背景：医保支付改革持续成为深化医改的工作重心，我国当前实行以DRG、DIP为代表的多元复合医保支付方式，并在多地进行试点工作，已经收获长足的发展，但仍存在一定的优化空间。政策实施的真实效果需要真实世界的数据进行评价，当前对于DRG和DIP对于控制医疗支出、提高医疗效率的对比研究尚浅。目的：通过对病案首页数据的基础分析，描述深圳市住院患者的基本情况，继而通过分析不同时期各医院住院患者在干预前后医疗费用、平均住院时长的差异，评价DRG与DIP对控制医疗费用与提升服务效率的效果。最后，结合实证研究结果为深圳市后续的医保支付改革提供量化的证据并提出政策建议。方法：本研究数据库提取自2016年1月1日至2023年6月30日深圳市各医院提交的西医病案首页，对患者人口学信息、住院费用相关信息以及费用结构采用描述性统计，对连续变量采取平均值（标准差）的形式进行分析。将住院总费用、患者自付金额、自付占比以及平均住院时长作为核心因变量进行回归分析，使用双重差分法（DID）对DRG实施效果进行分析，使用间断时间序列分析（ITSA）对DIP实施效果进行分析，使用多元线性回归对DRG与DIP实施效果进行对比。本研究中所有数据处理通过R软件（3.2.1）实现。结果：实行DRG之后，患者住院总费用的上升趋势大幅度减缓，自付金额显著下降，平均住院时长显著缩短。DIP实施之后，三甲医院患者住院总费用显著下降，自付金额与自付占比显著上升，平均住院时长显著缩短；三级医院患者住院总费用显著下降，自付金额和自付占比均显著上涨，平均住院时长显著缩短；二级医院患者住院总费用呈现下降趋势但影响并不显著，自付金额和自付占比显著下降，平均住院时长有所缩短但影响并不显著。在DRG与DIP同期实施效果对比中发现，相比于DRG，DIP对于患者住院总费用、自付金额与自付占比的控制效果更好，而DRG对于缩短平均住院时长的效果更加显著。结论：本研究发现DRG和DIP作为医保支付改革的有力手段，在保证医疗效果及服务量的前提下，均有助于控制住院费用、降低医保基金风险、提升医疗服务效率，在实施过程中有各自的优势。建议医院内部提高多维度管理水平、医保方结合大数据分析加强监管力度，配合多元复合医保支付方式最大化医改效果。
Background: The ongoing reform of healthcare payment systems remains a pivotal aspect in the profound advancement of China‘s medical reforms. Currently, China embraces a multi-dimensional, composite approach to health insurance reimbursement, exemplified by Diagnosis-Related Groups （DRG） and Diagnosis-Intervention Packet （DIP）, which have undergone substantial progress through pilot programs in multiple regions. The genuine impact of such policies necessitates evaluation using real-world data; at present, comparative studies examining the effects of DRG versus DIP on controlling healthcare expenditures and enhancing service efficiency are relatively underdeveloped.Objective: This study aims to describe the essential attributes of hospitalized patients in Shenzhen through a foundational analysis of the data from EHR （electronic healthcare records）. Further, it seeks to assess the impact of DRG and DIP in managing healthcare expenditures and improving efficiency by comparing the changes in medical costs and LOS （length of stay） at various benchmark hospitals. Ultimately, the study aims to furnish robust, quantifiable evidence that informs future iterations of the medical insurance reimbursement scheme in Shenzhen and culminates in evidence-based policy suggestions.Method: This study extracted data from the Electronic Health Records （EHR） in Shenzhen between January 1, 2016, and June 30, 2023. Further sample screening was conducted using distinct inclusion and exclusion criteria for evaluating the impact of DRG and DIP. Descriptive statistics were employed to analyze patient demographic information, hospitalization expense-related data, and cost structures. Continuous variables were analyzed in terms of mean value and standard deviation. The core dependent variables include total costs, out-of-pocket expenses, the proportion of out-of-pocket expenses, and length of stay were subjected to regression analysis. The impact of DRG was evaluated using a difference-in-differences （DID） model, while the impact of DIP was assessed through interrupted time series analysis （ITSA）, the comparative analysis of the impact between DRG and DIP was assessed by multiple linear regression model. All data processing was performed using R software version 3.2.1.Result: After the implementation of Diagnosis Related Groups （DRG）, the escalating trend of total costs of inpatients was markedly mitigated, accompanied by a significant reduction in out-of-pocket expenses and a conspicuous shortening of the length of stay. After the implementation of the Diagnosis-Intervention Packet （DIP）, there was a prominent decrease in total costs of inpatients in Grade III Class A hospitals, coupled with substantial increases in both the amount and proportion of out-of-pocket expenses, alongside a significant reduction in the length of stay. In the case of Level III hospitals, total costs also dropped significantly, with parallel escalations in out-of-pocket expenses and proportion of out-of-pocket expenses, and similarly shortened length of stay. For Level II hospitals, while total costs showed a declining tendency, the effect was not statistically significant; nonetheless, out-of-pocket expenses and their proportions decreased notably, and the length of stay showed a increasing tendency. The comparative analysis of impact of DRG and DIP showed that DIP showed a better effect of controlling total costs and out-of-pocket expenses, while DRG has a more significant effect on shortening the length of stay. Conclusion: This study found that DRG and DIP, as effective means of healthcare payment reform, both contribute to controlling healthcare expenditures, reducing healthcare fund risks, and improving healthcare service efficiency. It is recommended that hospitals enhance their multidimensional management capabilities, while medical insurance administrators should leverage big data analysis to strengthen regulatory oversight, optimizing the impact of healthcare reforms by keeping using a multidimensional medical insurance payment methods in the future.
