基于Apriori算法的糖尿病患者用药规律关联规则挖掘分析

Mining and Analysis of Medication Rules for Diabetic Patients based on Apriori Algorithm

目的挖掘糖尿病住院患者的电子病历信息,探析其高频用药组合,了解糖尿病患者的用药现状,为我国制定卫生政策提供科学依据。方法以糖尿病住院患者的电子病历信息为研究对象,合并诊疗数据,摘录用药信息并组建数据库,对药物进行频次统计分析,建立Apriori模型,进行关联规则分析。结果获得联合用药关联规则62则。控糖与防治并发症是糖尿病的主要治疗措施。结论通过对高频药物药性及关联规则分析,得出在治疗糖尿病时最常使用的药物,符合糖尿病治疗的用药特点,与糖尿病治疗的临床指南相符,从而为医学专业人员了解疾病的诊疗过程及用药规律提供了不同视角与可行性路径。

Objective To mine electronic medical record information of diabetic patients,explore the high-frequency combination of medications,understand the medication status of diabetic patients,and provide scientific basis for the development of health policies in China.Methods Electronic medical record information of diabetic patients was used as the research object,and diagnostic and treatment data were merged.Medication information was extracted and a database was established for frequency statistical analysis of medications.An Apriori model was constructed for association rule analysis.Results A total of 501 diabetic diagnosis records were obtained,and there were 20 medications with a usage frequency of more than 39 times.A total of 62 combined medication association rules were obtained.Glycemic control and prevention of complications are the main treatment measures for diabetes.Conclusion Through the analysis of high-frequency medication properties and association rules,the most commonly used medications in the treatment of diabetes were identified,which are consistent with the medication characteristics of diabetes treatment and clinical guidelines,providing different perspectives and feasible pathways for medical professionals to understand the diagnosis and treatment process and medication rules of the disease.