A process to validate prognostic factors for unanchored matching-adjusted indirect comparison of single-arm trials in oncology: a proof-of-concept study

• DOI: 10.57264/cer-2024-0235
• Publication Date: 2025-04-07T08:50:05Z
• Authors: Yao Yi; Yawen Jiang

Aim: The choice of covariates in unanchored matching-adjusted indirect comparisons (MAICs) single-arm cancer trials with time-to-event outcomes remains a challenge. Currently, there is lack systematic approach for validating the selection bias reduction MAIC. Materials & methods: This study proposes validation framework to evaluate appropriateness selected prognostic factors before their use process involves identifying potential from individual patient data and calculating risk scores using regression; artificially creating two groups that are unbalanced such predetermined hazard ratio (HR) between achieved; weights based on factors; running re-weighted Cox regression assess HR, value which should suggest balanced risks across indicate sufficiency being included. We also conducted proof-of-concept analysis simulated dataset showcase this process. Results: successfully stratified sample into pre-determined HR 1.8. When all were included weighting, was 0.9157 (95% CI: 0.5629–2.493), close one. one critical omitted covariates, became 1.671 1.194–2.340), significantly different Conclusion: Filling gap existing evidence synthesis literature, introduces structured data-driven covariate prioritization may be useful tool quantitative selection.