Vice Chair & Professor of Medicine, UNC | Balanced healthcare perspectives
In a new AMA survey, physicians reported feeling both excited and concerned about AI. Respondents (n=1,081) were most excited about tools that reduce administrative burden and most concerned that AI could tarnish doctor-patient relationships and jeopardize patient privacy. Only one-third have used AI, primarily for documentation, translation, and assisting diagnoses. To use AI more broadly, respondents wanted to be protected against liability, assured of data privacy, safety, and efficacy, and for the AIs to integrate with their EHR. https://bit.ly/4as0rLX My take: After decades of overpromising and underdelivering, the claim that better technology will necessarily improve healthcare is no longer credible. We do not want to be burned again. Having taken an oath to “do no harm,” we are (understandably) concerned about an emerging, incompletely understood, sometimes unreliable set of technologies. At the same time, many of us feel overburdened and are desperate for tools that will help us provide better care and get through our workdays more easily. We’d love to offload unrewarding, administrative drudgery, but it can be (surprisingly) hard to separate “admin” from “clinical” work. Of course, we’ve used different types of AI for years. But the term AI has become so broad that it’s lost its meaning. Ultimately, as a generally change-averse group that tends to work in slow-moving organizations, it will take meaningful incentives (not just largely unproven promises) to motivate us to change how we work and implement AI. For now, we'll continue to hold two opposing feelings -- excitement and concern -- at the same time. #healthcareai #physicianlife #healthcareonlinkedin