Characteristics and Treatment Patterns of Patients with Deep Vein Thrombosis and/or Pulmonary Embolism in the US Hospital Setting

Introduction:

Current practice guidelines in the United States (US) recommend that anticoagulation treatment for deep vein thrombosis (DVT) and/or pulmonary embolism (PE) is initiated with parenteral administration of low-molecular-weight heparin (LMWH), fondaparinux or low-dose unfractionated heparin (UFH),  followed by the use of warfarin or other vitamin K antagonist therapies (Kearon et al Chest 2012). Given the relatively recent introduction of novel oral anticoagulants (NOACs), it is essential to understand the demographic and clinical characteristics of DVT/PE patients receiving these therapies. The objective of this study was to evaluate characteristics and treatment patterns of patients with DVT and/or PE in the US hospital setting.

Methods:

This retrospective observational cohort study used hospital administrative claims data spanning 11/1/2011-12/31/2013 from more than 600 hospitals. Patients ≥18 years old with a primary discharge diagnosis of DVT and/or PE in the inpatient setting during the identification period (11/1/2012-12/31/2013) were selected. The first-observed hospitalization for DVT/PE during the identification period was defined as the index hospitalization, and thus the principal event for analysis. Patients with a secondary diagnosis of atrial fibrillation, atrial flutter, cardiomyopathy, or a coagulation disorder during the index hospitalization were excluded. Patients needed to have had a complete hospital stay defined as (1) observed admission and discharge dates and (2) a discharge status other than expired or unknown. Patients were categorized into mutually-exclusive treatment cohorts based on the anticoagulant (AC) treatment regimen received during the index hospitalization, including lack of receipt of any AC treatment. Parenteral anticoagulant (PAC) therapy was defined as LMWH, fondaparinux or UFH. Demographic, hospital, and clinical characteristics were assessed during the index hospitalization. A logistic regression was conducted to assess the likelihood of receiving rivaroxaban with PAC and/or warfarin (vs. rivaroxaban alone) adjusting for the following covariates: age, sex, race, geographic region, payer type, primary or secondary diagnosis of DVT and/or PE, history of DVT/PE, discharge status, receipt of thrombolytic drugs, presence of primary malignancy, presence of renal disease, attending provider specialty, hospital setting, and hospital size.

Results:

In this study, 46,214 patients met the selection criteria (mean age: 61 years; 53% female). Sixty-eight percent of all patients were white and 50% were Medicare beneficiaries. The majority of patients received a primary diagnosis of PE during the index hospitalization (54%), followed by those with a primary diagnosis of DVT without mention of PE (38%) and primary diagnosis of DVT and secondary PE (8%). The most common AC treatment during the index hospitalization was PAC + warfarin (70%), PAC alone (16%), PAC + rivaroxaban (6%), PAC + warfarin + rivaroxaban (4%), warfarin alone (1%), rivaroxaban alone (1%), and no AC therapy (2%). Among the patients who received rivaroxaban, more than 90% additionally received PAC. The mean [SD] Deyo-Charlson Comorbidity Index scores were lowest for the cohorts receiving rivaroxaban (rivaroxaban alone, 1.1 [1.7]; PAC + rivaroxaban, 1.2 [1.8]; and PAC + warfarin + rivaroxaban cohorts, 1.2 [1.6]). In logistic regression analyses, patients with a primary PE diagnosis (odds ratio (OR): 1.49; 95% CI 1.17-1.88), a primary DVT diagnosis with a secondary PE diagnosis (OR 2.07; 95% CI 1.31-3.26), receipt of thrombolytic drugs (OR: 12.45; 95% CI 3.95-39.23), presence of renal disease (OR: 1.61; 95% CI 1.02-2.54), and receipt of care in an urban hospital (OR: 1.53; 95% CI 1.16-2.00) were significantly more likely to receive rivaroxaban with PAC and/or warfarin compared to rivaroxaban alone.

Conclusion:

This study evaluated characteristics and treatment patterns among patients with DVT and/or PE in the US hospital setting, following the introduction of the NOACs. Our findings indicate that the standard of care (PAC + warfarin) remains the dominant treatment regimen despite the availability of newer NOACs. Future studies of treated patients with DVT and/or PE should examine whether these preliminary observations of treatment patterns change with increased clinical experience with NOACs.