Artificial Intelligence and Physician Deskilling in Gastroenterology

Artificial intelligence (AI) has shown great promise in helping physicians improve their diagnostic accuracy of important patient conditions, particularly in the realm of gastroenterology. AI has been proven to assist human physicians in better detecting small polyps (adenomas) during colonoscopy, which are at risk for turning into cancer if not removed early. However, a recent European study has shed light on the potential negative effects of over-reliance on AI among physicians.

The study, conducted in four different medical centers, tracked the adenoma detection rate (ADR) of physicians who performed colonoscopies with and without AI assistance. The results showed that while the ADR with AI assistance was slightly higher than without AI, the ADR significantly decreased when physicians who had become accustomed to AI support stopped using it. This phenomenon, known as medical “deskilling,” can lead to a decline in physicians’ skills when they no longer have access to AI assistance.

Researchers found that continuous exposure to AI decision support systems may cause clinicians to over-rely on AI recommendations, leading to a decrease in motivation, focus, and responsibility when making cognitive decisions without AI assistance. This raises concerns about the long-term impact of AI on physician skills and patient outcomes.

While this study focused on gastroenterology, similar deskilling effects may occur in other branches of medicine, such as radiology, as AI becomes more routinely adopted. It is essential for physicians, policymakers, and the general public to consider the implications of AI on medical practice and patient care.

Key Questions to Consider

1. How are patient outcomes and physician error rates affected by the use of AI in medical practice?
2. What methods can help physicians minimize deskilling and recover lost skills when AI assistance is discontinued?
3. Can AI be implemented in a way that enhances physician capabilities without leading to deskilling?

Deskilling is a complex issue that requires further research and exploration across multiple medical specialties. As AI technology continues to advance, it will be crucial to develop best practices that optimize patient outcomes while ensuring that physicians maintain their skills and expertise in the era of AI.