Skip to main content

DIME PROJECT

AI-Enabled Care Navigation Framework
for Health Systems

AI Care Nav icon
Where navigation breaks down & where AI can help, responsibly
AI Care Navigation

Impact-Utility-Trust
A checklist for evaluating AI-enabled care navigation solutions

The Impact–Utility–Trust Checklist helps health systems decide which AI use cases make care navigation more efficient, scalable, safe and personalized for patients, and which aren’t ready yet. It is built around three questions:

IMPACT

Does the potential gains for patients, care teams and the organization, and the strength of the evidence behind them, justify the investment and risk?

UTILITY

Does the capability remove friction from navigation workflows for patients and care teams alike?

TRUST

Under what conditions can this capability be deployed responsibly and kept that way over time?

Download the checklist (PDF)

Who this checklist is for

The checklist is written for the people inside a health system who decide whether to buy, pilot or scale an AI navigation tool. Patients and caregivers don’t fill it in, but their perspective runs through it – several items can only be answered by asking them, and those items are tagged. DiMe’s Playbook: Implementing AI in Healthcare recommends a patient or family representative on every AI implementation task force. This checklist assumes one is at the table. Questions about trust, burden, language, and whether a tool fits patients’ daily lives are best answered by the people living them.

STAKEHOLDERS & THE QUESTIONS THEY BRING
Will this measurably improve the financial metrics that matter, and can we attribute the savings?

Is this safe to deploy and will it measurably improve patient outcomes?

Will this make coordination easier, or just add another system to manage?
Can we integrate this safely into what we already have?
Will this make my next step easier, and can I reach a person when I need one?
Economic
Will this measurably improve the financial metrics that matter, and can we attribute the savings?
Clinical
Is this safe to deploy?
Operational
Will this make coordination easier, or just add another system to manage?
Technical
Can we integrate this safely into what we already have?
Patient & Caregiver
Will this make my next step easier, and can I reach a person when I need one?

When to use it

Use the checklist at three decision points:

DECISION POINT

What you’re deciding? Evidence you can expect

Is this worth piloting here? Vendor documentation, published studies, peer references, your own current-state data on the problem

What you’re deciding? Should we run a bounded pilot, and under what limits?

Is this worth piloting here? Retrospective or silent-mode testing on your data, usability testing with your patients and staff, a monitoring plan

What you’re deciding? Should this expand beyond the pilot?

Is this worth piloting here? Pilot results against your baseline, results by subgroup and channel, safety events, staff workload, cost

Vendor selection

What you’re deciding? Evidence you can expect

Is this worth piloting here? Vendor documentation, published studies, peer references, your own current-state data on the problem

Pilot Go/No-Go

What you’re deciding? Should we run a bounded pilot, and under what limits?

Is this worth piloting here? Retrospective or silent-mode testing on your data, usability testing with your patients and staff, a monitoring plan

Scale-up

What you’re deciding? Should this expand beyond the pilot?

Is this worth piloting here? Pilot results against your baseline, results by subgroup and channel, safety events, staff workload, cost

DiMe’s Playbook describes the testing methods behind each stage, from retrospective validation and silent mode through to a prospective pilot (DiMe’s Playbook: Implementing AI in Healthcare > Plan).

Next steps

This checklist, and the principles throughout the AI-enabled care navigation framework, are meant to structure discussion and sharpen your team’s strategy. Use this checklist as a complement to the broader care navigation framework, and DiMe’s Playbook: Implementing AI in Healthcare or other ecosystem resources as you procure and evaluate AI-enabled tools.

Back to project homepage