Objective: To explore the characteristics of the military hospitalized patients by clustering method, and to provide better medical service for the military patients. Methods: 2038 patients admitted to a military hospital between January 2005 and June 2016 were included, K-means clustering method is used to cluster military hospitalized patients according to pseudo F statistic index, the characteristics of different types of hospitalized patients were analyzed by chi-square test and variance analysis. Results: The military hospitalized patients were clustered into six groups: normal clinic patients, patients with modality, surgery patients with high value material, simple operation patients, patients with long length of stay, patients with examination and laboratory test. There were statistical significance between the patient's age, admission form, admission conditions, number of illness, length of stay, and the total cost of six groups. Conclusion: The clustering analysis of military hospitalized patients could be used to find the problems of the military medical health service, and the suggestions were put forward to improve the quality of medical service for military patients.%目的:通过聚类方法分析军队住院病人特征,提高军队病人的医疗保障水平.方法:提取某军队医院2015年1月1日到2016年6月30日军队住院病人2038例,采用K-means聚类方法进行聚类,使用卡方检验和方差分析分析住院病人特征.结果:军队住院病人聚为6类:正常就诊病人、以仪器治疗为主的病人、使用高值耗材的手术病人、小手术病人、压床病人、检查和化验病人.各类病人的年龄、入院方式、入院病情、诊断个数、住院日和住院总费用差异均有统计学意义.结论:通过对军队住院病人进行聚类分析,发现了军队医疗保障中存在的问题,对提升军队病人的医疗保障水平提出了建议.
展开▼
