HOME
Return to Table of Contents
173 OUTCOMES OF GI BLEEDS IN AN INNER CITY HOSPITAL NJ Samadder1, M Bragaza2, M Mashouf-Fard2, G Kandel2, B Salena1, JK Marshall1, EJ Irvine2 RATIONALE AND AIM: Acute bleeding from the upper or lower GI tract is a major cause of morbidity and mortality and an important cause for hospital admission. This audit was performed to examine the adherence to guidelines and quality of management of patients with GI bleeding attending an academic inner city hospital.
1Divisions of Gastroenterology, McMaster University, Hamilton, Ontario; 2St Michael's Hospital, Toronto, Ontario
METHODS AND STATISTICS: Consecutive patients presenting to St Michael's hospital with GI bleeding between September 1 and November 30, 2005 were identified by ICD-10 codes. Risk factors and potential outcome predictors were recorded on standardized data forms. Analysis was performed to examine predictors of outcome.
RESULTS: 98 patients (60 males), average age 63.6 yrs were identified. 78% were admitted for bleeding, 69% upper and 30% lower GI bleeds. 83% came through ER and 15% were referred from a community hospital. Admitting services were Medicine (50%), GI (15%), General Surgery (7%), Cardiology (7%), ICU (5%). Average length of stay was 12.9 days. 20% were in ICU (average ICU days 1.33). Mean risk stratification scores: Rockall Pre-Endoscopy (3.6/7), Post-Endoscopy (4.9/11) and Blatchford (10.7/23). Intermediate Risk of death based on Rockall and High risk based on Blatchford. Hemoglobin at admission was mean 86.2 mg/dl (range: 35-149 mg/dl). GI was consulted in 89% of cases and 83% had endoscopy in the Endoscopy Unit (90%), ICU (5.4%), ER (2.7%), OR (1.4%). Procedures included EGD (81.7%), Colonoscopy (25.6%), capsule endoscopy (1.2%) and double balloon enteroscopy (1.2%). 30% had hemostatic intervention. 13 patients (13.5%) developed complications prior to GI consult, including MI (10), shock (2), renal failure (2), thrombosis (1) and sepsis (1). 61.5% of patients received blood products prior to endoscopy, average units RBC 3.42. 82% of patients received a proton pump inhibitor (PPI) pre-endoscopy (59% iv infusion). 40% of patients were on PPI for >48hrs prior to endoscopy. Major findings included: Esophagitis (15.7%), Duodenal Ulcer (14.5%), Esophageal Varices (12.1%), Gastric Ulcer (10.8%), GI Cancer (8.4%), Colitis (7.2%). 86% and 73% had >1 risk factor for upper or lower bleeding with 52% taking a NSAID. Outcomes were: Discharged Alive (94%), Died (4%), Re-Bleed (17%), Readmitted for Bleeding (6%), Surgery (5%), Embolization (2%).
CONCLUSION: Most patients received appropriate GI consultation and endoscopy. We observed a higher rate of empirical iv PPI infusion and a longer than expected length of stay, likely reflecting a high risk population. Our rate of death was significantly lower than that predicted by Rockall scores suggesting appropriate clinical and endoscopic management.