联合胃肠内镜检查存在质量保证缺口。在当前规模下，这一缺口将带来机构性后果。
Combined GI Endoscopy Has a Quality Assurance Gap. At Current Volume, That Gap Has Institutional Consequences.
Dror Zur，博士，Magentiq Eye首席执行官
联合胃肠内镜检查已成为美国许多医疗机构的常规实践。当两种操作均有临床指征时，约65%的患者在同一时段或相近时间内接受两种检查：一次肠道准备、一次镇静、整合的诊疗流程。
效率逻辑看似合理，但运营分析并未充分考虑效率提升背后所隐含的风险。上消化道和下消化道在质量基础设施方面并非等效操作。结肠镜检查拥有数十年标准化绩效指标以及目前已有文献记录的AI辅助检测成果。上消化道检查在同等成熟度上两者皆无。这一缺口并非新问题，但医疗机构如今开展联合操作的规模却是新的。而规模会改变质量缺口在机构层面意味着什么。
规则正在制定中。就在当下。
联合操作中的结肠镜检查部分拥有标准化绩效基准、有记录的结果以及可供质疑时引用的既定记录。而在许多项目中，上消化道部分并无这些。这种不对称就是风险敞口，而大多数医疗机构的风险管理框架尚未明确指出这一点。
上消化道癌症漏诊率超过8%，部分估计在三年窗口期内可达11.3%。在69%的漏诊病例中，内镜医师已在后来确诊癌症的同一解剖部位记录到异常。失败模式并非检测问题，而是系统性覆盖问题：即撤镜前是否检查了所有必需的解剖标志。若联合操作的上消化道部分缺乏结构化质量项目，就没有机制来证明该标准已得到满足。
正是在此背景下，美国胃肠内镜学会于2025年初发布了新的上消化道质量指标，这是标准化该操作绩效测量最具综合性的尝试。临床界已认识到这一缺口并正致力于弥合。对医疗机构而言，运营层面的影响尚未得到充分理解。质量指标一旦发布并被专业学会采纳，往往遵循可预测的轨迹：它们影响资质认证标准，塑造评审期望，并建立风险和法律团队在结果受到质疑时援引的基准。现在就开始向这些标准靠拢的医疗机构，将比等待其获得正式约束力的机构处于更有利的位置。
规模将缺口转化为模式
联合操作中单一上消化道发现被漏诊，是针对个案的临床问题。而一家开展数千例联合操作却对上消化道部分缺乏明确质量保证机制的医疗机构，则属于另一类问题：该机构是否已建立相应流程以满足新兴的诊疗标准，并全程一致地应用于整个项目。前者是不良结局，后者是机构政策问题，两者在结果审查时具有不同的影响。
报销结构解释了为何此类模式会形成且未被纠正。美国内镜检查计费基本是二元的：诊断性操作有基础支付，若发生活检或切除则增加费用，但对于上消化道部分的彻底解剖标志覆盖并无财务信号。实践中，结肠镜检查承载着临床权重；上消化道检查先行完成且快速。若无标记不完整覆盖的机制，这种倾向会在项目执行的每一次联合操作中累积，直到某个结果迫使人提出质疑，才产生可见信号。
在规模效应下，这就是将已记录在案的临床缺口转化为一致的机构实践的过程。而一致的机构实践正是风险和法律团队需要辩护的对象。
抢先一步是什么样
AI辅助结肠镜检查获得机构认可，是因为质量指标和检测技术同步发展，创造了可审计的操作绩效记录，既服务于临床改进，也服务于机构问责。2024年一项涉及近24,000名患者的28项随机对照试验荟萃分析发现，AI辅助使腺瘤检出率提高20%，漏诊率降低55%。这一证据基础为医疗机构运营者提供了可辩护的采纳理由，也为风险团队提供了具体可引用的依据。
上消化道AI在该曲线上尚处早期，但其构建逻辑相同。首先是标志验证：在撤镜前实时确认所有解剖必需区域均已检查，从而建立此前不存在的基础覆盖记录。随后是癌前病变检测，包括巴雷特食管和胃黏膜肠化生中的异型增生。这一顺序经过深思熟虑，因为覆盖和检测不可互换；未系统检查的区域无法可靠地发现异常。
对于大规模开展联合操作的医疗机构而言，这一顺序所产生的是可审计的记录，证明上消化道检查达到既定标准，在诊疗现场为每次操作生成。这一记录在上消化道质量指标具有正式约束力之前就具有机构价值；一旦其具有约束力，价值将显著增加。现在就开始向此方向建设的医疗机构，将按照自身条件确立该标准，而不必等到规模使缺失成为更艰难的对话。
在窗口关闭前行动
整合胃肠操作是理性的运营决策，且依然如此。改变的是该决策如今所要求的内容。大规模联合胃肠内镜检查不仅是排程和吞吐量问题，更是质量保证问题，对于每次联合操作中的上消化道部分，大多数医疗机构尚未有完整答案。
标准正在制定中。证据正在形成。两者都不会等待。现在就将此视为机构风险问题的医疗系统领导者，而非留待日后考虑的临床问题，才是当被问及此事时能有可辩护答案的人。
关于Dror Zur
Dror Zur，博士，是Magentiq Eye的首席执行官，该公司是一个AI驱动的结肠镜平台，专注于提高胃肠内镜检查的操作质量。
Dror Zur, PhD, CEO of Magentiq Eye
Combined GI endoscopy is already standard practice across many US health systems. When both procedures are clinically indicated, around 65% of patients undergo them in the same session or close window: one prep, one sedation, consolidated throughput.
