Multiple Criteria Decision Analysis for HTA across four EU Member States: Piloting the Advance Value Framework

Abstract Multiple Criteria Decision Analysis (MCDA) has emerged as a likely methodology for Health Technology Assessment (HTA). However limited empirical evidence is available on its use by decision-makers and only as part of single-setting exercises, without cross-county studies available. This pilot study applies the Advance Value Framework (AVF), an MCDA methodology for HTA based on multi-attribute value theory, through a series of case studies with decision-makers in four countries, to explore its feasibility and compare their value preferences and results. The AVF was applied in the evaluation of three drugs for metastatic, castrate resistant, prostate cancer (abiraterone, cabazitaxel and enzalutamide in the post-chemotherapy indication). Decision conferences were organised in four European countries in collaboration with their HTA or health insurance organisations by engaging relevant assessors and experts: Sweden (TLV), Andalusia/Spain (AETSA), Poland (AOTMiT) and Belgium (INAMI-RIZIV). Participant value preferences, including performance scoring and criteria weighting, were elicited through a facilitated decision-analysis modelling approach using the MACBETH technique. Between 6 and 11 criteria were included in the value model of each country, allocated across four criteria domains; Therapeutic Benefit criteria conistently ranked first across countries in their relative importance. Consistent drug rankings were observed in all settings, with enzalutamide generating the highest overall weighted preference value (WPV) score, followed by abiraterone and cabazitaxel; dividing drugs' overall WPV scores by their costs produced the lowest “cost-per-unit of value” for enzalutamide, followed for abiraterone and cabazitaxel. These results contrast the HTA recommendations and pricing decisions in real life. Overall, although some differences in value preferences were observed between countries, drug rankings remained the same. The MCDA methodology employed could act as a decision support tool in HTA, due to the transparency in the construction of value preferences in a collaborative manner.