As health systems weigh how to deploy artificial intelligence amid growing hype, Ann Arbor-based Michigan Medicine is focused on using AI to support clinical outcomes.
Tim Calahan, MD, chief technology officer of Michigan Medicine, shared how the academic health system is approaching AI adoption, governance and technology prioritization during a live recording of the “Becker’s Healthcare Podcast” at the 10th Annual Health IT, Digital Health and RCM Meeting in October.
“There’s a lot of buzz, and there’s a lot of hype around AI,” Dr. Calahan said. “We’re really trying to use it not as a replacement for clinicians or for staff for that matter, but really to augment how staff perform their jobs today.”
One of the most impactful use cases has been ambient clinical documentation. Michigan Medicine recently completed a project using Microsoft DAX to enable ambient voice technology in clinical settings.
“We recently completed a project with Microsoft DAX, to where we’re doing ambient voice, within our clinical space,” Dr. Calahan said. “That has provided a great deal of benefit for clinicians in terms of engaging the patients and reducing the amount of time it takes for them to process some of their clinical work.”
The technology has had a meaningful impact on clinicians’ day-to-day experience. Dr. Calahan called the ambient technology “hugely transformative” for administrative workflows and has improved work-life balance.
The health system is also leveraging AI tools embedded within its Epic EHR, particularly to help clinicians manage patient communications more efficiently. Physicians can receive hundreds of messages per day and spend hours answering emails without extra support.
“Things like, message creation – being able to synthesize a message or a message that comes in – and then summarize that and provide a response back to patients has been super helpful for clinicians,” he said.
In addition to vendor-based tools, Michigan Medicine is beginning to develop its own AI capabilities.
“Whether we’re using those types of tools that are kind of embedded into the products that we have or whether we’re creating our own sets of tools, which we’re beginning to do a little bit of as well using some predictive analytics and things of that nature,” Dr. Calahan said. “It’s an exciting time, but we’re using it mainly to augment and enhance how we provide care throughout the healthcare system.”
But the health system isn’t moving full speed ahead. Governance and ethics have been central to the organization’s AI strategy. Dr. Calahan said Michigan Medicine and the University of Michigan moved early to establish guardrails around how generative AI would be used.
“Early on, Michigan Medicine and the University of Michigan at large recognized not only the power that generative AI could provide, but also the need to have proper governance and ethics in place,” he said.
That work included defining how AI tools would and would not be used, with an emphasis on maintaining human oversight. The leadership philosophy focused on AI that augments clinicians, not replaces them, and ensuring clinicians obtain patient consent. Operational discipline has also shaped how Michigan Medicine rolls out new technologies.
“As we’ve gone out with some of these initiatives, we’ve started very slowly to make sure that we are optimizing the workflow and making sure that the operational considerations are taking place, and then we scale,” he said. “We’ve been able to scale very successfully because of a lot of that preplanning and prework that we did at the very beginning of some of these work efforts.”
Looking ahead, Dr. Calahan said focus is the most important trait for health IT leaders navigating rapid technological change.
“What I’ve recognized is every healthcare organization is on their own journey and on their own maturity level as it relates to different types of technologies,” he said. “I think it’s important to understand the type of technology you want to get better at, and then how do you focus on making sure that you can execute on that particular effort.”
At Michigan Medicine, that focus has narrowed to a small set of priorities, including migrating and modernizing the health system’s infrastructure and leveraging the public cloud. The organization has also prioritized better integration with affiliates and a disciplined approach to AI adoption.
“It’s important to stay focused on those initiatives and not chase shiny objects,” he said.
