政策背景:为什么2026年是关键节点Why 2026 Is the Critical Inflection Point

中国预防医疗AI政策在过去三年经历了从探索到明确的转变。2023年以前,AI+医疗的政策讨论主要聚焦在影像诊断、辅助决策等治疗端;2024年起,随着慢病负担数据持续恶化和医保基金压力加剧,政策重心开始向"防"端移动。China's AI preventive medicine policy has shifted from exploration to clarity over the past three years. Before 2023, AI+healthcare policy discussions focused primarily on imaging diagnosis and treatment-side decision support. From 2024 onward, as chronic disease burden data worsened and healthcare fund pressure intensified, policy gravity began shifting toward prevention.

2026年是"十四五"医改收官之年,多项改革政策进入落地深水区。对于AI预防医疗从业者而言,理解这些政策的底层逻辑,是判断市场窗口的前提。2026 is the final year of the 14th Five-Year Plan for healthcare reform, with multiple reform policies entering deep implementation. For AI preventive medicine practitioners, understanding the underlying logic of these policies is prerequisite to identifying market windows.

核心判断:Core Assessment:AI预防医疗的商业化不是技术问题,而是支付问题。2026年的政策组合拳正在从支付端打开预防结算的通道——这是过去十年从未有过的结构性变化。AI preventive medicine commercialization is not a technology problem — it's a payment problem. The 2026 policy combination is opening the channel for prevention settlement from the payment side — a structural change unprecedented in the past decade.

四大政策驱动力Four Major Policy Drivers

驱动力 01 / Driver 01
健康中国2030:心脑血管死亡率降30%的硬指标Healthy China 2030: Hard Target of 30% Reduction in CVD Mortality
国务院《健康中国行动(2019-2030年)》明确要求到2030年心脑血管疾病死亡率较2015年基线下降30%以上。这一指标已分解至省级政府考核体系,对三甲医院产生了直接的绩效压力。医院必须拿出可量化的预防成效数据,而不仅仅是治疗数量。AI心血管风险预测与干预闭环,成为医院完成考核指标的技术路径。The State Council's Healthy China Action Plan (2019-2030) explicitly requires cardiovascular and cerebrovascular disease mortality to drop over 30% from 2015 baseline by 2030. This target has been decomposed into provincial government performance metrics, creating direct performance pressure on top-tier hospitals. Hospitals must produce quantifiable prevention outcome data, not just treatment volume. AI cardiovascular risk prediction and intervention loops become the technical pathway for hospitals to meet these targets.
驱动力 02 / Driver 02
DRG/DIP支付改革:预防从成本变利润DRG/DIP Payment Reform: Prevention Shifts from Cost to Profit
DRG(按疾病诊断相关分组)和DIP(按病种分值)支付改革是近年最深刻的医院商业模式变革。在按项目付费时代,医院的利润来自多做检查和手术;在DRG/DIP时代,医院的利润来自在打包付费额度内降低实际医疗成本——而降低心脑血管并发症率,是降低实际成本的最有效路径之一。这意味着医院预防科室第一次从成本中心变为潜在的利润贡献中心。DRG and DIP payment reforms represent the most profound hospital business model transformation in recent years. Under fee-for-service, hospital profits came from more tests and procedures; under DRG/DIP, profits come from reducing actual medical costs within bundled payment limits — and reducing cardiovascular complication rates is one of the most effective cost-reduction pathways. This means hospital prevention departments for the first time shift from cost centers to potential profit contributors.
驱动力 03 / Driver 03
NMPA真实世界证据(RWE)注册路径:AI预防产品的合规通道NMPA Real-World Evidence Registration Pathway: Compliance Channel for AI Prevention Products
国家药监局2021年发布《真实世界证据支持药物研发与审评的指导原则》,2023年进一步扩展至医疗器械领域。这意味着AI预防产品不再只能走传统RCT临床试验路径,可以通过真实世界证据支持二类医疗器械注册申报。对于心血管风险预测AI而言,这是进入医院临床和医保报销体系的关键合规通道。NMPA released the Guidance Principles for Real-World Evidence Supporting Drug Development and Review in 2021, further extended to medical devices in 2023. This means AI prevention products no longer must follow traditional RCT clinical trial pathways and can use real-world evidence to support Class II medical device registration. For cardiovascular risk prediction AI, this is the critical compliance channel for entering hospital clinical use and medical insurance reimbursement systems.
驱动力 04 / Driver 04
商业健康险创新试点:按效付费的制度基础Commercial Health Insurance Innovation Pilots: Institutional Foundation for Pay-for-Performance
银保监会(现国家金融监督管理总局)近年持续推动商业健康险产品创新,鼓励险企开发与健康行为和预防干预效果挂钩的浮动费率产品。部分试点地区已允许险企将PSM因果证据报告作为费率调整依据。这为保险公司从被动赔付者转型为主动健康管理者提供了制度支撑。The former CBIRC (now NFRA) has continuously promoted commercial health insurance product innovation, encouraging insurers to develop variable-rate products linked to health behaviors and prevention intervention outcomes. Some pilot regions already allow insurers to use PSM causal evidence reports as rate adjustment basis. This provides institutional support for insurance companies transitioning from passive payers to active health managers.

