The diagnostic accuracy of the faecal immunohistochemical test in identifying significant bowel disease in a symptomatic population.

Abstract

Provisional research on the faecal immunohistochemical test (FIT) for symptomatic colorectal patients has shown a high negative predictive value but has lacked long-term patient follow-up, raising the possibility of missed diagnoses of colorectal cancer (CRC). The aim of this work is to describe the long-term diagnostic accuracy of the FIT for CRC and significant bowel disease (SBD) in a symptomatic population in NHS Lanarkshire. From October 2016 to February 2019, all primary care referrals of symptomatic colorectal patients in NHS Lanarkshire were asked to provide a FIT. The baseline demographics, investigations and diagnoses for each patient were prospectively completed until February 2021. A FIT result of ≥10 μg haemoglobin (Hb)/g faeces was considered to be positive. A total of 5250 patients were identified (median age 62 years; 46% male; median follow-up 31 months) with 65.1% (3418) being FIT negative. The SBD rate was 6.2% and the CRC rate was 2.9% (151). The SBD rate was significantly higher in the FIT-positive group (13.8% vs. 2.2%; p < 0.001) and 32.9% of patients with FIT ≥ 400 μg Hb/g had SBD. The sensitivity of FIT ≥ 10 μg Hb/g for CRC was 87.4% and for SBD it was 76.9%. Specificity was 66.6% and 66.7%, and the negative-predictive value was 99.4% and 97.7%, respectively. Sensitivity for CRC could theoretically be increased to 94.8% if FIT-negative patients were to undergo flexible sigmoidoscopy. A FIT-only referral pathway for symptomatic colorectal patients will miss over 12% of cancers and over 23% of SBD. Theoretically, combining FIT-negative patients with flexible sigmoidoscopy increases the sensitivity for CRC. The FIT offers a mechanism for prioritizing patient access to investigations, particularly in resource-limited areas; however, further work to identify FIT-negative patients diagnosed with CRC is required.