How Hospitals Can Assess AI Tools: UVA Health and Clemson Introduce Total Mission Value Framework

What the Total Mission Value Framework Is

Andrew Taylor, an emergency physician and professor at the UVA Emergency Medical Center, and Arwen Declan, a clinical assistant professor at Clemson University, developed the Total Mission Value (TMV) framework—a structured, five-priority checklist designed to help hospital leaders evaluate AI tools across multiple dimensions, not just financial return.

The framework is built as a pyramid. Patient care sits at the top. Below it, the framework layers in ethics, safety, workflow impact, and broader organizational considerations—including guardrails against algorithmic bias.

The core idea is straightforward: an AI tool that saves money but compromises care quality, introduces bias, or overloads clinical staff isn’t actually a good investment.

Why This Matters for Hospital AI Procurement

Most AI procurement decisions in healthcare still lean heavily on cost-benefit analysis. That’s understandable—hospital budgets are tight and administrators need to justify spending. But cost savings alone don’t capture what a clinical AI tool actually does to patient outcomes, staff workload, or ethical accountability.

Taylor put it plainly: there are a lot of dimensions that need to go into the consideration of whether an AI tool is right for a given organization.

The TMV framework gives hospital leaders a structured way to ask harder questions before signing a contract—or before renewing one.

Declan described it as a “big picture” lens that leaders can use both when purchasing an AI product and when deciding whether to keep it running.

The Five-Priority Structure

While the full checklist details aren’t publicly exhaustive, the framework’s pyramid structure points to a clear hierarchy of values:

• Patient care — the top priority; any AI tool must demonstrably support, not compromise, care delivery
• Ethics — includes bias detection and guardrails to prevent discriminatory outcomes
• Safety — addresses clinical risk and the potential for AI tools to introduce new failure points
• Workflow impact — evaluates how the tool affects care providers day-to-day, with a focus on supporting staff rather than adding burden
• Organizational fit — considers whether the tool aligns with the hospital’s broader mission and operational context

The framework is explicitly designed to support care providers, not replace them or pile on additional administrative work.

The Risk the Framework Addresses Directly

One tension the creators flagged is worth highlighting: it’s entirely possible for an AI tool to cut costs while simultaneously introducing new clinical or ethical risks. A tool that automates triage documentation, for example, might reduce billing overhead but flag certain patient demographics differently based on biased training data.

Without a structured evaluation process, those risks can go undetected until they cause real harm.

The TMV framework is positioned as a pre-deployment and ongoing assessment tool—meaning hospitals can use it not just when buying, but when deciding whether to keep a tool in production.

What This Means for the Broader AI Tools Ecosystem

The introduction of the TMV framework reflects a growing recognition that AI governance in healthcare can’t be an afterthought. As more clinical AI tools enter the market—covering everything from diagnostic imaging to patient communication to administrative automation—the gap between what vendors promise and what tools actually deliver in real clinical environments is becoming harder to ignore.

Frameworks like this one shift the evaluation burden back toward the buyer in a structured way. Instead of relying on vendor benchmarks or pilot data curated by the company selling the product, hospital leaders get a mission-aligned checklist they can apply independently.

For anyone working in healthcare AI—whether you’re building tools, buying them, or advising on implementation—the TMV framework is a practical signal that the field is moving toward more rigorous, values-driven procurement standards.

The Practical Takeaway

If your organization is evaluating clinical AI tools right now, the Total Mission Value framework offers a useful starting point: stop leading with cost savings and start leading with patient care outcomes. Build your evaluation criteria outward from there—ethics, safety, workflow, and organizational fit.

The best AI tool for a hospital isn’t the cheapest one or the most technically impressive one. It’s the one that holds up across all five dimensions when you actually pressure-test it.