Community Health System

Laura Shue

Platform Transformation Strategist

Most nights, my spouse and I end our day with a brisk stroll around our neighborhood. This time is to connect on all the things a busy couple navigates: homework, household repairs, grocery shopping, negotiating logistics of two grade-school and one college-aged child. Being self-ascribed nerds, our nightly conversations inevitably turn to AI, digital transformation and disruption.

The definitions of “transformation” and “disruption” are just as crucial to any discussion of transforming the healthcare ecosystem as the mutual understanding of “who is taking our daughter to her appointment tomorrow” between spouses in household logistics.

American healthcare resides presently at the intersection of disturbance and interruptions. How do we honestly face our current realities in healthcare and hold onto the promise of generative and agentic AI? The answer is- we don’t transform through disruption alone – not if we want to be left without something worth having. We need leadership and commitment. We need humans.

As we walked around, I recalled the time I was hired as Director of Health Information Services at a large academic medical center in the Midwest. I was hired because of my analytic skills, leadership capabilities and my existing relationships with local physicians - and spent my first weeks on the new job feeling out of league. My command structure had quintupled. Electronic health record had just gone live the month before. Staff was now scrambling to make sense of their new world with a relatively inexperienced new leader.

I had no background in Health Information Management. I lacked technical subject expertise. Despite that, my team and I thrived. In short order, I was given additional leadership responsibilities and areas of functional oversight. How did I lead my team to success, given the unfamiliar terrain in all directions?

“AI tools are imperfect and require our critical thinking. Not all of us have to be experts in the underlying technology. Most of us just need to know how to use the tools to work more efficiently and effectively.”

At this point in our conversation, I turn to my husband and excitedly announced, “We have been here before! We can do this.” He looked back at me, puzzled by this remark in context of our discussion of various AI technologies. He didn’t spend six years as an HIM Director. And yet, as we continued, he helped me to unpack the core components of our success and how they apply to AI today.

Vulnerability: I was comfortable not having all the answers. Our team candidly approached failures as learning opportunities. AI tools are imperfect and require our critical thinking. Not all of us have to be experts in the underlying technology. Most of us just need to know how to use the tools to work more efficiently and effectively. That requires trial and error. It requires failures initially.

Coachability: I was willing to give and receive direct, helpful and honest feedback. We need teams and individuals who commit to success by providing uncomfortable but necessary feedback. We are also needed to coach generative and agentic AI systems via feedback and reinforcement learning.

Inclusivity: I was able to build a team of people with diverse subject expertise at various levels of professional experience and democratize decision-making processes. We need to integrate AI tools into our analytics workflows and use cases. We need to envision entirely new analytics workflows. Furthermore, we require people who know how things have always worked together at the same table with people who carry a vision for how things could work in the future.

Confidence: There were people who didn’t make the transition from the paper medical record world to the electronic medical record world. This included approximately 20% of my own staff who were unable to make the mental shift to the digital transition. The core work of Analytics for your organization does not change. With AI, you will simply shift how you do your work. It’s not that those people were incapable of the shift; just that they believed they couldn’t, and they never changed their minds.

As an industry, as a health system, as leaders and as individual contributors, we too can be at our best when we are our most afraid or uncertain. Because I had no investment in the way things had been done, I was able to see the wide-open door of possibilities of electronic workflows and data exchange. We need to think of AI tools in the same way, bearing in mind the lessons of 2010-2015 and EHR implementations. We have been here before. We can do this.