Selecting Cost-Effectiveness Methods for Health Benefits Package Design: A Systematic Approach

Background: Cost-effectiveness (CE) is a common prioritization criterion in health benefits package (HBP) design. However, to assess CE is a time- and data-demanding process, so most HBP exercises rely wholly or partially on global evidence. Extensive investment has been made in analyses, models, and tools to support cost-effectiveness analyses (CEAs) for HBPs. However, little attention has been paid to how national HBP assessors should both understand and select costeffectiveness estimates. A structured, national process to select assessment methods is essential for ensuring the accuracy, ownership, and transparency of HBP design. This can be supported by ‘adaptive’ health technology assessment (aHTA) principles, which focus on structured methodological choices based on the time, data, and capacity available. The objective of this paper was to apply aHTA framing to CEA methods selection for HBPs, and to make recommendations on how countries may consider systematically making these choices going forward.

Methods: We first reviewed the definitions and categorization of different aHTA methods. We then conducted a scoping review of previous HBP assessments to understand how CEA methods used in HBPs fit into the aHTA framework, and a follow-up survey of authors to fill gaps. Results of the literature review and survey were interpreted and narratively synthesized.

Results: We found that previous HBP assessments used four aHTA methods, sometimes simultaneously: expert opinion (n=3/20), review (n=12/20), model adaptation (n=6/20), and new model (n=2/20). The literature review and survey found that aHTA methods for HBPs take between 1-13 months; require different data sources depending on the method(s) used; and generally, require capacity in health economics, medicine, public health, and cost-effectiveness modelling. We supplement our report with a discussion of key considerations for methods selection.

Conclusion: Trading off time, data, and capacity needs for different CE assessment methods can help to support structured, local design of HBP assessments.