A Study Of Risk Prediction Model Of Birth Defects At Individual Level Based On System Evaluation

Objective:Sift the risk factors of Birth Defects based on literature system evaluation. At the same time, according to the principle of disease risk assessment, using the main risk factors of birth defects and the basic principle of the logistic model and Rothman-Keller model,birth defects in the individual risk assessment model at individual level is established. The model is used to identify people at high risk, so as to reduce the risk of birth defects, to provide reference basis for the primary prevention. The data is collected from six counties of Shanxi Province in 2006-2008 years. To explore the factual usage of the methods in the population of six counties of Shanxi Province.Method:The major risk factors of birth defects were collected by retrieval of literature’s Search,review and evaluate the Meta analysis of birth defects to find out the quantitative relationship between risk factors and birth defects. According to the calculation methods of the logistic model and Rothman-Keller model to establish the risk prediction model of birth defects. Predicted the birth defects and compared with the actual outcome, and evaluate the accuracy of the model.Results:20 major risk factors of birth defects were collected including: age of pregnant woman, passive smoking during pregnancy, the habits of eating fresh fruits and vegetables and proteins, taking folic acid during pregnancy, cold or fever during pregnancy, anemiaduring pregnancy, chronic diseases, the TORCH infection during pregnancy, negative mental stimulation, taking sedatives during pregnancy, taking antibiotics during pregnancy,family history of birth defects, abnormal pregnancy history, intermarriage, exposure to pesticides during pregnancy, exposure to harmful physical factors during pregnancy,exposure to pets during pregnancy, exposure to pollution residence, father smoking, father drinking often. Change the OR of factors to β and risk score conversion table of birth defects. Calculate the p value and the birth defects risk of the simulated data. Through the random sampling of the survey data of six counties in Shanxi province in 2006-2008 years,made the average incidence rate 5.6%. The actual incidence was 12.2% in high-risk group in the logistic model, and the actual incidence was 10.7% in high-risk group in the Rothman-Keller model. Compared with other groups, the difference was statistically significant. The OR value of the family history of birth defects is the biggest in the general survey data. 86.6% of population with family history was assigned to the high-risk group in the logistic model. And 100% of population with family history was assigned to the high-risk group in the Rothman-Keller model.Conclusion:In this study, our country’s secondary literature of birth defects were used, and the relationship between main risk factors and birth defects were found out. The model of risk prediction of birth defects were established which could be used to predict individual potential risk of morbidity of birth defects and screen high-risk groups. It will lay a foundation for health education, and provide reference for the etiology of the birth defects prevention, and improve the individual consciousness about birth defects. And it will provide a useful reference for other disease.