剖宫产手术预防性使用抗菌药的数据挖掘

目的 建立、比较和评价剖宫产手术抗菌药物预防使用的分类模型,为针对性干预打下基础。方法 应用数据挖掘软件PASW^® Modeler 13,建立分类模型,获得对抗菌药物预防性使用影响较大的变量(临床因素)。结果 由787例行"子宫下段剖腹产术"的病例数据获得的分类模型中,以贝叶斯网络,logistic回归和CHAID 3个模型总体较佳; 21个变量指标中,"失血量"是对该医院剖宫产手术抗菌药物预防性应用影响程度最大的因素。结论 数据挖掘技术,可以快速地建立反映剖宫产手术抗菌药物预防性使用的分类模型,为药物利用调查提供了新的分析方法。

Data mining of antibiotic prophylactic use for cesarean section patients

Objective To establish, compare and evaluate the classification models of antibiotic prophylactic use for cesarean section patients for the targeted intervention in future. Method PASW^® Modeler 13 was applied to establish classification models and to get the influential variables (clinical factors) in antibiotic prophylactic use. Results With the data of 787 cases, the classification models were established, in which, Bayesian networks, logistic regression and CHAID were better. In 21 clinical factors, blood loss was the most influential variable. Conclusion The data mining technique was able to quickly create models reflecting the use of prophylactic antibiotics use for cesarean section, which would provide a new analysis tool for drug use survey.