非综合征性唇腭裂高危因素条件Logistic回归与数据挖掘相结合的临床研究

Clinical study of conditional Logistic regression and date mining on risk factors of nonsyndromic cleft lip and palate

目的探讨非综合征性唇腭裂(NSCL/P)发病的主要危险因素;确立NSCL/P发病概率的预测模型,为优生网络的构建奠定基础。方法采用1∶1配对病例对照研究,病例组为年龄在0～12岁之间的NSCL/P患儿126例;对照组来源于同一机构门诊或病房或同一居住区符合配对条件的非唇腭裂患儿。根据危险因素编制调查表,对病例组与对照组患儿父母进行调查,数据经审核后录入Excel2003建立数据库。首先使用条件Logistic回归对资料进行单因素分析,再对单因素筛选的变量结合专业知识进行多因素分析,筛选主要危险因素并建立回归模型,根据危险因素分别建立分类树与LogitBoost算法的发病概率预测模型,采用ROC曲线对两模型进行评价,从而确立本研究中NSCL/P发病概率的预测模型。结果病例组与对照组作对照分析,进入条件Logistic回归模型的变量有:母亲孕期感染史(P=0.011)、家族遗传史(P=0.008)、母孕期饮食是否规律(P=0.005)、胎次(P=0.003)、母亲孕期异常情绪史(P=0.001)、父亲学历(P=0.000)。经ROC曲线评价,确立分类树模型可用来预测NSCL/P的发病概率。结论母亲孕期感染、家族遗传、母亲孕期饮食不规律、胎次、母亲孕期异常情绪是NSCL/P发病的促进因素;父亲学历是该病的保护因素。经ROC曲线评价,最终确立分类树模型为NSCL/P发病概率的预测模型。

Objective To discuss the main risk factors of nonsyndromic cleft lip and palate （NSCL/P）, to propose the reasonable predictive model for aristogenesis net. Methods A hospital-based 1：1 matched case-control study design was applied to our epidemiological study. 126 nonsyndromic cleft lip and palate children, the age in the 0 - 12 year-old were selected from some hospitals in Shandong province from Sep,2006 to Sep,2007 as collecting cases, and control cases were chosen from the same section or same living area with the matched patients. A self-administered questionnaire prepared according to risk factors of NSCL/P was used to collect information, that is, the parents of patients and controls of NSCL/P were asked some questions about these risk factors and the answers were filled in the questionnaire by investigators. After the investigation, these data were cleaned up by pair and input into computer to establish data base in Excel 2003. Carried on the multivariate stepwise regression analysis in univariate condition Logistic regression analysis foundation, discovered the risk factors from multitudinous possible influencing factors and establish the regression model. Established the classification tree and the LogitBoost predictive models separately according to the risk factors, to appraise the two models using the ROC curve, established the reasonable predictive model of NSCL/P. Results There were 6 risk factors in the model which were related to NSCL/P, maternal infectious history （P=0.011）, genetic family history （P=0.008）, maternal disorder diet （P=0.005）, birth order （P=0.004） ,maternal stress during the first trimester of pregnancy （P=0.001）, fathers＇ education level （P=0.000）. After the evaluation with ROC curve, established the classification tree model which was used to predict the morbidity of NSCL/P. Conclusions Maternal infectious history, genetic family history, maternal disorder diet, birth order and maternal stress during the first trimester of pregnancy, were promoting factors of NSCL/P, Fathers＇ education level was a protecting factor. According to the investigation, we established the classification tree model as the predictive model of the morbidity of NSCL/P.