Rethinking the ROI of AI

Clinical Report: Rethinking the ROI of AI

Overview

Myra Cherchio argues that ophthalmology practices should assess AI’s return on investment not only through labor and time savings, but also through its potential to support decision-making, expose assumptions, and improve workflows. Her examples describe using AI to pressure-test an employee incentive structure and map revenue-cycle processes, revealing gaps in documented procedures and ownership.

Background

Ophthalmology practices face rising labor costs, tightening reimbursement, and pressure to improve efficiency, according to the article. The author distinguishes automating repetitive, logic-based tasks from using AI to support work that depends on human judgment and connection. She compares AI adoption with electronic health record implementation, noting that technology can reproduce existing inefficiencies when work is not redesigned. The article presents AI as a tool for examining decisions and systems, rather than simply adding automation to current processes.

Data Highlights

The article provides practice examples but reports no numerical trial data or quantified outcomes. In the incentive-design example, the author modeled aggressive, moderate, and modest-success scenarios; in the revenue-cycle example, AI-assisted workflow mapping identified areas where procedures and ownership were insufficiently documented.

Key Findings

The author describes AI as useful for challenging assumptions, identifying gaps in reasoning, and considering alternative perspectives before high-stakes organizational decisions.
In an employee incentive-design exercise, AI was used to compare strategies, consider possible behavioral effects and unintended consequences, and develop a simpler structure linking rewards to desired behaviors.
Modeling different success scenarios helped frame a discussion about employee-controlled actions and their relationship to practice goals, according to the author.
Mapping the revenue cycle with AI exposed reliance on individual knowledge and areas where standard operating procedures or explicit ownership were lacking.
The author reports that the mapping exercise led to a comprehensive, living training resource and improved accountability among staff involved in the workflow.
The article argues that AI’s value may include improved decision quality and workflow design, in addition to reduced administrative effort and cost.

Clinical Implications

For practice leaders, the article describes using AI to clarify a problem and its constraints, invite sequential questions, challenge assumptions, and generate counterarguments before making consequential decisions. It also describes workflow mapping as a way to examine handoffs, documentation, and ownership; the article does not report quantified effects or clinical outcomes.

Conclusion

The article frames AI’s potential return in ophthalmology as extending beyond automation to decision support and organizational learning. Its examples are practice-level accounts rather than quantified evaluations of effectiveness.