Safety of esophagogastroduodenoscopy within 30 days of myocardial infarction: A retrospective cohort study from a canadian tertiary centre

Abstract

BACKGROUND: Patients who experience myocardial infarction (MI) are at risk of gastrointestinal (GI) bleeding complications. Endoscopic evaluation may lead to cardiopulmonary complications. Guidelines and studies regarding the safety of endoscopy in this population are limited. OBJECTIVE: To evaluate the safety of endoscopy in a retrospective cohort of post-MI patients at a Canadian tertiary centre. METHODS: Using hospital diagnostic/procedure codes, the charts of patients meeting the inclusion criteria of having ST elevation MI or non-ST elevation MI, and GI bleeding detected at endoscopy were reviewed. The information retrieved included demographics, medical history, medications, endoscopy details and cardiopulmonary/GI events. RESULTS: A total of 121 patients experienced an MI and underwent endoscopy within 30 days. However, only 44 met the inclusion criteria and were reviewed. The mean age of the patients was 75 years, and 55% were female. The mean hemoglobin level was 86 g/L, and 38 of 44 patients required a transfusion. Comorbidities included hypertension (82%), diabetes (46%), heart failure (55%), stroke (21%), lung disease (27%), previous MI (46%), cardiac bypass surgery (30%), history of GI bleed (25%), history of ulcer (18%) and ejection fraction <50% (48%). The median number of days to endoscopy after MI was three. Complications included seven patients with acute coronary syndrome, one with arrhythmia, one with respiratory failure, one with aspiration pneumonia and two with perforation. Age, hemoglobin level or timing of endoscopy did not significantly predict a complication. CONCLUSIONS: Patients with GI bleeding after MI often have comorbidities and are on antiplatelet agents. Endoscopy is a valuable tool in the diagnosis and management of bleeding complications, but must be weighed against the potential risk of other complications, which in the present study occurred in more than 25% of procedures. ©2012 Pulsus Group Inc. All rights reserved.