The H5N1 bird flu virus is changing. It emerged last week that it has acquired two mutations that suit it better to infecting human cells. It has not triggered a pandemic yet, and may never do, but these are the sort of developments we would see if the worst- case scenario were unfolding.

Britain is among the countries that are best prepared for this. The Government’s contingency plans have won international praise, and the decision to stockpile 14.6 million doses of the antiviral agent Tamiflu means that a first line of defence is in place.

The lack of room for complacency, though, was highlighted by this week’s report from the Royal Society and the Academy of Medical Sciences. It found that while the Tamiflu order is a necessary measure, it is not sufficient. There is a good case for buying a lot more, so it can be used preventively. And supplies of a second antiviral, Relenza, are also needed, as there are signs that H5N1 could become resistant to Tamiflu. An alternative weapon is essential.

Another way of protecting against a pandemic, of course, is vaccination. But as the report cautioned, this is fraught with difficulty. Production capacity is limited by demand for seasonal flu jabs, currently about 350 million a year. It would be difficult to cover more than a small fraction of the world’s 6.5 billion people.

On top of that, an effective vaccine can be designed only once the precise pandemic strain is known. Safety testing, regulatory approval and manufacturing mean a delay of seven to nine months before anyone can be immunised. Millions would be dead before the first injection is given.

The first problem will probably be impossible to solve. It is impractical for industry or governments to build vaccine factories and mothball them. The priority should be research into adjuvants, additives that get more response from vaccines at lower doses.

The second issue also looks intractable. But as Professor Nick White, a flu specialist at the Wellcome Trust, pointed out in the Natural History Museum’s annual science lecture, that is true only up to a point. Some manufacturing delays are inevitable, but weeks or even months might be saved by cutting corners on research.

Under normal circumstances, it is right and proper that vaccines are assessed rigorously and produced to the highest standards to prevent side effects. A pandemic, though, is not a normal circumstance. Professor White argued persuasively that trading off a little more risk for speed might be a bargain worth making. “Have we become too risk-averse to move quickly?” he asked. “We need to think as a society about underwriting scientists to run risks.” A saving of a month would not allow a vaccine to be used in the first phase of a pandemic, but it might make all the difference against its second wave. The vaccine might itself cause deaths, but this must be set against lives saved by speed. The balance of risks and benefits is not yet clear, but this is certainly a debate worth having.

It is not one that has much engaged the Government so far. As Professor White said, it would get fuller consideration if ministers were to accept another of the Royal Society’s recommendations: the appointment of an expert scientist as “flu czar”.

The Chief Medical Officer and the Chief Scientific Adviser have done fine work on pandemic preparedness, but both have significant other responsibilities. They cannot be expected either to be influenza specialists or to give the issue the time it warrants. The UK is fortunate in that many of the world’s leading authorities on pandemic flu are based here: Professor Neil Ferguson of Imperial College London and Sir John Skehel of the National Institute for Medical Research are two of them.

It is critical that their voices are heard at the highest level. Mark Henderson is Science Editor of The Times