This project aims to evaluate novel strategies for finding the few remaining cases in the end-game of the global lymphatic filariasis elimination campaign. It will take a data-driven modelling approach using GPS data from community disease surveys to simulate the whole population and strategies for finding infections.

The WHO 2030 roadmap for neglected tropical diseases (NTDs), lists thirteen diseases as targeted for eradication or elimination. However, as these campaigns make progress towards elimination, it becomes increasingly difficult to find the people with remaining infections. This is particularly true for diseases like lymphatic filariasis (LF), where infected people might not show the any of the debilitating symptoms that characterise the disease until years or decades after their first infection. Moreover, many NTDs are highly focal, with infected people clustered within villages, neighbourhoods, and households. This clustering of infection over space presents both a challenge and an opportunity; a challenge because even large surveys can fail to miss the locations with residual infections; an opportunity because once an infected person has been found we are likely to find more infected people if we look to people who are located near to them. Efforts to find remaining infections should therefore take into account and leverage this spatial clustering. However, field trials of spatial search strategies can be costly to implement and difficult to interpret if few or no infected persons are identified. Therefore, theoretical or computational tools are required to identify appropriate strategies and interpret the outcome of field trials.

In this project, the student will use data from community and school-based surveys for LF in Samoa and other countries with ongoing programmes to eliminate the disease. By combining the results of surveys for LF with information on demographics and spatial distribution of human populations, the student will generate a synthetic population of infected and non-infected people distributed across households and space. This synthetic population will then be used as a platform to compare alternative search strategies in terms of coverage (number of remaining infections identified) and effort (number of tests).

The project would be suited to a student with good quantitative and programming skills, such as those with a background in statistics, mathematics, computer science, or allied discipline. Familiarity with R and mapping software (e.g. ArcGIS) would be an asset for the student. The length and scope of the project could be varied to suit the needs of the student, e.g. a 6-unit PhB project, honours thesis, or as one component of a higher-degree research (PhD, MPhil) project.

Partnerships

This project will be conducted in conjunction with the ODeSI team based at the University of Queensland (http://bit.ly/46NKn40).