Another potential pill scare?
A number of newspapers have carried a report of research conducted at the Mayo clinic concerning adverse effects of the very early forms of the contraceptive pill. The piece in the Independent is fairly typical:
Pill linked to breast cancer in daughters. The oral contraceptive pill may increase the risk of breast cancer being passed down the generations, researchers have found. An American study of 426 families in which at least one member had been diagnosed with breast cancer has shown that the daughters and sisters of the women affected were three times as likely to develop the disease if they had taken the Pill than daughters and sisters who had not taken it.
The author of the piece, Jeremy Laurance, who is the Independent's health editor, contrasts this 'finding' with the view of Professor Beral of the Imperial Cancer Research Fund: "We looked at family history and whether the Pill had an effect [on breast-cancer risk in close relatives] and found it was exactly the same as for everyone else. Having a family history of breast cancer increases the risk by one and a half to two times [in close relatives] and current Pill use multiplies that by 20 per cent. It's a 20 per cent increase not a three-fold increase and that is known and based on firm evidence."
All of this sounds like good, balanced journalism, and it is, up to a point. The study, however, related only to early (pre 1975) forms of the contraceptive pill, which contained significantly higher levels of oestrogen than the current versions. This was not made terribly obvious in the American medical journal article JAMA, on which the report was based, and it was not mentioned in the Independent until the very last paragraph. The headline similarly failed to make this clear. The press release from the Mayo Clinic, however, where the researchers involved are based, was clearly titled Early oral contraceptive formulations linked to breast cancer risk. In the second paragraph they emphasised:
"If there is a risk associated with current formulations, we probably don't have the data to detect it," says Thomas Sellers, Ph.D., Mayo Clinic Cancer Center epidemiologist and senior investigator of this study. "What our data suggest is that it is early formulations with high doses of estrogen and progestins that pose a risk, not later ones. More importantly, these results don't apply for women at average risk for breast cancer, who should not interpret the study as reason to change their contraceptive practices."
The Mayo Clinic clearly anticipated the potential for another 'pill scare' and took steps in their release to avoid an inappropriate interpretation being made. British newpapers, unfortunately, were less sensitive.
The SIRC / RI Guidelines on the reporting of health and science issues specifically highlight the need to anticipate the effects of reports and emphasise the need to put appropriate caveats in articles at an early stage, rather than relegating them to little more than footnotes which many people will not even read. And that applies to medical journals as well as the more popular media.
We have already seen the damage that can result from unfounded scares about the contraceptive pill. (See, for example, The side effects of health warnings and Naming and Praising.) We should, therefore, be very aware of the need not generate another one.