Faecal immunochemical tests for patients with symptoms suggestive of colorectal cancer: An updated systematic review and multiple‐threshold meta‐analysis of diagnostic test accuracy studies

Résumé fourni par la source

AIM: Extending faecal immunochemical tests for haemoglobin (FIT) to all primary care patients with symptoms suggestive of colorectal cancer (CRC) could identify people who are likely to benefit from colonoscopy and facilitate earlier treatment. The aim of this work was to investigate the diagnostic accuracy of FIT across different analysers at different thresholds, as a single test or in duplicate (dual FIT). METHOD: This systematic review and meta-analysis searched 10 sources (December 2022). Diagnostic accuracy studies of HM-JACKarc, OC-Sensor, FOB Gold, QuikRead go, NS-Prime and four Immunodiagnostik (IDK) tests in primary care patients were included. Risk of bias was assessed (QUADAS-2). Statistical syntheses produced summary estimates of sensitivity and specificity at any chosen threshold for CRC, inflammatory bowel disease and advanced adenomas separately. Sensitivity analyses investigated reference standard and population type (high, low or all-risk). Subgroup analyses investigated patient characteristics (e.g. anaemia, age, sex, ethnicity). RESULTS: Thirty-seven studies were included. At a threshold of 10 μg/g, pooled results for sensitivity and specificity (95% credible intervals) for CRC, respectively, were: HM-JACKarc (n = 16 studies) 89.5% (84.6%-93.4%) and 82.8% (75.2%-89.6%); OC-Sensor (n = 11 studies) 89.8% (85.9%-93.3%) and 77.6% (64.3%-88.6%); FOB Gold (n = 3 studies), 87.0% (67.3%-98.3%) and 88.4% (81.7%-94.2%). There were limited or no data on the other tests, dual FIT and relating to patient characteristics. CONCLUSION: Test sensitivity at a threshold of 10 μg/g highlights a requirement for adequate safeguards in test-negative patients with ongoing symptoms. Further research is needed into the impact of patient characteristics and dual FIT.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.