Overview of the COLOFIT study

Objective There are 44 000 new cases of colorectal cancer (CRC) and 16 000 deaths in the UK annually. Most CRCs are diagnosed following symptom development; however, symptoms are poorly predictive of CRC. Around 700,000 patients are referred on lower GI pathways in England annually, though fewer than 5% have CRC, placing a huge demand on diagnostic services. Faecal immunochemical testing (FIT) has been introduced to assist in the referral pathway of patients; however, this ignores other epidemiological risk factors. Methods We sought to establish whether additional variables could refine a FIT-based approach to lower GI pathways through undertaking a systematic review; creating the COLOFIT algorithm from a large UK dataset incorporating several additional variables; testing the algorithm; examining health-economic aspects of using the COLOFIT algorithm; and studying patient and practitioner experiences of FIT pathways. Results COLOFIT demonstrated a lack of well-validated models involving FIT. COLOFIT used a rigorous process to derive and validate a model including FIT and other parameters. Using the model would reduce the number of colonoscopies required with a very small reduction in CRCs diagnosed. The model was validated in internal and external populations. Those populations need to have similar FIT positivity and cancer detection rates to those in the study populations. Health economic analysis demonstrated COLOFIT to be cost effective. Health professionals and patients were broadly acceptant of FIT. COLOFIT may disadvantage some younger patients. Conclusions The COLOFIT algorithm can be used to improve the referral of patients with suspected CRC and is cost-effective.