How Sanford Health is embedding artificial intelligence into its EHR to detect disease early and provide personalized care
Electronic health records hold so much patient data that clinicians often struggle to extract the data that really matters at the point of care.
Jeremy Cauwels, chief medical officer at Sanford Health, said in an interview at the Reuters Total Health conference in Chicago last month that one of the most promising aspects of artificial intelligence is its ability to help make sense of data and provide rapid insights so doctors can act faster and with more confidence.
He noted that his health system is embedding artificial intelligence into its electronic medical records to automate evidence-based care recommendations. For example, Sanford deployed a chronic kidney disease module that tracks when patients move from one stage of the disease to another and automatically prompts primary care physicians to the correct labs based on National Kidney Foundation guidelines.
Cauwels noted that the tool has even helped his own 76-year-old father keep on track with his monitoring schedule.
“Last year, he moved from CKD stage 2 to CKD stage 3 due to age,” Cowells explained. “Even if he had health care at home, I wouldn't consider asking for it and making changes. This is a reminder of how we can improve quality without requiring our brain power to do it.”
The model doubled the number of people with diabetes receiving recommended chronic kidney disease tests and tripled the rate of early diagnosis. Cowells noted that this allows for early intervention and helps reduce expensive dialysis costs, which is significant for rural patients who live far from dialysis centers.
He also described a new artificial intelligence-driven colon cancer screening tool that Sanford plans to embed into its electronic medical record in January.
Because current gastroenterology guidelines only classify patients as “normal” or “high risk,” Sanford built a model using 85 different variables to generate a personalized risk score for each patient.
This allows clinicians to tailor their conversations and recommendations, Cowells explained, which is especially important in the Midwest, where colon cancer rates are among the highest in the country.
“If you screen appropriately and catch it, it's absolutely treatable — but it's happening to younger and younger people, and if you don't catch it early, it's an extremely lethal or life-shortening cancer,” he said.
By turning complex data into clear and timely guidance, AI can help doctors detect disease earlier and tailor treatment to individual patients, thereby improving the quality of care without increasing cognitive load, Cauwels said.
Photo: Volha Rahalskaya, Getty Images
