Many physicians are viewing artificial intelligence as a double-edged disruptor: Its high potential for improving physicians' day to day workflow is paired with high potential for abuse.
"Artificial intelligence has potential to completely disrupt the way healthcare is delivered in the foreseeable future with potential improvements in efficiencies as well as diagnostic accuracy," Brian Curtin, MD, an orthopedic surgeon at Charlotte, N.C.-based OrthoCarolina, told Becker's. "However, the potential is also there for further erosion of the physician-patient relationship and overreliance on algorithm-based clinical decision-making that may in some instances result in worse patient care and outcomes."
This "slippery slope" requires oversight to be implemented successfully, he added.
AI can refine operations as healthcare policy has become increasingly complex, with prior authorizations and documentation requirements burning out providers.
"AI can streamline care by reducing physician administrative burden, improving diagnostic accuracy, eliminating redundancy and minimizing physician errors," Sean Moroze, MD, an orthopedic surgeon affiliated with Ascension Medical Group Sacred Heart Bay Orthopedics in Panama City, Fla., told Becker's. "The introduction of AI into electronic medical records will occur piecemeal, with often undetectable change. However, in the coming decade, such technology will be commonplace and come to be expected by physicians and other providers."
Physicians and staff spend an average of two whole workdays per week completing prior authorizations, taking them away from patients and patient care for about 14 hours a week, according to a March 14 report from the AMA that surveyed 1,001 practicing physicians.
Additionally, 84 percent of employed physicians said administrative burdens from commercial payers and government insurance programs had an effect on their employment decision, according to the AHA survey.
Many physicians are wary of artificial intelligence, however, because of a lack of oversight.
"My concern is the use of AI in healthcare needs to be evaluated and governed before unleashing its potential in clinical care," Matt Mazurek, MD, assistant clinical professor of anesthesiology at St. Raphael's Campus of Yale New Haven (Conn.) Hospital, told Becker's. "It is an unproven technology, and solving the problem of ownership of how it is used and to what extent it is used requires our systems to slow down and be very deliberate."
If implemented responsibly, it could alleviate some of healthcare's issues of burnout, but Dr. Mazurek said he worries its downsides are potentially being overlooked.
"The use of AI is the next revolutionary event not only in healthcare but in all aspects of our lives," he said. "I hope the technology is used properly and appropriately for the right reasons. Slowly incorporating the technology will assist us in addressing both ethical and legal issues. The potential for using AI in nefarious ways also needs to be appreciated."