Objectives <p>Patients who receive hip or knee replacement surgery should be anticoagulated to prevent thrombosis-related events, such as pulmonary embolism. The Pulmonary Embolism Prevention after Hip and Knee Replacement (PEPPER) is a large pragmatic trial studying which anticoagulant (aspirin, warfarin, or rivaroxaban) is optimal. As an adjunct to this study, we examined the role of patients’ preferences in the tailoring of therapy, focusing on identification of subgroups or preference phenotypes.</p> Methods <p>We constructed a multimedia conjoint analysis (CA) survey based on anticoagulants’ beneficial and adverse effects at the expected probabilities being studied in the PEPPER trial. We recruited 192 hip and knee post-surgery patients at the Medical University of South Carolina who were eligible for the PEPPER trial and were studied 1 to 7 months after surgery. K-means clustering was used to characterize heterogeneity in patients’ preferences.</p> Results <p>Across the studied population, expected risks of major adverse effects (bleeding, venous thrombosis, and pulmonary embolism) were, on average, rated as being of similar importance, with somewhat greater weight being placed on avoiding risks of pulmonary embolism. However, few patients had values near the average for the population, with patients grouping in three distinct, minimally overlapping segments or preference phenotypes: thrombosis-focused values (aligns with rivaroxaban treatment), balanced values (bleeding and thrombosis-focus (aligns with aspirin)), and out-of-pocket-cost focused values (aligns with aspirin or warfarin).</p> Conclusions <p>In the post-knee/post-hip replacement setting, a CA survey revealed that patients value the risks and benefits of anticoagulation differently, with three distinct preference phenotypes that have potential implications for individualizing therapy.</p>