【摘要】 目的探讨判别统计方法在新生儿先天性肾上腺皮质增生症(CAH)筛查中的应用价值。方法对上海交通大学医学院附属新华医院新生儿筛查中心2008年至2011年CAH筛查资料进行回顾分析,根据CAH筛查指标17羟孕酮(17-OHP)传统切值30 nmol/L,随机选择17-OHP正常组11 130例及阳性组398例为研究样本。采用非标准化Fisher判别方法,建立包含出生孕周、体重及17-OHP浓度三项因素的判别公式及阳性判别切值;并用已确诊的24例CAH患者及381例假阳性样本进行验证。结果建立判别函数公式Z=(-0.019·x1)+(-0.00012·x2)+0.055·x3+0.471;阳性判别切值为1.314。验证显示24例CAH患者判别结果 100%符合传统方法;降低了27%假阳性率。结论融入多因素的判别分析法用于新生儿CAH筛查可显著降低假阳性率,简便可行。更多还原

【Abstract】 Objective To evaluate the discriminant method using for the neonatal congenital adrenal hyperplasia(CAH) in neonatal screening. Methods The retrospective CAH screening data during 2008-2011 from our center were investigated. The source data was separated into 2 group based on the traditional CAH screening indicators 17-OHP cut-off 30 nmol/L. 11 130 cases with 17-OHP value<30 nmol/L were selected randomly as the normal group, 398 cases including congenital adrenal hyperplasia patients(CAH) with 17-OHP value ≥30 nmol/L were set as the positive group. A discriminate equation and a new positive cutoff which involving the neonatal gestational age, weight, and 17-OHP value were set up by using of the unstandardized Fisher analysis program. All positive samples including 24 CAH patients and 381 false positive cases which were validated by the traditional 17-OHP cut off were re-evaluated with new set criterion. Results A discriminant equation base on the source data were established as Z=(-0.019·x1)+(-0.00012·x2)+0.055·x3 +0.471, and the new positive cut off was set as 1.314. The data of 24 real CAH patients was verified with new set discriminating formula, and the result was confirmed as 100%(24/24) agreed as its previous diagnosis. 27%(156/573) false positive cases were change into negative after the data was re-evaluated by new method. Conclusion The multi-factor discrimant analysis for neonatal CAH screening will expect to significantly reduce the false positive rate, therefore it show more reasonable and feasible in neonatal screening.更多还原