Estimating the impact of social structure on epidemics and predicting the impact of targeted interventions

Lead Research Organisation: Imperial College London
Department Name: School of Public Health


Communicable diseases like flu or SARS transmit through close inter-personal contacts. So, the risk that you get ill during an outbreak is likely to be influenced by who you mix with. Understanding how people interact with each other may therefore be the key to designing efficient control policies. Examples of public health interventions which are triggered by the structure of the social network include those which target households (for example, all household members are treated with antivirals such as Tamiflu when a member is sick) or specific age groups. Consider for example vaccination against seasonal influenza. A first strategy is to vaccinate the elderly as they constitute the main risk group for severe disease and mortality. However, vaccination of children has been suggested as a better policy to minimise mortality overall, since children are the most important transmitters of flu. To assess whether vaccinating children or the elderly is likely to be more effective, it is important to precisely assess how children interact with each other and with other age groups.

If you had flu, would you be able to say where and by whom you got infected? The frequent difficulty in answering this question is what makes it so challenging to determine ?who acquires infection from whom? and thus assess the effectiveness of the different public health strategies. A first aim of this study will be to develop a set of relatively simple mathematical and statistical tools that can be used to make such an evaluation. On an operational level, those tools will make it possible to gain insight on the potential impact of interventions, to monitor the efficacy of control measures and to support decision making in real-time during an outbreak. This work could for example feed the UK surveillance system designed for pandemic influenza; and should increase the capacity of politicians and other decision makers to make the correct choices at the appropriate moments.

By comparing the route of transmission (?who acquires infection from whom?) for different infections and to other sociological indicators (?who talks to whom?), we will also try to answer fundamental questions on the complex nature of transmission. For example, what types of social contact can lead to transmission? Talking? Hugging? Answers to those seemingly simple questions may have major implications in terms of future disease prevention.

Technical Summary

Quantitative epidemiology is now expected to have a key role in preparing for and responding to novel infectious disease outbreaks. The first aim of this project is to develop new statistical and mathematical modelling methods for use in responding to an emerging epidemic of a directly transmitted pathogen, in a situation where little is known about key epidemiological determinants of spread, and thus where analysis must be responsive and rapid. Recent examples of such a situation include the 2003 SARS epidemic or the 2001 UK foot-and-mouth virus epidemic, and the current focus is on highly pathogenic avian influenza or deliberately release smallpox.

In these examples, a key feature of epidemic spread is granularity of spread linked to structure in the host population. Demographic and social data are routinely recorded during epidemic surveillance in an outbreak, and the aim here is to ensure that best use of made of those data in the public health response to outbreaks, and to improve data collection protocols where possible. Specifically, the analysis will focus on incorporating household and age structure in analyzing outbreaks of human pathogens such as SARS or influenza.

Models which guide the public health response will be developed and assessed. Public health response such as isolation, quarantine or drug treatment will commonly target households rather than individuals as a way of extending the reach of interventions to high risk individuals. Age is also an important risk factor for many infections, and similarly interventions such as vaccination can be targeted at specific age groups. The new suite of methods developed here will inform these detailed policy options in real-time as an epidemic unfolds.

The second aim of the project, closely allied to the first, is to advance the state of understanding of these models and of their key parameters based on extensive literature searches for data. The aim is to compare different infections with similar routes of transmission but differences in natural history and immunity, and thus to disentangle the effects of social structure and host-pathogen biology on transmission and spread. This feeds into the first aim in that these newly inferred parameters will inform model and parameter choice in the epidemiological response to a new pathogen.

The project will involve a multi-disciplinary team and will employ a statistician and a mathematical modeller who will work closely together, and thus who will both acquire skills which underpin this dynamic research area.


