The current threat of swine flu doesn't justify a gamble on a vaccine that has not been fully tested.

A mass vaccination programme moves ever closer. Orders have been placed; priority groups identified. There will be enough swine flu vaccine to inoculate the entire population, starting with NHS staff, in an attempt to halt the spread of the disease and save lives.

Is all this really necessary? To start with, swine flu is far milder than we first feared, so the case for vaccinating millions of healthy adults against a disease that is no more unpleasant than a bad cold is questionable. There is a stronger argument for vaccinating those at greater risk, such as those with lung, heart or kidney disease, those with suppressed immune systems (such as those on cancer treatment), pregnant women and children under 5 - but only if the vaccine works and is safe. But there are serious doubts about this.

Rushing the vaccine on to the market means we will have no idea how effective it is, although we do have a body of research on the effectiveness of flu vaccines in general, which gives some idea of what we might expect from the swine flu vaccine. Provided that we have matched the vaccine well with the virus, it is likely to be up to 80 per cent effective in healthy adults, the group at least risk from the virus.

A number of trials have looked at the effect of flu vaccination on children's asthma and have failed to demonstrate any benefit; one trial even suggested that the vaccine made asthma worse. There is no good evidence that the vaccine helps those with chronic health problems or pregnant women. However, we do know that the immunisation offers no more than a modest benefit in the elderly; indeed, the effectiveness of the vaccine is known to decrease sharply after 70 years of age.

The first vaccines are expected to arrive in the UK by the end of next month. It will be some weeks later before they have gone through the minimal safety testing necessary to consider offering them to the general population. Realistically, it is unlikely that sufficient doses will arrive to vaccinate substantial numbers until the end of the year.

Even putting aside the daunting logistical problems of administering millions - or even tens of millions - of vaccines over a short period of time (everyone needs two doses), it is quite possible that most of the population will have come across the virus by then and developed natural immunity (always stronger and longer-lasting than vaccine-derived immunity) and so be in no need of the vaccine.

Perhaps the biggest concern is the speed at which the vaccine is being rushed out. Research for my book, The Truth about Vaccines, taught me how vaccines are increasingly being released on to the market with little testing of either safety or effectiveness, against infections that are rarely the threat that the Department of Health or pharmaceutical companies (who are finding the vaccine business an increasingly lucrative market) claim.

To be properly tested for safety a vaccine needs to be given to tens of thousands of people and followed up for several months to detect uncommon but serious side-effects. This is clearly not going to happen with the swine flu vaccine, which is being fast-tracked at unprecedented speed. The little safety testing that does occur is likely to be in healthy people, and not those with health problems who are in greatest need of the vaccine, but probably also at greatest risk from side-effects.

We have experience of mass vaccination against swine flu from which lessons should be learnt. In America in 1976 a vaccine was offered to the whole population to prevent the spread of an epidemic of swine flu. Millions were rapidly immunised, but the vaccination campaign was stopped after a rise in cases of Guillain-Barrรฉ syndrome (GBS) among recipients of the vaccine. GBS is an autoimmune disorder that causes paralysis of the arms or legs or, rarely, the whole body; the sufferer usually makes a complete recovery, but some suffer permanent paralysis and a few die. Research later estimated that there was one case of GBS caused by every 100,000 swine flu vaccines given. If the current vaccine caused a similar rate of cases, then we could expect hundreds of people to get GBS, some of whom will suffer permanent paralysis or die.

Vaccinating a large proportion of the UK population with an "experimental" swine flu vaccine will be a huge gamble. It may save lives and, more likely, prevent healthy adults from having to take a few days off work. It may also cause serious side-effects and deaths. It is a gamble that the current threat of the virus does not justify.

There is a case for offering the vaccine to those at risk, but not without informing them of the uncertainty over the vaccine's safety and effectiveness. To vaccinate the whole population would be a huge and foolhardy experiment for which there is currently no scientific rationale.

Dr Richard Halvorsen is a Central London GP and medical director of BabyJabs, a children's immunisation service. His book The Truth about Vaccines is published on August 7