计算机病例分型模型的研究与应用  

Research and use of the computerized cases classification model

作　　者：覃国强[1] 杨云滨[2] 莫嘉颖[3] 郭睿 

机构地区：[1]东莞市人民医院,广东东莞523059 [2]广东医学院 [3]华南师范大学 [4]广东医学院附属松山湖医院

出　　处：《中华医院管理杂志》2013年第12期939-941,共3页Chinese Journal of Hospital Administration

基　　金：东莞市软科学研究项目（东科[2011174号）

摘　　要：目的建立计算机病例分型方法，为完善病种分类分型体系以及按病种的质量与费用的管理提供依据。方法对875份病例进行定义分型，并分为两种情况进行分析：在非手术情况下，筛选出病种诊断数、入院时的严重程度、患者年龄等8个判别参数；在手术情况下，筛选出诊断数、手术量、入院时的严重程度等5个判别参数。然后使用Fisher判别得出函数判别方程，实现计算机判别分型。结果计算机病例分型与手动分型的吻合率约为86．2％。所得出的函数判别方程准确率较高，这说明了计算机病例分型的判别效果比较好。结论计算机病例分型的判别效果令人满意，因此将其应用在临床实践中，能更好地进行医疗质量或医疗费用管理。Objective To build the method of computerized case classification, for the purpose of perfecting the diseases typing and classification, and supporting the management based on the quality and expenses of the disease. Methods 875 cases were defined into two types. Eight parameters were selected for non-surgery cases, namely the disease diagnosis, severity at admission, and age. For surgery cases, five parameters were defined, namely the disease diagnosis, operative quantity, and severity at admission. Then Fisher function was called into play to obtain the function descriminant equation, realizing computerized classification. Results The rate of matching was 86.2~between computerized classification and manual classification. The high accuracy of function descriminant equation proves the satisfactory outcomes of the classification. Conclusion The computerized classification is satisfactory in its outcomes, and therefore it can better quality of care and cost management of diseases in clinical practice.

关 键 词：计算机分型 病例分型 判别分析 

分 类 号：R197.3[医药卫生—卫生事业管理]