成分数据预测模型在住院费用构成中的应用

Application of Compositional Data Forecast Model in Medical Expenses Structure and Tendency

目的:利用HIS系统信息,研究不同支付方式下医疗费用结构的变动规律,为制定医疗费用结构控制措施提供理论依据和方法学基础。方法:运用球坐标变换方法,以上海市某综合性三甲医院2007—2010年住院患者费用的变动情况为例进行分析。结果:不同支付方式下的药品费和手术费均占较大比重,且自费患者比医保患者更高。随时间变化,药品费构成呈下降趋势,手术费构成呈上升趋势。自费患者除输血费转换角与时间无线性回归外,其他各项费用均有统计学意义;医保患者除治疗费和输血费转换角与时间无线性回归外,其他各项费用均有统计学意义。结论:可以通过HIS系统进行数据挖掘。采用球坐标变换用于医疗费用成分数据资料是可行的。社会医疗保险体系正逐步步入正轨,医保患者住院费用得到了一定的控制。

Objective： To analyze the medical expenses structure and tendency among inpatients, and to master the constitution of inpatients＇ medical expenses under different ways of payment through HIS. Methods： Analyzing and comparing medical expense of inpatients in a hospital in Shanghai through hyperspherical transformation. Results： The drug expenses and operation expense constitution is the largest, and the expense constitution is higher among self-paid inpatients than among medical insured inpatients, Drug expense constitution show a descending trend while operation expense constitutes shows an ascendant trend. The regression models are significant except blood transfusion expense among self-paid inpatients while the regression models are significant except treatment and blood transfusion expense among medical insured inpatients. Conclusion： Data mine could be done based on HIS. It is feasible that hyperspherical transformation is used in medical expense constitute analysis. The development of social insurance system progressed smoothly and the medical expense of basic medical insured has been controlled to a certain extent.