OBJECTIVES: This project compares different perspectives on the ways choice sets at individual or subgroup levels on study designs (Huttin, ISPOR 2021) can address current limitations in clinical diagnosis current emphasis on correlation rather than causality analysis (Liebman, 2021) and their impact on health policy choices (Huttin and Hausman, 2021; Hahn, Hausman and Lustig, 2020).

METHODS: various approaches exist especially between econometric statistical models and trained data models including machine or deep learning methods to improve diagnostic spaces, but these may add bias to the analysis of real-world data. Dr M Liebman will present his analytics approach using multiple sclerosis as an example where inadequacy in addressing the complexity of the patient, of the disease and of clinical practice have negative impact on the patient, physician, payer, and pharma development. Prof C Huttin will show adaptive design from the perspective of the payer, for reimbursement policies: a selection of adaptive pricing and coverage policy making processes to various new data elements are reviewed in different countries.

RESULTS: on the first case, enhancing MS phenotyping can address critical decision points in the patient journey to identify inappropriate testing and treatments thus improving outcomes while reducing unnecessary costs. Value based pricing per indication using adaptive designs can be quite successful, but several factors can impede their performance especially IT infrastructures.

CONCLUSIONS: The authors demonstrate how to use adaptive study designs both for technology development and various reimbursement policies for economic systems. They identify ways to combine various new data elements to improve clinical diagnosis and policy making processes, as an additional perspective to accrued clinical adaptive trials including cost effectiveness analysis (Flight et ales, 2021: Chick et ales, 2017; Claxton et ales, 2012).