What do we do?

We use mathematical and statistical tools to understand patterns of disease incidence, and the effects of heterogeneity, in time and space.

Design and implementation of surveillance for endemic disease

Models are only as good as the data upon which they are based … and the value of the policy recommendations based on those models are even more so dependent on the quality underlying data. Rather than being passive consumers of infectious disease surveillance data, modelers can play an important role in designing the surveillance systems necessary to develop efficient and effective control for infectious disease. We are working with ministries of health and agriculture to develop surveillance strategies to efficiently allocate vaccination effort to control or eliminate endemic diseases (measles and rubella in humans; foot and mouth disease in livestock) and to detect and rapidly respond to outbreaks while they are still small.

Early in the development of a surveillance and control program, infrastructure and data resolution may be limited. We use an adaptive approach to build models that are first tailored to the current data systems but can be used to develop evidence-based allocation strategies for surveillance-system improvements; e.g increasing resolution, introduction of new technologies.

Measles dynamics in Low and Middle countries

Measles virus still kills over 100,000 children each year worldwide. Attempts to eradicate the disease through mass vaccination are hampered by both logistical and epidemiological challenges. Interactions between demographic rates, seasonal drivers, and accessibility of preventive vaccination mean that measles persists in the populations most vulnerable to the consequences of infection.

In collaboration with World Health Organization, Medecins Sans Frontiers, US-CDC, and the Bill and Melinda Gates Foundation we are investigating local and regional dynamics of annual measles epidemics around the world. We have worked directly with ministries of health in China, Democratic Republic of Congo, Ethiopia, India Niger, Nigeria, Madagascar, Malawi, Pakistan, and Zambia to develop vaccination strategies to minimize mortality and morbidity due to measles. We are using time series analysis and epidemic models to investigate:

• The nature of the strong annual seasonality in incidence at the regional scale

• Local variation in the scale of measles outbreaks

View a recent seminar on this topic. 

Statistical methods for Evaluating Vaccination Programs

Disease incidence data are often gathered at spatial and temporal scales that are coarse relative to scales considered by quantitative epidemiological models of host-pathogen systems (e.g. case counts are generally reported over discrete time intervals, while many classic epidemic models employ differential calculus, which makes predictions in continuous time). Furthermore, observed data often suffer from incomplete reporting, imperfect diagnosis, measurement error and other biases. One of the great challenges in quantitative epidemiology is to develop statistical models that provide a coherent link between theory and data. For over 10 years we have developed methods used by the World Health Organization to estimate the global burden of measles mortality.

We analyze time series as partially observe Markov processes to both estimate the burden of infection and fit dynamical models of measles transmission. We use the resulting estimates to evaluate the performance of past vaccination programs and use the fitted models to project the potential impact of alternative vaccine strategies.

Scaling within-host immune dynamics to populations

The rapid clearance or long-term persistence of parasites within hosts isdetermined by the interaction of both parasite life-history characteristics and the immune response of the host to infection. Variation along this axis has implications for the rate of parasite shedding, the accumulation of transmissible stages in the environment, and the encounter rate and transmission rate in naive hosts. Thus, the host immune system is a critical regulator of the cycle of infection and transmission that determines large-scale patterns of parasite distribution andburden at the population scale.

I work with Dr. Isabella Cattadori to study the impact of interactions between worm life-history characteristics and host immune response on population-level transmission processes.  We combine lab-scale experiments in a rabbit/worm model with long-term temporal observations of worm burden and distribution in wild populations of rabbit to quantify the role of within host processes in determining population scale processes.

Dynamics of directly transmitted pathogens on host networks

I use simulation and analytical techniques to investigate how the spread of disease in social networks of hosts is affected by heterogeneities in contacts and local restrictions on transmission. These have important implications for the scaling of transmission across networks of different size and geometries — and can even lead to structural evolution of the network itself (as hosts are removed by mortality or acquired immunity).