Reduced Emergency Room Utilization for Initiation of Anticoagulation with Rivaroxaban Versus Low Molecular Weight Heparin for Treatment of Cancer-Associated Thrombosis

Background: Venous thromboembolism (VTE) is a leading cause of mortality and morbidity in cancer patients. The current standard of care is to treat cancer-associated VTE with Low Molecular Weight Heparin (LMWH), which is more effective than vitamin K antagonists, such as warfarin. However, LMWH injections are painful, expensive, and a burden in healthcare resource utilization. Further, many of these patients are referred to an emergency room for education on self-injection techniques, an added health care cost. The Hematology/Anticoagulation Management Service at Memorial Sloan Kettering Cancer Center is developing rivaroxaban as a safe and effective alternative to LMWH for cancer-associated VTE. In this report, we demonstrate a significant reduction in the referral of patients to the MSKCC Urgent Care Center (UCC), our in-house emergency room, to initiate anticoagulation, resulting in a significant reduction in resource utilization.

Methods: As a Quality Assessment Initiative, we track all patients with cancer-associated VTE at MSKCC receiving rivaroxaban since January 2014, and have a similar database of cancer-associated VTE from June through December 2013, treated with enoxaparin. For this utilization of resources study we evaluated where anticoagulation was initiated for treatment of a new pulmonary embolism or lower extremity deep vein thrombosis, specifying either a single outpatient visit, two outpatient visits on the same day, a telephone call, or a visit to the UCC. When patients had a second outpatient visit on the same day as diagnosis of the VTE, the second visit was for patient education on injection technique and insurance authorization. The site of anticoagulation initiation was by the judgment and discretion of the physician managing the patient's cancer. Patients who developed a VTE during a hospital stay or were managed at an outside emergency room were not included in this analysis, as our program has no influence in those settings. Statistical analysis was with the Chi-square test. Comparison of safety and efficacy of rivaroxaban and LMWH is the subject of a separate study.

Results: As anticipated, changing from a parenteral anticoagulant to rivaroxaban (an oral agent) resulted in significant changes in practice (Table). Significantly fewer rivaroxaban patients required visits to the UCC than with enoxaparin (p=0.009). Fewer patients required a second outpatient, as well. When viewed in the aggregate of UCC or second outpatient visit, 82% of patients treated with enoxaparin required additional medical resources for initiation of anticoagulation, beyond a single outpatient visit, which decreased to 59% with rivaroxaban (p<0.001). Of note, 11% of rivaroxaban patients were initiated with a phone call only, typically after a recent medical visit and outpatient imaging. The reduction in UCC utilization was confined to weekday hours when the outpatient clinics are open. In all cases with enoxaparin and rivaroxaban, initiation of anticoagulation was in the UCC on weekends and between the hours of 6 PM and 8 AM on weekdays.

Discussion: In addition to the burden of morbidity and mortality, management of cancer-associated VTE with LMWH is painful to the patient and expensive to the healthcare system. In our QAI we have been developing rivaroxaban as an oral alternative. Safety and efficacy are being analyzed separately. In addition to the markedly lower cost of rivaroxaban compared with enoxaparin, and patient quality of life preference, we also demonstrate a significant reduction in the healthcare resources associated with initiation of anticoagulation. Cancer patients tend to be higher risk and more complex than general medical patients and it remains appropriate for some to be evaluated in an emergency room for diagnosis and management. Despite this, we observed a reduction in emergency room visits for patients in this setting. One limitation to our findings is that our study was within a single institution, with a devoted Hematology/Anticoagulation Management Service. It would be appropriate to perform a similar analysis in other institutions, including non-cancer patients, to confirm these findings.