Appraisal of the faecal haemoglobin, age and sex test (FAST) score in assessment of patients with lower bowel symptoms: an observational study.

ABSTRACT

BACKGROUND: Many patients present in primary care with lower bowel symptoms, but significant bowel disease (SBD), comprising colorectal cancer (CRC), advanced adenoma (AA), or inflammatory bowel disease (IBD), is uncommon. Quantitative faecal immunochemical tests for haemoglobin (FIT), which examine faecal haemoglobin concentrations (f-Hb), assist in deciding who would benefit from colonoscopy. Incorporation of additional variables in an individual risk-score might improve this approach. We investigated if the published f-Hb, age and sex test score (FAST score) added value. METHODS: Data from the first year of routine use of FIT in primary care in one NHS Board in Scotland were examined f-Hb was estimated using one HM-JACKarc FIT system (Kyowa Medex Co., Ltd., Tokyo, Japan) with a cut-off for positivity ≥10 µg Hb/g faeces. 5660 specimens were received for analysis in the first year. 4072 patients were referred to secondary care 2881 (70.6%) of these had returned a FIT specimen. Of those referred, 1447 had colonoscopy data as well as the f-Hb result (group A) 2521 patients, also with f-Hb, were not immediately referred (group B). The FAST score was assessed in both groups. RESULTS: 1196 (41.7%) of patients who returned a specimen for FIT analysis had f-Hb ≥10 µg Hb/g faeces. In group A, 252 of 296 (85.1%) with SBD had f-Hb > 10 µg Hb/g faeces, as did 528 of 1151 (45.8%) without SBD. Using a FAST score > 2.12, which gives high clinical sensitivity for CRC, only 1143 would have been referred for colonoscopy (21.0% reduction in demand) 286 of 296 (96.6%) with SBD had a positive FAST score, as did 857 of 1151 (74.5%) without SBD. However, one CRC, five AA and four IBD would have been missed. In group B, although 95.2% had f-Hb < 10 µg Hb/g faeces, 1371 (53.7%) had FAST score ≥ 2.12 clinical rationale led to only 122 of group B completing subsequent bowel investigations a FAST score > 2.12 was found in 13 of 15 (86.7%) with SBD. CONCLUSIONS: The performance characteristics of the FAST score did not seem to enhance the utility of f-Hb alone. Locally-derived formulae might confer desired benefits.