mv

The project has six research objectives which will be achieved through collaboration by the ten participants:

1. To improve characterisation of population contact and travel patterns in models. This will involve a number of strands, including (a) using detailed individual-level information from population surveys of contact patterns to build more realistic representations of contact networks relevant for influenza transmission; (b) to use such data to quantify heterogeneity in contacts (all the way up to potential ‘behavioural superspreaders’); (c) to update and/or obtain data on population density, age-structure, household composition and school/workplace/hospital sizes of all EU member states; (d) to collect data on population movements within and between individual Members States, both regular and other journeys where such data exist; (e) to collect and integrate data on contacts during minority activities (i.e. leisure, shops, restaurants/bars).
2. To evaluate behavioural responses to epidemics and social acceptance of restriction measures. It is likely that people change their behaviour and contact patterns in response to epidemics, particularly lethal ones. This may modify model predictions, since behavioural changes in people at risk are not generally assumed by existing models. Experience from the 1918 epidemic indicates that characterizing such changes may prove critical for improving model realism. Spontaneous behaviour change by the population which reduces transmission risk in a pandemic may enhance the impact of organised control measures. However, if such behaviour change occurred in past pandemics, current estimates of pandemic transmissibility may be over-optimistic, as those estimates implicitly assume no behaviour change occurred during past pandemics. We will collate data from past epidemics (e.g. SARS) and pandemics to quantify the potential effect of spontaneous behaviour changes in a pandemic, and undertake social research to gather data on how people think they would behave in a pandemic, and their likely acceptance of potentially intrusive and disruptive control measures.
3. To develop a suite of models for the spatio-temporal spread of a new influenza pandemic. We will develop two core model types: compartmental models (with and without spatial structure) and individual-based simulations. The different modelling approaches have dual merits: on the one hand, simulations can reproduce in detail individual behaviour, and easily accommodate data from behavioural surveys but this is at the price of making validation and parameter estimation challenging; on the other hand, differential equation models with few parameters can be fitted, with standard statistical methods, to existing data on actual epidemics (such as seasonal flu) – but at the cost of ignoring important and relevant biological and population detail. Integrating the two approaches will have the benefit of being able to obtain broad-brush estimates of the rate and pattern of global spread of a pandemic, while using more sophisticated simulations for modelling of pandemic spread and control at smaller scales.
4. To estimate model parameters and test model adequacy using data on seasonal flu and endemic diseases. Key to the potential utility of models is the quality of the input parameters, and the reliability of model assumptions. Hence, our goal is (a) to obtain improved estimates of key epidemiological parameters from the analysis of past epidemics (both pandemics and seasonal influenza) and to demonstrate the adequacy of models of population movement to reproduce the spatiotemporal patterns of spread of past epidemics; (b) to design protocols for future epidemiological studies to measure key unknown parameters relating to the natural history and transmission of influenza and the effectiveness of non-pharmaceutical interventions; (c) to develop parameter estimation methods further, with emphasis on real-time estimation from outbreak data.
5. To evaluate the impact of intervention options for containing and mitigating a pandemic influenza outbreak. We will analyse several intervention options, including those outlined in the call, to assess their potential impact on limiting spread and/or disease during a pandemic. A range of different assumptions regarding the biological and epidemiological characteristics of the pandemic virus will be explored. This work will contribute to pandemic planning by identifying effective combinations of control measures (both medical and non-pharmaceutical), giving insight into optimal deployment of resources and providing estimates of the logistical requirements for different policy options. In addition, methods for refining pandemic response strategies in real-time and assessing the impact of interventions during a pandemic will be developed.
6. To develop efficient, extensible and usable individual-based simulation models. Individual-based simulation models are increasingly the model of choice for pandemic planning, and in future may be used – in conjunction with real-time data analysis – for prediction and to refine control policies in the face of an outbreak. Key to the effective deployment of such models is efficiency, extensibility and ease of use. Efficiency is a trade-off between RAM memory consumption, depending on the scale of the simulation (regional, national, continental or world-wide), and the computational cost of the algorithms. This WP will develop innovative algorithms for reducing the computational demands of spatial simulation models and giving optimal performance of single compute nodes with multiple cores and moderate RAM. Extensibility will be achieved by developing the first truly modular simulation code with a clear and simple object schema. Ease of use will be achieved by integrating simulation engines with databases to store parameters and output, and with GIS and statistical systems for clear visualisation of output.

The spread of influenza A(H1N1)v made the FluModCont research very topical. However, the emergency has changed the project timeline and is forcing the partners to use methods and models still in development and not fully validated. 
The evaluation of behavioural responses and social acceptance of restriction measures has become focused on the present influenza A(H1N1)v spread. The main focus of WP4 is now the real time parameter estimation from collected data on the influenza A(H1N1)v, more than the analysis of data from past epidemics of seasonal influenzaor from other endemic infections. Similarly, we are now developing simulation models that work now, while leaving the objective of WP6 of obtaining efficient, extensible and easy to use software to a later stage. The development of WP5 is occurring under the guidance of public health authorities, and the measures analysed will certainly be related to the severity of the disease.