OBJECTIVES: Patients with symptoms suggestive of colorectal cancer (CRC) presenting to primary care were referred to secondary care if their symptoms were “high risk”. Using fecal immunochemical tests for hemoglobin (FITs) in these patients could reduce referrals and colonoscopy waiting lists, and enable earlier treatment. This systematic review synthesized test accuracy data and estimated impact on referrals and missed diagnoses.

METHODS: We searched ten sources (December 2022). We included FIT diagnostic accuracy studies on nine different FIT analyzers, and studies on the use of the tests in duplicate (Dual FIT), conducted in patients presenting to primary care with symptoms suggestive of CRC. We assessed risk of bias with QUADAS-2. Syntheses of sensitivity and specificity which unified data across thresholds were planned for each test. Summary data and estimates of referrals and missed diagnoses at key thresholds were generated. Sensitivity analyses investigated the population recruited and type of reference standard used.

RESULTS: Across16 HM-JACKarc studies (threshold range 2-400 µg Haemoglobin/g faeces (µg/g)), pooled sensitivity and specificity ranged from 95.9% to 46.3% and from 97.7% to 65.1%. In 11 OC-Sensor studies (range 4-200 µg/g) values respectively were 94.2% to 54.2% and 97.3% to 62.7%. In 3 FOB Gold studies (range, 2-150 µg/g) the values were 91.4% to 73.9% and 96.4% to 78.1%. There were limited or no data on the other analyzers. Dual FIT increased sensitivity but decreased specificity compared to a single test.

CONCLUSIONS: FIT is an effective risk stratification tool. However, safeguards are needed to retest patients with persistent or worsening symptoms because even at the lowest thresholds some CRC cases are missed. Higher thresholds increase miss rates but significantly reduce the need for costly colonoscopies, potentially reducing waiting lists and speeding up diagnoses. Research recommendations include more data on Dual FIT, patient characteristics, and comparative accuracy.