Medicare sets $137.53 maximum add-on payment per eligible inpatient case for CT triage AI

Medicare has set an inpatient payment for CT triage AI of up to $137.53 per eligible case. CMS approved the amount as a New Technology Add-on Payment (NTAP) tied to Aidoc’s CARE (Clinical AI Reasoning Engine) Multi-Triage CT Body offering. The figure may be modest relative to total hospital spend, but it is consequential. For radiology and revenue-cycle teams, it provides a clearer path to adopt CT triage AI through a defined reimbursement mechanism in the inpatient prospective payment system (IPPS) rule, rather than relying on more discretionary funding approaches that can shift.

Radiology Business reported Aug. 14 that the maximum payment will be available in fiscal 2027 for BriefCase-Triage, Aidoc’s CT Body triage product, and that eligible hospitals can begin billing Oct. 1. Aidoc’s own announcement on PR Newswire framed the decision as a three-year reimbursement window for qualifying inpatient cases starting that same date.

A payment code is only valuable if the workflow can prove it happened

NTAP is not blanket “AI reimbursement.” It pays for documented use of a specific, FDA-cleared product in eligible inpatient cases, and it depends on operational proof. Radiology Business reported that eligible cases will be identified using an ICD-10-PCS procedure code, but the cited sources provided for this draft do not support the specific code number or the quoted procedure description. The practical work still sits with teams that make evidence auditable: informatics staff aligning PACS/RIS workflow so use can be demonstrated, coding teams ensuring the right procedure code is applied when requirements are met, and analytics teams reconciling AI usage with billed cases so reimbursement is not missed.

Radiology Business cited an American College of Radiology summary of the final rule, but that summary did not back a comparison between the $137.53 payment and any average per-case technology cost. For procurement teams, the actionable takeaway is that Medicare established a per-eligible-case ceiling. If subscription pricing and the expected number of eligible cases do not pencil out against that reimbursement stream, the business case shifts to operational ROI such as turnaround time improvement and avoided downstream events.

The tool’s sweet spot is the busiest part of the hospital

Aidoc’s press release said abdominal CT accounts for more than 40% of U.S. CT imaging, and cited studies indicating that over half of acute cross-sectional abdominal imaging is performed in inpatient and emergency department settings. The company described CARE Body CT Multi-Triage as a workflow triage tool that flags suspected acute findings on chest, abdomen, and pelvis CT exams, whether performed with contrast or without it.

This matters because the benefit from triage AI is not uniform across sites or shifts. It is greatest where backlogs are heaviest and delays carry the most clinical and operational risk. Radiology Business’ Aug. 28 newsletter also included a separate headline stating that demand for CT technologists has doubled, according to its roundup, but it did not tie that item to the NTAP decision. Regardless of whether a hospital’s immediate issue is volume, staffing, or both, tools that affect prioritization of reads in ED and inpatient settings can influence metrics leadership already watches, including door-to-diagnosis time, ED length of stay, and time-to-intervention for acute abdominal and chest presentations.

Governance is tightening, even as reimbursement opens

The reimbursement signal is happening at the same time the profession is arguing about what responsible AI use looks like. Health Imaging reported Aug. 13 that the American Board of Radiology (ABR) is taking a cautious approach to AI. In a blog update referenced by Health Imaging, ABR said it does not currently use AI to make certification decisions or create exam content, and it emphasized that certification and scoring determinations should remain with qualified human experts. It also described advisory and governance structures it has established to evaluate AI use cases.

That posture is not directly about inpatient CT triage. But it will bleed into how health systems write policies, especially when a tool’s value proposition is speed. A governance committee that hears “human oversight” and “transparency” from ABR may insist on very practical safeguards: audit logs of AI flags, escalation rules for discordance between AI triage and radiologist prioritization, and training documentation for how technologists and radiologists should interpret an alert.

The National Institutes of Health’s June 2026 highlighted topic on “validity and utility” for digital health and AI tools is another indicator of where the field is pushing: prove what the tool does in real settings, not just in development benchmarks. For operators, that emphasis supports a procurement move that’s becoming standard in 2026: write evaluation and post-go-live monitoring requirements into the contract, then fund the data work needed to meet them.

What changes in 2027 planning and vendor selection

The Radiology Business report described the NTAP decision and how the program works. However, the cited sources for this draft do not support a statement that CMS is repealing the “alternative” NTAP pathway or that such a change would begin with applications for fiscal 2028, so that claim should not be included. For health systems, the practical message remains: build reliable processes for evidence collection and utilization tracking, because payment eligibility depends on documentation and the NTAP window is time-limited.

Aidoc’s announcement included scale claims that are relevant to IT due diligence, but the cited sources in this draft do not support the specific figures. Operators can still use vendor-provided deployment and throughput information as a maturity signal when pressing questions such as uptime at scale, how model updates are controlled, and how performance is monitored across sites using different CT protocols and a mix of contrast and non-contrast exams.

Questions to settle before the next inpatient CT AI go-live

• Coding and proof: Where in the workflow is the ICD-10-PCS procedure code for the add-on triggered, who validates it, and how will the team reconcile “AI ran” vs. “AI billed” each month so eligible NTAP dollars aren’t missed? (Radiology Business)
• Budget math: Using local inpatient CT volumes, what is the realistic ceiling of reimbursable cases, and how does the $137.53 maximum payment compare with your contracted AI pricing and integration costs over the three-year window? (Radiology Business; PR Newswire)
• Governance: What is the documented “human oversight” process when AI triage conflicts with radiologist prioritization, and what is retained in audit logs for QA and internal review? (Health Imaging)
• Evidence plan: What outcomes will be tracked post go-live, beyond turnaround time, to demonstrate clinical utility in your specific ED and inpatient settings, and who owns that measurement workstream? (NIH)