When it comes to weaving AI into daily hospital work, executives are rushing the field while their employees hesitate on the sidelines.
A few health system leaders have told Becker's they have concerns about unintended harm from artificial intelligence's miscalculations, but many are charging ahead with using generative AI such as ChatGPT.
New Orleans-based Ochsner Health plans to soon launch a limited ChatGPT technology that helps clinicians answer emails; Rochester, Minn.-based Mayo Clinic is using Google Cloud's generative AI to quickly find patient information in EHRs; and Chapel Hill, N.C.-based UNC Health streamlines daily tasks with an internal version of ChatGPT.
"This surge in AI-enabled solutions harbors immense potential to enhance both operational and clinical efficiencies and improve early disease detection and clinical decision-making," said Richard Mulry, president of New Hyde Park, N.Y.-based Northwell Holdings, adding that investments in workforce education and cloud infrastructure are vital.
As some studies show minor deviations between AI and human results for duties such as taking notes and making decisions, researchers and hospital employees are urging caution.
One reason front-line healthcare workers are worried is because they think their bosses will use generative AI as a reason to lay off workforces that are already stretched thin, The Washington Post reported Aug. 10.
Melissa Beebe, RN, an oncology nurse at UC Davis Medical Center, told The Wall Street Journal that AI once misdiagnosed a patient for sepsis, but she drew blood — which increases patient's costs and can expose them to infections — because she faces discipline if she's wrong.
Ms. Beebe said she feels "moral distress" when her expertise is pitted against AI.
Healthcare workers also argue there's no equal replacement for strong provider-patient relationships and reading body language.
Nurses are trained to look beyond vital signs and monitor a patient's countenance and smell unnatural odors, Michelle Mahon, the assistant director of nursing practice at the National Nurses United union, told the Post. "AI can't do that," she said.