Broadstreet team develops method for reconstructing patient-level data from KM curves

Cost-effectiveness analyses and indirect treatment comparisons are important areas of health economics and outcomes research and when conducting these analyses, individual patient data (IPD) can provide valuable insight. Often, however, aggregate survival data are only published in the form of Kaplan-Meier (KM) curves and actual patient-level data are not provided. In order to be able to capture these data from KM curves for use in analyses, Broadstreet’s Basia Rogula, Greta Lozano-Ortega and Karissa Johnston have developed an algorithm for reconstructing IPD from such curves.

The team recently published the technique in MDM Policy & Practice. After testing and validating the algorithm (which requires total patient count, the coordinates of the drops in survival, and the times of the marked censoring points) they found that IPD can reliably be reconstructed from KM survival curves, when extractable censoring times are marked and easily distinguishable. Use of the algorithm will allow health researchers to reconstruct IPD more closely by incorporating censoring times exactly as marked.