The VMI aims to use computational models and simulations of measles disease data to address measles control strategies, in particular to determine optimal characteristics of the measles vaccine (target product profile, TPP). VMI work on measles is done in collaboration with government in Niger and currently, a new partnership is being developed in Nigeria. Models of measles dynamics in Niger were used to assess the regional variation in seasonality and measles persistence. Despite significant annual variation in magnitude, measles outbreaks in Niger consistently begin in the dry season and decline at the onset of the seasonal rains.
We previously proposed that human movement, in response to agricultural cycles, drives seasonal fluctuation in density and thus measles transmission in Niamey, the capital of Niger. Using surveillance data from the 38 districts and urban centers of Niger, we have addressed the question of the human versus environmental drivers of measles seasonality through a comparative study of the estimated seasonal variation in transmission across the country.
A key issue in measles control is the ability to provide measles vaccine prior to the expected age of first exposure. Previous work has shown that the median age of measles infection in Niger is two years. Variation in population density and access to health services present a challenge to achieving high coverage at young ages within and between districts. Understanding this regional variation in health accessibility is key to optimizing the distribution of vaccine services and regional demand on the vaccine supply chain. Regional variation in age specific vaccination rates in urban settings with high access to health services has been compared to rural settings with poorer access to services to identify potential gaps in the current vaccine distribution strategies.
Understanding the appropriate scale with which to represent populations when modeling measles transmission dynamics and control is critical to ensure models adequately represent observed dynamics while remaining computationally tractable. While agent-based microsimulations offer unrivalled flexibility and the ability to represent the spatial distribution and mobility of populations at great detail, they are also orders of magnitude computationally more intensive than compartmental ‘patch’ (also called metapopulation) models. The Princeton and Imperial groups are collaborating on a project to compare these two model formulations.
Measles disease dynamics models will be linked to vaccine supply chain models. This linkage allows us to explore how altering vaccine and supply chain characteristics will affect measles transmission. Parameter sweeps will elucidate the relationship between different characteristics of the supply chain and measles control. By incorporating economic parameters we will be able to determine the incremental economic value of adding, altering, and removing components of the supply chain. Results from models and simulations of measles dynamics and the vaccine supply chain in Niger support policy making on measles vaccination strategies and on the effects on the cold chain of potential introduction of new vaccines in Niger. The VMI is making an effort to extend these models from Niger to other countries in the region to better understand cross-border epidemic dynamics.