Introduction Colonoscopy is often performed in patients undergoing investigation for unexplained diarrhoea. Obtaining colonoscopic biopsies for persistent diarrhoea is an auditable JAG standard. The aims of this audit were (1) To determine the diagnostic yield of colonoscopy in patients undergoing investigation for diarrhoea. (2) To determine the rate at which biopsies are undertaken in patients with a “normal” colonoscopy. (3) To assess for variations in biopsy sampling amongst endoscopists.
Methods An analysis was performed of all colonoscopies with the indication of diarrhoea, undertaken in 2010. Interrogation of the electronic endoscopy reporting tool, looked at endoscopist discipline, findings at endoscopy, if biopsies were taken, number of biopsies and biopsy sites, and corresponding histology results.
Results A total of 609 patients were identified in whom the indication for colonoscopy was diarrhoea. The mean age was 57 years (range 14–90 years) with 40.4% male and 59.6% female. Caecal intubation was achieved in 565 patients (92.8%) with terminal ileal intubation recorded in 231/609 patients (37.9%). Overall, biopsies were taken in 545/609 patients (89.5%). The median number of biopsies taken per procedure was 10.5 (range 1 – 22), with a median number of 5.5 from the left side of the colon and 4 from the right side. Colonoscopic appearances were abnormal in 295/609 (48.4%) patients with isolated proximal disease in 36/295 (12.2%). The most common endoscopist abnormality was diverticular disease in 149 patients (24.4% overall), followed by polyps in 115 (18.8%), suspected inflammation in 67 (11%), suspected tumour/cancer in 11 (1.8%), and melanosis coli in 1 patient (0.2%). Of the patients with suspected mucosal inflammation, histology revealed features of IBD in 25 (4.1%) of patients with isolated right-sided inflammation in 5 (0.8%) and terminal ileum alone in 2 (0.32%). Of the 609 colonoscopies, 261 (42.9%) were referred as a 2-week wait urgent suspected cancer referral, yet a diagnosis of cancer was made in only 4 cases (1.53%). Of the 7 cancers detected, 6 (85.7%) were located in the left colon.
Of the 314 «normal» colonoscopies, biopsies were taken in 268 (85.4%) patients and histology confirmed microscopic colitis in 15 (4.77%) and mucosal inflammation in 27 (8.6%). There was variation in the frequency and number of biopsy specimens obtained: GI physicians 91.59% (median number 10.5), GI surgeons 84.39% (median number 8.5), Nurse/GP Endoscopist 92.45% (median number 8) and non-GI physicians/surgeons 92.8% (median number 5).
Conclusion Although abnormal findings are not uncommon in patients undergoing colonoscopy for symptoms of diarrhoea, yield for cancer is low. There is variation in practise among endoscopists in obtaining biopsy samples in the setting of diarrhoea and normal colonoscopy
Disclosure of Interest None Declared.