Colonoscopy screening with additional AI could save $290 million annually

A researcher reports that the use of artificial intelligence (AI) during colonoscopy examination could be a cost-saving strategy that could also enhance the prevention of colorectal cancer (CRC) and mortality.

Among a simulated group of patients at average risk of developing CRC, compared with no screening, the relative reduction in CRC incidence was 44.2% with colonoscopy screening without AI tools and 48.9% with colonoscopy screening using AI tools, for gains An additional 4.8%. Additionally, compared to no screening, the relative reduction in CRC mortality was 48.7% for a colonoscopy examination without AI versus 52.3% for a colonoscopy examination using AI, for an additional 3.6% increase, reported Yuichi Mori, Ph.D. Medicine, from Showa Yokohama Northern University. hospital in Japan.

AI detection tools were also associated with savings of $57 per patient after cutting discounted costs from $3,400 to $3,343, which continued in secondary colonoscopy modeling analysis, Morey said in a presentation at Gastroenterology week (DDW). The results were simultaneously published in Lancet Digital Health.

“Across the US population, implementation of AI detection during colonoscopy screening resulted in an additional annual prevention of 7,194 [CRC] “Cases and 2,089 associated deaths, saving $290 million annually,” Morey and colleagues wrote.

“We are able to find a cost reduction as a whole, which is very surprising because the primary use of AI increases costs, but it can be achieved through the preventive effect of colorectal cancer,” Morey said at a press conference at DDW.

“Using AI to detect polyps leads primarily to increased costs because it can increase the detection of polyps and adenomas, and may increase the number of polyp resections, and thus can increase the number of control colonoscopies,” Morey explained. “However, this kind of increase can be mitigated by the benefits that come from the use of AI, that is, the effect of cancer prevention with increased ADR. [adenoma detection rates] by AI. So it is very important how the use of AI contributes to the healthcare system in terms of cost-effectiveness.”

The study had some limitations, including the fact that the authors hypothesized “a linear relationship between the effect of cancer prevention and an increase in ADR, [and] there ongoing discussion About whether there is a file ADR threshold effect in preventing cancer. Also, the authors hypothesized the same increase in the detection rate of high-risk adenomas as that of low-risk adenomas using AI to detect polyps, although the detection rate of advanced adenomas was not shown in [a] Previous dimensional analysis. “

Morey and colleagues used Markov model simulations in a hypothetical group of 100,000 US patients who underwent a colonoscopy exam with or without AI every 10 years, beginning at age 50 and ending at age 80. The patients had no personal or family history of CRC, adenomas, or inflammatory disease Bowel or hereditary CRC syndrome.

They stated that “the costs of AI tools and the cost of terminal treatment for screening for disease have been estimated at annual discount rates of 3%.”

The authors reported that based on an assumption of 60% screening uptake, colonoscopy reduced the incidence of CRC from 6.0% per 100,000 to 3.3% per 100,000, which corresponds to an absolute reduction of 2,638 cases per 100,000 and a 44.2% relative reduction. for non-examination. Compared with colonoscopy without AI, the application of AI reduced the incidence of CRC from 3.3% to 3.0% cases per 100,000 people, and the CRC mortality rate from 1.2% to 1.1% per 100,000 people, they reported.

“This corresponds to an additional 0.3% (8.4% relative decrease) decrease in [CRC] Incidence and 0.1% absolute decrease (6.9% relative decrease) from [CRC] Mortality compared to colonoscopy without artificial intelligence”.

Mori’s group also found that AI reduced costs related to CRC treatment by 8.2%, from $1,636 to $1,502 per person, although this was “partly offset by the cost of implementing the AI ​​that increased screening costs from $1,764” US$1841 to US$1841 per person (also including monitoring. Colonoscopy and treatment of adverse events).

“I would argue that the use of AI during a colonoscopy exam may be cost-effective in the United States,” Morey concluded.

Morey said his group has plans for a large randomized trial in Europe and Japan, with long-term follow-up of CRC as the primary end point.


The study was funded by the European Commission and the Japan Society for the Promotion of Science.

Morey revealed his ties to Olympus and the Cybernet System. The co-authors revealed multiple relationships with industry.