政策机会与落地挑战对照Policy Opportunities vs. Implementation Challenges

政策Policy对AI预防医疗的机会Opportunity for AI Prevention主要落地挑战Key Implementation Challenge
健康中国2030Healthy China 2030医院必须采购预防能力Hospitals must acquire prevention capabilities缺乏标准化效果评估框架No standardized outcome evaluation framework
DRG/DIP预防直接影响医院利润Prevention directly impacts hospital profits需要可结算的因果证据链路Requires causal evidence chain for settlement
NMPA RWE路径NMPA RWE PathwayAI预防产品有合规落地通道AI prevention products have compliant deployment path数据质量和隐私合规要求高High data quality and privacy compliance requirements
健康险创新试点Health Insurance Pilots按效付费制度初步成型Pay-for-performance mechanism taking shape试点范围有限,全国推广需时Limited pilot scope, national rollout takes time

政策窗口期:现在到2027年的关键动作Policy Window: Critical Actions from Now to 2027

政策窗口期通常在3-5年内。"十四五"收官和"十五五"开局的交接期(2026-2027年),是AI预防医疗玩家卡位的黄金窗口。错过这个时间窗口,等"十五五"政策体系成熟后,先发者的标准制定优势将难以追赶。Policy windows typically last 3-5 years. The handover period between the closing of the 14th Five-Year Plan and the opening of the 15th (2026-2027) is the golden window for AI preventive medicine players to establish position. Miss this window, and once the 15th Five-Year Plan policy framework matures, early movers' standard-setting advantages will be difficult to catch up with.

  • 医院端:布局DRG/DIP预防价值核算标准,成为医院绩效体系的基础设施Hospital side: Position in DRG/DIP prevention value accounting standards, become infrastructure for hospital performance systems
  • 保险端:推动PSM因果证据报告成为险企精算部门的标准输入格式Insurance side: Drive PSM causal evidence reports to become standard input format for insurer actuarial departments
  • 监管端:参与NMPA RWE评估标准制定,将自身方法论纳入监管认可框架Regulatory side: Participate in NMPA RWE evaluation standard development, embed methodology into regulatory recognition framework
  • 药企端:成为药企RWE申报的数据基础设施合作方Pharma side: Become data infrastructure partner for pharma RWE registration applications
Key Takeaway

2026年的政策组合——健康中国指标压力、DRG/DIP利润釡合规)双向推动AI预防医疗商业化。关键不在于政策本身,而在于谁能最先建立把预防效果转化为可结算证据的标准化能力。The 2026 policy combination — Healthy China target pressure, DRG/DIP profit restructuring, NMPA RWE pathway, health insurance innovation pilots — for the first time simultaneously pushes AI preventive medicine commercialization from both demand side (hospitals, insurers) and supply side (regulatory compliance). The key isn't the policies themselves, but who can first establish standardized capability to convert prevention outcomes into billable evidence.

常见问题FAQ

健康中国2030的心脑血管指标如何传导到医院采购决策?How do Healthy China 2030 cardiovascular targets translate into hospital procurement decisions?

指标通过省级卫健委分解为医院年度绩效考核,直接影响院长的职业评价。三甲医院通常需要每年向卫健委提交心脑血管防治专项报告,AI预防平台能够自动生成符合上报格式的效果数据,显著降低医院的合规成本。Targets are decomposed through provincial health commissions into hospital annual performance assessments, directly affecting hospital president career evaluations. Top-tier hospitals typically must submit annual cardiovascular prevention reports to health commissions; AI prevention platforms that auto-generate compliant reporting data significantly reduce hospital compliance costs.

DRG/DIP改革全国推广进度如何?What is the national rollout progress of DRG/DIP reform?

截至2025年底,全国超过90%的统筹区已启动DRG或DIP付费试点,其中约60%已进入正式实施阶段。预计2026年底前,绝大多数三甲医院将全面进入DRG/DIP结算体系。As of end-2025, over 90% of pooling regions have launched DRG or DIP payment pilots, with approximately 60% entering formal implementation. By end-2026, the vast majority of top-tier hospitals are expected to fully enter the DRG/DIP settlement system.

AI预防产品申请NMPA二类医疗器械证需要哪些关键材料?What key materials are needed for AI prevention products to apply for NMPA Class II medical device registration?

核心材料包括:①算法性能验证报告(AUC、敏感性、特异性等);②真实世界临床数据集(通常需要≥1000例);③软件安全性测试报告;④数据隐私合规声明(需符合《个人信息保护法》和《医疗数据安全规范》)。RWE路径下,真实世界数据集可部分替代传统RCT要求。Core materials include: ①Algorithm performance validation report (AUC, sensitivity, specificity, etc.); ②Real-world clinical dataset (typically ≥1,000 cases required); ③Software safety test report; ④Data privacy compliance declaration (must comply with Personal Information Protection Law and Medical Data Security Standards). Under the RWE pathway, real-world datasets can partially substitute traditional RCT requirements.