The United Kingdom hosted a pandemic influenza modeling workshop in London in 2005. This meeting brought together modelers from all of the Global Health Security Initiative (GHSI) member states and the risk managers from the Risk Management and Communications Working Group. This workshop forms part of the extensive international work being carried out by the Pandemic Influenza Working Group (PIWG) of the GHSI.

The next influenza pandemic, when it begins, will pose considerable challenges to public health services in delivering our response plans. The situation in south-east Asia – and in particular the apparent changes in the virus recently observed - sends a clear signal that response plans should be completed as soon as possible.

Much work has been accomplished: the World Health Organization’s pandemic influenza plan has been revised and our own individual country plans have been updated and tested. The PIWG has been very active in determining the requirements for vaccines and antivirals.

Anticipating what might happen is the key to robust plans. Modeling is playing a key role in determining the range of likely trajectories of the next pandemic, both globally and nationally. Predictions of speed of spread and possible attack rates provide the information needed to have our capabilities and response assets fit for purpose. Modeling cannot precisely predict the exact details of the next pandemic but utilizing information about previous pandemics, international travel and transmission dynamics modeling can provide scenarios in which our medical countermeasures and control of infection countermeasures can be tested to provide the best possible defense to limit the effect on public health.

International collaboration is a component of the global response and the international media will certainly compare the responses in each of our countries. This workshop therefore allowed us to try and answer key issues facing the response effort.

1. Control options within country of origin

What chance is there that intensive use of antivirals combined with measures to increase social distance might contain a pandemic in its earliest stages before widespread geographic spread beyond the source area (e.g. South East Asia) has occurred?

If ultimately unsuccessful, what delay to global dissemination might such policies provide?

2. International spread and concerted control options

How quickly, and with what geographic pattern, would an uncontrolled pandemic be expected to spread globally?

What would be the effect on transmission of restricting travel (and exit and entry screening) between affected and unaffected countries, and how effective would such measures need to be to delay spread significantly?

3. National spread and public health mitigation and control options

Once the pandemic reaches our respective countries, our individual response efforts will clearly shift from trying to slow international spread and prevent entry to mitigation of the impact of the pandemic on the population and economy of our nations. Key issues remain to be addressed concerning the optimal manner in which to apply limited resources in at a national level.

Could antivirals be used in the early stages of a national epidemic to prevent or significantly slow the spread, and how might antivirals be optimally targeted to achieve this? Given a limited antiviral stockpile, what is the optimal strategy for using antivirals to limit mortality, morbidity and/or hospitalizations in (a) the general population, and (b) essential workers (e.g. healthcare, emergency services)?

How does the total effectiveness (in terms of reductions in transmission, morbidity or mortality) of a policy of antiviral use scale with the size of the stockpile available (e.g. equivalent of treatment courses for 1, 5, 10, 25% of the population), and does the ‘optimal’ policy change as a function of the number of doses stockpiled?

Data needs and real-time modeling

At a global, regional and national scale there will be a need to track the progress of the pandemic, analyze patterns of spread, and to use models to predict future short- and medium-term trends. The data needs of real-time epidemiological analysis and modeling need careful consideration: the desire for detailed statistics needs to be balanced against the difficulty of implementing real-time national and international surveillance and data collation systems in the heat of a pandemic which is almost certain to overwhelm healthcare systems’ standard surge capacity.

Making the best use of the limited subset of data that is likely to be available will therefore be key. Consensus would therefore be desirable on the most appropriate statistical and mathematical methods for real-time analysis during a pandemic.

The papers presented here give a key insight into these questions and provide modeling information on which risk managers can plan their responses.

1. Strategies for mitigating an influenza pandemic ( N. Ferguson, et. al.)
2. Delaying the International Spread of Pandemic Influenza (WJ Edmunds, et. al.)
3. Network models for influenza transmission and control (M. Schwehm, et. al.)
4. Transmission parameters and real-time estimation methods (S. Cauchemez, et. al.)
5. Real Time Modeling of Pandemic Influenza (Gani, et. al.)
6. The use of antivirals to reduce morbidity from pandemic influenza (Gani, et. al.)
7. Estimates of the Basic Reproductive Number for 1918 Pandemic Influenza in the United States (Lipsitch, et. al.)