中医药科研思路与方法(讲义合成)加目录[1]

三、诊断试验

?对新的诊断技术的试验与评价

?也适用于中医新辨证方法试验与评价

金标准：

?金标准是进行诊断试验的基础和前提；

?金标准是：明确肯定或排除某种疾病最可靠的诊断方法，如： 活体

组织检查、手术探查、尸体解剖、特殊影像学检查（CT、 MRI、血管造影）、放射性标记扫描、、长期随访。

病原学和免疫学检查一般不作为诊断疾病的金标准。

研究目的：

?评价一项非“金标准”诊断方法的诊断效能。

研究方法：

?研究对象：病例或社区人群

?分组：用金标准诊断，分为患者组与非患者组。 ?样本大小的确定：方法同前。

?结果测量：严格执行测量条件，保证测量准确性；主观性指标（如影像诊断）盲

法判定。

界值确定：

?呈正态分布者，可以用 X±1.96s 结合临床确定界值。

?百分位法：不管是否为正态分布，均可用95%百分位数值作为界值点。 ?两组分布交叉对应点

?受试者工作性曲线（receiver operator characteristic curve 简称ROC曲

线）： 多次试验，试定不同界值，用真阳性率和假阳性率作图得出曲线，决定最佳临界点。

46

结果分析：

设计表格与结果

------------------------------------------------------------- 金标准诊断

待评价诊断 ----------------------------- 合计 患病 非患病

------------------------------------------------------------- 阳 性 真阳性（a） 假阳性（b） a+b 阴性 假阴性（c） 真阴性（d） c+d

------------------------------------------------------------- 合计 a+c b+d a+b+c+d

-------------------------------------------------------------

灵敏度(sensitiviti, Se)：

真阳性率，指正确检出患者的能力。 Se=[a/(a+c)]*100%

敏感性愈高，假阳性率愈高，欲降低漏诊率时重点考察。

特异性(specificity,Sp)：

真阴性率，反映鉴别非患者的能力。 Sp=[d/(b+d)]*100%

特异性愈高，假阴性愈高，欲降低误诊率时重点考察。

误诊率(misdiagnosis rate)：

即假阳性率(α)

α=[b/(b+d)]100%=100%-Sp

漏诊率(missed diagnosis rate)：

即假阴性率（β）

β=[c/(a+c)]100%=100%-Se

准确度(accuracy,Ac)：

即诊断结果与金标准的诊断结果一致的程度 Ac=[(a+d)/n]*100%

约登指数（Youden index,YI）：

亦称正确指数

YI=Se+Sp-1=1-α-β

47

预测值（predictive value,PV）：

可能的试验后概率

阳性预测值(positive predictive value,PV(+))--真正有病的概率

PV(+)=[a/(a+b)]*100%

阴性预测值(negative predictive value）真正无病的概率

PV(-)=[d/(c+d)]*100%

减少假阳性和假阴性对诊断的干扰，帮助准确诊断。

拟然比（likelihood ratio,LR)：

患病人群中诊断结果的概率与无病人群中诊断结果概率之比

阳性拟然比，患病组真阳性率和未患病组假阳性率的比值(positive likelihood ratio,LR(+))

LR(+)=[a/(a+c)]/[b/（b+d）]=Se/(1-Sp)

主观正确判断阳性的可能性是错判阳性可能性的倍数。

阴性拟然比，患病组假阴性率和未患病组假阳性率的比值(negative likelihood ratio,LR(-))

