A universal threshold of 10 μg/g for faecal haemoglobin (f-Hb) in isolation does not meet the aims of NICE guidance in the urgent suspected colorectal cancer pathway.

Abstract

The positive predictive value (PPV) of a test at a particular threshold depends on the prevalence of disease in the population tested. Because the prevalence of colorectal cancer rises with age and iron deficiency anaemia (IDA), applying a universal faecal haemoglobin (f-Hb) threshold of 10 μg/g would yield variable PPV across the population being referred on the urgent suspected cancer (USC) pathway. We used the number needed to scope (NNS) and the cancer conversion rate (CCR) as indicators of the PPV of f-Hb to examine this hypothesis. Between 2021/22 and 2024/25, 192,456 patients had an f-Hb result recorded in General Practice. 84,423 USC referrals were recorded; 41,952 were linked to f-Hb results. f-Hb levels were grouped into bands and analysed against cancer diagnoses, colonoscopy activity, demographics, iron deficiency anaemia (IDA), CCR and NNS to detect one cancer. CCR and NNS varied markedly with age and f-Hb. In patients <40 years, NNS was 68 at f-Hb ≥ 10 μg/g and 40 at ≥100 μg/g, with CCR <3% at all thresholds. In those 40-49 years, NNS was 17 at ≥10 μg/g and 10 ≥ 60 μg/g; CCR remained <3% below 100 μg/g. In 50-69 years, NNS was 16 at ≥10 μg/g, 10 at ≥40 μg/g; CCR exceeded 3% from ≥40 μg/g. In ≥70 years, NNS was 7 at ≥10 μg/g, 6 at ≥20 μg/g; CCR >3% from ≥20 μg/g. f-Hb testing increased disproportionately in patients <50 years. CCR increased with age, male sex and IDA. Applying a single f-Hb ≥ 10 μg/g threshold at all ages has not optimised the USC pathway. Age and risk-adjusted f-Hb thresholds are needed to maintain cancer detection while restoring efficiency.