The extent of the threat from the H5N1 influenza strain has been underlined this week by news from Indonesia. Two deaths have brought the country’s toll from avian flu to six and the region’s total to 65. Another 54 cases are suspected.

The virus remains exceptionally lethal, killing more than half of those it infects. The devastating Spanish flu was fatal only about 5 per cent of the time. If the H5N1 strain mutates so that it can spread readily from person to person, a comparable pandemic is likely. Even if it does not, another strain will oblige before long: the world is overdue another virulent bout of influenza.

Britain has thus been wise to order in the Tamiflu, which reduces symptoms and prevents onward transmission. But even if the UK stockpile were big enough to provide for every citizen, which it is not, it cannot be the only or even the main line of defence. Flu is too devious an opponent for that.

If a strain of H5N1 flu starts to spread among people in South-East Asia, a race against time will begin. According to computer models from Imperial College London and Emory University in Atlanta, the authorities will have about three weeks from the first reported cases to contain an outbreak.

Should they fail within this period, the game will be up. The proliferation of the virus might subsequently be slowed but it will not be stopped. Even drastic quarantine and suspension of air travel would not prevent it spreading around the world. Countries like Britain, with good stocks of antivirals, might have reduced death tolls, though it is possible that a slowly spreading H5N1 strain might evolve resistance to Tamiflu. But it would still stretch into the tens, if not the hundreds, of thousands.

It is therefore essential that measures are established to ensure that any cluster of human cases in South-East Asia is rapidly detected and contained. There is some welcome movement in this direction; Thailand has spent £100 million on bird flu and now boasts a surveillance system in which any severe respiratory infection is instantly reported to health authorities. Roche, which makes Tamiflu, has also donated three million courses to a World Health Organisation “mobile stockpile” that can be deployed anywhere within three days.

In other Asian countries where a pandemic could begin, however, surveillance remains patchy, and rich Western nations are doing too little to help. The UN Food and Agriculture Organisation announced this week that governments have so far pledged just $20 million (£11 million) of the $150 million needed for monitoring and containing the disease. Britain is not among the donors confirmed so far.

This is not good enough from countries that can afford to provide assistance, and neither is it in their best interests. By providing the funds and technical support needed to track bird flu in Asia, the developed world will in the long run be protecting itself.

This is the only way in which a pandemic might potentially be prevented: everything else is damage limitation.

An antiviral stockpile is necessary but there is a danger that it could encourage a false sense of security, much as the Maginot Line did for the French in 1939. The flu virus will always get through, unless it is tackled at source. Helping the countries in the line of fire to do just that should now be an urgent priority, for the sake of public health at home as well as abroad.

Mark Henderson is the Times science correspondent

Times Recommends

• DIY video helps the WI get to grips with sex
• Artists urge nation to save Titians
• Time to go home, Iraqi leader tells Britain