LR(-)=[c/(a+c)]/[d/（b+d）]=(1-Se)/Sp

表示试验错判阴性可能性是正确判断阴性可能性的倍数。

超脱了正常与异常的界限；可重复使用，多用于难以诊断的疾病。

?如 45岁女性患者，因间歇性胸痛就诊，可能患有冠心病、食管及上消化道疾病、

情绪紧张行胸痛。

?文献记载45岁女性患冠心病的概率为1%。

?验前比=验前概率/（1-验前概率）=0.01/（1-0.01）=0.01。 ?患者症状为典型心绞痛，其拟然比为100。

验后比=验前比*拟然比=0.01*100=1；

验后概率=验后比/（1+验后比）=1/（1+1）=0.5（50%） 患冠心病的概率为50%。

?进行心电图运动试验，ST段压低2.2mm（拟然比为11） ?验后比=验前比*拟然比=1*11=11

?验后概率=验后比/（1+验后比)=11/(1+11）=1/（1+11）=0.91（91%）。

?患冠心病的概率为91%。

48

?检测血CPK大于80单位，拟然比为7.75。 ?验后比=11*7.75=85.25

?验后概率=验后比/（1+验后比）

=85.25/（1+85.25）=98.8% ? 患冠心病的概率为98.8% ?可以确诊冠心病。 诊断方法比较

?用U检验

U=(p1-p2)/s (p1-p2)

P1、p2代表两个率，可以是灵敏度或特异性或准确度。 s (p1-p2)为两个率相差的标准误。

s (p1-p2)= √pq(1/ni+1/mi)

?i为n或m的下角标,在比较灵敏度时， i =1；在比较特异性时，?q=1-p

?代表灵敏度时，p=(a1+a2)/(n1+m1) ?代表特异性时, p=(d1+d2)/(n2+m2)

?n m 分别代表四格表中，列的合计数。

?当U值<1.96,P>0.05. ?当U值>1.96,P<0.05.

四、临床研究

（一）临床流行病学调查和回顾性病因学研究

应用调查研究的方法。

（二）前瞻性病因学研究 1、配对对照设计 2、长期随访纪录

49

。 i =2

3、全面收集有关信息 4、严格掌握条件

5、准确评价病因效果

（三）临床评价研究 1、纳入病例严格 诊断标准 纳入标准 排除标准

2、慎重考虑对照

空白对照：新发生的疾病

缺乏有效疗法的疾病

病情长期稳定，传统不做治疗的疾病（耳聋、近视眼

等）

标准对照：用目前通用的公认的治疗方法（中药、西药及其他疗法）

配对对照：自身疗前疗后对照

异体配对对照

相互对照：同时评价几种疗法

联合治疗的对照安排，要充分考虑交互影响问题。

3、严格遵守医德

全面的安全性评价为基础 伦理委员会通过 患者知情同意 损害补偿机制

4、适当安排指标 基础指标 关键性指标

安全性指标（毒副作用）观察要全面 （临床严重事件的报告制度）

5、避免主、客观偏倚 盲法设计

安慰剂的安排

6、科研设计类型的选择

配对设计（自体、异体）

完全随机设计：单因素、多水平。

50

中医药科研思路与方法

（讲义）

天津中医药大学 ： 徐宗佩 博士，博导、教授、研究员

1

目 录

第一篇 概 论 ·····················································································································7

一、进入研究生学习阶段的几个转变···············································································7 二、科学研究··················································································································7 三、科学研究的方法与过程·····························································································7 四、医学科学研究···········································································································7

1．特点： ················································································································7 2．类型 ···················································································································8 3．要 求················································································································ 14 五、中医科研的特点 ····································································································· 14 六、中医科研的注意事项 ······························································································ 14 第二篇 医学研究层次的基本方法 ······················································································· 15

一、观察法 ··················································································································· 15

1.科学观察的任务 ·································································································· 15 2.科学观察的研究方法 ··························································································· 15 3.直接观察常常是科学发现的源泉 ·········································································· 15 二、实验法 ··················································································································· 15

1.比较实验 ············································································································· 16 2.预初实验 ············································································································· 16 3.断决实验 ············································································································· 16 4.正式实验 ············································································································· 16 5.模型实验： ········································································································· 16

第三篇 医学科学研究选题的种类 ······················································································· 17

一、调查研究性质········································································································· 17 二、实验观察性质········································································································· 17 三、实验研究性质········································································································· 17 四、资料分析性质········································································································· 17 五、经验体会性质········································································································· 17 第四篇 选 题 ··················································································································· 18

一． 选题的产生与来源 ································································································ 18

