药物流行病学--实验流行病学

《药物流行病学--实验流行病学》由会员上传分享，免费在线阅读，更多相关内容在教育资源-天天文库。

1、实验流行病学ExperimentalEpidemiology第五章广东药学院姚振江学习目的实验流行病学的定义实验流行病学的特点实验流行病学的分类实验流行病学的设计实验流行病学的强势与弱势推荐参考文献吴涛,詹思延,李立明.流行病学实验研究发展历史.中华流行病学杂志2004;25(7):633-636.…5第五章实验流行病学研究(experimentalstudy)历史发展1747年苏格兰医师JamesLind首次对比研究了用桔子、柠檬及其他干预治疗坏血病的疗效5第五章实验流行病学研究OnMay20,1747heisreadytobeginthefirstclinicalnutrit

2、ionexperiment.group1drankonequartofcideradaygroup2gargledwithsulfuricacidgroup3hadtwospoonfulsofvinegar,3timesadaygroup4drank1/2-pintseawateradaygroup5drankbarleywatergroup6atetwoorangesand1lemonadayHowdidheconducthisexperiment?Whathappened?Onlysailorswhoatetheorangesandlemonfeltbetter.Infact

3、,inonlysixdaystheyfeltgreatandwereabletostartworkingagain.Theothersailorsintheexperimentfeltworse.1816年法国Hamilton医师首次报道了爱丁堡的一项大型对照实验，评价放血疗法的效果，这是迄今为止有关采用交替法产生对照的最早记载之一。‘Ithadbeensoarranged,thatthisnumberwasadmitted,alternately,insuchamannerthateachofushadonethirdofthewhole.Thesickwereindiscri

4、minatelyreceived,andwereattendedasnearlyaspossiblewiththesamecareandaccommodatedwiththesamecomforts.Onethirdofthewholeweresoldiersofthe61stRegiment,theremainderofmyown(the42nd)Regiment.NeitherMrAndersonnorIeveronceemployedthelancet.Helosttwo,Ifourcases;whilstoutoftheotherthird[treatedwithbloo

5、dlettingbythethirdsurgeon]thirtyfivepatientsdied."1898年丹麦医师Fibiger通过半随机对照实验验证血清治疗白喉的效果。8outof239patientsintheserumtreatedgroupand30 outof245inthecontrolgroupdied.Noformalstatisticalanalysiswasperformedbut"noobjectioncanberaisedagainstthestatisticalsignificanceofthenumbers,"whichweredeemedcorr

6、ectbyaninspectorofthesickbenefitassociation(in1900 Pearsoninventedthe2test,10whichwouldhaveshownP=0.0003).Therateofserumsicknesswashighat60%.Tables-2-by-2unstratified11:44:19,2011-4-22

7、+-

8、Total-------+-------------+-------+

9、8231

10、239-

11、30215

12、245-------+-------------+-------Total

13、38446

14、484Testso

15、fsignificanceFisherexacttest(onetailed):0.000187Fisherexacttest(twotailed):0.000289Uncorrectedchi-square:13.24p-value:0.000275YatescorrectedChi-square:12.04p-value:0.000522Measuresofexposureeffect[95%CI]Riskratio:0.27[0.13,0.58]Oddsratio:0.25