Scott Kaatz1, An-Chen Fu2, Azza AbuDagga3, Joyce LaMori4, Brahim K Bookhart5, C V Damaraju6, Hiangkiat Tan7, Jeff Schein8, Edith Nutescu9. 1. Hurley Medical Center, 1 Hurley Plaza, Flint, MI 48503, USA. Electronic address: skaatz1@hurleymc.com. 2. HealthCore, Inc., 800 Delaware 52, Wilmington, DE 19801, USA. Electronic address: AFu@healthcore.com. 3. HealthCore, Inc., 800 Delaware 52, Wilmington, DE 19801, USA. Electronic address: AAbudagga@healthcore.com. 4. Janssen Scientific Affairs, LLC, 1000 Route 202, Raritan, NJ, 08869, USA. Electronic address: JLamori@its.jnj.com. 5. Janssen Scientific Affairs, LLC, 1000 Route 202, Raritan, NJ, 08869, USA. Electronic address: BBookhar@its.jnj.com. 6. Janssen Research and Development, 920 Route 202, Raritan, NJ, 08869, USA. Electronic address: CDamara1@its.jnj.com. 7. HealthCore, Inc., 800 Delaware 52, Wilmington, DE 19801, USA. Electronic address: JTan@healthcore.com. 8. Janssen Scientific Affairs, LLC, 1000 Route 202, Raritan, NJ, 08869, USA. Electronic address: JSchein@its.jnj.com. 9. University of Illinois College of Pharmacy and Hospital and Health Sciences System, 833S. Wood St. MC 886, Rm 164, Chicago, IL 60612, USA. Electronic address: enutescu@uic.edu.
Abstract
INTRODUCTION: This retrospective observational study examined whether anticoagulant treatment duration varies by risks of venous thromboembolism (VTE) recurrence and bleeding. MATERIALS AND METHODS: VTE patients naïve to anticoagulants were identified from the HealthCore Integrated Research Database between 06/01/2007 and 09/30/2011 and categorized into three groups: provoked, cancer-related, and unprovoked VTE. Treatment duration was from initiation to discontinuation of anticoagulation, based on a 60-day gap in prescription fill unless there was an international normalized ratio test every 42 days. Bleeding risk was estimated using RIETE score, and VTE risk categories were based on ACCP guidelines. Kaplan-Meier curves and Cox proportional hazards models were used to evaluate association between VTE recurrence/bleeding and anticoagulation duration. RESULTS: Of 2002 patients identified (52.3% males, mean age 57 ±15 years), 21.4% had provoked, 16.4% had cancer-related, and 62.1% had unprovoked VTE. Average anticoagulant treatment duration was 294 ± 261 days. After adjusting for demographics and clinical characteristics, provoked and cancer-related VTE patients were 32% (95% CI=14-54%, P<0.001) and 35% (95% CI=7-70%, P=0.013) more likely, respectively, to discontinue anticoagulants than unprovoked VTE patients. No differences were observed between provoked and cancer-related VTE patients. Patients with an intermediate/high bleeding risk were 26% (95% CI=14-36%, P<0.001) less likely to discontinue treatment than those with a low bleeding risk. CONCLUSIONS: The observed anticoagulation duration for VTE may not be concordant with guidelines, due to the challenge of counterbalancing risks of VTE recurrence and bleeding. Further studies are needed to explore this.
INTRODUCTION: This retrospective observational study examined whether anticoagulant treatment duration varies by risks of venous thromboembolism (VTE) recurrence and bleeding. MATERIALS AND METHODS:VTEpatients naïve to anticoagulants were identified from the HealthCore Integrated Research Database between 06/01/2007 and 09/30/2011 and categorized into three groups: provoked, cancer-related, and unprovoked VTE. Treatment duration was from initiation to discontinuation of anticoagulation, based on a 60-day gap in prescription fill unless there was an international normalized ratio test every 42 days. Bleeding risk was estimated using RIETE score, and VTE risk categories were based on ACCP guidelines. Kaplan-Meier curves and Cox proportional hazards models were used to evaluate association between VTE recurrence/bleeding and anticoagulation duration. RESULTS: Of 2002 patients identified (52.3% males, mean age 57 ±15 years), 21.4% had provoked, 16.4% had cancer-related, and 62.1% had unprovoked VTE. Average anticoagulant treatment duration was 294 ± 261 days. After adjusting for demographics and clinical characteristics, provoked and cancer-related VTEpatients were 32% (95% CI=14-54%, P<0.001) and 35% (95% CI=7-70%, P=0.013) more likely, respectively, to discontinue anticoagulants than unprovoked VTEpatients. No differences were observed between provoked and cancer-related VTEpatients. Patients with an intermediate/high bleeding risk were 26% (95% CI=14-36%, P<0.001) less likely to discontinue treatment than those with a low bleeding risk. CONCLUSIONS: The observed anticoagulation duration for VTE may not be concordant with guidelines, due to the challenge of counterbalancing risks of VTE recurrence and bleeding. Further studies are needed to explore this.