Propensity Score Matching: A Real-World Approach for Comparative Effectiveness in Oncology (Motivated by the Multicenter Study on Chemotherapy Protocols for Burkitt Lymphoma by Rajendra et al.)

This paper presents a detailed analysis of Burkitt lymphoma treatment outcomes in low- and middle-income countries (LMICs), utilizing real-world data and the Propensity Score Matching (PSM) statistical method. The study, motivated by the multicenter research on chemotherapy protocols for Burkitt lymphoma by Rajendra et al. (2025), compares the effectiveness of methotrexate-based and EPOCH-based chemotherapy regimens in adolescents and adults with Burkitt lymphoma across India. PSM addresses the confounding biases typically present in observational studies by balancing covariates like age, HIV status, and disease severity. The results showed no significant differences in survival outcomes between the two treatment protocols, with event-free survival at 68% and overall survival at 70% for methotrexate and 72% for EPOCH. The document emphasizes the importance of detailed patient-level clinical data and robust matching algorithms for ensuring valid comparisons in non-randomized settings. Additionally, the paper discusses the strengths and limitations of PSM in oncology, especially in resource-limited settings. It provides insights into future directions, such as incorporating machine learning and longitudinal data to improve treatment effect estimation. Ultimately, the study highlights the potential of PSM in enhancing the credibility of real-world evidence for treatment decisions in LMICs.