ColonFlag for identifying people at risk of colorectal cancer

The technology described in this briefing is ColonFlag. It uses routinely available datasets to help identify people who are at high risk of developing colorectal cancer.

The innovative aspects are that ColonFlag uses so‑called big data and machine learning methods.

The intended place in therapy would be as an addition to current investigations, such as faecal immunochemical tests, to help identify people who may need referral for suspected colorectal cancer.

The main points from the evidence summarised in this briefing are from 4 observational studies including nearly 3.5 million patient records. They show that ColonFlag may provide an additional means for identifying people at risk of colorectal cancer, alongside standard screening and diagnostic tests.

Key uncertainties around the evidence or technology are that there is no published evidence detailing the resource consequences of, or patient outcomes from, using ColonFlag.

The cost of ColonFlag (excluding VAT) varies depending on the size of the population served per installation. For example, the annual estimated per capita cost in year 1 of an installation covering a population of 300,000 is £1.30. The resource impact would be an initial increased cost to the NHS including the time spent managing the system.