基于风险调整的城镇居民基本医疗保险门诊统筹按人头付费测算模型研究

With the launching of outpatient pooling fund of the urban residents' basic medical insurance, an increasing number of areas have adopted capitation payment. It is a scientific and practical issue to find a way to establish a standard for capitation payment according to the risk and medical expenses of different groups served by different designated medical institutions and the financing level of outpatient pooling fund. Based on the theories and methods of Actuarial Science of medical insurance at home and abroad, this study will summarize the similarities and differences of capitation payment of outpatient pooling fund in the urban residents' basic medical insurance in different areas of China. Multilevel models will be used to seek and compare the risk factors of outpatient health service expenditure in different designated medical institutions. What's more, four-part model, propensity score matching and other approaches will be adopted to analyze the outpatient services expenditure and relevant reimbursement expenditure of the outpatient pooling fund in the designated medical institutions. Based on these upon, the total financing allocation standard of outpatient pooling fund among different designated medical institutions will be established. Thus, we will build a capitation payment measurement model based on risk-adjusted method. Meanwhile, according to the changes of outpatient pooling fund and reimbursing strategy, the preliminary model will be adjusted and modified. Finally, a theoretical model, which can be tested by and applied to practice, will be established to provide calculating basis for defining a standard for capitation payment of outpatient pooling fund of the urban residents' basic medical insurance among different areas.

随着我国城镇居民基本医疗保险门诊统筹的普遍开展，越来越多的统筹地区采用按人头付费的支付方式。如何根据不同定点医疗机构服务人群的风险、医疗消费支出差异和门诊统筹的筹资水平确定适宜的按人头付费支付标准是一个亟待解决的科学问题和现实问题。本研究拟借鉴国内外医疗保险精算学的有关理论和方法，通过总结我国各地城镇居民基本医疗保险门诊统筹按人头付费模式的异同，运用多层统计分析模型寻找和比较不同定点医疗机构服务人群门诊医疗消费支出的风险因素，结合四部模型法、倾向得分匹配法等方法分析各定点医疗机构服务人群的门诊医疗消费支出和相应的门诊统筹基金补偿支出，并以此为依据确定出门诊统筹筹资总额在各定点医疗机构中分配的标准，从而建立基于风险调整的按人头付费测算模型。同时根据门诊统筹筹资和补偿方案的变化，对上述模型进行适当调整和校正，最终构建出能够为实践所检验和应用的理论模型，为各地确定按人头付费支付标准提供计算依据。

随着我国医药卫生体制改革不断深化，基本医疗保险支付方式改革已经越来越重要，对基层医疗卫生机构提供的普通门诊实行按人头付费是其中重要的组成部分。本项目针对各地按人头付费实践中往往未根据不同参保人就医和费用的风险差异，制定有差别的人头费标准，这一普遍存在的问题展开研究，寻找可能的按人头付费风险调整因素及风险调整方法。本研究以广东省深圳市、河北省唐山市、湖北省崇阳县和湖北省大冶市作为典型调查的样本城市，对其城镇居民基本医疗保险参保对象的社会经济状况、普通门诊服务利用和医疗费用情况、基层医疗卫生机构的服务提供以及医保经办机构的运行情况进行了抽样调查。根据获取的数据特征，主要运用Logistic回归和广义线性模型分别建立了普通门诊服务利用和医疗费用的拟合模型，最终提取了性别、年龄、参保档次、是否患有慢性病或大病等可能的风险调整因素，进而使用分组计算法和两部模型法对不同风险组人群的医疗费用进行了预测。同时结合各风险组人数、年龄组合并等，进一步将初步提出的按人头付费风险组进行了适当压缩，并根据各风险组的预期医疗费用和实际补偿比，计算了各风险组的人头费用及相对权重。本研究中发现的风险调整因素，可供各地开展按人头付费风险调整时参考；研究使用的方法也可为各地具体测算时所运用；各风险组的相对权重也可为今后进行动态调整时所借鉴。本研究的政策建议包括进一步完善基本医疗保险信息系统，如增补门诊诊断等必要信息；根据各地差异，因地因时地制定系统策略，分步骤将风险调整因素逐渐纳入到按人头付费中等。随着我国城乡居民基本医疗保险的建立，如何在城乡一体化的情况下进一步做好按人头付费中的风险调整，还有待于后续进一步研究。