Posted By :  

A recent observational study published in The Lancet Gastroenterology & Hepatology suggests that healthcare professionals may experience a dip in their ability to detect pre-cancerous growths—specifically adenomas—when AI assistance is removed. Researchers compared colonoscopy performances before and after the introduction of an AI tool and found that the adenoma detection rate without AI decreased from approximately 28.4% to 22.4%, even though AI-supported procedures maintained a 25.3% success rate.

Experts warn this “de-skilling” effect—where over-reliance on AI impairs independent diagnostic capability—could extend beyond colonoscopies to other clinical settings. Dr. Catherine Menon notes that physicians accustomed to AI support may perform worse if systems fail or are unavailable, for instance due to cyber-attacks or technical glitches. Critics also highlight potential contributing factors: for instance, increased workload after AI adoption might lead to fatigue and reduced vigilance, rather than AI itself being the sole cause . Still, the study raises important concerns about maintaining human expertise amid growing reliance on technology.

To safeguard patient outcomes, the authors recommend striking a careful balance: continue leveraging AI for its advantages, but avoid letting automation replace fundamental clinical skills. One suggested strategy is to schedule regular non-AI periods—so physicians remain proficient without technological support—and to conduct more behavioral research exploring how AI affects decision-making and attention.