All-cause and non-colorectal cancer 1-year mortality among people having a faecal immunochemical test in primary care for investigation of symptoms of suspected colorectal cancer: an English cohort study

Background: High blood concentrations in stool, as determined by the faecal immunochemical test (FIT), in individuals screened for colorectal cancer are associated with increased mortality. To our knowledge, no studies to date have assessed whether this association is present in symptomatic individuals. Our objective was to estimate the risk of all-cause and non-colorectal cancer 1-year mortality in adults who have undergone a FIT after symptomatic presentation to primary care.

Methods: In this regional retrospective cohort study, all adults (age ≥18 years) with symptoms of colorectal cancer who were tested with FIT in primary care in Nottingham, UK, between Nov 1, 2017, and Nov 30, 2022 were included. Patient data were retrieved from existing medical records. Associations between FIT and mortality were examined using Cox regression to produce adjusted hazard ratios (aHRs) comparing those with FIT results of 10 μg of haemoglobin per g or higher against those with results lower than 10 μg of haemoglobin per g. Standardised mortality ratios (SMRs) were derived using English data from the UK Office for National Statistics.

Findings: 49 889 patients were included following their first FIT to investigate symptoms of colorectal cancer. The median age was 65 years (IQR 53–79); 27 912 (55·9%) patients were female and 21 977 (44·1%) were male. In the year following a symptomatic FIT, 1971 (4·0%) patients died, including 864 (8·3%) of 10 352 patients with a FIT result of 10 μg of haemoglobin per g or higher and 1107 (2·8%) of 39 537 with a FIT result lower than 10 μg of haemoglobin per g. After adjusting for age, sex, and year, there was evidence of increased all-cause mortality (aHR 1·96 [95% CI 1·79–2·14]) and non-colorectal cancer mortality (1·70 [1·55–1·88]) in those with a FIT result of 10 μg of haemoglobin per g or higher compared with those with a FIT result lower than 10 μg of haemoglobin per g. The overall cohort had higher mortality than the general English population (SMR 1·50 [95% CI 1·44–1·57]). The magnitude of the increase in mortality following a FIT result of 10 μg of haemoglobin per g or higher was greater in younger than in older patients and in female than in male patients compared with those with a FIT result lower than 10 μg of haemoglobin per g (HR for all-cause mortality was 3·83 [2·86–5·14] in female patients and 2·91 [2·21–3·85] in male patients at age 50 years vs 2·27 [1·98–2·60] in female patients and 1·73 [1·53–1·95] in male patients at age 80 years).

Interpretation: Further understanding of the cause-specific mortality associations with symptomatic FIT has the potential to transform diagnostic pathways and subsequent treatment—particularly in younger and female patients, for whom the greatest excess all-cause mortality risks were observed at 1 year of follow-up.

Funding: None.