Outcomes and resource utilization during and after oral anticoagulant-related bleeding in chronic kidney disease (ORACLE-CKD): A population-based cohort study

Abstract Background: In the United States, an estimated 37 million people have chronic kidney disease (CKD), which increases the risk of atrial fibrillation (AF) and venous thromboembolism (VTE). Oral anticoagulant (OAC) treatment of these common conditions is complicated by a high risk of bleeding. Data on outcomes following OAC-related bleeding in CKD are limited. We aimed to evaluate clinical outcomes, healthcare utilization, and factors associated with mortality among adults with CKD hospitalized for OAC-related bleeding. Methods: We conducted a population-based cohort study using linked administrative health databases in Ontario, Canada which has a single-payer, publicly funded healthcare system including prescription drug coverage for people aged ≥65 years. We accrued patients ≥66 years who were hospitalized for bleeding and had an OAC (apixaban, dabigatran, edoxaban, rivaroxaban, or warfarin) dispensed preceding admission (April 2012 to March 2022). Patients were followed from the discharge date of their index hospitalization to the first of death, loss of OHIP coverage, or 365 days of follow-up. The exposure of interest was severity of CKD categorized as none/mild (no eGFR recorded or eGFR ≥60 mL/min/1.73m²), moderate (eGFR 30-59 mL/min/1.73m²), or severe (eGFR <30 mL/min/1.73m² or receiving outpatient hemodialysis prior to the index admission). The primary outcome was all-cause mortality within 90 days of hospital discharge. Secondary outcomes were mortality during index hospitalization, 1-year mortality, bleeding, thromboembolism (myocardial infarction, ischemic stroke, transient ischemic attack, deep vein thrombosis, pulmonary embolism), hospital readmission, and emergency department (ED) visits. Outcomes were defined using validated ICD-10 codes associated with ED or hospital admission as most responsible diagnosis. Crude outcomes were reported descriptively according to CKD severity. We examined associations between baseline covariates and mortality using multivariable Cox regression to calculate adjusted hazard ratios (HR) and 95% confidence intervals (CI). Results: Among 29,503 cohort members, 4,584 had no CKD (15.5%), 9,996 had mild CKD (33.9%), 11,558 had moderate CKD (39.2%) and 3,365 had severe CKD (11.4%). The mean age was 81 years and 45% were female. Most patients were treated for AF (65%) and were receiving factor Xa inhibitors (50%), followed by warfarin (41%). Gastrointestinal bleeding was most common (59%), followed by intracranial (15%) and genitourinary (11%). Prior to index, 482 (2% of) patients were receiving renal replacement therapy. Median duration (interquartile range [IQR]) of hospitalization was 4 days (2-8) for patients with none/mild CKD, 5 days (3-9) for moderate CKD and 6 days (3-10) for severe CKD. In-hospital mortality was similar across CKD categories (none/mild 9%, moderate 9% and severe 10%). Within 90 days of hospital discharge, mortality was highest in severe CKD patients (16%), followed by moderate CKD (12%) and no/mild CKD (9%). Mortality at 1 year was highest in severe CKD patients (39%), followed by moderate CKD (28%) and no/mild CKD (20%). Within 1-year post-discharge, 80% of severe, 72% of moderate, and 68% of none/mild CKD patients visited an ED, while 68% of severe, 55% of moderate and 49% of none/mild CKD patients were re-hospitalized. Baseline characteristics independently associated with an increased risk of 90-day mortality were cancer (HR 2.90; 95%CI 2.56-3.30), residence in long-term care (HR 2.15; 95%CI 1.92-2.41), intracranial bleeding (HR 1.97; 95%CI 1.69-2.30), severe CKD (HR 1.40; 95%CI 1.24-1.57), congestive heart failure (HR 1.34; 95%CI 1.23-1.45), Elixhauser comorbidity score ≥4 (HR 1.32; 95%CI 1.20-1.45), and dementia (HR 1.24; 95%CI 1.12-1.36). Conclusion: Patients hospitalized with OAC-related bleeding are at high risk of death during and immediately following hospitalization. Mortality risk was greater among patients with CKD. Within 1 year of hospital discharge, three-quarters of patients with moderate or severe CKD visited the ED, and over half were readmitted to hospital. Risk factors for 90-day mortality included severe CKD, intracranial bleeding, residence in long-term care, congestive heart failure, dementia, cancer and high comorbidity burden. These novel data highlightthe burden of OAC-related bleeding to both facilitate shared decision-making regarding OACs and inform the development of strategies to mitigate adverse outcomes.