1． 来自自身实践的课题 ······················································································· 18 2． 来自他人实践的课题 ······················································································· 18 3． 来自客观要求的课题 ······················································································· 18 4． 来自协作要求的课题 ······················································································· 18 5． 来自上级指定的课题 ······················································································· 18 6． 来自公开招标的课题 ······················································································· 18 二． 选题的途径 ·········································································································· 19 三． 选题的方法 ·········································································································· 19

1.假说形成法 ········································································································· 19 ２.移植法··············································································································· 19 ３.旧题发挥法········································································································ 19 四． 初选、筛选 ·········································································································· 19

１.指导思想 ··········································································································· 19 ２.方法适当 ··········································································································· 19 五．定题······················································································································· 19 六． 选题的注意事项 ··································································································· 20

2

七． 研究题目的表述 ··································································································· 20 第五篇 中医科研的假说 ······································································································· 21

一． 假说的定义 ·········································································································· 21 二． 假说的性质 ·········································································································· 21 三.假说的作用··············································································································· 21 四． 假说的形成 ·········································································································· 21

1．孕育阶段 ·········································································································· 21 2.形成阶段 ············································································································· 22 五． 假说的检验 ·········································································································· 22 六、假说的类型 ············································································································ 22 七、中医学中的假说 ····································································································· 23 八、确定假说的注意事项 ······························································································ 23 第六篇医学文献的检索········································································································· 24

一、文献检索的意义 ····································································································· 24 二、医学文献的特点 ····································································································· 24 三、医学文献基本情况·································································································· 24

1.级别 ···················································································································· 24 2.类型 ···················································································································· 24 四、医学文献检索········································································································· 25

1.检索工具 ············································································································· 25 2.检索方法 ············································································································· 25 3.检索步骤 ············································································································· 25 4.检索途径 ············································································································· 25 五、计算机检索 ············································································································ 26 六、中医文献检索········································································································· 26 第七篇中医文献综述的撰写·································································································· 27

一、概论······················································································································· 27 二、目的和意义 ············································································································ 27 三、特点······················································································································· 27 四、内容分类················································································································ 27 五、结构······················································································································· 27

1.题目 ···················································································································· 27 2.作者 ···················································································································· 27 3.内容提要,关键词·································································································· 28 4.前言: ··················································································································· 28 5.正文 ···················································································································· 28 6.总结 ···················································································································· 28 7.参考文献 ············································································································· 28 六、写作步骤················································································································ 28

1.确定题目 ············································································································· 28 2.搜集文献—检索 ·································································································· 28 3.阅读文献 ············································································································· 28 4.制定、修改提纲 ·································································································· 28 5.再次查阅文献 ······································································································ 28 6.正式写作 ············································································································· 28 7.修改---反复修改 ·································································································· 28 七、注意事项················································································································ 28

3

第八篇 科研设计················································································································ 29

一、概论······················································································································· 29 第九篇 常用实验设计方法 ··································································································· 38

一、配对设计················································································································ 38

1.自身对照设计 ······································································································ 38 2.异体配对对照 ······································································································ 38 二、交叉(配对)设计 ······································································································ 38 三、完全随机设计········································································································· 38 四、随机区组设计········································································································· 38 五、拉丁方设计 ············································································································ 38 六、正交设计················································································································ 39 七、析因设计················································································································ 40 八、序贯设计················································································································ 40 第十篇 几种常见场合的科研设计 ························································································· 41

一、调查研究················································································································ 41

（一）调查研究使用范围 ······················································································· 41 （二）要求 ············································································································ 41 （三）设计要点 ····································································································· 41 二、病因学研究 ············································································································ 41

1、几个基本概念 ··································································································· 41 2、病因学研究方法 ································································································ 42 三、诊断试验················································································································ 46

金标准：················································································································ 46 研究目的： ············································································································ 46 研究方法： ············································································································ 46 界值确定： ············································································································ 46 结果分析： ············································································································ 47 四、临床研究················································································································ 49

（一）临床流行病学调查和回顾性病因学研究························································ 49 （二）前瞻性病因学研究 ······················································································· 49 （三）临床评价研究 ······························································································ 50 五、动物实验················································································································ 51

