Presentation (CTI)

OBJECTIVES: To demonstrate the feasibility and validity of using reconstructed randomized controlled trial outcomes to support model-based cost-utility analyses when individual patient-level data are unavailable.
METHODS: A model-based cost-utility analysis was conducted from a healthcare payer perspective using a partitioned survival model with a lifetime horizon. Published KM curves from a recently approved therapy evaluated in a pivotal RCT were digitized and reconstructed using the Guyot iterative algorithm to generate pseudo-IPD. Parametric survival models (including Weibull and log-normal distributions) were fitted to reconstructed data and selected based on statistical fit and clinical plausibility. Health state utilities and direct medical costs were sourced from peer-reviewed literature and standard reimbursement schedules. Outcomes were expressed as quality-adjusted life years (QALYs) and costs (2024 USD), with incremental cost-effectiveness ratios (ICERs) calculated. Deterministic and probabilistic sensitivity analyses were performed to assess parameter uncertainty.
RESULTS: Reconstructed survival data closely reproduced published trial survival estimates, supporting their suitability for long-term extrapolation. The economic model generated plausible ICER estimates consistent with previously reported evaluations of comparable therapies. Sensitivity analyses identified survival extrapolation assumptions and drug acquisition costs as key drivers of cost-effectiveness. Cost-effectiveness acceptability curves demonstrated stable decision uncertainty across commonly applied willingness-to-pay thresholds.
CONCLUSIONS: This study demonstrates that survival curve reconstruction techniques provide a practical and reproducible approach for conducting model-based economic evaluations when access to proprietary IPD is unavailable. The findings support the use of reconstructed trial data in HTA submissions and early value assessment of newly approved therapies, aligning with ISPOR good research practices.