Agent-based Modelling for Ebola Risk Reduction: Understanding and Mitigating the Potential for Global Transportation-fuelled Epidemics (ABMERR)

Outbreaks of Ebola have gripped the world's attention in recent years, from the Western Africa Ebola epidemic of 2013-2016 to the ongoing outbreak in Democratic Republic of Congo (DRC). The first outbreak of the Ebola virus was recorded in 1976 in what was then called Zaire (now the Democratic Republic of Congo); 318 cases were identified, of which 280 people died. Since that time, changes to infrastructure, travel patterns, population density, and climate have led to larger and more frequent epidemic outbreaks of the disease. The Western Africa Ebola epidemic saw cases spread as far as Italy, Spain, and the United States. The outbreak in DRC is now in its 11th month with no signs of abating and extensive disruptions and cost for local people. Humanitarian groups including Médecins Sans Frontières (MSF) and the British Red Cross (BRC) are organising local responses to the crisis, but the magnitude of the problem is so great that there are no established techniques for resolving it. Humanitarians are in urgent need of operational intelligence in order to fight increasingly large epidemics.
To that end, this fellowship aims to develop a simulation of the spread of Ebola through a region as a function of human behaviour and movement. I will undertake this work in conjunction with partners from MSF's Manson Unit for Innovation and BRC's GIS team. The project will create and make use of an agent based model (ABM) to simulate the interactions and choices of individual people living, working, and moving within Ebola-effected areas. By capturing the ways in which Ebola is transmitted between people through funerary practices, caring relationships, and the seeking of medical care, the simulation will help the project's humanitarian partners understand how different interventions may influence the spread of disease in these social contexts. The research will particularly focus on analysing and visualising the results in ways that are accessible to the humanitarian project partners.
The work will make use of open source data to estimate populations and recreate transportation networks. These synthetically generated datasets will be fed into the ABM, giving more insight into where outbreaks may flare up and where they may travel next based on appropriate local behaviours and choices. In turn, the software developed as part of the project will be made open source. By making the simulation software freely available, the research will be accessible to the humanitarian partners, but also government organisations and other researchers. The toolkits will be developed with the needs of non-specialists in mind, so that they are suitable for users with potentially sensitive datasets and without high-powered machines or constant access to the internet. The project aims to strengthen the ties between academics and humanitarians, and to make cutting edge methodologies available to non-specialists.
The fellowship will provide resources to help humanitarians and government to understand their options in the fight against Ebola. It will also make these new tools available to other researchers, propagating outward and facilitating new work. Finally, it will allow me to establish myself as an internationally recognised leader in the field of ABM for humanitarian modelling, allowing me to push this line of research forward in the future.

By investing in British research capacity in agent-based modelling (ABM), the project would deliver substantial academic and social benefits across a range of disciplines and timescales.

ACADEMIC: The development of the simulation proposed here will have tangible benefits for a number of different parties, including:
Agent-based modellers: ABM researchers will benefit from the development of cutting edge behavioural frameworks, extensive engagement with simulation exploration, and publication of toolchains for research. The focus on open-source research will make the code developed immediately available to other researchers to extend for their own purposes. The state of the art of agent-based modelling will be pushed forward by this work. This funding of basic research excellence will cascade long-term.

Data scientists: The tools and techniques developed to populate the model with agents and transportation networks will be useful to simulators working with synthetic populations or transportation in underserved areas. Again, the open source nature of the project means that component parts can be spun off and used for distinct purposes with ease.

Public health researchers: Public health researchers and medical scientists often grapple with questions of the efficiency versus effectiveness of various interventions - the theoretical impact an intervention might have versus the reality in situ. By capturing the choices made by individuals within the epidemic context, public health researchers can better understand the options available to them.

Postdoctoral researchers: The two postdoctoral researchers funded as part of this project will further their training and be supported in their development into independent researchers. Their skills will give rise to the next generation of proposals in UK research. Their ability to develop collaborative research programmes with humanitarian partners will be enhanced by this experience.

SOCIAL: The project would serve the needs of both organisations and communities more broadly. In particular:

Humanitarians and governments: MSF and BRC are dealing with an outbreak of Ebola in the Democratic Republic of Congo, currently in its 11th month. This has led to a total of 1287 deaths confirmed by the DRC Ministry of Health at the time of submission. They also dealt with the Western Africa Ebola epidemic of 2013-2016, in which 11316 people were killed out of a total of 28639 cases reported by the WHO. The economic costs were staggering: World Bank estimates that $2.2 billion USD was lost in 2015 during the Western Africa Ebola epidemic, while the US, UK, and Germany spent more than $3.6 billion USD to help fight the outbreak. The scale and complexity of the problem are without precedent, and belie top-down solutions. By helping humanitarians and governments to target their interventions, the project aims to decrease the suffering associated with epidemic disease. The workshop and final symposium will facilitate this. The intelligence generated by this project is designed to be maximally accessible to nonspecialists, supporting government and organisations to realise and exploit the potential of ABM. Better informed policy can save lives and allow health spending to be diverted to the other issues which face these communities.

Individuals living in Ebola-affected areas: The use of the project findings are intended to shorten and proscribe the impact of epidemics; this will impact not only the individuals who are now not exposed to disease, but also those whose livelihoods depend on free movement. Individuals who are suspected to have been exposed to the disease are often quarantined for 48 hours, which can massively impact the wellbeing of families dependent on daily employment. Ending epidemics sooner will impact the health of individuals not only by decreasing the number of fatalities from Ebola, but by decreasing disruption and its associated dangers to their welfare.