Burden-of-Illness Associated with Bleeding-Related Hospitalizations in Atrial Fibrillation Patients: Findings from the Nationwide Readmission Database.

Anne Beaubrun,Belinda Lovelace,Benjamin Miao,Charles Michael White,Craig I Coleman,Kelly Mcneil-Posey,Mark J Alberts,Monique Miller,Olivia S Costa,William Frank Peacock

doi:10.1055/s-0040-1716549

Abstract

Introduction A paucity of contemporary data examining bleeding-related hospitalization outcomes in atrial fibrillation (AF) patients exists. Methods Adults in the Nationwide Readmissions Database (January 2016–November 2016) with AF and hospitalized for intracranial hemorrhage (ICH), gastrointestinal, genitourinary, or other bleeding were identified. Association between bleed types and outcomes were assessed using multivariable regression (gastrointestinal defined as referent) and reported as crude incidences and adjusted odds ratios (ORs) or mean differences with 95% confidence intervals (CIs). Results In total, 196,878 index bleeding-related hospitalizations were identified in this AF cohort (CHA2DS2VASc score ≥2 in 95.1%), with 70.8% classified as gastrointestinal. The overall incidences of in-hospital mortality, need for post-discharge out-of-home care, and 30-day readmission were 4.9, 50.8, and 18.2%, respectively. Multivariable regression suggested traumatic and nontraumatic ICHs were associated with higher odds of in-hospital mortality (OR = 3.99, 95% CI = 3.79, 4.19; OR = 13.09, 95% CI = 12.24, 13.99) and need for post-discharge out-of-home care (OR = 2.92, 95% CI = 2.83, 3.01; OR = 2.74, 95% CI = 2.59, 2.90), and increases in mean index hospitalization length-of-stay (8.31 days, 95% CI = 8.03, 8.60, 6.27 days, 95% CI = 5.97, 6.57) versus gastrointestinal bleeding. Genitourinary and other bleeds were associated with lower mortality (OR = 0.37, 95% CI = 0.25, 0.55; OR = 0.59, 95% CI = 0.53, 0.64) and reduced length-of-stays (−2.84 days, 95% CI = − 2.91, −2.76; −2.06 days, 95% CI = − 2.11, −2.01) versus gastrointestinal bleeding. Genitourinary bleeds were also associated with a reduced need for post-discharge out-of-home care (OR = 0.86, 95% CI = 0.77, 0.97). Conclusion The burden of bleeding-related hospitalizations was notably driven by relatively rare but severe and life-threatening ICH, and less morbid but more frequent gastrointestinal bleeding. There is need for continued research on bleeding risk factors and mitigation techniques to avoid bleeding-related patient hospitalizations.

Highlights

A paucity of contemporary data examining bleeding-related hospitalization outcomes in atrial fibrillation (AF) patients exists
In one real-world study of patients with AF who were treated with oral factor Xa inhibitors, those who were hospitalized with a major bleed of any type were found to have mean hospital stays exceeding 5 days and inpatient mortality rates of approximately 3%, though mortality risk was notably higher for patients with intracranial hemorrhage (ICH)
Real-world study of nearly 200,000 AF patients hospitalized for bleeds, we found that bleeding-related hospitalizations are associated with differing levels of detrimental clinical consequences across various bleed types

Summary

Introduction

A paucity of contemporary data examining bleeding-related hospitalization outcomes in atrial fibrillation (AF) patients exists. In one real-world study of patients with AF who were treated with oral factor Xa inhibitors, those who were hospitalized with a major bleed of any type were found to have mean hospital stays exceeding 5 days and inpatient mortality rates of approximately 3%, though mortality risk was notably higher for patients with intracranial hemorrhage (ICH) (approximately 14%).. The availability of contemporary data on this topic would allow clinicians and other decision-makers to better understand the absolute and relative risks of detrimental outcomes across various bleeding-related hospitalization types. This study sought to evaluate incidence rates and subsequent outcomes associated with bleeding-related hospitalizations, overall and by bleeding subtype, among patients with AF. Though the dataset used for this analysis does not contain data on anticoagulant use, we limited the present analysis to AF patients as they are considered higher risk for thrombosis and should be treated with oral anticoagulation in all but the lowest risk patients.

Methods

Results

Discussion

Conclusion