Artificial intelligence (AI) is seeing rapid integration into healthcare, yet a recent survey from Wolters Kluwer Health reveals significant concerns among healthcare professionals. Many clinicians fear that over-reliance on AI tools might hinder their critical thinking and decision-making abilities.
The survey indicated that approximately 75% of doctors and 70% of nurses now utilize AI at least weekly, a notable increase from just 38% of doctors and 46% of nurses in the previous year. Dr. Peter Bonis, Chief Medical Officer at Wolters Kluwer, attributed this rise to greater exposure and familiarity with AI technologies. Importantly, he noted that these tools are effectively meeting previously unmet needs in the sector. However, 74% of clinicians expressed apprehension regarding potential skill deterioration, particularly as AI takes on more roles in diagnostics and treatment processes.
AI is fast becoming a central asset in healthcare, with expectations that it could alleviate workforce shortages and streamline various tasks, including patient documentation and data analysis. The transition from initial trials to regular AI usage is evident, as shown by the report, which analyzed input from over 350 healthcare professionals and more than 250 patients. For instance, 38% of physicians and 32% of nurses reported interacting with AI multiple times a day, a significant rise from just 10% and 16% the previous year.
Healthcare professionals are applying AI in diverse areas. More than half of the doctors indicated they use AI to summarize medical literature or for data analysis, while 44% leverage AI-driven scribe tools, which document physician-patient interactions and generate clinical notes. Patients, too, are engaging with AI, with over half using it to investigate medication side effects or to gather information regarding their diagnoses. Approximately 40% are currently utilizing or are open to employing AI to clarify medical jargon or interpret test results.
Despite the rising adoption of AI tools, clinicians remain apprehensive about their implications for professional competency. While research specific to deskilling in clinical settings is limited, findings from other sectors suggest a potential detriment to fundamental skill development, according to Dr. Bonis. Clinicians risk mastering tasks incorrectly and might experience a decline in pre-existing abilities.
Moreover, the issue of AI hallucinations—instances where AI tools generate inaccurate information—adds to the unease, with about 75% of clinicians highlighting this as a key concern. Although 73% expressed confidence in their ability to detect erroneous responses, this still leaves a significant portion uncertain about identifying hallucinations. Dr. Bonis noted that the complexity of discerning inaccurate medical information presents challenges. For instance, AI may incorrectly reference primary sources or cite valid research while overlooking other studies that contradict its findings.
Awareness of AI governance policies among clinicians appears limited. The survey revealed that only 27% of healthcare professionals were informed about how their organizations are managing AI governance. Among those aware, 63% understood the implications of privacy regulations like HIPAA on AI usage. However, only 35% were familiar with guidelines concerning the accuracy and reliability of AI data, and just 22% reported that their employers had policies outlining the responsibilities of clinicians in relation to AI tools.
Dr. Bonis emphasized the evolving nature of AI integration in healthcare, noting the uncertainty regarding accountability for the repercussions on care delivery and the associated risks. As the adoption of AI tools becomes more widespread, ongoing learning and adaptation will be essential for healthcare systems.
