AI in Colonoscopy: Promising Yet Problematic
The role of artificial intelligence (AI) in healthcare is expanding, but recent findings highlight potential risks. A new study conducted by European researchers looked into how AI impacts colonoscopy procedures, especially in the detection of tumors.
This research involved clinicians from four endoscopic centers in Poland, where AI tools for detecting polyps were introduced at the end of 2021. Colonoscopy sessions at these centers were randomly assigned to either have AI support or not.
The quality of the colonoscopy was assessed by analyzing tumor detection rates, specifically looking at adenoma detection rates (ADR) three months before and after the AI implementation. From September 2021 to March 2022, 1,443 patients underwent non-AI-supported colonoscopy before and after the AI tools were introduced.
Interestingly, the study revealed a “significant” decline in detection rates, dropping from 28.4% before AI deployment to 22.4% afterward. These findings were shared in the Journal Lancet Gastroenterology and Hepatology.
Researchers noted that exposure to AI, along with patient age and gender, were key factors influencing tumor detection rates. They concluded that ongoing reliance on AI could actually lower adenoma detection rates during standard colonoscopy, adversely affecting medical outcomes.
Harvey Castro, an AI-focused emergency physician based in Texas, commented on the study, describing it as significant. He pointed out a healthcare paradox: AI can enhance cancer detection, but its absence might detract from a doctor’s ability to identify key issues when these tools aren’t accessible.
He also stressed that even a small reduction in detection rates could have serious consequences for cancer survival across many patients. Castro emphasized the importance of addressing the 6% reduction revealed by the study, stating, “Even small changes in the detection of adenomas can alter cancer outcomes.”
Experts advocate for a careful integration of AI rather than an outright rejection. Castro sees AI as potentially transformative in saving lives, noting the need for balanced approaches that enhance rather than diminish the role of clinicians. He insisted that the future of medicine should involve both doctors and technology working in concert to provide the best possible patient care.
In his view, achieving this balance requires intentional training and design strategies. He concluded by highlighting the irreplaceable value of human insight in medicine, reminding us that the physician’s judgment remains essential.
