Title: Use of Real-World Evidence (RWE) to Validate a Trial-based Health Economic Model
Abstract: Decision makers often rely on health economic models populated with clinical trial data to inform initial assessments about treatment selection, coverage, and reimbursement. To date, there have been few (if any) published model re-analyses using real-world evidence (RWE). The purpose of this study is to 1) assess real-world health and economic outcomes associated with oral anticoagulant vs. low molecular weight heparin (LMWH) as prophylaxis for venous thromboembolism (VTE) in patients undergoing total hip (THR) or knee (TKR) replacement, and 2) compare results of a health economic model populated with clinical trial data vs. RWE. Patients who underwent THR or TKR between July 2011 and June 2012 were identified in a US commercial insurance claims database. Patients were required to be continuously enrolled 3 months pre-/post-index and were excluded if treated with multiple anti-coagulants within 10 days post-index. A propensity score matching technique was employed to reduce selection bias. Patient characteristics, inpatient-related VTE events and health care costs were determined. A health economic model previously parameterized with clinical trial data was repopulated and reanalyzed using inputs derived from the claims study. A total of 14,880 patients were identified (7,440 oral anticoagulant, 7,440 LMWH). In both groups, mean age was 59 and 53% were female. Compared with LMWH, oral anticoagulant use was associated with fewer symptomatic VTE events over 1-year. When repopulated with clinical inputs from claims data, the model projected similar VTE event differences as trial-based model (-0.023 vs. -0.015). Costs (per patient/year) in oral anticoagulant and LMWH groups were consistent across the trial-based model ($385 vs. $1,011), claims-based model ($437 vs. $1,290), and direct reported results from claims analysis ($506 vs. $1,125). Use of RWE is a practical and objective way to validate a trial-based health economic model. Future work should consider study design issues and practical use of results.