Development and Validation of Algorithms for Identifying Lines of Therapy in Multiple Myeloma using Real-World Data - Supplementary material

Supplementary Table 1. Criteria for identifying CRAB symptoms and SCTSupplementary Table 2. Algorithm accuracy at predicting specific regimens in LOT1: aggregate populationSupplementary Table 3. Algorithm accuracy at predicting specific regimens in LOT1: SCT cohortSupplementary Table 4. Algorithm accuracy at predicting specific regimens in LOT1: non-SCT cohortSupplementary Table 5. Algorithm performance for regimens identified in LOT2, as verified by clinician review of unstructured information from electronic health recordsSupplementary Table 6. Algorithm performance for regimens identified in LOT3, as verified by clinician review of unstructured information from electronic health recordsSupplementary Table 7 Algorithm accuracy at predicting specific maintenance therapies used in LOT1: aggregate populationSupplementary Table 8. Algorithm accuracy at predicting specific maintenance therapies used in LOT1: SCT cohortSupplementary Table 9. Algorithm accuracy at predicting specific maintenance therapies used in LOT1: non-SCT cohortSupplementary Figure 1. Decision rules for identifying lines of therapy. A. Aggregate population. B. SCT cohort. C. Non-SCT cohort. FLT = first-line therapy; LOT = line of therapy; SCT = stem cell transplant.<br>