EGS P12 An audit of local management of upper gastrointestinal bleeds in a UK district general hospital

Abstract

Abstract Background Acute upper gastrointestinal bleed (AUGIB) is associated with a 10% in-patient mortality and this outcome has not significantly improved over the past 50 years (1). The UK's National Institute for Health and Care Excellence (NICE) has developed guidance for the management of patients presenting with AUGIB, which highlights several priorities that should be addressed such as: a) risk assessment using the Blatchford and Rockall scores, b) the optimal timing of endoscopy, c) management of non-variceal bleeds (NVB's), d) management of variceal bleeds (VB's), and e) guidance on controlling bleeding and preventing re-bleeding in patients on non-steroidal anti-inflammatory drugs, aspirin or clopidogrel (1). Methods This study aimed to assess the compliance of AUGIB management in a district general hospital against the NICE quality standards. In the first cycle, we retrospectively reviewed case notes from 52 patients who presented with AUGIB and were admitted between September 1 and December 31, 2019. A re-audit was performed between December 1, 2020 to March 31, 2021 where we reviewed the case notes of 22 patients. We allowed for a 1-year interval between the first and second cycle where we educated doctors about the NICE quality standards of AUGIB management. The primary outcome of the study was the compliance of local management of AUGIB in a district hospital compared to NICE quality standards. Results Intervention indications in the first cycle were: haematemesis in 26.9% (14/52); melaena in 50% (26/52); both haematemesis and melaena in 23.1% (12/52). Indications in the second cycle were haematemesis in 64% (14/22); melaena in 18% (4/22); both haematemesis and melaena in 8% (4/22). Compliance with discontinuing aspirin during AUGIB was 60% (3/5) in the first cycle, and 80% (4/5) in the second. Blatchford risk score was used in 12.5% (5/40) in the first cycle, and 50% (11/22) in the second. Among haemodynamically unstable patients, endoscopy was performed immediately after resuscitation in 100% in both the first (4/4) and second (3/3) cycles. Among haemodynamically stable patients, 91.4% (32/35) had endoscopy <24 hours in the first cycle, and 53% (10/19) in the second. For NVB's, proton pump inhibitors were used in 74.4% (29/39) in the first cycle and 77% (17/22) in the second. For NVB endoscopy, there was 100% compliance in using clips with/without adrenaline in both cycles. Dual therapy with thermal coagulation and adrenaline was performed in 100% (1/1) of patients in the first cycle, and none in the second. For VB's, 100% of patients in the first (4/4) and second (5/5) cycle were given terlipressin, antibiotics, and band ligation. Conclusions In our district hospital, pre-endoscopic management of AUGIB such as PPI use, and time to endoscopy were adequate, but did not fully meet the quality standards stipulated by NICE. Endoscopic management for AUGIB, however, was fully compliant with NICE standards.