Multi-Attribute Ranking Tool for Vaccines: Strategic Priority Setting to Support Vaccine Development
Methods: Using axiomatic principles of multi-criteria decision making (specifically, multi-attribute utility theory), SMART Vaccines includes 28 attributes in eight broad categories (e.g. health, economic, demographic, policy, scientific and business) and allows additional user-defined attributes. The multi-attribute utility model was implemented and tested in MATLAB. A population process model simulates demographic changes, with necessary data and proxies obtained from a number of sources (including World Health Organization and the Global Burden of Diseases). Preferential weights applied by users on the attributes determine the final rank order of the vaccine candidates.
Results: Output results—SMART Scores—combine computed values and user weighted entries, making the scores meaningful only to the users of the software. We evaluated ten vaccine candidates for hypothetical development in the United States and South Africa, and are pursuing use case analyses in collaboration with public and private vaccine decision makers to determine the software efficacy, and its usability across different scenarios. Built-in sensitivity analysis tools enable users to carry out a number of “what if” scenarios comparing specific health, demographic, economic and vaccine impacts to arrive at a decision independently or cooperatively with other users.
Conclusion: Test results, sensitivity analyses, and initial applications of SMART Vaccines indicate the need for a broad community driven effort to advance the software and data development efforts. Enhanced applications of the software are possible to help compare public health and other interventions beyond vaccines, and notably as an educational tool in academic public health courses.
See more of: Poster Abstract Session