AI-assisted colonoscopy improves cancer prevention while increasing cost and patient burden

AI-assisted colonoscopy has increased the proportion of patients requiring extensive monitoring by 35% in the US and 20% in Europe, which may improve cancer prevention but increase costs and patient responsibility.

An important part of the costs and burdens for patients with polyps is colonoscopy monitoring after polyp resection. Current colonoscopy guidelines recommend frequent observation of patients with polyps.” Yuichi Morey, Masters, Ph.D., From the Clinical Efficacy Research Group at the University of Oslo in Norway, colleagues wrote in Clinical gastroenterology and hepatology. Recently, the use of artificial intelligence to detect polyps has been shown to be increasing [adenoma detection rate] From monoscopic about 12%. While there is a benefit of increased ADR, there is also an increased burden associated with more intense colonoscopy monitoring.”

In a pooled analysis of nine randomized controlled trials in China, Italy, Japan, and the United States, Morey and colleagues compared colonoscopy with and without AI in 5,796 patients (51% men; median age, 53 years). The primary outcome was the proportion of patients recommended to undergo subsequent intensive monitoring, which was defined as 3-year monitoring based on guidelines from the ASGE, the European Society of Gastrointestinal Endoscopy, and the Japanese Society of Gastrointestinal Endoscopy.

A total of 2894 patients underwent AI-assisted colonoscopy and 2902 patients underwent standard colonoscopy. The researchers reported a higher incidence of adverse drug reactions among patients in the AI ​​group compared to the non-AI group in all trials.

When following US and Japanese guidelines, the proportion of patients recommended for intensive monitoring increased from 8.4% (95% CI, 7.4-9.5) in the non-AI group to 11.3% (95% CI, 10.2-12.6) in the AI ​​group with a difference An absolute risk ratio of 2.9% (95% CI, 1.4-4.4) and a hazard ratio of 1.35 (95% CI, 1.16-11.57). The proportions also increased when European guidelines were followed, from 6.1% (95% CI, 5.3-7) to 7.4% (95% CI, 6.5-8.4) with an absolute difference of 1.3% (95% CI, .01-2.6) and a hazard ratio of 1.22 ( 95% CI, 1.01-1.47).

Among patients who underwent colonoscopy for colorectal cancer screening, the proportion of patients who recommended intensive surveillance increased from 8.1% (95% CI, 6.1-10.5) to 10.8% (95% CI, 8.6-14.4) with an absolute difference of 2.7% (95% CI, -0.5 to 5.9) and a risk ratio of 1.32 (95% CI, 0.95-1.84) when US and Japanese guidelines were followed. When European guidelines were followed, it increased from 6% (95% CI, 4.3-8.1) to 6.6% (95% CI, 4.8-8.8) with an absolute difference of 0.6% (95% CI, -2 to 3.2) and a risk ratio 1.09 (95% CI, 0.72-1.64).

Morey and colleagues conclude, “Our study demonstrated the effect of AI on baseline risk stratification, shifting a large proportion of patients into higher-risk categories with little effect on the proportion of patients in low-risk groups.” This, in turn, prompts more intensive monitoring after polyp removal, which may lead to more effective cancer prevention.

They continued, “Such surveillance strategies must take into account the balance between high efficacy, on the one hand, and endoscopic ability and risk of overdiagnosis, on the other.” “Large, population-based trials with long-term follow-up will bring clear answers to these important questions.”