Diagnostic performance of 23 candidate 1-of-n rules.

Sensitivity, specificity, and likelihood ratio negative (LRN) from the bivariate random-effects meta-analysis model. Rule is at least one of the indicated symptoms. C, current cough; H, haemoptysis; F, fever; S, sweats; W, weight loss.ap-Value >0.05 for the same sensitivity of the CFSW rule and the indicated rule. All specificities are significantly different from that of the CFSW rule.LRN, likelihood ratio negative.本数据集提取自双变量随机效应荟萃分析模型（bivariate random-effects meta-analysis model），包含敏感性、特异度及阴性似然比（likelihood ratio negative, LRN）。其判定规则为至少存在一项指定症状：C代表现症咳嗽（current cough）、H代表咯血（haemoptysis）、F代表发热（fever）、S代表盗汗（sweats）、W代表体重减轻（weight loss）。CFSW判定规则与对应指定规则的敏感性无统计学差异（P值>0.05）；所有指定规则的特异度均与CFSW规则的特异度存在显著统计学差异。LRN即阴性似然比（likelihood ratio negative）。