宫颈癌筛查试验评价方法的研究  

A Methodology Study on Assessment of Cancer Screening Test

作　　者：高国福[1] 赵方辉[1] 武燕萍[1] 戎寿德[1] 章文华[2] 潘秦镜[3] 李凌[3] 乔友林[1] 

机构地区：[1]中国医学科学院中国协和医科大学肿瘤研究所流行病学研究室,北京100021 [2]Department of Gynecological Oncology,Cancer Hospital,CAMS and PUMC,beijing100021 [3]Department of Pathology,Cancer Hospital,CAMS and PUMC,beijing100021

出　　处：《中国医学科学院学报》2002年第6期573-576,共4页Acta Academiae Medicinae Sinicae

摘　　要：目的分析诊断金标准、盲法和临界值选择对筛查试验评价的影响,促进循证医学理论在筛查试验中的应用。方法利用1999年完成的一项1997人的以病理组织学诊断为金标准的宫颈癌随机盲法筛查试验结果,模拟诊断金标准、盲法和临界值的作用,计算检出率、灵敏度、特异度等评价指标,绘制受试者工作特性(ROC)曲线,比较ROC曲线下面积。结果没有诊断金标准时无法区分诊断试验间的优劣,不使用盲法会使筛查试验的诊断价值人为地增高,在不同的临界值水平筛查试验的灵敏度和特异度不同。结论诊断金标准、盲法的使用和选择恰当的临界值对提高筛查试验的质量有重要作用,绘制ROC曲线是确定筛查试验临界值和评价筛查试验的有效方法,应在医学科研中大量运用循证医学的理念。Objective To analyze the effect of gold standard,blind comparison and diffe rent cut -points choosing on screening techniques assessment,and to promote th e application of evidence-based medicine theory in screening study.Methods A s creening study for cervical cancer in rural China in1999,where1997women had been tested for pathology as gold standard and simulating situations withou t gold standard,blind comparison and under different cut -points.Indices suc h as detectable rate,sensitivity and specificity were calculated for each techn ique.Receiver operating characteristics(ROC)curves were drawn and areas under ROC curves between screening techniques were tested.Results Without gold stan dard,diagnostic techniques could not be evaluated correctly,and without the b lind comparison,the sensitivity and specificity of the tests would be subjectiv ely increased.Furthermore,use of different cut -points led to different sens itivities and specificities of test.Conclusions Gold standard,blind compariso n and perfect cut -points can improve the quality of screening test and drawing ROC curves is an effective way to confirm cut -points and evaluate diagnostic te chniques.It is necessary to enforce the application of evidence-based medicine theory in scientific research.

关 键 词：宫颈癌 循证医学 筛查试验 模拟分析 ROC曲线 诊断金标准 盲法 

分 类 号：R737.33[医药卫生—肿瘤]