When it comes to AI investments, boardrooms across every health system are starting to ask the same question:
“Is this worth what we’re putting into it?”
It’s a fair question. As investments in AI continue to accelerate, boards and executive leaders are seeking clearer, measurable evidence of value to guide investment and scaling decisions. The challenge is no longer accessing more technology, but demonstrating the value it delivers.
While ROI focuses on financial outcomes, it does not account for other measurable impacts on clinical, workforce, and patient outcomes. This is where a broader focus on “value” becomes essential to capture those additional dimensions. However, without structure, the term “value” can also blur accountability.
The Ambient AI scribe boom clearly illustrates this challenge. Physician satisfaction scores are high, but many organizations have not yet demonstrated measurable financial gains. The efficiency improvements are real, but the captured time often is not converted into additional billable encounters, limiting measurable financial impact.
This is why we believe health systems need a repeatable, evidence-based method to measure the value of AI, not just ROI.
A Structured Way to Think About Value: Levers
Our approach defines four primary value levers that reflect the measurable impact an AI initiative can create:
- Clinical Outcomes: Improved quality, safety, and patient outcomes, supporting more consistent care delivery and stronger performance against clinical standards
- Financial Value: Cost reduction, increased revenue, and margin protection, translating operational and clinical improvements into sustainable financial performance
- Workforce Experience: Reduced cognitive load and burnout, supporting clinician retention and mitigating recruiting and turnover pressures
- Patient Experience: Better access and communication, strengthening patient experience and reducing leakage
How We Think About Value?
To avoid the trap of “every initiative must be ROI positive”, a taxonomy that clarifies what kind of value each initiative creates can help:
These value levers may be enabled by capabilities such as efficiency gains and new insights generated by AI. Examples include time saved through workflow automation or AI-generated patient chart summaries. These capabilities create value only when they translate into measurable improvements across clinical, financial, workforce, or patient outcomes.
This framework helps clarify that value in healthcare AI can take many forms, some measurable in dollars, others in outcomes or satisfaction, but all connected. A structured taxonomy allows leadership to prioritize, measure, and scale what truly drives impact.
Example in Practice: CLABSI Chart Abstraction LLM
One example of this framework in practice is a quality and safety-focused AI initiative in partnership with a large academic medical center. The CLABSI (central line-associated bloodstream infection) chart abstraction LLM was part of a broader effort to reduce the time and burden associated with reviewing complex clinical cases related to hospital-acquired conditions.
When a potential CLABSI is identified, Infection Preventionists (IPs) must determine whether the case meets the CDC’s National Healthcare Safety Network (NHSN) reporting criteria. This requires reviewing an average of 150+ pages of labs, notes, and clinical timelines to verify whether the case qualifies as a reportable CLABSI. The process can consume several hours each week for every IP on staff and often pulls time away from rounding, education, and prevention work.
In response, the CLABSI chart abstraction LLM streamlines the review process by synthesizing the relevant patient data, applying NHSN rules, and generating a preliminary case summary in minutes. IPs can then review, validate, and make the final determination.
When viewed through the value lever framework, the expected impact of this tool becomes measurable and actionable.
A structured value framework translates these levers into measurable ranges. For example, improvements in documentation quality reduce rework, and reductions in CLABSI events can prevent penalties and avoid high-cost adverse events, supporting both Clinical Outcomes and Financial Value levers. Improved efficiency can also be considered an enabler, translating into faster review cycles that, when converted into avoided FTE time or reduced labor effort, contribute directly to the Financial Value lever.
From Value Estimation to Value Evidence
As organizations mature, value measurement must evolve from estimation to evidence. The goal is a repeatable value management process that can predict, measure, and refine the impact of every initiative.
Five Practices to Build Evidence-Based Value
- Document Value Levers
Use a value lever framework to define expected outcomes for every initiative - Capture a Before and After
Estimate conservative value ranges before launch and measure against real-world results - Track Adoption
Assess influence on workflows, throughput, and decision-making - Focus on Lessons Learned
Validate assumptions early and treat pilot outcomes as lessons, not failures - Embed Value in Governance
Value management should live within governance structures to ensure accountability and sustained proof of value
The Bottom Line
If we believe AI can improve both care and financial performance, then the missing piece is not more technology, but a structured, evidence-based approach to value management.
Ready to develop a comprehensive AI value measurement plan for your organization? Contact Impact Advisors to learn how we can help you demonstrate and amplify the impact of your healthcare AI initiatives.
