新生儿持续肺动脉高压危险因素的Logistic回归分析

Logistic regression analysis of risk factors for persistent pulmonary hypertension of newborn infants

目的 探讨新生儿持续肺动脉高压（PPHN）发病的危险因素.方法 对41例PPHN新生儿进行1∶1配对病例对照研究,采用Х^2检验、t检验或方差分析等方法进行单因素分析,选择单因素分析中P值小于0.05的因素进行多因素Logistic回归分析,建立PPHN发病的Logistic回归预测模型,并对模型进行初步评价.结果 多因素Logistic回归分析显示胎粪吸入综合征（OR =52.391,95％ CI：3.899～703.959）、孕母糖尿病（OR=49.758,95％CI：3.544～698.686）、新生儿肺炎（0R=49.637,95％CI：6.143 ～401.100）、宫内发育迟缓（OR=26.666,95％CI：2.997～237.282）、新生儿肺透明膜病（OR=15.200,95％CI：2.862～ 80.726）与PPHN发病明显相关.预测回归模型的拟合优度的Х^2检验结果为23.681,模型检验有统计学意义（P＜0.01）.结论 胎粪吸入综合征、孕母糖尿病、新生儿肺炎、宫内发育迟缓、新生儿肺透明膜病为PPHN发病的主要危险因素.

Objective To explore the risk factors for persistent pulmonary hypertension of the newborns （PPHN）.Methods An 1 ∶ 1 matched case-control study was conducted in 41 cases of PPHN.Chi-square,t-test or ANOVA were used to perform univariate analysis.All factors for P values less than 0.05 were used as independent variables for multivariate analysis of Logistic regression,and a predictive model of Logistic regression for the genesis of PPHN was established,and the model was evaluated preliminarily.Results By multivariate analysis of Logistic regression,meconium aspiration syndrome （OR =52.391,95 % CI：3.899-703.959）,matemal diabetes （OR =49.758,95 % CI：3.544-698.686）,neonatal pneumonia （OR =49.637,95% CI：6.143-401.100）,intra-uterine growth restriction （OR =26.666,95% CI：2.997-237.282）,hyaline membrane disease （OR =15.200,95 % CI 2.862-80.726） were significantly related to genesis of PPHN.Goodness of fit for predictive model of Logistic regression was checked by Chisquare and Х^2 value was 23.681.It indicated model check was statistically significant （P 〈 0.01）.Conclusions Meconium aspiration syndrome,maternal diabetes,neonatal pneumonia,intra-uterine growth restriction,and hyaline membrane disease are major risk factors for genesis of PPHN.