Data-Related Challenges in Cost-Effectiveness Analyses of Vaccines

Abstract

Cost-effectiveness analyses (CEAs) are often prepared to quantify the expected economic value of potential vaccination strategies. Estimated outcomes and costs of vaccination strategies depend on numerous data inputs or assumptions, including estimates of vaccine efficacy and disease incidence in the absence of vaccination. Limitations in epidemiologic data can meaningfully affect both CEA estimates and the interpretation of those results by groups involved in vaccination policy decisions. Developers of CEAs should be transparent with regard to the ambiguity and uncertainty associated with epidemiologic information that is incorporated into their models. We describe selected data-related challenges to conducting CEAs for vaccination strategies, including generalizability of estimates of vaccine effectiveness, duration and functional form of vaccine protection that can change over time, indirect (herd) protection, and serotype replacement. We illustrate how CEA estimates can be sensitive to variations in specific epidemiologic assumptions, with examples from CEAs conducted for the USA that assessed vaccinations against human papillomavirus and pneumococcal disease. These challenges are certainly not limited to these two case studies and may be relevant to other vaccines.