（一）实验动物 ····································································································· 51 （二）对照全面 ····································································································· 51 （三）指标丰富 ····································································································· 51 (四）实验结果的解释留有余地 ··············································································· 51

第十一篇 中医科研发展思路 ································································································ 54

一、总体思路： ············································································································ 54

（一）中医科学研究要满足中医学科自身发展的需求。·········································· 54 （二）适应和满足现代社会的需求 ········································································· 54 二、中医科研方法学 ····································································································· 54

问题： ··················································································································· 54 科研思路： ············································································································ 54 避免······················································································································· 54 三、中医基础理论········································································································· 55

1．中医语言 ·········································································································· 55 2．中医生命观 ······································································································· 55

4

四、中医诊断技术········································································································· 55 五、中医辨证理论········································································································· 56 六、病因病机理论········································································································· 56 七、中医医史研究········································································································· 57 八、方剂理论················································································································ 57 九、中医治疗方法········································································································· 57 十、针灸推拿学科········································································································· 57

1、经络的研究 ······································································································· 57 2、腧穴的规范化研究 ···························································································· 58 3、针灸推拿机理研究 ···························································································· 58 4、针灸推拿学临床研究 ························································································· 58 十一、中医临床疗效评价体系研究 ················································································ 58 十二、中医药疗效机制研究··························································································· 59 十三、几点想法 ············································································································ 59 第十二篇 开 题 报 告 ······································································································· 60

一、概念······················································································································· 60 二、主要内容················································································································ 60 三、注意事项················································································································ 60 四、课题论证················································································································ 60 第十三篇 医药科技论文的写作··························································································· 61

一、撰写科技论文的目的 ······························································································ 61 二、对科技论文的要求·································································································· 61 三、资料的处理 ············································································································ 61 四、材料的表现方法 ····································································································· 62 五、撰写论文················································································································ 62

(一)医药科技论文的种类 ························································································ 62 (二)对医药科技论文的评价 ····················································································· 63 (三)医学论文的一般格式 ························································································ 63 (四)撰写过程 ·········································································································· 63 (五)临床病例分析 ··································································································· 67

5

第十篇 几种常见场合的科研设计

一、调查研究

（一）调查研究使用范围

用于流行病、地方病、病因学、卫生学及社会卫生学等方面的研究。

（二）要求

不施加任何措施，只调查和记录现场已显示的结果或条件。

（三）设计要点

1、确定调查的目的、范围、对象。 2、设计调查表格

内容包括：基本情况如调查地区、气 象条件、调查者姓名，被调查者姓名、性别、年龄及联系方式等。需要调查的内容如患病情况、影响因素、临床表现、中医证候及治疗情况及效果等。 3、确定调查数量

血清流行病学调查： 例 一般临床流行病学调查 1000例

其他相应例数。

4、收集资料时要遵循“随机抽样”、“齐同对比”原则，必要时进行配对调查，以避免偏性 。

5、最好事先进行小规模预查

6、事先培训

明确调查目的意义， 统一方法、仪器、标准、要求，编制和掌握“调查须知”和“填表须知”。

7、确定适当的统计方法。

二、病因学研究

1、几个基本概念 病因

危险因素：暴露后增加某种疾病发生危险性的有关因素。 直接病因：发生某种疾病必不可少的因素。 间接病因：促使发病或使病情加重的因素。

41

?单因----单果 ?单因----多果 ?多因----单果 ?多因----多果

2、病因学研究方法 第一步：提出病因假说

来源于病例观察与报告：

如：Gregg注意到Australia某地区新生儿白内障病例增多，且：

病例分布地区广、时间集中—共同因素；78例中，76例出生即有病----病因胚胎期暴露；晶体中心核纤维受损----发病于妊娠早期；母亲妊娠早期均有风疹史。提出母亲妊娠初期感染风疹是新生儿白内障的可能病因。

?来源于基础(实验室）医学研究：

?来源于流行病学或现状调查：

如各地区在不同时间调查均发现香烟消耗量、煤烟浓度与肺癌发病正相关。提出吸烟与空气污染是肺癌发病病因的假说

第二步：验证病因假说

?病例对照研究

病例对照研究模式

---------------------------------------------------------- 病例组 对照组 合计

---------------------------------------------------------- 有暴露因素 a b a+b 无暴露因素 c d c+d

