Laszlo, HE;
Seward, E;
Ayling, R;
Lake, J;
Malhi, A;
Hackshaw, A;
Stephens, C;
... Machesney, M; + view all
(2020)
Quantitative faecal immunochemical test for patients with 'high risk' bowel symptoms: a prospective cohort study.
MedRxiv: Cold Spring Harbor, NY, USA.
Preview |
Text
Pritchard-Jones_2020.05.10.20096941.full.pdf Download (1MB) | Preview |
Abstract
Objectives: To evaluate whether quantitative measurement of faecal haemoglobin (f-Hb) using faecal immunochemical testing (FIT) can be used to rule out colorectal cancer (CRC) for patients who present to primary care with ‘high risk’ symptoms defined by national guidelines for urgent referral for suspected cancer (NICE NG12). / Design: Prospective cohort study carried out between April 2017 and March 2019. / Setting: 59 GP practices in London and 24 hospitals in England. / Participants: Symptomatic patients in England referred to the urgent CRC pathway who provided a faecal sample for FIT in addition to standard investigations for cancer. / Main outcome measures: CRC was confirmed by established clinical and histopathology procedures. f-Hb (μg per gram of stool) was measured in a central laboratory blinded to cancer outcome. We calculated sensitivity (percentage of patients with CRC who have f-Hb exceeding specified cut-offs); false-positive rate [FPR] (percentage of patients without CRC whose f-Hb exceeds the same cut-offs); and positive predictive value [PPV] (percentage of all patients with f-Hb above the cut-offs who have CRC). / Results: 4676 patients were recruited of whom 3596 patients were included (had a valid FIT test and a known definitive diagnosis). Among the 3596, median age was 67 years, 53% were female and 78% had colonoscopy. 90 patients were diagnosed with CRC, 7 with other cancers, and 3499 with no cancer found. f-Hb did not correlate with age, sex or ethnicity. Using f-Hb ≥4μg/g (lowest limit of detection), sensitivity, FPR and PPV were 87.8%, 27.0% and 7.7% respectively. Using f-Hb ≥10μg/g, the corresponding measures were 83.3%, 19.9% and 9.7%. 15 patients with CRC had f-Hb below 10μg/g. If FIT had been used at thresholds of 10μg/g or 4μg/g, 1 in 6 or 1 in 8 patients with cancer respectively would have been missed. If the absence of anaemia or abdominal pain is used alongside f-Hb 10 μg/g, only 1 in 18 cancers would be missed but 56% of people without CRC could potentially avoid further investigations including colonoscopies. / Conclusions: In our study, if FIT alone had been used to determine urgent referral for patients with ‘high risk’ symptoms for definitive cancer investigation, some patients with bowel cancer would not have been diagnosed. If used in conjunction with clinical features, particularly in the absence of anaemia, the efficacy of FIT is significantly improved. With appropriate safety netting, FIT could be used to focus secondary care diagnostic capacity on patients most at risk of CRC.
Archive Staff Only
 |
View Item |
