Assessing empirical thresholds for investigation in people referred on a symptomatic colorectal cancer pathway: a cohort study utilising faecal immunochemical and blood tests in England

Abstract

AbstractObjectiveTo quantify risk of colorectal cancer (CRC) at empirical FIT cut offs, across age, haemoglobin and platelet strata in current diagnostic pathways.DesignCohort study of all people who were referred on a symptomatic CRC diagnosis pathway from primary care with a FIT test in Nottingham, UK between November 2017 and 2021 with 1-year follow-up for cancer and death. Heat maps showed the cumulative 1-year CRC risk using Kaplan-Meier estimates. We estimated the number of investigations that could potentially be re-purposed if a threshold of ≥3% 1-year risk of CRC was instigated.ResultsDuring the study period 514 (1.5%) colorectal cancers were diagnosed following 33694 index FIT tests with available blood tests. Individuals with a FIT ≥10 μg Hb/g faeces had a greater than 3% risk of CRC, except patients under the age of 40 years (CRC risk 1.45% (95% CI 0.03-2.86%)). Non-anaemic patients with a FIT <100 μg Hb/g faeces had a CRC risk of less than 3%, except those between the age of 70-85 years (5.26% 95% CI 2.72-7.73%). Using a ≥3% CRC threshold in patients < 55 years calculated using FIT, age and anaemia would allow 160-220 colonoscopies per 10000 FIT tests to be used for other pathways, at the cost of missing 1-2 CRCs.ConclusionsCRC risk varies by FIT, age and anaemia status when fHb levels are below 100 μg Hb/g faeces. Tailored cut offs for investigation on a CRC pathway could reduce the number of investigations needed at a 3% CRC risk threshold.What is already known on this topicThe National Institute for Health and Care Excellence (NICE), the joint Association of Coloproctology of Great Britain & Ireland and the British Society of Gastroenterology guidelines and the Welsh Government recommend a FIT cut off of ≥10 μg Hb/g faeces for investigation of CRC on an urgent pathway based on an expected colorectal cancer risk threshold of 3%. However, empirical evidence of this threshold in practice and the impact of age, anaemia and thrombocytosis upon it is lacking.What this study addsPeople who had a FIT test in primary care in Nottingham between 2017 and 2021 had a 1- year risk of colorectal cancer of just 1.5%.Non-anaemic patients over 70 years old do not meet the 3% threshold set by NICE for urgent investigation until they have a FIT greater than 40 μg Hb/g faeces.Patients under 40 years of age only meet the 3% threshold for investigation when they have a FIT ≥100 μg Hb/g faeces and are anaemic.How this study might affect research, practice or policyWe estimate that by using a stratified approach to meet the ≥3% risk of CRC threshold that includes FIT, age and anaemia rather than a single cut off for FIT of ≥10 μg Hb/g faeces will allow the optimum number of cancers to be diagnosed for the investigations undertaken.This study assessed existing empirical categorisations of FIT, age and anaemia. Ideally, further optimisation and validation of pathways could be achieved by deriving cut offs and strata using continuous modelling of FIT, age and blood test results.The balance of investigations required, cancers diagnosed and missed is crucial to consider when attempting to optimise diagnostic accuracy and health service provision in the real world. Consensus among all stakeholders needs to be reached on the threshold (risk of CRC) at which investigation should be triggered, taking all these factors into account.