---------------------------------------------------------- 合计 a+c b+d N

----------------------------------------------------------

?计算相对危险度（relative risk, RR） 和比数比 （odds ratio, OR）

RR=[a/(a+b)]/[c/(c+d)]=[a(c+d)]/[c(a+b)] OR=(ad)/(bc)

OR值的检验

?用X2检验测定OR是否有显著性

42

X2=[(ad-bc)2(n-1)]/[(a+b)(c+d)(a+c)(b+d)]

n=a+b+c+d;自由度=1

病例对照研究的优点

?适用于少见疾病，无需大样本； ?适用于有很长潜伏期的疾病；

?所需时间短、省人力物力，易得出结果； ?医德问题少；

?可调查分析多种因素； ?可使用病史资料。

病例对照研究的缺点

?有回忆偏倚；

?对照组有选择偏倚；

?不能直接确定因果关系，只能计算相对危险度，提供病因分析线索。

队列对照研究

队列对照研究（前瞻性调查）设计模式

------------------------------------------------------------- 病例组 对照组 合计

------------------------------------------------------------- 有暴露因素 a b a+b 无暴露因素 c d c+d

------------------------------------------------------------- 合计 a+c b+d N

-------------------------------------------------------------

?计算相对危险度（relative risk, RR）

RR=[a/(a+b)]/[c/(c+d)]=[a(c+d)]/[c(a+b)] RR值的检验

?用X2检验测定RR是否有显著性

X2=[(ad-bc)2(n-1)]/[(a+b)(c+d)(a+c)(b+d)]

n=a+b+c+d;自由度=1

43

计算AR

?归因危险度（attributable risk,AR); 又称为特异性危险度或超额危险度

(excess risk):

AR=[a/(a+b)]-[c/(c+d)]=暴露组发病（死亡）率-非暴露组发病（死亡）率

计算AR%

队列对照研究的优点

?前瞻性研究，直接估计因素与发病的联系和联系的程度； ?可控制选择性偏倚；

?可了解致病因子的致病过程；

?可在某一阶段研究一个以上的疾病的结局。

队列对照研究的缺点

?研究罕见病时需随访较多人和较长时间，化费较多； ?确定疾病结局时，可能有偏倚； ?可能存在心理因素干扰； ?存在失访(drop out)问题。 ?

随机对照研究

?随即分组，施加因素，对照观察； ?易产生医德问题； ?特殊情况可用；

如：早产儿可因缺氧产生脑损害致智力障碍，曾普遍用高浓度氧疗，但出现晶状体后纤维组织增生致失明，为证实两者关系，将早产儿随即分组，分别使以高浓度和低浓度氧疗从而肯定了这种联系。

几种方法论证能力比较

?随机对照研究论证能力最强； ?队列对照研究论证能力较强； ?病例对照研究论证能力不太强； ?描述性研究论证能力最弱。

病因的论证

?排除虚假联系和间接联系

虚假联系：可能为两组调查方法不同或仪器条件不同等原因导致。

44

间接联系：为未知混杂因素所致，应注意排除。如随年龄增长，肿瘤患病率增高，同时白发增多，白发与肿瘤为间接联系，应予排除。

因果关系的确定

?美国公共卫生署咨询委员会1964年提出5条标准确定因果关系：

1、联系的一致性：不同研究产生一致的联系； 2、联系的强度：越大，因果说服力越大；

3、联系的特异性：特定的暴露因其特定的疾病。 4、时间先后：先因后果；

5、联系的有理性：符合理论或在动物模型中确立。

对病因学研究结果的评价方法 ?是否为源于人体的研究结果； ?因果关联性的强度如何； ?结论的重复性如何； ?因果时间先后；

?是否有 剂量----效应 关系； ?是否有流行病学意义； ?是否有生物学意义； ?是否有特异性关联；

?是否与已证明的因果关系有雷同性。

45

?如：高血压病，一期不宜作为研究对象；二期可以。 ?患者不宜作为研究对象的情况：

并发症；危重状态；各种疗法无效；不能配合者。

3.受试对象标准化

患者：诊断标准；纳入标准；排除标准。 动物：标准物种及生理、病理状态。 细胞、菌种：标准。

4.受试对象数量计数方法

5.受试对象的集中性与代表性

力求一致，减少个体差异----集中性。 纯化对象，考虑构成的均衡性—代表性。

效果反应：

?定义:

