Incremental Health Care Burden of Bleeding Among Patients with Venous Thromboembolism in the United States

BACKGROUND: The health care and economic burden of venous thromboembolism (VTE) has been evaluated in regard to acute VTE, VTE recurrence, and some VTE complications, such as postthrombotic syndrome, but the cost burden attributed to bleedings is not well understood. OBJECTIVE: To evaluate health care resource utilization and costs associated with major bleeding (MB) and clinically relevant nonmajor bleeding (CRNMB) among a large U.S. commercially and Medicare-insured population with VTE. METHODS: Patients (≥18 years of age, continuously insured) with a diagnosis of VTE between January 1, 2008, and December 31, 2011, were identified from the Truven Health Analytics Commercial and Medicare MarketScan databases. Patients who did not have any bleedings during the study period were grouped into a no-bleedings cohort and a random date after VTE diagnosis was selected as the index date. VTE patients who experienced MB within 1 year of the initial VTE diagnosis were grouped into a MB cohort, and patients without MB but with CRNMB were grouped into a CRNMB cohort. Baseline patient demographics and clinical characteristics were determined for study cohorts. All-cause and bleeding-related health care resource utilization and costs (inflation adjusted to 2013 level) during a 12-month follow-up period after the index date of the initial bleeding event were measured and compared. Descriptive statistics were used to evaluate differences in demographics, clinical characteristics, and unadjusted health care resource utilization and costs of patient cohorts. Multivariable generalized linear models were used to evaluate incremental health care costs of bleedings after adjusting for key patient characteristics. RESULTS: Among the 112,885 patients identified with a VTE diagnosis, 14% (n = 15,897) had MB and 14% (n = 15,842) had CRNMB; 72% (n = 81,146) had neither of these events occur during the study period. The mean ages of the MB and CRNMB cohorts were both 63.6 years, while the mean age of the no-bleedings cohort was significantly lower at 59.6 years (P < 0.001). Mean Charlson Comorbidity Index scores were highest for the MB cohort (3.2), followed by those of the CRNMB cohort (2.5) and the no-bleedings cohort (1.6). The MB cohort had the greatest proportion of patients with an initial VTE event of pulmonary embolism only (23.5%), followed by that of the CRNMB cohort (20.2%), and the no-bleedings cohort (16.7%). For MB and CRNMB cohorts, the total mean bleeding-related inpatient and outpatient costs during the follow-up period were $49,779 (SE = $820) and $2,187 ($89), respectively. After adjustment for key patient characteristics, the estimated mean differences in total bleeding-related medical costs were $45,367 ($1,853) for patients with MB and $2,140 ($88) for patients with CRNMB versus patients with no bleedings. CONCLUSIONS: Among patients with VTE diagnosis in the United States, approximately 28% have a bleeding event within 1 year of VTE diagnosis, and about half of these patients experience MB. Patients with MB have greatly elevated health care costs. Disclosures: Sponsorship for this study and article processing charges were funded by Bristol-Myers Squibb and Pfizer. Amin is a consultant for Novosys Health, which received funding from Bristol-Myers Squibb and Pfizer in connection with conducting this study and development of this manuscript. Bruno is an employee of Bristol-Myers Squibb and owns stock in the company. Trocio is an employee of Pfizer and owns stock in the company. Lin and Lingohr-Smith are employees of Novosys Health, which has received research funds from Bristol-Myers Squibb and Pfizer in connection with conducting this study and development of this manuscript. Amin, Bruno, Trocio, and Lin designed the study, contributed to the interpretation of data and revising of the intellectual content and the manuscript, and approved the final version to be published. Lin, Lingohr-Smith, and Amin additionally assisted in the development of the cost model and contributed to the writing of the manuscript. Lingohr-Smith also assisted in the interpretation of data.