Bing Chen, MD1, Michael S. Smith, MD, MBA2; 1Mount Sinai Morningside and West, New York, NY; 2Mount Sinai West & Mount Sinai Morningside Hospitals, New York, NY
Introduction: Patients with atrial fibrillation (AF) are predisposed to gastrointestinal bleeding (GIB) due to anticoagulation use. Evaluation frequently requires endoscopy to make a diagnosis and, in some cases, provide treatment. This study aimed to investigate the safety of endoscopy in hospitalized patients with atrial fibrillation. Methods: The 2016 National Inpatient Sample database was analyzed using ICD-10 codes to identify all hospitalizations with a primary or secondary diagnosis of AF. Those patients were subdivided based on whether any endoscopic procedure was performed, as evidenced by listing of a concurrent procedure code. Subgroup analysis for three presumed indications (upper GIB, lower GIB, and anemia) was performed. Multivariate regression was used to adjust for age, gender, incomes, insurance, hospital characteristics, Charlson comorbidity index, and presumed indications. The primary outcome was in-hospital all-cause mortality. STATA 15 (College Station, TX) was used for analysis. Results: A total of 3,656,792 patients were admitted to a hospital in 2016 with a diagnosis of AF, of which 219,539 (6.00%) underwent endoscopy (Table 1). Patients in the endoscopy group had a slightly higher Charlson comorbidity index compared to the non-endoscopy group. The three most common presumed indications for endoscopy were upper GIB (32.62%), lower GIB (27.22%), and anemia (14.29%). In comparison to the non-endoscopy group, the endoscopy group had significantly lower overall mortality (3.65% vs 5.27%), with an adjusted odds ratio of 0.42 (p< 0.001, Table 2). In the subgroup analysis, patients who had an endoscopy had significantly lower mortality in all three subgroups: upper GIB (4.04% vs 11.03%), lower GIB (2.58% vs 12.64%), and anemia (3.27% vs 5.74%) (all p< 0.001). Discussion: In hospitalized AF patients, performing endoscopy was associated with significantly lower mortality. This study suggests that AF should not be a deterrent to endoscopy when it is indicated. However, the study is limited in that we cannot say how many patients could not safely have an endoscopy even when it was a potentially life-saving measure. Lower mortality could therefore reflect a bias toward selecting only healthier patients for endoscopy, even though the Charlson index was higher in the cohort undergoing endoscopy. Prospective studies are needed to validate these findings.
Table 1: Characteristics of AF Patients Having Endoscopy vs. No Endoscopy
Table 2: Subgroup Analysis of Mortality in AF Patients Having Endoscopy vs. No Endoscopy
Disclosures: Bing Chen indicated no relevant financial relationships. Michael Smith indicated no relevant financial relationships.