The Department of Health announced yesterday that women who take certain newer brands of the contraceptive pill - the grimly named Femodene, Minulet, Triadene, Tri-Minulet, Marvelon and Mercilon - should go back to see their doctor. They used to think that these low-dose pills, which combine both oestrogen and two particular types of the hormone progesterone, reduced the risk of blood clots in the veins (compared to the older oestrogen- only varieties). It turns out that they were wrong. For this particular side-effect, the older ones are better after all.
Grounds for switching pills? Probably. After all, if you are going to take a drug for years, you might as well take the one that is marginally safer. Grounds for forgetting that pill tonight? Definitely not.
So what is this new risk? For the one and a half million women taking the newer, low-dose combined pills at the moment, about 450 are likely to get blood clots in their veins each year. Of those 450 women, about nine are likely to die from them.
If they all switched to other brands, such as the older oestrogen-only pill, the number contracting thrombosis would fall to around 250, and deaths would be reduced to five. That is four lives saved, and 200 women spared a lot of discomfort. But bear in mind that numerous women take the pill, so these risks are relatively small. And even if all of these women came off the pill completely, there would still be around 75 cases a year, and one or two deaths.
Even more important, if they all stopped taking the pill tomorrow and became pregnant instead, their chance of getting thrombosis would double.
Given the nature of the figures, there are certainly no grounds for panic. And it is unclear whether everyone should switch to other pills. Many women have reactions to the older pills. And the relative risks of other side-effects, such as heart attack or breast cancer, are still unknown.
Which raises another question: why do we not know more about the various side-effects yet? Thrombosis was first raised as a concern 35 years ago. In this latest case the answer may lie partly in the fact that it's hard to study low-risk side-effects until there are large numbers of women taking the pills, and enough of them suffering from thrombosis. It is also true that post-sales research into the pill has been more extensive than into any other drug. But when you consider that so many millions of women take this drug, that they are healthy rather than sick and that they take it for such long periods - contraceptive pills may merit even more research than they already attract.
In the end, millions of women will carry on taking the pill. And they are right. The risks remain overwhelmingly outweighed by the benefits of choosing when to have a baby.