【摘要】 目的了解不同指标对危重型手足口病的预测能力。方法收集2009年1-12月在郑州市某儿童医院住院的455例重型及危重型手足口病资料,用logistic逐步回归筛选危重型手足口病的影响因素,建立预测模型,用ROC曲线评价不同指标对危重型手足口病的预测能力及最佳界值。结果综合了发热持续时间、血白细胞计数和空腹血糖的危重型手足口病预测模型的曲线下面积(AUC)为0.819,界值为0.085,灵敏度为72.7%,特异度为93.1%,阳性似然比为10.54,阴性似然比为0.29,一致率为88.59%,kappa值为0.67,阳性预测值为75.00%,阴性预测值为92.31%,均优于单一指标。结论联合外周血白细胞计数、空腹血糖和发热持续时间的回归模型有助于提高对危重型手足口病的预测能力。更多还原

【Abstract】 Objective To explore the predictive power of different indexes for critical hand-foot-mouth disease( HFMD). Methods Medical records of 455 severe and critical HFMD cases hospitalized in some children hospital in Zhengzhou city from January to December in 2009 were collected,and the influencing factors of critical HFMD were screened and the predictive model was set up with logistic regression. Then predictive power of different indexes for critical hand-foot-mouth disease( HFMD) was evaluated and the cut-off points were determined with receiver operating characteristic curve( ROC curve).Results The area under ROC curve( AUC) of the predictive model for critical HFMD combined with fever duration,white blood cell count and fasting blood glucose,cut-off point,sensitivity,specificity,Youden’s index,positive likelihood ratio,negative likelihood ratio,Kappa,concordance rate,positive predictive value and negative predictive value was 0. 819,0. 085,72. 7%,93. 1%,65. 8%,10. 54,0. 29,0. 67,88. 6%,75. 0% and 92. 3%,respectively. All the data above surpassed that of either parameter alone. Conclusion The model of logistic regression combined with white blood cell count,fasting blood glucose and fever duration is helpful to improve the predictive power for critical HFMD.更多还原