基于大数据的DIP病种分值支付方法利用真实世界的临床病案数据,对“疾病诊断”和“治疗方式”进行穷举聚类形成自然分组。然而,目前医院病案书写不规范、不统一等数据源头问题可能产生分组重复、分组偏离临床指南及路径等问题,从而导致病例入组路径不清晰,也不利于医保基金管理。本研究以结直肠癌为例,通过广泛的临床专家咨询及医院数据分析,梳理了广州市2020版DIP分值表在实践中遇到的问题及其与临床实际情况存在偏差的地方,并提出基于临床指南及诊疗路径的分组优化建议。这些建议经广州市医保局采纳,得以在广州市2022版分值表中落地实践。本研究旨在通过广州市的经验分享,为DIP支付方法的有效落地及其他统筹地区的DIP分组优化提供参考依据。

Big data Diagnosis-Intervention Packet (DIP) aggregates "diagnoses" and "interventions" into logical categories using data from real-world hospital discharge record. However, problems concerning data sources, such as non-standardized and inconsistent writing of medical records, can lead to duplicate clustering and deviation from clinical treatment guideline and pathways, resulting in unclear case enrollment paths and making it difficult to distribute medical insurance funds. Using colorectal cancer as an example, through comprehensive consultation with clinical experts, this study identifies the problems encountered in practice with the Guangzhou 2020 version of the DIP Grouping and how it differs from actual clinical situations while also making optimization suggestions based on clinical treatment pathways. These suggestions were adopted by the Healthcare Security Administration of Guangzhou and implemented in the newly published DIP Grouping List for 2022. Through the sharing of Guangzhou's experience, this study aims to provide references for DIP grouping optimization in other regions and support the efficient and effective DIP payment method implementation.