Prior to the availability of a specific oral factor Xa (fXa) inhibitor reversal agent, off-label replacement agents including 4-factor prothrombin complex concentrates (4F-PCC) were used to treat life-threatening hemorrhage including gastrointestinal (GI) bleeds. We compared outcomes of US Medicare patients hospitalized for oral fXa-inhibitor associated GI bleeds treated with or without 4F-PCC. We utilized a 20% random sample of Medicare fee-for-service beneficiaries who had Part A and D (medication) coverage. We included adult (≥ 18 years) patients experiencing their first hospitalization (index event) for major GI bleeding identified via ICD9/10 codes, between 10/2013 and 9/2017. Serious GI bleeds were defined by ICD9/10 codes developed based on the Cunningham and Joos et al. algorithms. Inclusion criteria were continuous medical and prescription coverage and a claim for apixaban, edoxaban, or rivaroxaban during the 6-months prior to the index event. Patients were divided into two cohorts based on treatment with or without 4F-PCC during the index GI bleed hospitalization. Fresh frozen plasma, cryoprecipitates, packed red blood cells, activated PCC, recombinant factor VIIa, desmopressin acetate, tranexamic acid and vitamin K administration were allowed in either cohort. Use of more than one agent in either group was defined as concomitant therapy. Endpoints included the 1) intensive care unit (ICU) treatment, 2) total hospital length of stay, 3) discharge disposition and 4) 30-day readmission. Among 8,169 patients hospitalized for GI bleeding, 62 (0.75%) patients were treated with 4F-PCC. Mean age was 78.4 years and 58.8% were female. Because of the low numbers the two groups could not be propensity matched. Overall in-hospital mortality was 1.7%. Concomitant therapy was 56.5% 4F-PCC-treated vs. 45.3% non-4F-PCC patients. Incidence of ICU was 71.0% 4F-PCC vs 39.4% non-4F-PCC patients. Median (Interquartile range) hospital stay was 6 (4, 10) for 4F-PCC vs 5 (4, 6) days non-4F-PCC patients. Discharge disposition home was 45.2% for 4F-PCC vs 57.8% non-4F-PCC patients. All-cause readmission at 30 days was 18.0% for 4F-PCC vs 18.8% non-4F-PCC patients (Table 1). In a large descriptive analysis of hospitalized Medicare patients with oral fXa-inhibitor associated major GI bleeds, those treated with 4F-PCC had higher concomitant medication use, longer LOS, and lower frequency of discharge to home. Whether this is a result of off-label use of 4F-PCC or other variables in these complex patients is uncertain and deserves further study.Table 1Characteristics of US Medicare patients hospitalized for oral fXai-associated GI bleedsAll (N=8,169)4F-PCC Treated (N=62)Non-4F-PCC Treated (N=8,107)Indication for fXa inhibitors (%)Atrial fibrillation75.879.075.8Deep vein thrombosis/venous thromboembolism24.025.824.0Any concomitant medication (%)45.456.545.3Packed RBCs (%)43.853.243.7Characteristics of index hospitalizationLength of stay in days (Median; interquartile range)5.0 (4.0-6.0)6.0 (4.0-10.0)5.0 (4.0-6.0)Index hospitalization as ICU (%)39.771.039.4Discharge status (%)Home57.745.257.8Discharge to a location other than home (including in-hospital death)*42.354.842.2*Data on in-hospital death alone was not reported by treatment cohort in order to comply with the Centers for Medicare and Medicaid Services’ Cell Size Suppression Policy Open table in a new tab
Read full abstract