摘要
由经验性的引深为伦理性的应用危险因素辅助诊断在于研讨其真实性,即其灵敏度(sen)和特异度(spe),分别可借助病例对照研究2x2表计算,由此又可计算比值比(OR)、推导sen和阳性似然比(LR+)。spe对LR+的影响显著大于OR值。危险因素的sen与spe也呈互补关系,参考暴露剂量提高诊断价值实质足增大了spe。sen和spe是来自样本的估计,也有它在总体中的可信区间。危险因素辅助诊断的价值还受着验前概率、两种以上危险因素联合使用、修饰因子等因素的制约。
To logically,not empirically use the risk factors for theassistance in the diagnosis of diseases
rclies on the study ofthe validity in respect of risk factors,1.e its sensitivity(sen)and specificity
(spe).By the aid of the tabled 2x2from case- control study ,the sen and Spe can be estimated.
Thus sen and positive likelihood(LR+ )can be also in-ferred. The influence of spe on LR+ iS
more remarkblethan that of OR, if spe was less than 50 precent, OR doesnot distinctly influence
on LR+ evcn though OR is increasedto 100. There is a compensatve relationship between
senand spc,When two risk factors are used as narallel or seriesconnection, the sen or spe will
bc promoted.The rising ofthe diagnosis value with increasing the dose of cxposure infact
expends the spc of risk factor.The confidence intervalof sen and spe must be estimated in
population sincc it is de-rived from the sample,The prior disease probability influ-ences the
utility of risk factors in diagnosis of discase bymeans of Bayes'theorem,In addition.it is also
influencedby the modification factor.
出处
《中国慢性病预防与控制》
CAS
1994年第2期77-79,共3页
Chinese Journal of Prevention and Control of Chronic Diseases
关键词
危险因素
诊断
流行病学
临床
Risk factor Diagnosis ValiditySensitivity
Specificity OR Bayes' theoren Modifica-tion,factor