How AI Can Improve the EHR and Bedside Medicine

In a recent JAMA Network Viewpoint feature, Stanford Medicine 25’s Abraham Verghese and Sonoo Thadaney readily endorse artificial intelligence (AI) in medicine, describing its promise as “undeniable.” But Verghese and Thadaney also emphasize there is an overlooked opportunity, specifically how AI can “help clinicians deliver better and more humanistic care.”

In the essay, they reference the current electronic health record (EHR). Since it was originally intended for billing purposes, it fails to paint a holistic picture of a patient. Verghese and Thadaney call upon innovators who they feel are “surely capable of something better.” Their recommendation is for technology that can help physicians understand “what sort of a patient has a disease,” rather “than what sort of a disease a patient has.” They outline two examples:

·       An EHR able to prepare a “clinical team with previsit material” presented “in a vivid useable form with graphics and animation.”

·       Natural Language Processing technology and video that capture clinical encounters.

With an enhanced understanding of the patient, rather than a sole focus on the disease, Verghese and Thadaney say that providers can prioritize the actual patient and ask more informed questions. Ultimately, this allows “clinicians to better engage in” their care at the bedside.

Doctors are burdened with considerable time restraints. As noted by the authors, this can create “angst, depression, and disenfranchisement” among physicians. But that’s only part of the story. The authors describe the “desirable attributes of humans who choose” to care for others naming empathy, expression, generosity, advocacy and a commitment to the greater good. These characteristics of bedside medicine are critical, and Verghese and Thadaney say the true “breakthrough” will be AI that can help “clinicians nurture and protect these qualities.”

Subscribe to our mailing list