G G Robins, M S Sarwar, M Armstrong, M E Denyer, S Bush, T Hassan, S M Everett

Evaluation of the need for endoscopy to identify low-risk patients presenting with an acute upper gastrointestinal bleed suitable for early discharge
Postgrad Med J 2007; 83: 768-772 [Abstract] [Full text] [PDF]

j.schrieber{at}doctors.org.uk Joe Schrieber, et al.

We read with interest the paper entitled “Evaluation of the need for endoscopy to identify low-risk patients presenting with an acute upper GI bleed suitable for early discharge” Postgraduate Medicine Journal December 2007; 83: 768-722. The aim of this paper was to audit the safety of two similar protocols used to manage low-risk upper GI bleeding within one trust in Leeds. Both protocols advised admission to a clinical decisions unit (CDU) whilst one advocated endoscopy during the admission compared to early discharge and outpatient endoscopy/follow up in the other group. The rationale for the audit is seemingly sound due to the high incidence of Upper GI bleed and the use of two slightly different protocols within the same primary care trust. However, the primary objective of the study was to assess safety of the two policies - although safety was not actually assessed or discussed. Although the authors did imply that very few patients actually required intervention at endoscopy and relevant morbidity and mortality was zero, this end point was not proved. The study only used small numbers of patients and although clear exclusion criteria were set out, the authors failed to produce data regarding all patients with the numbers that were included or excluded and the outcomes for each group and still included data from the >60age group in the results, without explanation. Similarly there is no mention of the failure to adhere to the afore mentioned protocol in ‘Hospital A’ where all patients should undergo Oesophagogastroduodenoscopy (OGD) before discharge. Only 74% of this cohort did in fact have an OGD and no exclusion criteria were discussed with reference to the remaining 26%. We also feel the selection of such an Ultra-low risk group aids the authors in demonstrating the fact that immediate OGD’s are unnecessary, and feel that a cost analysis for each protocol was necessary.