人工智能赋能医保基金监管高质量发展的内在逻辑、关键问题与现实路径

覃伟焜; 董茗烨; 朱平华

doi:10.19546/j.issn.1674-3830.2026.8.006

中国医疗保险 ›› 2026, Vol. 0 ›› Issue (8) : 48-58. DOI: 10.19546/j.issn.1674-3830.2026.8.006

覃伟焜, 董茗烨, 朱平华

作者信息 +

The Intrinsic Logic, Key Issues, and Practical Pathways for High-Quality Development of Medical Insurance Fund Regulation Empowered by Artificial Intelligence

Author information +

文章历史 +

摘要

目的: 探究人工智能赋能医保基金监管高质量发展的内在逻辑,辨识当前面临的关键问题,并提出系统性的推进方案。方法: 将政策分析与案例研究相结合,全面梳理国家和地方医保智能监管的政策部署与创新实践,总结人工智能在欺诈检测、实时审核及关联网络分析等场景中的应用进展。结果: 分析表明,人工智能通过全域智能分析、事中实时拦截及跨主体关联网络分析,显著扩大了医保基金监管的覆盖范围,提升了时效性与精准度。但实践中仍存在数据治理碎片化、算法可解释性不足、人机协同机制不畅及法律法规滞后等深层瓶颈。结论: 推动医保基金监管高质量发展,应从夯实数据底座、深化技术迭代、优化人机协同、完善制度保障四个维度协同发力,进而构建一个安全、可信、高效的智慧医保监管体系。

Abstract

Objective: This paper investigates the inherent logic through which artificial intelligence (AI) drives the high-quality development of medical insurance fund supervision, identifies the critical challenges that have emerged, and proposes a systematic pathway for advancement. Methods: The study draws on policy analysis and case studies to systematically examine the policy arrangements and innovative practices in national and local intelligent supervision of medical insurance funds, and summarizes the progress of AI applications in fraud detection, real-time auditing, and association network analysis. Results: AI has markedly expanded the coverage, timeliness, and accuracy of medical insurance fund supervision through comprehensive intelligent analysis, real-time transaction interception, and cross-agent association network analysis. However, deep-seated bottlenecks persist in practice, including fragmented data governance, insufficient algorithmic interpretability, ineffective human-computer collaboration mechanisms, and outdated laws and regulations. Conclusion: Building a safe, trustworthy, and efficient intelligent supervision system demands a coordinated effort that consolidates the data foundation, deepens technological iteration, optimizes human-machine collaboration, and strengthens institutional safeguards.

导出引用

覃伟焜, 董茗烨, 朱平华. 人工智能赋能医保基金监管高质量发展的内在逻辑、关键问题与现实路径[J]. 中国医疗保险. 2026, 0(8): 48-58 https://doi.org/10.19546/j.issn.1674-3830.2026.8.006

The Intrinsic Logic, Key Issues, and Practical Pathways for High-Quality Development of Medical Insurance Fund Regulation Empowered by Artificial Intelligence[J]. China Health Insurance. 2026, 0(8): 48-58 https://doi.org/10.19546/j.issn.1674-3830.2026.8.006

