Abstract Background In patients presenting with acute coronary syndrome (ACS) and anaemia, esophagogastroduodenoscopy (EGD) is commonly performed to exclude an upper gastrointestinal source of bleeding prior to perform percutaneous coronary intervention (PCI). However, the diagnostic yield and clinical impact of EGD in this setting is uncertain. Purpose We aimed to investigate the clinical yield and impacts of EGD on the 1-year clinical outcomes in ACS patients with concomitant anaemia who underwent PCI. Methods We retrospectively analysed electronic health records of ACS patients who underwent PCI at 16 public hospitals from 2014-2018. Patients with anaemia (defined as haemoglobin (Hb) <11dl/L) and EGD before PCI during the same admission were identified. The primary outcomes were bleeding events including clinical GIB (as diagnosed by EGD), need for urgent EGD, need for transfusion, Hb drop of >3 dl/L and Hb <8 dl/L. Secondary outcomes were 12-month 4-point MACE including myocardial infarction (MI), stroke, subsequent PCI or CABG, or cardiovascular-related death. Propensity-score models with inverse probability of treatment weighting (IPTW) against EGD were developed for the comparisons. A weighted Cox Proportional Hazard Model was performed to estimate adjusted hazards ratio (aHR) with corresponding 95% confidence intervals (CI). Results Of 13,793 patients, 3625 (25%) had anaemia on admission and 12.8% (n=466/3625) underwent EGD. Most EGD findings were not accountable for overt GIB, with gastritis, gastric polyps and gastric erosion being the 3 most common EGD diagnoses. After IPTW, clinical outcomes of 466 patients with EGD were compared with 3159 patients without EGD. 12-month bleeding event rates were higher among EGD patients compared to non-EGD patients (59.7% vs. 27.6%, p< 0.01; aHR = 3.01, 95% CI 2.58-3.52, p<0.005). Overall, 12-month GIB rate was low 1.0%, with 4.1% in EGD patients and 0.57% in non-EGD patients (aHR=7.99, 95% CI 4.13-15.45, p<0.005). The 12-month cumulative MACE rate overall was 39.9% and no significant difference between EGD and non-EGD patients. Conclusion Our real-world data demonstrated that the presence of anaemia is common among ACS patients. The yield of EGD in identifying sources of upper GI bleeding was low and there was no difference in 12-month clinical outcome among EGD patients. Although EGD was associated with higher rates of 12-month bleeding events, the cause was unclear with very low incidence of upper GI bleeding events. Therefore, the benefits of routine EGD prior to PCI in ACS patients with anaemic is unclear.
Read full abstract