“Patients are healthcare’s most valuable resource and the best hope for the future of our healthcare system.”

Dr. Danny Sands (combining quotes from John F. Kennedy and Warner Slack)

There has been some progress in devising and deploying artificial intelligence in clinical medicine and healthcare, but there has been very little emphasis on how patients would perceive this new paradigm in clinical practice. This paper from JAMA Open Network is a presentation of a nationally representative survey to understand the public perceptions of the use of artificial intelligence in clinical diagnosis as well as treatment. The survey used a probability-based representative panel that was weighted to correct for biases in sampling.

The results of the survey from close to 1,000 respondents showed that most patients believed that AI would make health care much better or somewhat better (11% and 45%, respectively) with relatively few that feel AI would render health care worse or much worse (4% and 1% respectively). In addition, the majority of those surveyed feel that it is very or somewhat important (66% and 30%, respectively) that they be informed of AI playing a big role in their diagnosis or treatment. Interestingly, some of the respondents were very uncomfortable (31%) or somewhat uncomfortable (41%) with receiving an AI-derived diagnosis that was 90% accurate of the time but not able to explain its rationale. So both patients as well as clinicians remain concerned about AI not being able to be explained to some extent (yet probably much more comfortable with other technologies that remain somewhat mysterious, such as MRI and pacemakers). The comfort level varied as expected with the type of clinical application (chest radiograph vs cancer diagnosis). Again, patients have similar concerns with AI in healthcare (misdiagnosis, privacy breaches, less time with clinicians, and higher costs) as clinicians.

This manuscript is enlightening in informing us all of the patients’ perception of AI in healthcare. Perhaps it would be even more interesting to have a parallel survey of clinicians with more or less the same questions. Overall, it is a very useful piece to understand where we currently are with patients in the AI in health arena.

Read the full paper here: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2792374