Efficacy and accuracy of faecal sampling by a digital rectal examination for faecal immunochemical testing

Abstract

Aim A digital rectal examination (DRE) during routine assessment for patients with abdominal symptoms provides an opportunity to obtain faeces from the glove for faecal immunochemical testing (FIT). Here, we compared sampling via DRE to the standard faecal sampling by patients. Method Patients were recruited to a prospective observational cohort study between July 2019 and March 2020. Patients provided a sample for the FOB Gold Wide® which was compared to a further sample taken at clinic via DRE. Clinicians reported whether they obtained a ‘good’ sample filling all the grooves, a ‘poor’ sample filling some of the grooves or no faecal sample. Cohen’s kappa was used to compare percentage agreement around a negative threshold of <10 μg haemoglobin/g of faeces. Sensitivity for serious bowel disease (SBD) was calculated. Results Of 596 patients who underwent attempted DRE sampling, there were 258 (43.3%) ‘good’ samples, 117 (19.6%) ‘poor’ samples and 221 (37.1%) with no sample to wipe in the grooves. Cohen’s kappa dropped from 0.70 to 0.30 for the ‘good’ and ‘poor’ samples, respectively. Of those with DRE samples and definitive diagnostic outcomes, the sensitivity for SBD dropped significantly from 76.0% to 41.7% between ‘good’ and ‘poor’ samples, respectively ( p = 0.041). Conclusions A ‘good’ sample obtained by DRE provides comparable results to samples obtained by patients. This creates potential benefit in speed and ease of testing for patients. However, not all DRE sampling attempts are successful, and the clinician must be satisfied that enough faeces is obtained to wipe adequately into all grooves.