At a Sept. 16 panel at Becker’s 11th Annual IT + Revenue Cycle Conference, three health system technology leaders were asked to share an example of a project’s specific ROI. Chif Umejei, CIO at New York City-based NewYork-Presbyterian, answered with a single patient.
Mr. Umejei’s team built a diagnostic algorithm with cardiologists at New York City-based Columbia University that reads an EKG for signs of undiagnosed structural heart disease. The clearest proof of its value, he said, was not a population-level statistic but a patient who came in coughing up blood and was having trouble breathing, got an EKG as part of a routine workup and was flagged by the algorithm.
“This person had structural heart disease,” Mr. Umejei said. “The person had a viable heart transplant.”
He said arguments about cost savings or efficiency gains matter less to him than outcomes such as that one.
“We could talk about dollars, cents all the time,” he said, “but we had a meaningful impact in that person’s life.”
Nari Gopala, chief digital officer at Rochester, Minn.-based Mayo Clinic, pointed to something less dramatic but more systemic: the discharge summary. Historically, he said, those documents functioned as an exhaustive record of everything that happened during a hospital stay — useful as an archive, not as guidance.
“We dumped everything on you,” Mr. Gopala said. “It’s like every single thing that we did.”
For Mayo’s destination-medicine patients, who typically return home to a local primary care provider for follow-up, that approach left a gap: the receiving physician often could not tell what needed to happen next.
Mr. Gopala’s team rebuilt the summaries to feature more prominently what a patient came in with, how it was diagnosed, and precisely what follow-up was required. The result, he said, was “a significant decrease in terms of readmissions.”
Thomas Bentley, BSN, RN, chief information and digital transformation officer at Columbus-based Ohio State University Wexner Medical Center, pointed to something more incremental but broadly adopted: AI-drafted visit summaries for physicians and end-of-shift care plan updates for nurses.
He estimated the summaries save clinicians 15 to 20 minutes of chart review per encounter, and said uptake has been unusually strong: Clinicians use the AI draft “north of 70%” of the time it is offered. For nurses, he said the tool addresses a specific complaint: “I didn’t go to school to learn to document in the EMR,” he said clinicians tell him. “I went to school to care for patients.”
Note: This story was updated at 6:26 p.m. Central time Sept. 16 to correct the reason the NewYork-Presbyterian patient was being treated.