处理因素作用于受试对象所产生的反应或效应,称效果反应。效果反应一般通过指标来反映。

?关于指标的一些要求:

1.指标的关联性----必须扣题

2.明确指标的性质、种类、数目与条件 计数、计量；主观、客观 数目适当；条件严格掌握。 3.指标客观化 主观指标客观化 4.指标标准化

观察、操作、时间记录---标准。 5.灵敏、精确、特异

灵敏----迅速反应变化，用不同样品校正。

精确----现象与本质数量相应，用同一样品校正。 特异----无关因素干扰少。

重 复：

?重现性:在同样条件下,能重复显示同样或类似实验结果的性质。 ?重复数:为使实验结论可靠,受试对象(样本)必须达到足够的数量。

重复的目的

–①使均数与差值逼真

31

–②使实验误差(标准误)确定合理 –③使统计判断正确

–④排除偶然因素的干扰

决定重复数的因素

–①组间均数差值或百分率差值的大小 –②标准差的大小 –③反应指标的性质 –④显著性检验的要求 –⑤实验结果的可能性 –⑥实验设计类型

重复数的估计方法： ?经验法 ?临床实验

–难治性疾病:5-10例 –危重病症:30-50例

–一般性疾病:100-500例

–血清流行病学调查:300-600例 –一般流行病学调查:1000例以上 ?动物实验

–犬、猫等大型动物：5-20只 –兔、豚鼠等中型动物：10-20只 –大、小鼠等小型动物：15-30只

计算和查表法：

?两样本率比较实验的样本数

–

n=[(p1*q1+p2*q2)/(p1-p2)2 ]*(zα+zβ)2 –n为样本数

–p1、p2分别为两样本率的估计值 –q1=(1-p1),q2=1-p2)

–zα、zβ为正态分布曲线下面积的相应正Z值和负Z值

举 例

?两种中成药治疗乙肝病毒携带者HBsAg转阴率分别为50%、30%，拟取α=0.05、1-β(把握度)=0.9,问正式实验时每组所需病例数(双侧实验)。

?P1=0.5,p2=0.3,q1=1-0.5=0.5,q2=1-0.3=0.7,查表: zα=1.96, β=0.10, zβ=1.28

n=[(0.5*0.5)+(0.3*0.7)]*(1.96+1.28)2==121

32

zα、zβ值表

----------------------------------- zα

α、β ------------- zβ 单侧 双侧

----------------------------------- 1% 2.32 2.58 2.32 5% 1.65 1.96 1.65 10% 1.28 1.65 1.28 20% 0.84 1.28 0.85 -----------------------------------

以计量资料为指标的实验样本数估计

?配对实验

n=(S/D)*(zα+zβ)

2

2

S为标准差的估计值，D为差值的估计值

举 例

预实验某中成药治疗再生障碍性贫血可升红细胞2g/dl, s/D约2.75,拟取α=0.05、β=0.05,正式实验需多少例患者?

s/D=2.75, zα=1.65,zβ=1.65 n=2.752× (1.65+1.65)2=83

? 两组成组比较实验的样本数

2 2 2

n=2*(S/D)× (zα+zβ)+0.25× (zα)

S为两组合并的估计值,D两组均数的差值(X1-X2) 举 例

预实验可知升血散治疗白细胞减少症可升高白细胞2000个/微升,对照药黄芪为1000个/微升,标准差为1800个/微升,拟取α=0.05(双侧)、β=0.10,问正式实验时各组例数?

S=1800.D=2000-1000=1000, zα=1.96 zβ=1.28 n=2*(1800/1000)2*(1.96+1.28)2+0.25*1.962=69

? 随机区组比较样本数

2

n=2*(Mse/D)*(Q+zβ) Mse为误差的均方 D为组间差值

33

一般取β=0.05,zβ=1.65.

