THE idea that bird flu will mutate or combine with a human flu virus, enabling it to spread from person to person, is now all too familiar. The result could be a pandemic of H5N1 flu that kills a large proportion of those it infects. Yet if we act promptly when the first cases are detected, the effects could be equally dramatic 鈥� the pandemic could be nipped in the bud. So say two international groups of researchers who have modelled the emergence of H5N1 flu and measures to treat it.
The groups used computers to model the network of contacts between people in Thailand. One study, led by Neil Ferguson of Imperial College London and published in Nature (DOI: 10.1038/nature04017), includes the entire Thai population. The other, published in Science (DOI: 10.1126/science1115717) and led by Ira Longini of Emory University in Atlanta, Georgia, focuses on the 500,000-strong rural population of the Nang Rong district. The models simulate contacts between people within households, schools and workplaces, for example, and the random encounters that take place when people travel and carry out everyday activities.
On the face of it, their findings are reassuring. The researchers try a variety of interventions, from treating infected people and their close contacts with oseltamivir (Tamiflu) to giving the drug to everyone in a neighbourhood, closing schools and workplaces, confining everyone to their homes and even setting up quarantine zones. The models show that different combinations of these techniques can quash an outbreak, with only a few hundred people ever being infected.
Advertisement
Yet these models come with their own health warnings. The predictions work only under certain assumptions, such as accurate detection of patients and their contacts, swift treatment with oseltamivir, and cooperation of local people with, for example, quarantine measures. But in the real world, a country may be reluctant to admit it has an outbreak, patients and their contacts will be missed and people will try to flee quarantine. Murphy鈥檚 Law says that the real outbreak will take place near a national frontier, so the health agencies of two countries will have to work together.
Nevertheless, modelling an outbreak before it happens is a step in the right direction. When foot and mouth disease appeared in the UK in 2001, tactics were initially made up ad hoc. With prior planning, a strategy of vaccinating animals in wide rings around sites of infection might have worked. But lack of foresight left the government with only one option, the dreadful slaughter of 6 million animals.
A critical factor in modelling epidemics is a virus鈥檚 鈥渂asic reproduction number鈥� 鈥� how many people each patient infects. The higher it is, the stronger any interventions need to be, but the number is never known until an outbreak begins. Once it is known, however, it can be plugged into the models so that public health planners can see exactly what they are up against.
鈥�3 million courses of oseltamivir would be needed to halt an outbreak. The WHO has 120,000鈥�
Ideally, of course, we would have a vaccine to stop the flu doing its worst. Although a specific vaccine against the pandemic virus cannot be made until the strain emerges, several generic vaccines against H5 strains are in the pipeline. Longini鈥檚 team found that even giving people a relatively ineffective vaccine before an outbreak would help to stop or delay the virus鈥檚 spread. Yet vaccinating even the most at-risk populations in south-east Asia would be a logistical and financial nightmare.
And it gets worse. This week, bird flu reached Russia, and Kazakhstan reported its first suspected case of human infection, so the area over which bird flu could 鈥済o human鈥� is growing fast. In this case, it looks as though drugs may be the best option until a specific vaccine can be made.
Ferguson calculates that 3 million courses of oseltamivir would be needed to halt an outbreak. Longini, working with a smaller area and a virus with a lower basic reproduction number, puts the figure at about 1 million. Thailand, South Korea, Hong Kong and Singapore all have orders in for oseltamivir, though how much is not known. That leaves huge gaps in the other countries where bird flu has been detected.
The WHO has only 120,000 courses, which is worrying. According to the models published this week, these stocks will stop nothing. If the outbreak is not contained at source, the entire world will find itself fighting the virus at immense cost in human lives.
