Introduction Despite emergency readmission being a NHS quality indicator and unplanned admission after elective endoscopy a GRS auditable outcome; there is a lack of audit standards and difficulties in attaining these data due to limitations in hospital IT systems. Commercial analytical reporting tools that use NHS administrative data can generate local reports from HES and may serve as a starting point for clinical audit.

Aim To determine the prevalence, demographics, aetiology and outcomes of emergency readmissions following elective endoscopy.

Methods From 1.4.08 to 31.3.09 emergency admissions within 6 days of a day-case upper or lower GI endoscopy (ERCP excluded) were identified from analyses of NHS administrative data at RLUH, AUH, SRH, and SMUH. Using a standard proforma, each readmission was audited using case notes, endoscopy database, and the hospital informatics systems. Two senior doctors determined the validity of the admission subdividing them into; unrelated, possible and definite with the latter two categories defining an endoscopy readmission.

Results Of the 29 868 daycases, 235 admissions were identified, of which 147 were endoscopy readmissions. The following data all relates to endoscopy readmissions.

Demographics Male:Female ratio=1:1; Mean age (SD): 65.1 years (17). Readmission rate overall was 0.49% (0.56% for upper and 0.42% for lower GI procedures) with highest rates for therapeutic uppers (1.75%) > therapeutic lowers (0.56%) > diagnostic upper (0.47%) > diagnostic lower (0.38%). Aetiology The most common reasons for admission were vascular events (26.5%) > Symptoms such as pain (17%), GI haemorrhage (17%), suspected/confirmed perforation (17% (confirmed 6.1%))> respiratory (10.2%) causes >bowel obstruction (4.8%): Investigations 71% required simple blood tests and/or x-rays and 26%, cross-sectional imaging. Treatment 84.4% required only observation, 4.8%; surgery, 2.7%; repeat endoscopy, and 7.5% IV antibiotics. Outcomes 24.5% were short-stays discharges from either AED, or medical/Surgical assessment units. Mean LOS was 8.2 (13) days and 30-day mortality was 6.8%.

Conclusion Commercial analytical reporting tools combined with local interpretation and multicentre collaboration was essential for determining useful information regarding endoscopy readmissions. Readmission rates predictably varied with the procedure type (0.38–1.75%) with therapeutic procedures having the highest. While most patients required simple observation/supportive care, LOS was considerable and mortality rate among readmitted cases was significant.