A comparison of the faecal haemoglobin concentrations and diagnostic accuracy in patients suspected with colorectal cancer and serious bowel disease as reported on four different faecal immunochemical test systems

Objectives Faecal immunochemical tests for haemoglobin (FIT) are used in colorectal cancer (CRC) screening programmes and to triage patients presenting with symptoms suggestive of CRC for further bowel investigations. There are a number of quantitative FIT analytical systems available. Currently, there is no harmonisation or standardisation of FIT methods. The aim of the study was to assess the comparability of numerical faecal haemoglobin concentrations (f-Hb) obtained with four quantitative FIT systems and the diagnostic accuracy at different f-Hb thresholds. Methods A subgroup of the National Institute for Health and Care Excellence (NICE) FIT study, a multicentre, prospective diagnostic accuracy study were sent four FIT specimen collection devices from four different FIT systems or two FIT devices for one FIT system. Faecal samples were examined and analysis of results carried out to assess difference between methods at thresholds of limit of detection (LoD), 10 mu g haemoglobin/g faeces (mu g/g) and 100 mu g/g. Results 233 patients returned specimen collection devices for examination on four different systems; 189 patients returned two FIT kits for one system. At a threshold of 100 mu g/g the sensitivity is the same for all methods. At lower thresholds of LoD and 10 mu g/g differences were observed between systems in terms of patients who would be referred and diagnostic accuracies. Conclusions The lack of standardisation or harmonisation of FIT means that differences are observed in f-Hb generated on different systems. Further work is required to understand the clinical impact of these differences and to minimise them.

Automated summary

This study assessed the comparability of faecal haemoglobin concentrations obtained with four quantitative FIT systems at different thresholds. The results showed differences between systems at lower thresholds, highlighting the need for further understanding of their clinical impact and efforts to minimize these differences.