The below was presented as an oral abstract presentation at NCRI Cancer Conference 2019.

Accelerating early identification of cancer in primary care using an artificial intelligent driven solution.

Abstract

Background

By 2020, one in two people in the UK will be diagnosed with cancer at some point in their lifetime. When diagnosed early (in Stages 1 or 2), over 80% of patients will survive for five years or more. Despite this, 46% of patients continue to be diagnosed in the late stages of the disease (Stages 3 or 4), with less than 34% of these patients surviving to five years.

C the Signs is a multi-platform digital tool that uses artificial intelligence (AI) to accelerate early identification and the management of patients at risk of cancer. The tool supports identification of which cancer(s) a patient is at risk of, and what investigation or referral a patient may need, in less than 30 seconds.

Method

C the Signs was made available across three Clinical Commissioning Groups in the East of England between 2017 to 2018, covering an estimated population of 1,054,000 patients. The tumour types, pathways and recommendations were monitored at a user level. Clinicians were surveyed at the end of the study period to determine confidence, acceptability, usability and benefit. National Cancer Registry and Analysis Service (NCRAS) data was used to determine the impact of the tool on cancer detection rates and emergency presentation rates.​

Results

286 healthcare professionals took part in the study, representing 85 GP practices. 2,084 patients were risk assessed during the study period, triggering 5,121 cancer pathways. 19% of pathways recommended were urgent cancer referral pathways (known as 2-week-waits), with the remainder being direct-access cancer investigations. In the study areas, cancer detection rates improved by an average of 6.40% from the previous year, compared to 0.21% in the rest of the East of England and 0.59% in England as a whole (P< 0.05). Emergency presentation rates reduced by 7.09% in the study areas, compared to 5.75% and 4.49% in East of England and England respectively.

Conclusion

C the Signs is the first AI-driven tool that has had a statistically significant impact on cancer detection rates - through early identification and referral of patients at risk of cancer in primary care.