36 Gordon Square
London WC1H 0PD

T +44 (0) 20 7958 8251

Preparing for a Flu Pandemic: Why Mathematical Modelling Can Help

The insights drawn from the mathematical modelling of pandemic influenza were revealed at an engaging presentation at LIDC. Dr Nim Pathy, of the Department of Zoology at the University of Oxford, referred to computer simulations which show how an aggressive use of the limited drug stockpile would minimise the number of flu cases. He also showed how such a strategy would delay the peak of an outbreak. The event on 26 November was the latest event in the Social Science of Infectious Disease seminar series run by LIDC and the London School of Economics and Political Science (LSE). Dr Pathy said: “We are approaching the same questions but with different methodologies and different languages. There is a lot that mathematical models can offer the social sciences, and vice versa”.

The cost of flu
The talk, entitled Antiviral Drug Stockpiles for Pandemic Influenza: What Can Epidemiological Modelling Offer Policymakers?, began with an explanation of the current impact of flu, also known as seasonal flu. Flu affects five to 15 per cent of the population and is responsible for 3-4,000 deaths every year in the

UK. It accounts for 10 per cent of sick leave in Europe and is estimated to cost the US economy $90 billion a year. Social and economic disruption, particularly regarding air travel, was also a key feature of SARS, which spread rapidly from China to countries worldwide during the outbreak of late 2002-2003.  

Evolution of flu and bird flu
Dr Pathy highlighted how every seasonal flu strain originates from one of the flu pandemics, namely the Spanish influenza (1918), Asian influenza (1957), Hong Kong influenza (1968) and Russian influenza (1977). He continued by charting the development of highly pathogenic H5N1, also known as bird flu. Wild birds act as a reservoir for the virus, and H5N1 has spread since 2003 when it became established in the poultry population in South-east Asia. It has jumped the species barrier from poultry to humans on 387 occasions and 245 people have died from the virus, but it cannot yet be efficiently transmitted (some transmission may already have occurred) from human to human. Whether future mutations of H5N1 will be passed from human to human, and how effectively, remains to be seen.

Controlling a pandemic
Antiviral drugs have been developed to counteract bird flu and the UK’s supply of Tamiflu – the preferred antiviral drug – is sufficient to treat 25 per cent of the population. This stockpile is the main weapon against a pandemic as it will take at least six months to develop a vaccine against a highly pathogenic human-to-human virus which currently does not exist.

Dr Pathy then showed how a series of pandemic scenarios would progress if Tamiflu was distributed at different rates. He showed how the quickest usage of the stockpile, although it could lead to its early depletion, would be the most effective strategy in mitigating the pandemic. He said: “Models always entail simplifications, but they can inform about transmission dynamics. By shortening the infectious period and reducing infectiousness, antiviral drugs can influence the course of infection in the community”.

The conversation which followed the presentation raised many questions and considerations about the usefulness of the mathematical model and other concerns related to a pandemic. The possibility of improving the mathematical model by incorporating social distancing was discussed, as well as the cost and political commitment required to increase the reserves of Tamiflu. There was agreement that evidence from ethnographic study, for example, could serve alongside modelling as equal, complementary strands of a coherent pandemic plan.
By Guy Collender, Communications Officer, LIDC

Further Reading

Antiviral Drug Stockpiles for Pandemic Influenza: What Can Epidemiological Modelling Offer Policymakers? , PowerPoint slides from presentation delivered by Dr Nim Pathy at LIDC on 26 November 2008