Introduction Greater than 20% of all colonoscopies are associated with suboptimal bowel preparation,¹ resulting in lower caecal intubation rates, longer procedure times, decreased adenoma detection rates and higher costs.² Inadequate bowel cleansing is more common in hospitalized patients.³

Aims/Background To critically analyse the utilisation of inpatient colonoscopy in a tertiary care university hospital.

Method We performed a retrospective review of our electronic database for inpatient colonoscopies over a two month period from September to October 2012. Patient demographics, clinical details and key performance indicators were reviewed. Univariate and bivariate statistical analyses were performed using Mann Whitney and Fisher's exact tests, where appropriate.

Results Fifty six inpatient colonoscopies were performed during the study period. Twenty colonoscopies (36%) were normal and polyps were found in 15 (27%). There were 2 cancers (4%) and one high grade adenoma (2%). Fourteen (25%) colonoscopies were incomplete. Bowel preparation was suboptimal in 31 (56%) of cases. The reason for admission to hospital related primarily to a GI complaint in 23 (41%) of cases. These patients had significantly shorter median overall lengths of stay (10 versus 18 days, p=0.007) and time to colonoscopy (4 versus 10 days, p<0.0001) when compared with patients admitted for non-GI reasons. No difference was seen in rates of pathology encountered in these patients but there was a trend towards better cancer detection rates (3/23, 13% versus 0/33, 0%; p=0.06).

Conclusion We observed very high rates of inadequate bowel preparation and relatively low caecal intubation rates for inpatient colonoscopy. Cancer detection rates may be higher in patients admitted with GI complaints, although this needs to be confirmed by expansion of the cohort. The data suggest inpatient colonoscopy should be used sparingly and perhaps reserved for patients with primary GI complaints.