The efficiency logic is sound, but what the operational analysis hasn’t fully accounted for is the risk that sits inside that efficiency gain. Upper GI and lower GI are not equivalent procedures in their quality infrastructure. Colonoscopy has decades of standardized performance metrics and a now-documented AI-assisted detection record. Upper GI has neither at equivalent maturity. That gap isn’t new, but the scale at which health systems are now running combined procedures is. And scale changes what a quality gap means institutionally.
The Rules Are Being Written. Now.
The colonoscopy component of a combined session has standardized performance benchmarks, documented outcomes, and an established record to point to if a result is questioned. The upper GI component, in many programs, does not. That asymmetry is the exposure, and most health system risk frameworks have not yet named it.
Upper GI cancer miss rates exceed 8%, with some estimates reaching 11.3% over a three-year window. In 69% of missed cases, the endoscopist had already recorded an abnormality at the exact anatomical site where cancer was later diagnosed. The failure mode is not detection, but systematic coverage: whether every required anatomical landmark was examined before the scope was withdrawn. Without a structured quality program for the upper GI component of combined sessions, there is no mechanism to demonstrate that standard was met.
That is the context in which the American Society for Gastrointestinal Endoscopy published new upper GI quality indicators in early 2025, the most comprehensive attempt to standardize performance measurements for the procedure. The clinical community has recognized the gap and is working to close it. The operational implications for health systems are less well understood. Quality indicators, once published and adopted by professional societies, tend to follow a predictable trajectory: they inform credentialing standards, shape accreditation expectations, and establish the benchmark that risk and legal teams reach for when outcomes are questioned. Health systems building toward those standards now are in a stronger position than those waiting for them to carry formal weight.
Volume Turns a Gap Into a Pattern
A single missed upper GI finding in a combined session is a clinical question about an individual case. A health system running thousands of combined sessions without a defined quality assurance mechanism for the upper GI component is a different category of question: whether the institution had the processes in place to meet an emerging standard of care, applied consistently, across its entire program. One is an adverse outcome. The other is an institutional policy question, and the two land differently when results are reviewed.
The reimbursement structure explains why that pattern forms and goes uncorrected. US endoscopy billing is largely binary: a base payment for the diagnostic procedure, with an increment if biopsy or resection occurs, and no financial signal for thorough anatomical landmark coverage in the upper GI component. In practice, the colonoscopy carries the clinical weight; the upper GI exam is completed first and quickly. Without a mechanism to flag incomplete coverage, that tendency accumulates across every combined session a program performs without producing a visible signal until an outcome forces the question.
At scale, that is what converts a documented clinical gap into a consistent institutional practice. And consistent institutional practices are what risk and legal teams are asked to defend.
What Getting Ahead of It Looks Like
AI-assisted colonoscopy gained institutional traction because quality metrics and detection technology were developed in parallel, creating an auditable record of procedural performance that served both clinical improvement and institutional accountability. A 2024 meta-analysis of 28 randomized controlled trials involving nearly 24,000 patients found a 20% increase in adenoma detection rate and a 55% reduction in miss rate with AI assistance. That evidence base gave health system operators a defensible basis for adoption and gave risk teams something concrete to point to.
Upper GI AI is earlier on that curve, but it is being built on the same logic. Landmark verification comes first: real-time confirmation that all anatomically required areas were examined before scope withdrawal, creating a structured coverage baseline that did not previously exist. Detection of precancerous conditions, dysplasia in Barrett’s esophagus and gastric intestinal metaplasia, follows from that baseline. The sequencing is deliberate because coverage and detection are not interchangeable; you cannot reliably catch what you have not systematically examined.
What that sequence produces, for a health system running combined sessions at scale, is an auditable record that the upper GI examination met a defined standard, generated at the point of care, for every procedure performed. That record has institutional value before upper GI quality indicators carry formal weight. It has considerably more once they do. Health systems that build toward it now are establishing that standard on their own terms, before volume makes its absence a harder conversation to have.
Act Before the Window Closes
Consolidating GI procedures was a rational operational decision, and it remains one. What has changed is what that decision now requires. Combined GI endoscopy at scale is not just a scheduling and throughput question. It is a quality assurance question, and for the upper GI component of every combined session being run, most health systems do not yet have a complete answer to it.
The standards are being written. The evidence is forming. Neither will wait. Health system leaders who treat this as an institutional risk question now, rather than a clinical consideration for later, are the ones who will have a defensible answer when it is asked of them.
About Dror Zur Dror Zur, PhD, is the CEO of Magentiq Eye, an AI-powered colonoscopy platform focused on improving procedural quality in GI endoscopy.
Reader Interactions