10 25 50

publication icon
Cauchemez S (2016) Unraveling the drivers of MERS-CoV transmission. in Proceedings of the National Academy of Sciences of the United States of America

publication icon
Cauchemez S (2012) Methods to infer transmission risk factors in complex outbreak data. in Journal of the Royal Society, Interface

Description Contribution to Ebola response
Geographic Reach Multiple continents/international 
Policy Influence Type Implementation circular/rapid advice/letter to e.g. Ministry of Health
Impact The methods and software developed during this grant have recently been used to provide reports to WHO, US CDC, UK Govt and various international stakeholders on the ongoing Ebola epidemic. Some of this work has also been reported in the NEJM by a WHO collaborative team including a former postdoc employed on this grant. This work has served as part of the motivation for increased international responses in summer 2014 to the epidemic.
Description Contribution to H1N1 pandemic response
Geographic Reach Multiple continents/international 
Policy Influence Type Participation in a national consultation
Impact Staff working on this project (recruited approx 6 months ago) have analysed winter H1N1 pandemic in Southern Hemisphere countries, results of which have been communicated to UK Department of Health and WHO committees via other members of the MRC Centre for Outbreaks at Imperial
Description Contribution to WHO HIV testing and treatment guidelines
Geographic Reach Multiple continents/international 
Policy Influence Type Membership of a guidance committee
Impact The change in treatment guidelines took place in 2013. The PI and one postdoc gave advice and were named in the WHO guidelines document. The new guidelines resulted in a much number larger people being eligible for antiretroviral treatment, globally, and these changes are being rolled out globally.
Description Short course in infectious disease epidemiology
Geographic Reach Multiple continents/international 
Policy Influence Type Influenced training of practitioners or researchers
Impact The PI, all of the co-Is, and the postdoctoral fellows, teach on a short course in infectious disease epidemiology delivered annually. The course introduces important methods and concepts to a wide range of infectious disease practitioners, ranging from very junior to extremely senior (heads of units in public health agencies), and some extremely influential (a nobel laureate attended in 2011). The course is directed by Prof Christophe Fraser.
Description PANGEA-HIV: a consortium for HIV phylogenetics in sub-Saharan Africa
Amount $6,000,000 (USD)
Organisation Bill and Melinda Gates Foundation 
Sector Charity/Non Profit
Country United States
Start 11/2013 
End 10/2017
Description PopART - HPTN071 - Population Antiretroviral Therapy for HIV Prevention
Amount $1,000,000 (USD)
Organisation National Institutes of Health (NIH) 
Sector Public
Country United States
Start 08/2012 
End 08/2017
Title New software 
Description A new software package for analysis of epidemics in real time has been created, and released as both an easy to use Excel 'macro' and as an R package. 
Type Of Material Technology assay or reagent 
Year Produced 2012 
Provided To Others? Yes  
Impact Only just released so too early to tell. 
Description Collaboration with CDC, Atlanta USA 
Organisation Centers for Disease Control and Prevention (CDC)
Department Influenza Division
Country United States 
Sector Public 
PI Contribution We will be modelling the surveillance response to the 2009 and future pandemics, and in particular working on optimising the estimation of disease severity early on during a pandemic. We have met face-to-face several times, and are in the process of model development.
Collaborator Contribution Our collaborators at the CDC will provide intellectual input into models of surveillance during a pandemic, and will provide epidemiological and virological surveillance data into a range of pathogens.
Impact Nothing yet
Start Year 2009
Description Health Protection Agency, UK 
Organisation Public Health England
Department Centre of Infectious Disease Surveillance and Control
Country United Kingdom 
Sector Public 
PI Contribution Together with partners at the HPA and the CDC, we have agreed to model different surveillance strategies during an early pandemic to help determine disease severity.
Collaborator Contribution Our partners at the HPA will provide intellectual input into the formulation of a mathematical model of disease surveillance during the early phases of a pandemic.
Impact None yet
Start Year 2010
Description Wellcome Trust Sanger Institute 
Organisation The Wellcome Trust Sanger Institute
Country United Kingdom 
Sector Charity/Non Profit 
PI Contribution Modelling and analysis
Collaborator Contribution Sequencing and analysis
Impact Papers, grant applications
Start Year 2010
Description School project 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Type Of Presentation Workshop Facilitator
Geographic Reach Local
Primary Audience Schools
Results and Impact Activity for 60 year 6 primary school children on antibiotic resistance.

I think my son was not too embarrassed. The other children enjoyed it, and hopefully were inspired.
Year(s) Of Engagement Activity 2012