : the standard deviation among subjects in the same group医学统计学10.Experimentdesign；Surveydesig Check Table 13-2, or calculate byExample 13-1 A pilot study shows, the sample mean andstandard deviation of pulse among male patients with leadpoisoning are 67/min and 5.97/min respectively. To testwhether the population mean of this kind of patients islower than the mean of normal males (one-side test), howmany cases are needed? Given , take (Check Table 13-2, / 0.85, n=14)1. Comparison between the mean and a given number医学统计学10.Experimentdesign；SurveydesigCheck Table 13-2, or calculate byExample 13-2 Comparing mean reductions of blood sedimentationbetween drug A and B. To test whether the effects of two drugs aredifferent (two-side), how many cases are needed? A pilot study shows, .Take , . (Check Table 13-3, / 0.60, n=60) 2. Comparison between two means of two independent 2. Comparison between two means of two independent samplessamples医学统计学10.Experimentdesign；Surveydesig3. Comparison between two frequencies of two3. Comparison between two frequencies of twoindependent samplesindependent samples Check Table 13-4, or calculate byExample 13-3 Comparing two chemotherapies for lymphoma, how many cases are needed? The pilot study shows, the remission rates are . Given . (Check Table 13-4, n=35) 医学统计学10.Experimentdesign；Surveydesig13.2 Experimental design1.Why? 2. To plan and arrange subject selection, treatment3.assignment, data collection and statistical analysis 4. To make sure validity, reproducible and economy.5.2. Types of research6.Experiment: animal experiment, clinical trial, community intervention trial Survey Both need well design ! 医学统计学10.Experimentdesign；Surveydesig(1)Subject: lSubject could be: gene, protein, cell, tissue, animal, patient or healthy population.lSubject should be clearly defined, and homogeneous. EligiblesubjectsTreatmenteffectsControl 3. Three elements of experimental design医学统计学10.Experimentdesign；Surveydesig2) Treatment: lA measure used to intervene a life process.lTreatment factor: drug Levels of treatment factor: dose A, dose B lTreatment factor and non-treatment factors should be identified. Non-treatment factors: age, gender, disease status, weather, environment lTreatment should be standardized:3) Effect: lChange caused by treatment. lEffect measurement should be objective, accurate, precise, specific and sensitive.医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig(1) Completely random designlRandomly allocate the subjects into two or more groupslIndependently sampling from two or more populationsExample 13-4 Randomly allocate 10 animals into two groups. (Using Table 13-1, read 0-9 only) A: 1, 4, 6, 8, 9 B: 2, 3, 5, 7, 10 4.Commonly used experimental designs 医学统计学10.Experimentdesign；SurveydesiglExample 13-5 Randomly allocate 15 animals into three groups. A: 4, 6, 8, 11, 15 B: 3, 5, 9, 12, 14 C: 1, 2, 7, 10, 13Data analysis: t test; rank sum testAdvantage: Simple; stableDisadvantage: If there are many confounders,they may not be balanced by randomization only医学统计学10.Experimentdesign；Surveydesig(2) Paired designExample 13-6 Randomly allocate 8 pairs ofsubjects into two groups (read 0-7 only, odd for AB) A: 1-1, 2-2, 3-2, 4-1, 5-2, 6-1, 7-2, 8-1; B: the restlData analysis: Paired t test; signed rank sum testl Advantage: Well control the non-treatment factors, Save sample sizel Disadvantage: Difficult to perform in practice Used for short period studies 医学统计学10.Experimentdesign；Surveydesig(3) Random block designExample 13-7 Four groups of subjects; fourtreatments: A, B, C, D.医学统计学10.Experimentdesign；SurveydesiglData analysis: Analysis of variance for random block designlAdvantage: The individuals in same block are quite similar so that the comparison within block is more sensitive than that between completely randomized groups.lDisadvantage: if any one individual observation is missed, the data analysis in the block will be very difficult. 医学统计学10.Experimentdesign；Surveydesig(4) Cross-over design 1st period wash-out 2nd periodGroup 1 A none BGroup 2 B none AExample 13-8 lIf there are 16 patients, then randomly allocate them into group 1 and 2 (same as Completely randomizeddesign).lIf there are 8 pairs of patients, then randomly allocate two subjects within each pair into group 1 and 2 (same as paired design).医学统计学10.Experimentdesign；SurveydesiglData analysis: Analysis of variance for cross-over designlAdvantage: Well control the non-treatment factors, save sample size; Everyone receives both treatments EquitylDisadvantage: Assume the subjects keep the same in 1st period and 2nd period - it can only be used for chronic diseases; Wash-out period is required the treatment should be stopped during wash-out period 医学统计学10.Experimentdesign；SurveydesigBlinding and PlacebolNecessary for clinical trial.lBlinding (Masking) - To reduce the bias caused by psychological effect of knowing the treatment Single blinding: Any patient does not know what treatment is taken. Double blinding: Both patient and physician do not know the treatment.lPlacebo To ensure blinding Everything should be the same to the treatment group, except that it does not contain any effective components of the treatment. Special skills are needed for blinding and placebo. 