What Psychiatrists Think About Artificial Intelligence

August 6, 2019

Results of the first global survey of psychiatrists on the impact of future autonomous technology.

NEWS BRIEF

About 50% of psychiatrists in a recent survey believe that artificial intelligence (AI) and machine learning will significantly transform the way they work, but only 4% think that future autonomous technology will replace them. Sermo, a global networking platform open to licensed physicians, conducted the survey in partnership with psychiatry and health technology researchers at Duke University School of Medicine and Harvard Medical School.1,2

The pilot survey included 791 psychiatrists in 22 countries. Respondents were asked to assess the likelihood that future technology would be able to replace-not just assist-physicians in performing complex tasks. Psychiatrists were also asked about the benefits and risks of future technologies in mental health and the impact on their daily work.

Among the salient findings of the survey:
• Only 4% of psychiatrists felt that future technology would make their jobs obsolete
• Only 17% believed technology is likely to replace a human’s role in providing empathetic care
• More women (48%) psychiatrists than men (35%) were uncertain that the benefits of AI and machine learning would outweigh the risks
• More US psychiatrists (46%) than those in other countries (32%) were uncertain that the benefits of future autonomous technology would outweigh the risks

The majority of psychiatrists also indicated that future technology would be unlikely to replace physicians for complex tasks such as a mental status examination (67%), assessing the risk of violence (58%), and determining the need for hospitalization (55%).

There were only two tasks that the majority felt technology would likely replace:
• Providing patient documentation, such as updating medical records (75%)
• Synthesizing patient information to reach diagnoses (54%)

“It is time for us to stop thinking about AI as a battle of machines versus humans. We need to instead focus on how AI can optimize and improve clinicians’ abilities to deliver better care,” said study author Murali Doraiswamy, MBBS, Professor of Psychiatry and Medicine at Duke University School of Medicine, in a press statement.

The skepticism and uncertainty expressed by physicians in the survey could result from several factors. One possibility is that they place high value on human interaction and personalized professional analysis. Another possible explanation is that respondents are underestimating the pace of technological change.

“The findings from this survey also raise questions about the preparedness of the profession to navigate technological change in the delivery of patient care,” said study author Charlotte Blease, PhD, Research Fellow in General Medicine at Harvard Medical School, in a press statement.

While physicians were skeptical about the prospects of AI and machine learning replacing them, one in two psychiatrists felt that future technologies would significantly transform their jobs. Psychiatrists also predicted that AI and machine learning could help ensure more accurate diagnosis, reduce administrative burden, provide 24/7 monitoring, individualize drug targets to reduce adverse effects, and integrate new streams of data from wearables and genetics.

Survey respondents had many ethical and safety concerns about AI, however. “This should be a high priority for research, since even a single line of bad code could have serious repercussions,” cautioned Dr Doraiswamy.

The full scientific report can be accessed at https://arxiv.org/abs/1907.12386.

References:

1. Doraiswamy PM, Blease C, Bodner K. Artificial intelligence and the future of psychiatry: insights from a global physician survey. https://arxiv.org/abs/1907.12386. Submitted July 29, 2019. Accessed July 31, 2019.

2. First global survey of psychiatrists on the impact of artificial intelligence [press release]. New York, NY: Sermo; July 30, 2019.