Oral contraceptives and cancer.

Many of this Journal’s older readers will be familiar with the Oxford-Family Planning Association (Oxford-FPA) contraceptive study and will, indeed, have made important contributions to data collection. Accordingly, I was delighted that the Journal Editor had chosen the most recent publication from the study1 as the subject for a Journal Review. I am also grateful to Dr Mills for taking so much trouble to produce a succinct summary of a complex paper.2 There are, however, one or two points about the review to which I would like to draw readers’ attention. First, I would like to stress that the majority of the women in the study were followed up individually until mid-1994, although individual follow-up for a substantial subgroup of women ceased earlier than this. With regard to cancer registrations and death notifications, all women (save for those who emigrated) were followed up until the end of 2004 using information provided by the National Health Service Central Registries to supplement data collected during the course of individual follow-up. Second, the Journal Review does not include any confidence intervals for the rate ratios (no doubt in the interests of saving space). This is, perhaps, of concern mainly for invasive cervical cancer where the Oxford-FPA findings were considerably more unfavourable than has been described in most other studies. As the Oxford-FPA study included only 59 cases of this disease (with only six cases in the reference group who never used oral contraceptives), confidence intervals around the rate ratios were wide. Finally, while the population studied was certainly of higher social class than the general population, it was not “predominantly Social Class I”. The paper only gives the proportion of women from Social Classes I and II combined and this figure was 41%. These are relatively minor points that do not detract from the substance and conclusions of the careful review prepared by Dr Mills.