医学统计学10.Experimentdesign；Surveydesig13.3 Survey design1. Survey Observe the existing process Without intervention Well designlExample for surveys: Health condition survey Epidemiological survey Etiologic survey Clinical follow up survey Sanitary survey .医学统计学10.Experimentdesign；Surveydesig 2. Design (1) The purpose of survey, clear (2) The population concerned, well defined. (3) Space, time and sample size, specified (4) Observed unit, well decided individual? Family? Class? (5) The questionnaire, well designedl Items, carefully chosen l Language, clear and specificl Possible answers, well coded in advance医学统计学10.Experimentdesign；Surveydesig (6) Data collectionlDirect observation Measurement, observation, test, count lInterview Form filling Group meeting Telephone or internet ( the response rate ?)医学统计学10.Experimentdesign；Surveydesig(1) Overall survey (Complete survey) Example: National Census To get the population parameters directly There is no sampling error, but non-sampling error is relatively high (2) Sampling survey Sampling Observation Statistical inference knowledge about population Advantage: Efficient Disadvantage: Complicate in design, implementation and analysis 3. Classification of survey医学统计学10.Experimentdesign；Surveydesig Simple random samplinglEvery individual has same probability to be sampled.lSuitable for small population. Systematic samplinglAccording to the sequence of individuals, to sample subjects in a fixed interval.lWhen population has a sequence number ( such as ID code), it is convenient to carry out. Methods of random sampling医学统计学10.Experimentdesign；Surveydesig Stratified samplinglThe population is stratified according some factors that may influence the results of study and then the individuals in strata were randomly sampled.lSome important confounding factors could be controlled by stratified sampling. Cluster samplinglIf the individuals belong to certain unit ( such as community, school, class, city, county), we may directly sample the unit other than individuals. 医学统计学10.Experimentdesign；Surveydesig Stratified sampling Systematic sampling Simple random sampling Cluster samplingSampling error医学统计学10.Experimentdesign；Surveydesig(3) Typical survey (case survey) Advantage: Only a few typical subjects are observed, who well reflect the main characteristics of same kind of subjects. Disadvantage: There is no ground for any statistical inference. 医学统计学10.Experimentdesign；Surveydesig(4) Case-control survey (retrospective study) Good for rare diseases: Outcome Exposure (Cause?) “Case”: Patients “Control”: Non patients, main confounders are similar to the patients Comparing the frequency (or intensity) of exposure 1:1 case-control; 1:m case-control group group case control医学统计学10.Experimentdesign；Surveydesig(5) Cohort survey (Follow-up study, prospective study) Exposure (Cause?) Outcome Different groups of exposure Comparing the outcomes of different groups: Does different condition of exposure leads to different outcome?医学统计学10.Experimentdesign；SurveydesigSummary1. Commonly used Experimental Designs Completely random design Paired design Random block design Cross-over design 医学统计学10.Experimentdesign；Surveydesig2. Content of Survey Design Purpose Population concerned Space, time and sample size Observed unit Questionnaire Data collection 医学统计学10.Experimentdesign；Surveydesig3.Classification of survey Overall survey Sampling survey Sampling error Stratified sampling Systematic sampling Simple random sampling Cluster sampling Typical survey Case-control survey Cohort survey医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig医学统计学10.Experimentdesign；Surveydesig更多精品文档，请点击：更多精品文档，请点击：医学统计学10.Experimentdesign；Surveydesig 人的一生，总是难免有浮沉。不会永远如旭日东升，也不会永远痛苦潦倒。反复地一浮一沉，对于一个人来说，正是磨练。因此，浮在上面的，的，不必骄傲；沉在底下的，更用不着悲观。必须以率直、谦虚的态度，乐观进取、向前迈进。 松下幸之助 医学统计学10.Experimentdesign；Surveydesig成功人生八大心成功人生八大心态医学统计学10.Experimentdesign；Surveydesig一、成功者的心态（老板的心态）一、成功者的心态（老板的心态）天生天生赢家家为成功而生成功而生责任任老板的心老板的心态医学统计学10.Experimentdesign；Surveydesig二、学习的心态二、学习的心态学学习是通向未来的是通向未来的护照照空杯空杯归零零低低头以身作以身作则盗版盗版学到，做到，教到学到，做到，教到医学统计学10.Experimentdesign；Surveydesig三、合作的心态三、合作的心态环境改境改变人人1+121+12雁群雁群飞翔翔-合作合作天天时-地利地利-人和人和医学统计学10.Experimentdesign；Surveydesig四、付出的心态四、付出的心态医学统计学10.Experimentdesign；Surveydesig五、积极五、积极 的心态的心态平常心平常心平等心平等心包容心包容心医学统计学10.Experimentdesign；Surveydesig六、吃苦耐劳的心态六、吃苦耐劳的心态医学统计学10.Experimentdesign；Surveydesig七、坚持的心态七、坚持的心态医学统计学10.Experimentdesign；Surveydesig八、感恩的心态八、感恩的心态感恩国家、社会、董事感恩国家、社会、董事长、推荐人、推荐人伙伴伙伴父母父母另一半另一半子女子女医学统计学10.Experimentdesign；Surveydesig