Evaluating The Cost-Effectiveness of Repeated Serological Biomarker Screening Compared to the Khorana Risk Score for Guiding Thromboprophylaxis Prescription in Patients Receiving Chemotherapy

Abstract

Background: The Khorana Risk Score (KRS), measured prior to treatment initiation, is the internationally recommended standard for guiding thromboprophylaxis prescriptions based on venous thromboembolism (VTE) risk in patients beginning chemotherapy. However, it performs poorly in certain cancer types. As VTE risk evolves over chemotherapy regimens, repeated prothrombotic biomarker screening (RBS) has been proposed as an alternative prescribing guide. We aim to quantify the health benefits and cost implications of this approach, in a cost-utility framework. Methods: Predictive cost-utility analysis was performed from an Irish health-system perspective, comparing RBS with standard KRS cut-offs which guide thromboprophylaxis prescribing in patients beginning chemotherapy. Outcomes were measured as quality-adjusted-life-years. A decision tree model was constructed to compare these two strategies and assess their impact on prescribing, VTEs, major bleed rates, and ultimately on survival, and quality-of-life. Deterministic and probabilistic sensitivity analyses were conducted, accounting for parameter uncertainties. Results: Improvements to test-sensitivity led to fewer VTEs overall, while improvements to specificity reduced major bleeds through more targeted thromboprophylaxis. Additional RBS-associated costs may be offset by reduced costs for treating VTE-related events. However, findings depend on numerous uncertain parameters, including population-level VTE risk, sensitivity/specificity of RBS, and long-term health outcomes. Conclusion: Improved VTE risk prediction enabling targeted thromboprophylaxis may yield better health outcomes by reducing VTE occurrences and minimising major bleeds associated with anti-coagulant therapy. Additional testing costs may be offset by reduced VTE treatment costs. Further research is needed to refine the accuracy of RBS for VTE risk prediction, key uncertain parameters in our analysis.