Introduction Colonoscopy in the elderly and co-morbid is not without risk. Clinicians may feel that bowel preparation as an inpatient in this patient population may be safer and more effective. However, to the best of our knowledge the rationale, efficacy and complication rate of inpatient colonoscopy has not been documented. We present the results of a pilot study looking at inpatient bowel preparation for facilitating colonoscopy.

Method A retrospective case note review of all patients admitted for inpatient bowel preparation prior to colonoscopy over a 12 month period at one district general hospital was performed. Additional information was extracted from established databases (Unisoft, Track Care and Endoscopy Suite registers). Data collected included patients’ demographics, past medical history, reason for inpatient preparation, type of ward the preparation was carried out in and type of bowel preparation used.

Results Between January and December 2014, 63 patients (median age - 72 years) underwent 73 attempts at inpatient colonoscopy. (During the same time period, 2,458 outpatient colonoscopies were performed: inpatients = 2.5% of total). 47 patients underwent colonoscopy following emergency admission. Indications included anaemia (44%), PR bleeding (19%) and altered bowel habit (15%). Overall, 34% (n = 21) of inpatient colonoscopies were incomplete. Failure was attributed to inadequate bowel preparation in 27% (n = 17). Bowel preparation failed in 18% of females and 8% of males. Where bowel preparation was considered inadequate, the referring ward was medical (32%) and surgical (11%). Additionally, where bowel preparation failed, Moviprep had been prescribed in 18%, Kleenprep in 7%, Picolax in 5% and Fleet in 4%. Only 5% had a successful inpatient colonoscopy when previous outpatient preparation had failed. Only 3 of 10 patients having a second attempt at inpatient colonoscopy were successful. Of the original 63 patients who underwent this inpatient examination, only 6 were diagnosed with cancer (one patient went on to elective surgery). The “all cause” 30-day mortality following inpatient colonoscopy was 6.3%.

Conclusion Although this is a small cohort, the initial results are worrying. Inpatient colonoscopy often fails and the number diagnosed with significant pathology which is amenable to treatment is small. There is also a very real 30-day mortality rate which may have been contributed to by this invasive procedure. While our initial findings may be peculiar to our own unit where we have high levels of co-morbidity, this pilot suggests that further work is required nationally to clarify the indications/risks/results in this small but problematic patient group.

Disclosure of interest None Declared.