Doctors’ True Thoughts on AI Unveiled

  • A new survey from the American Medical Association (AMA) has revealed that two thirds of participating physicians see benefits in the application of AI in their fields, while one third doubts its merits.
  • The main advantage cited was a reduction in administrative burden, while concerns were voiced over patient privacy and the possible negative impact on the physician-patient relationship.
  • Only 38% of respondents reported currently using AI in their practice, with uses including documentation, translation services, and diagnostic aid.

In an increasingly digital healthcare environment, the use of artificial intelligence (AI) is becoming more endemic. However, opinions among physicians differ on the potential benefits and disadvantages of leveraging AI in their practices.

In a recent AMA survey, just two thirds of the participating physicians responded that they see worth in AI applications, while the remainder perceived the technology as lacking benefit or introducing new risks. The survey, conducted in August, collected responses from 1,081 practicing physicians nationwide.

Among those polled, 41% expressed mixed feelings about AI, balanced equally with excitement and skepticism. The dominant advantage credited to AI was a reduced administrative burden. Predominantly, physicians anticipate AI tools to aid them with time-consuming documentation and procedures requiring prior authorization. Current solutions delivering this kind of support include Augmedix, Nuance, Cohere Health, and Surescripts.

Another application of AI of interest to physicians was in supporting diagnosis. Companies like Aidoc and PathAI are presently providing this type of technology. However, at the time of the survey, only 38% of respondents were incorporating AI into their practices. These users primarily employed AI for documentation, translation services, and diagnostic assistance.

Looking forward, physicians expressed interest in adopting AI solutions within the next five years to facilitate with chart summaries, demand prediction, and workforce forecasting.

On the flip side, concerns were raised regarding patient privacy and the potential impact on the patient-physician relationship. Ethical questions are arising about who should gain access to patient data, how the information should be used, and to what extent patients are informed about these processes. Additionally, the introduction of AI is perceived by some as risking a dehumanization of healthcare, potentially damaging the trust and empathy integral to the patient-physician dynamic.

The survey suggests that to encourage more widespread adoption of AI among physicians, vendors should focus on improving their data privacy assurances and provide indemnities such as malpractice insurance and a release from liability for model errors.