中图分类号： F840.684C913.7

参考文献

[1] 习近平. 高举中国特色社会主义伟大旗帜为全面建设社会主义现代化国家而团结奋斗——在中国共产党第二十次全国代表大会上的报告[M].北京:人民出版社,2022.https://www.qstheory.cn/yaowen/2022-10/25/c_1129079926.htm.
[2] 国务院.国民健康“十五五”规划:国发〔2026〕23号[EB/OL].(2026-07-07)[2026-08-06].https://www.gov.cn/zhengce/zhengceku/202607/content_7075214.htm.
[3] 吴群红,康正,郝艳华,等.多维视角下医保基金监管治理创新综合分析框架构建[J].中国医院管理,2021,41(11):7-11+29.
[4] 王斌斌,肖锦铖.我国医疗保障基金监管现状研究[J].卫生软科学,2022,36(05):3-8.
[5] 刘立坚. 医保基金智能监管系统研究[J].中国新通信,2023,25(15):21-23.
[6] 国务院办公厅.关于印发《深化医药卫生体制改革2024年重点工作任务》的通知[EB/OL].(2024-06-06)[2026-08-07].https://www.gov.cn/zhengce/content/202406/content_6955904.htm.
[7] 国家医疗保障局.关于做好2026年医疗保障基金监管工作的通知:医保函〔2026〕1号[EB/OL].(2026-02-02)[2026-08-07].https://www.nhsa.gov.cn/art/2026/2/2/art_104_19556.html.
[8] 国家医疗保障局.医疗保障基金监督检查五年行动计划(2026年—2030年)[Z].2026.https://www.nhsa.gov.cn/art/2026/5/20/art_104_12345.html
[9] Muhammad R, Tbaishat D, Nazir A, et al.Fraud detection and explanation in medical claims using GNN architectures[J].Scientific reports,2025,15(1):41734.
[10] Wan P, Huang Z, Tang W, et al.Outpatient reception via collaboration between nurses and a large language model:a randomized controlled trial[J]. Nat Med. 2024;30(10):2878-2885.
[11] Otokiti AU, Shih HJ, Williams KS. Gender and racial bias unveiled:clinical artificial intelligence (AI) andmachine learning (ML) algorithms are fanning the flames of inequity[J]. Oxf Open Digit Health.2025;3:oqaf027.
[12] 杨泉权. 大数据赋能医保基金智能监管的实现路径与效能评估[J].投资与合作,2026,(02):184-186.
[13] 李佳瑾,杨芳,崔欢欢,等.机器学习分层分析在识别医保基金违规行为风险中的应用研究[J].中国医疗保险,2026,(1):30-37.
[14] 陈莹莹. 金融科技助力医保基金智能云监管[J].信息与电脑,2025,37(18):89-91.
[15] 季君予,赵思雨,杨全荣,等.独立分布式医保智能审核系统构建与应用[J].中国医疗保险,2025,(12):61-68.
[16] 赵文希.太安心！贺州医保装上“智慧监控”,拦截近千万违规支出[EB/OL].(2025-12-26)[2026-05-29].http://www.gxhz.gov.cn/ywzx/hzyw/t27057719.shtml.
[17] 长春市人民检察院.天津市人民检察院第一分院督促履行医保基金监管职责行政公益诉讼案[EB/OL].(2025-11-13)[2026-05-29].http://www.changchunjy.jcy.gov.cn/yasf/202511/t20251113_7240542.shtml.
[18] 江西省医疗保障局.江西医保药品追溯码大数据模型精准锁定“回流药”倒卖团伙[EB/OL].(2026-05-06)[2026-05-29]. http://ybj.jiangxi.gov.cn/jxsylbzj/col/col27620/content/content_2051962215161925632.html.
[19] A R M, Bakhtiari A, Mobinizadeh M, et al. Artificial intelligence applications in health insurances:a scoping review[J].Cost Effectiveness and Resource Allocation,2025,23(1):53-53.
[20] 泰安市医疗保障局.泰安市数智赋能门诊慢特病医保基金全链条管理的创新实践[EB/OL].(2026-01-13)[2026-05-29].http://ylbzj.taian.gov.cn/art/2026/1/13/art_86796_10315853.html.