Q为两组均数在P=0.05时应为标准差的倍数,此外,可查Q值表

Q 值 表(α=0.05)

---------------------------------------------------------- 组数 3 4 5 6 7 8 9 10 ---------------------------------------------------------- Q值 3.4 3.8 4.0 4.2 4.4 4.5 4.6 4.7 -----------------------------------------------------------

举 例

四种中成药降低血清转氨酶预实验得知误差均方为30单位/dl,组间差值8单位/dl, 取α=β=0.05.

Mse=30,D=8,查表Q=3.8, zβ=1.65 n=(30/82)*(3.8+1.65)2==28

实验组与对照组样本数的设置

t=ㄧX1-X2ㄧ/√Sc× (n1+n2)/(n1*n2)

n1=n2时检验效率最高

对 照：

?对照原则就是“齐同对比”原则,即除被试因素外,实验组与对照组其他条件尽

量相同。

1.对照的必要性 ?自然环境无法控制 ?疾病本身可以自愈 ?疾病存在波动性

?新出现的疾病，自然病程未确定。 ?有些疾病与休息、营养有关 ?心理影响

?治疗措施的不良反应和并发症

需要对照来抵消以上因素的影响。

2.对照的形式

?按时间 –同期对照

34

–历史对照 ?按对照物 –空白对照 –实验对照 –标准对照 –参考值对照

按对照方式

?配对对照 ?交叉对照 ?相互对照 ?潜在对照

3.对照的选择

?根据实验目的选择对照形式

目的在于排除干扰

如：复方丹参方的疗效判定—与标准药物对照—硝酸甘油 注射剂疗效----与其他剂型对照 与速效救心丸比较----相互对照

?保证对照的可比性

均衡性是可比性的基础，保证均衡

?力求对照完善

随 机：

?1.目的

保证组间的齐同性,减少抽样误差

保证统计处理结果有显著性,随机是数理统计的先决条件 2.方 法

?简单法

–占阄,抽签,掷硬币等 ?随机表法 –二组样本分配 –多组样本分配

二组样本分配练习

?总数(N=)20,每组(n=)10

35

?任选随机数字表一行或一列,如第2--4列,去除大于N的数字,结果为

17 17 17 20 6 8 13 9 13 8 6 8 15 2 17 18 6 12 3 4 10 7

?根据编号依次对应随机数字,奇数为一组,偶数为另一组,如果其中一组已满,停止,余下的样本归另一组.

?结果:第4、5、6、10、11、12、14、16、17、18号归乙组，其余归甲组。

多组样本分配

?用总样本数N处随机数字表的任意一行或一列,余数为1—n(n为每组样本数)时归入第1组,余数为(n+1)--2n时归入第2组,余数为(2n+1)--3n时归入第3组,??余数为?—N时归入最后1组.

?各组例数不等时,仿此法均衡.

?如:将15名患者分为3组,N为15,n=15/3=5

?从随机数字表第16行第6个数字开始,以15处随机数字.得:

15例分3组的分配示例

------------------------------------------------------------------------ 病人编号 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 随机数字 33 35 72 67 47 77 34 55 45 70 08 18 27 38 90 15除后余数 3 5 12 7 2 2 4 10 15 10 8 3 12 8 15 初步分组 a a c b a a a b c b b a c b c

------------------------------------------------------------------------

?a组为第1、2、5、6、7、12号病人，6例，b组5例，c组4例

?用a组总例数6除下一个随机数字16余4，将a组第4例病人转至c组，各组均衡。

?随机排列表法 ?计算器法

–两组：按INV或 SHIFT RAN键，数字〉=0.500者归依一组,<0.500者归另一组. –多组:参见程序型计算器说明书

?计算机法

–各种统计软件均有随机方法

3.随机性检验

?当Z或U值<1.96,P>0.05时,是随机的. ?当Z或U值>1.96,P<0.05时,不是随机的.

