How clinicians are using generative AI

For the survey, Wolters Kluwer partnered with Ipsos to survey 312 healthcare professionals across the United States between Feb. 26 and March 24. The healthcare professionals were divided into six categories: nursing, pharmacy, administration, physician, allied health, and medical librarian.

Currently, around 42% of clinicians say they use generative AI routinely at work, once a week or more. There were also some variations among different healthcare professionals, with medical librarians using generative AI the most (53%) and allied health professionals using the technology the least (26%).

When asked about how they wanted to use generative AI to improve their organizations, most respondents said addressing staffing shortages (82%), generating administrative efficiencies (77%), and reducing burnout (76%). 

"Every day, we see physicians struggling with workflow and the pressure to do more with less. This can be the straw that breaks the camel's back and results in exodus from the workforce," said Nisha Mehta, a radiologist and founder of Physician Side Gigs. "AI has been a welcome game changer for many early adopters who are seeking to grow and scale, both clinically and outside the exam room."

However, there continues to be a gap between what healthcare organizations want to achieve with generative AI and what they're currently capable of implementing — something the report says could be "a warning sign that [organizations] are more likely to focus on quick, immediate, small-scale wins instead of planning for a fully GenAI-enabled enterprise."

"Policies, procedures, and solid roadmaps will be essential for making GenAI a force for good in the healthcare community," the report said. However, only 18% of organizations have published policies on authorized uses for generative AI, and 20% require staff to follow formal training on the technology.

Among those who have published policies for generative AI, most have implemented data privacy rules (64%), transparency about how AI tools are trained (55%), and rules about the ethical implications of using generative AI (51%). 

"As with any new technology being introduced in the healthcare setting, effective governance and change management are essential for the successful and safe adoption of Gen AI tools," said Holly Urban, VP of strategy, clinical effectiveness at Wolters Kluwer Health. "Healthcare organizations must establish robust methodologies for evaluating, selecting, implementing, and monitoring these tools."

"Given the vast array of Gen AI tools available and their associated risks, employing well-defined evaluation methodologies will ensure good organizational fit for Gen AI tools," Urban added. "These tools aren't set-it-and-forget-it due to the risks of overfitting and model drift, so ongoing monitoring is critical."