Introduction Automated quantitative faecal immunochemical tests (FIT) for haemoglobin (Hb) include a faecal sampling device (with preservative buffer) and analytical system. In colorectal cancer screening programmes faecal samples are usually collected at home and returned to the laboratory for testing. The manufacturers state a minimum of one hour incubation before analysis, to allow time for Hb release from the sample into the buffer. We investigated the Hb concentration from one to 48 hours using Hb-positive faeces using three analysers.

Methods Hb-positive faecal samples were taken from stools provided for calprotectin analysis, mixed for two minutes using wooden mixing sticks to distribute the Hb and loaded on to the sampling devices according to the manufacturers’ instructions, with wiped tips. Seven samples were set up in duplicate for the Extel Hemo Auto MC Collection Picker (Kyowa Medex Co. Ltd), FOB Gold (Sentinel Diagnostics), and OC-Sensor Autosampling Bottle 3 (Eiken Chemical Co.) and an additional four samples were set up on the latter two analysers as the first was unavailable. Samples were incubated at room temperature (15–23^oC) and the Hb was measured at 1, 2, 3, 24, 48 hours using the HM-JACKarc, SentiFIT 270 and OC-Sensor PLEDIA, respectively, ensuring the samples were first mixed and allowed to stand for 15 minutes.

Results The percentage difference of the mean of the duplicate results from 1, 2, 3, 24 and 48 hours was calculated.

Conclusion Changes in Hb concentration over 48 hours showed variation for all three FIT systems possibly due to the difference in nature of the sample. Overall the Hb concentration increased to 24 hours then decreased using the SENTIFIT and PLEDIA by which time the faeces appeared to be distributed in the buffer compared with the HM JACKarc for which faeces remained on the sample stick for some samples at 48 hours. The average and minimum percentage differences for the SentiFIT 270 were lower than the PLEDIA with the highest values for HM JACKarc possibly due detection of different Hb degradation products over time. Further work is needed to assess a bigger sample group and whether samples prepared sooner after collection would show the same differences.

Disclosure of Interest None Declared