盲 法：

?1.正确地选择对照物

36

?2.双盲设计的确定 ?3.单盲设计的选择 ?4.非盲设计

均 衡：

?1.均衡与分层

–分层是增强均衡性的手段 ?2.分层的依据

–按主要影响因素分层,次要影响因素用随机来处理 ?3.均衡性检验 举 例

?中、西药治疗白血病对照实验已有17例分层情况(预定20例) ?----------------------------------------------- ? 性别 年龄 病 情 病 程 ? ------------------------------------ ? 男 女 中 青 轻 中 重 <=1年 >1年 ?----------------------------------------------- ?中药组 4 5 6 3 4 3 2 5 4 ?西药组 5 3 4 4 2 3 3 3 5 ?组差 1 2 2 1 2 0 1 2 1 ?-----------------------------------------------

?总组差=12,如来1男青年,病情重,病程1.5年,将其分入西药组,则总组差增至16,分入中药组则使之降为8.

医学科研设计的注意事项： ?1.合理地选择对照 ?2.正确执行盲法 ?3.适量安排重复 ?4.切实做到随机 ?5.认真搞好分层

37

第九篇 常用实验设计方法

一、配对设计

?先将条件相同或相似的受试对象配成对子,而后按随机原则给于每对中的个体施

以不同的处理。

1.自身对照设计 –自身前后对照 –自身左右对照

2.异体配对对照

二、交叉(配对)设计

?组别 第1次实验 第2次实验 ?第一组 观察组 对照组

?第二组 对照组 观察组

?病情相对稳定

?实验间隔为药物代谢的6—8个半衰期

三、完全随机设计

四、随机区组设计

五、拉丁方设计

?三种疗法(甲、乙、丙）治疗某病实验设计

?---------------------------------------------------------- ?年龄 老年 中年 青年

?---------------------------------------------------------- ?急性 A B C ?亚急性 B C A ?慢性 C A B

?----------------------------------------------------------

38

不完全拉丁方设计（乔敦方）

三种药物（甲、乙、丙）；每组动物给药二次

------------------------------------------------------------

分组 1组 2组 3组

------------------------------------------------------------- 第1次给药 A B C 第2次给药 B C A

六、正交设计

Lm （Kn ）

L 代表正交 ，m为实验次数 ，K未水平数 ，n为因素数。

3

L8(2)

------------------------------------------------------------- 因素及水平 次序

------------------------------- A B C

-------------------------------------------------------------

第1次 A1 B1 C1 第2次 A2 B1 C1 第3次 A2 B1 C2 第4次 A2 B2 C1 第5次 A1 B2 C1 第6次 A1 B2 C2 第7次 A1 B1 C2 第8次 A2 B2 C2

-------------------------------------------------------------

39

七、析因设计

乙药

----------------------------------------------------------- 用 不用

用 用甲用乙 用甲不用乙 甲药

不用 用乙不用甲 不用甲不用乙

-------------------------------------------------------------

八、序贯设计

序贯图曲线方程求法 ?1.单向质反应试验 –U: Y=a+b*n – L: Y=-a+b*n ?2.单向量反应试验 –U: Y=a*δ+b*n

–L: Y=a*δ+b*δ*n，δ为标准差

? 3.双向质反应实验

–U: Y=a1+b*n M:Y=-a2+b*n

–L: Y=-a1+b*n M':Y=a2-b*n， M、M'为两者差异不显著的界限

? 4.双向量反应试验

–U: Y=a1*δ

+b*n M:Y=a2*δ+b*n

–L: Y=-a1*δ-b*δ*n M':Y=a2*δ-b*δ*n 其 中

?a=log[(1-β)/α]/(log[p1*(1-p0)]/[p0*(1-p1)])

?b=log[(1-p0)/(1-p1)]/(log[p1*(1-p0)]/[p0*(1-p1)])

? p1为接受水平,p0为拒绝水平。

?α为假阳性率水平,β为假阴性率水平,一般α=β=0.05或0.01

40

百度搜索“77cn”或“免费范文网”即可找到本站免费阅读全部范文。收藏本站方便下次阅读，免费范文网，提供经典小说综合文库中医药科研思路与方法(讲义合成)加目录[1]在线全文阅读。