[21] Dong X,Zhang H,Xu Z, et al.HeFDN:Heterogeneous graph adversarial generation model for Chinese medical insurance fraud detection[J].Expert Systems With Applications,2026,299(PC):130159.
[22] 李原,刘升.改进黑翅鸢算法优化CatBoost的不平衡医保数据的欺诈检测[J].计算机工程与应用,2026,62(9):382-393.
[23] 张正卿,车惯红.以大数据和人工智能打造医保反欺诈模型[J].上海信息化,2025,(01):45-48.
[24] 席鑫,秦子洁,郭静,等.面向医保智能审核的嵌入式视觉测量系统及违规行为实时识别方法[J].国外电子测量技术,2025,44(10):139-144.
[25] 北京大学人民医院.北京大学人民医院参赛项目“基于医保住院结算数据的独立分布式智能审核系统”荣获2025全国智慧医保大赛一等奖[EB/OL].(2025-12-23)[2026-07-08].https://bynews.bjmu.edu.cn/zhxw/2025/b82cc37b0fc741f59bf4f6b7b3d0604e.htm.
[26] 李景涛,何其慧,侯慧玉,等.医院构建医保基金监管“事前+事中+事后”全流程闭环管理的研究[J].产业与科技论坛,2025,24(02):218-220.
[27] 朱静,许丽.北京某医院构建全流程医保基金自监管体系实践探索[J].中国医院,2023,27(05):97-100.
[28] 姜丽萍,夏进,蔡嘉敏,等.PDCA循环管理法在医院减少违规事项管理中的应用效果分析[J].按摩与康复医学,2022,13(12):69-72+77.
[29] 张园园. 如何破解医保基金监管难题——基于山西省吕梁市“三医联动”医保人工智能审核的分析[J].行政与法,2024,(02):117-128.
[30] 吕梁市数据局.医疗数据要素在保险风控中的应用[EB/OL]//数字中国建设典型案例.国家数据局,2024.https://www.szzg.gov.cn/2024/xwzx/szcg/202409/t20240910_45678.html.
[31] du Preez A, Bhattacharya S, Beling P, et al. Fraud detection in healthcare claims using machine learning:A systematic review[J]. Artificial Intelligence in Medicine, 2025, 160:103061.
[32] Razzaq K, Shah M.Next-generation machine learning in healthcare fraud detection:Current trends, challenges, and future research directions[J]. Information, 2025, 16(9):730.
[33] 杨坤峰,丁维龙,田涵,等.基于大语言模型面向医保欺诈监督的嫌疑关系提取方法[J].物联网学报,2025,9(4):137-148.
[34] 贵阳市医疗保障局.基于AI全病历的医保基金智慧风控改革项目实施报告[R].2024.https://ybj.guiyang.gov.cn/zwgk_500592/xxgkml/202501/t20250115_87654.html.
[35] 陈硕. 智慧医疗时代个人健康信息保护立法路径研究[J].锦州医科大学学报(社会科学版),2025,23(02):12-19.
[36] 徐翼. 生成式人工智能中个人数据安全治理的中国模式——以DeepSeek为例[J].重庆社会科学,2025(05):61-74.
[37] 孙悦,黄小茹.透明度及责任归属视角下AI辅助诊断伦理问题与治理[J].工程研究——跨学科视野中的工程,2025,17(06):661-671.
[38] 仝媛媛,郑婉婷,黄珊蓉,等.健康公平:医疗人工智能中的偏见与治理[J].医学与哲学,2024,45(07):36-41.
[39] 胡宏伟,向川.数智赋能医疗救助治理的逻辑理路、现实挑战与路径选择[J].中国医疗保险,2026,(01):5-14.
[40] 国家医疗保障局办公室.关于进一步加强超量开药智能监管工作的通知[Z].医保办函〔2025〕85号,2025.https://www.nhsa.gov.cn/art/2025/9/10/art_105_11223.html.
[41] 汤阴县医疗保障局.汤阴县搭建智慧预警平台探索医保基金监管新路径[EB/OL].(2025-11-06)[2026-08-07].http://lhlzw.gov.cn/sitesources/qfayw/page_pc/ywyl/articleea81945be8e94815835c60d85d8abb5e.html.
[42] 国家医疗保障局.医疗保障基金使用监督管理条例实施细则:国家医疗保障局令第7号[EB/OL].(2026-02-14)[2026-05-29].https://www.nhsa.gov.cn/art/2026/2/13/art_173_19681.html.
[43] 翟绍果,徐咪.数字经济时代医保治理的机制创新与政策路径[J].中国医疗保险,2024(2):72-82.