Article Abstract

Interpretation of faecal haemoglobin concentration data in colorectal cancer screening and in assessment of symptomatic patients

Authors: Callum G. Fraser


Faecal immunochemical tests for haemoglobin (FIT) are used in asymptomatic colorectal cancer screening and in assessment of patients presenting in primary care with lower gastrointestinal symptoms. Many qualitative and quantitative FIT are available. Qualitative FIT give dichotomous results and the most important characteristic for interpretation of test results is the C50, often described as the cut-off. Qualitative FIT have different C50 and thus give different clinical outcomes in screening. These are often stated to be simple, but both their performance and interpretation of results are fraught with difficulties. Quantitative FIT have many advantages. In screening, the faecal haemoglobin concentration (f-Hb) used as cut-off can be decided for each endeavour. As the f-Hb cut-off is increased, positivity rate, neoplasia detection rate and sensitivity decrease, while positive predictive value, specificity and interval cancer proportion increase. Different FIT give different outcomes and comparisons are best done using a consistent positivity rate. Many factors affect f-Hb and, although controversial, it has been suggested that sex and age stratified f-Hb cut-offs would have advantages. In consequence, current attention is becoming more focused on risk-scoring approaches incorporating f-Hb and on application of precision medicine strategies. FIT must have a low f-Hb in assessment of the symptomatic. The detection capability of FIT requires harmonisation with regard to terminology, determination and application in both academic research and routine clinical practice. The limit of quantitation is the lowest f-Hb that can be determined when some predefined analytical performance specifications (APS) for the characteristics of bias and imprecision (or total error) are satisfied. Proposals for interim APS are documented. The reporting of numerical data on f-Hb requires further consideration and creation and promulgation of consensus guidelines by international professional bodies.