Early flagging of colorectal cancer
A smart computer algorithm that interrogates patient data to identify patients at high risk of colorectal cancer has been reviewed by NICE (https://www.nice.org.uk/advice/mib142/chapter/Summary).
The ColonFlag algorithm uses existing datasets containing age, sex and blood test results to generate a risk score for a person. People with a high-risk score can be referred for further assessment, potentially before they show any symptoms.
ColonFlag works best when multiple blood tests, taken at different times, are available. The algorithm uses decision trees and gradient boosting (which increases the prediction power of decision trees) to assess trends in the blood test parameters and patient characteristics.
ColonFlag software would be installed on the central computer system within a healthcare organisation or network. If the score is above a defined threshold, the software will show that a person is at high risk of having colorectal cancer.
The NICE Medtech innovation briefing reviewed evidence from four observational studies including nearly 3.5 million patient records. According to NICE, ColonFlag may provide an additional means for identifying people at risk of colorectal cancer, alongside standard screening and diagnostic tests.