慢性肺心病心电图的多元回归分析

Diagnostic Value of Electrocardiogram in Cor Pulmonale by Means of Multiple Regression

本文对335例心电图30项指标进行多元回归判别分析.其诊断敏感性从现行诊断标准的77.2%提高到96.1%(p<0.01),特异性从92.1%提高到99.3%(P<0.01),表明应用心电图多因素判别分析将慢性肺心病的现行诊断指标与有争议的参考指标有机地结合,可有效地克服各项指标的假阳性和假阴性,提高对慢性肺心病诊断的敏感性和特异性.

In order to raise diagnostic sensitivity,specificity and correct classification rate of electrocardiogaram in cor pulmonale,electrocardio- grams from 127 patients withcorpulmonale,57 patients with pulmonary emphysema and 151 normal subjects were studied.By using 30 items,a mul- tiple regression formula was established by means of stepwise regression. The result showed that multiple regression analysis raised the sensiti- vity from 77.2% judged by conventional diagnostic criterion to 96.1% (P&lt;0.01);specificity,from 92.1% to 99.3%(P&lt;0.01);correct classifica- tion rate,from 85.3% to 97.8%(P&lt;0.01)