Presentation (CTI)

OBJECTIVES: Each pharmaceutical company uses its own individual patient data and indirect comparison methods to adjust its population to that of a competitor's trial, with results favoring each company's treatment. This study aims to analyze and resolve the conflicts arising from differences in the definition of patient populations in cross-trial treatment effect comparisons using population-adjusted indirect comparison methods.
METHODS: This study introduces duality theory from operations research and formalizes the problem of selecting the decision target population into a dual optimization problem. The primal problem is defined as finding an optimal adjustment method that minimizes the discrepancy between the study population and the target population characteristics. The corresponding dual problem analyzes from another perspective: finding a weighted scheme for the target population characteristics that maximizes the adaptation of the two study populations to the decision target population.
RESULTS: The analysis shows reveals that the core of the result conflict lies not in the differing conclusions drawn from indirect comparisons between different target populations, but in the differences in the definition of the decision target population. Duality theory enables the systematic identification of biases in population definitions and optimizes adjustment methods to reduce population discrepancies. The solution to the dual problem helps identify better weighting schemes for the target population, providing a theoretical basis for resolving the conflicts in population-adjusted indirect comparisons.
CONCLUSIONS: This study uses duality theory to uncover the underlying causes of population conflicts in population-adjusted indirect comparisons analysis and proposes solutions, including introducing third-party standard target populations, optimizing adjustment weights, and conducting sensitivity analysis. These approaches offer new insights for improving population-adjusted based decision support and highlight the importance of defining the target population in drug economic evaluations.