Background Unintentional weight-loss is a common clinical encounter. Frequently, these patients are referred for colonoscopy to rule out lower gastrointestinal pathology.
Aim Determine the diagnostic yield in colonoscopies performed for unintentional weight-loss (WL-O) versus weight-loss and associated GI symptoms (WL-GIs).
Method Retrospective analysis of colonoscopies performed in our centre (May 2013-July 2016). Data was obtained from the Endoscopic Reporting System. Baseline characteristics were established. For sub-analysis, we stratified our cohort into four age groups (<30 years, 30-54years, 55-74years and >75 years).
Results Of 5290 colonoscopies performed, 240 met our inclusion criteria (WL-O n=83, WL-GIs n=157). Baseline demographics were similar in both groups (WL-O Male 54%, mean age 56 years (SD ±16.6) compared with WL-GIs Male 52%, mean age 58 years (SD ±17.2)). Caecal intubation rates were similar in both groups (88% WL-O, 86% WL-GIs). Overall, colonoscopy was normal in 37.9% (64/169). Diverticulosis was detected in 24.9% (42/169), Colitis in 3% (5/169) and adenoma/polyp in 35.4% (85/240). For diagnostic yield between WL-O and WL-GIs, advanced adenoma detection rate was 7.2% (n=6) versus 8.3% (n=13) and CRC detection rate was 2.4% (n=2) versus 2.5% (n=4). CRC plus advanced adenomas was 9.6% (n=8) versus 10.8% (n=17)(p=0.774). Notably, in the WL-O group no high risk pathology was detected in colonoscopies in patients>75 years compared with n=3 in the WL-GIs group.
Conclusions For both groups diagnostic yield was low, comparable to the asymptomatic generapopulation. Diagnostic yield is extremely low in colonoscopies for weight-loss only in patients>75 years. Weight-loss may not be a valid indication for colonoscopy, particularly in >75 years.