An analysis of the utility of quantitative faecal immunochemical tests in screening and symptomatic populations

  • Paula Jane McDonald

    Student thesis: Doctoral ThesisDoctor of Philosophy


    Background: It has been demonstrated, in 4 large randomised control trials (RCT), that screening for colorectal cancer (CRC) using annual or biennial guaiac faecal occult blood tests (gFOBT) reduces mortality and incidence. The faecal immunochemical test (FIT) uses technology that is analytically more sensitive and specific for human haemoglobin (Hb) than gFOBT.

    Methods: An evaluation of the OC-Sensor Diana quantitative FIT analyser and prospective analysis of a single estimate of faecal haemoglobin concentration (f-Hb) in two clinical settings; the Scottish Bowel Screening Programme and patients referred from primary care to endoscopy services.

    Results: Uptake, in the cohort offered screening with FIT as a first-line test, was 4.8% higher than that seen contemporaneously in the Scottish Bowel Screening Programme. This returned to pre study levels when the study ceased and gFOBT was reintroduced. The cohort offered quantitative FIT had a positivity of 2.4% compared to 2.1% in the programme overall. Clinical outcomes, during the evaluation period, in the study cohort and the screening programme were similar.

    40,125 participants returned a FIT sample device and 38,720 had their f-Hb measured. An observational study of f-Hb by sex and age, using the 97.5th percentile as a potential upper reference limit, and 90% confidence intervals (CI) showed 519 ng Hb/ml buffer (90% CI: 468 – 575) for men and 283 ng Hb/ml buffer (90% CI: 257 – 316) for women. When the data was partitioned by age quintile, f-Hb increased with age in both sexes.

    Quantitative FIT and endoscopy were completed by 280 patients referred from primary care for endoscopy (median age: 63 years, range: 18 to 84 years), 59.6% were female. Six (2.1%) participants had CRC, 23 (8.2%) high-risk adenoma (HRA: >3 adenomas or any > 1 cm), 31 (11.1%) low-risk adenoma (LRA), and 26 (9.3%) inflammatory bowel disease (IBD) as the most serious diagnosis. Those with CRC had median f-Hb of > 1000 ng Hb/ml buffer. Using f-Hb with a cut-off of 50 ng Hb/ml buffer, negative predictive values of 100%, 94.4%, 93.4% and 93.9% were found for CRC, HRA, LRA and IBD.

    Conclusions: The introduction of quantitative FIT into population screening and symptomatic settings has the potential to optimise referral for endoscopy for those who have evidence of small amounts of bleeding, thereby improving outcomes and reducing the current burden on endoscopy services.
    Date of Award2016
    Original languageEnglish
    SupervisorBob Steele (Supervisor) & William Bartlett (Supervisor)


    • Faecal immunochemical test

    Cite this