Sanford Health is integrating AI into its EHR to convert extensive patient data into practical insights. New tools — including a chronic kidney disease module and a personalized colon cancer screening model — are enabling clinicians to detect conditions earlier and deliver more individualized care, said Jeremy Cauwels, the system’s chief medical officer.
Electronic health records contain massive amounts of patient information — often more than clinicians can easily interpret during a visit.
AI’s ability to sift through that data and surface timely, meaningful insights is becoming one of its greatest benefits, said Jeremy Cauwels, chief medical officer at Sanford Health, in an interview at Reuters’ Total Health conference in Chicago.
He explained that Sanford is integrating AI into its EHR to automate evidence-based care recommendations. For example, the health system now uses a chronic kidney disease module that monitors disease progression and automatically alerts primary care providers with the appropriate lab orders based on National Kidney Foundation guidelines.
Cauwels shared that the tool even helped keep his own 76-year-old father on the correct monitoring path.
“Over the past year, he moved from stage two to stage three chronic kidney disease due to age,” Cauwels said. “Even with a family in healthcare, I wouldn’t have thought to ask about updating his monitoring. The system reminded us how to improve quality without relying on our own memory.”
According to Cauwels, the model has doubled the number of diabetes patients receiving recommended kidney disease tests and tripled early detection rates. Earlier identification supports faster intervention and can help prevent costly dialysis — especially important for rural patients who face long travel distances to dialysis centers.
He added that Sanford also plans to roll out an AI-driven colon cancer screening tool within the EHR in January.
Since existing gastroenterology guidelines classify patients only as “normal” or “high-risk,” Sanford developed a model that uses 85 variables to create a personalized risk score for each patient.
This approach allows clinicians to offer more precise guidance and discussions — a critical need in the Upper Midwest, where colon cancer rates rank among the nation’s highest, Cauwels said.
“It’s absolutely preventable if you screen properly and catch it early — but it’s showing up in younger patients, and when it’s missed, it can be highly fatal or severely life-limiting,” he noted.
By turning complex data into clear, timely recommendations, AI supports earlier detection and individualized care — boosting quality without adding to clinicians’ cognitive load, Cauwels added.
