Sensitivity Specificity PPV NPV课件.ppt

灵敏度，特异度，阳性预测值，阴性预测值Sensitivity, Specificity, Positive Predictive Value, Negative Predictive Value;思考题1;思考题2;思考题3;思考题4;灵敏度和特异度 Sensitivity Specificity;100人中有30人患结核病 There are 100 people with 30 having TB 用皮肤试验（TST）来判断谁患病，谁没有患病 A test designed to identify who has the disease and who does not 现在要来看看这种方法有多好 We want to evaluate how good the test is;; 灵敏度=P (检测+ | 疾病+) Sensitivity = P (test+ | disease+) 特异度=P (检测- | 疾病-) Specificity = P (test- | disease-);通常是根据某种“金标准”来确定 Usually the ‘true’ disease status is determined by some ‘gold standard’ method 例：用QuantiFERON-TB Gold 判定是否患有TB;对特定检测方法，灵敏度? ? 特异度?；特异度? ? 灵敏度? For a specific test, sensitivity increases as specificity decreases and vice versa 将TST 5mm 定为（+）： 灵敏度很高（绝大多数TB感染病例都会被这种方法发现） 特异度较低（许多TST（+）者都不是真正的TB感染病例） 将TST 20mm 定为（+）： 特异度很高（绝大多数TST（+）的都是真正的TB感染） 灵敏度很低（用这样的判定标准会漏掉许多TB感染病例）;结核病病例定义Use TST for TB case definition;一种检测方法具有极高的灵敏度与特异度，它是不是就是一种十全十美的方法？ Is a test perfect if it has high sensitivity AND high specificity? 例如：某100万人群中，HIV阳性率是0.1% Suppose there are 1 million people with 0.1% infected with HIV 一种监测HIV病毒的方法灵敏度与特异度均为99.9% A test can identify HIV infected people with 99.9% sensitivity and 99.9% specificity;;;;阳性预测值，阴性预测值 Positive Predictive Value Negative Predictive Value ;患病率越低，阳性预测值越低，阴性预测值越高The lower the prevalence, the lower the PPV, and the higher the NPV 患病率越高，阳性预测值越高，阴性预测值越低The higher the prevalence, the higher the PPV, and the lower the NPV;思考题1;思考题2;思考题3;思考题4