Abstract
Introduction
Many pharmacologic agents have been used for venous thromboembolism (VTE) prophylaxis after elective total hip arthroplasty (THA). Rivaroxaban was the first novel oral anticoagulant approved for THA patients, but its actual efficacy and safety in clinical practice, beyond randomized trials, is unknown.
Materials and Methods
This is a retrospective study, using the Truven Health MarketScan database, of anticoagulation medication prescriptions after elective THA, in both commercially insured (CI) and Medicare supplement insured (MS) patients, from 2010 to 2015. After exclusions, there were 83,179 CI and 50,534 MS patients available for analysis. There were 12,876 new users of warfarin (W) and 10,892 new users of rivaroxaban (R) in CI patients, and 7,416 new users of W and 4,739 new users of R in MS patients. We asked the following questions: (1) What were the trends and predictive factors for anticoagulant use after elective THA? (2) What was the actual clinical efficacy: frequency of deep vein thrombosis (DVT) and pulmonary embolism (PE), and frequency of adverse events within 90 days with the two most commonly used oral agents, rivaroxaban and warfarin, from June 2011 to September 2015? Data was analyzed for each anticoagulant by odds ratios using logistic regression models with stabilized inverse probability treatment weighting.
Results
There was a change in use of anticoagulants after R approval. Use of W decreased from approximately 50% each in 2010 in both insurance cohorts to 10% in CI patients and 30% in MS patients in 4th quarter 2015. The use of R increased from 0 to 33% in both cohorts from 2011 to 2015. In the multivariate analysis, in CI patients, females had lower odds of getting R, and patients in Western region had higher odds of getting R; in MS patients, increasing age had reduced odds of getting rivaroxiban, but Western region and surgery in 2015 had higher odds. Patients with capitated insurance plans and renal impairment had lower odds of R initiation, but a history of cardiovascular disease or hypertension had higher odds. In 90 days after THA, patients given R had significantly lower odds ratio of both DVT and PE in both CI patients (DVT: 1.54 with W, 0.54 with R; PE: 2.12 with W, 0.73 with R) and MS patients (DVT: 3.01 W, 1.73 R; PE: 4.09 W, 1.88 R). With logistic regression analysis, users of W had significantly higher odds ratio of both DVT (CI 2.63 and MS 1.78) and PE (CI 2.60 and MS 2.09) than R. There was no significant difference in rates of bleeding between W and R, but W had higher odds ratio than R of prosthetic joint infection (PJI) in both CI (1.574) and MS (1.790) cohorts.
Conclusions
There has been an increase in VTE prophylaxis with R, and a decrease in both W and LMWH use after elective THA over four years. Patient factors, insurance type, and comorbidities were associated with this change. In actual clinical efficacy, R had lower odds ratio of both DVT and PE than W, and bleeding risks were similar. The association of W with an increased odds ratio of PJI compared to R requires further study.
