Library

Notes: Summary. Mortality from epithelial ovarian cancer is falling in women under 55 years of age in England and Wales. The decline does not appear to be a treatment effect nor to be attributable to changes in the rate of oophorectomy. Case-control studies have shown that high parity and oral contraceptive use are protective against the disease. We suggest that the decrease in mortality is compatible in timing and magnitude with exposure to oral contraceptives. No obvious effect on mortality attributable to parity was apparent in this analysis. Oral contraceptives may prove to be a widely acceptable means of preventing ovarian cancer, providing they do not increase breast cancer risk.

Notes: Summary. Up to the end of 1989, 206 parous women in the Oxford Family Planning Association contraceptive study had been referred to hospital with a first episode of pelvic inflammatory disease. Of these, 65 suffered from definite disease described as acute, 81 from definite disease not described as acute and 60 from ‘other disease’. Considering all forms of disease together, referral was less common in those aged 25–29 and in those aged 45 or more than in those aged 30–44. Referral was more common in those of low social class, in those who smoked and in those who married young. All these factors were taken into account in analyses considering the effects of contraceptive methods. In these analyses, women currently using the contraceptive pill, the diaphragm, the sheath, female sterilization or an intrauterine device (IUCD) were compared with those currently using other methods or no method of contraception. IUCD ex-users were, however, placed in a separate category, irrespective of their current method of contraception. The relative risks obtained in these analyses, with 95% CI, were as follows: contraceptive pill 0.5 (0.2-0.9), diaphragm 0.6 (0.3-1.2), sheath 1.2 (0.6-2.4), female sterilization 0.7 (0.3-1.5), non-medicated IUCD 3.3 (2.3-5.0), medicated IUCD 1.8 (0.8-4.0), IUCD ex-users 1.3 (0.7-2.3). These data suggest that oral contraceptives, the diaphragm and female sterilization protect against pelvic inflammatory disease and that IUCDs increase the risk. Medicated devices, however, appear to carry only about half the risk of non-medicated devices, and the elevation of risk in IUCD ex-users appears to be small. Special analyses examined the risk associated with use of a Dalkon Shield. Among women currently using an IUCD (of any kind), those who had used a Dalkon Shield (at any time) had nearly five times as great a risk of hospital referral for pelvic inflammatory disease as those who had never used a Dalkon Shield (relative risk 4.7, 95% CI 2.1-9.0).

Notes: Objective To examine patterns of hysterectomy in the Oxford-Family Planning Association (Oxford-FPA) study in relation to age, parity, social class and calendar period (–1974, 1975–79. 1980–84, 1985–89).Design The Oxford-FPA study is a large scale prospective study of 17 032 women recruited from 1968–74 and still under observation.Setting Seventeen family planning centres throughout England and Scotland.Subjects At recruitment the 17 032 women were all white, British, married, aged 25–39 and willing to co-operate. In addition, they were using the pill or an intrauterine device or a diaphragm as their method of contraception. Main outcome measure Hysterectomy rates per 1000 woman-years of observation in various subclasses of the data.Results Up to the end of 1989, 1885 (11.1%) of the 17 032 women in the study were known to have undergone hysterectomy. Fibroids were the most common cause followed closely by menstrual disturbances in the absence of fibroids (hereafter referred to as ‘menstrual disturbances’). Social class had a modest influence on hysterectomy rates. Hysterectomy for fibroids, prolapse, endometriosis and ‘other reasons’ showed little trend with calendar period while hysterectomy for menstrual disturbances and for cancer showed a sharp increase with calendar time especially at ages 30–39. Hysterectomy generally tended to increase with age and showed a strong relation to parity; in particular, hysterectomy for fibroids fell with parity and hysterectomy for menstrual disturbances rose sharply with parity. Using lifetable methods, it was estimated that almost 20% of the women in the study would have had a hysterectomy by age 55.Conclusions The results give insights into factors affecting hysterectomy rates. Of particular interest is the modest influence of social class, the strong influence of parity and the rise in rates with calendar time at ages 30–39 for those undergoing hysterectomy for menstrual disturbances or cancer, but since the cohort is not directly representative of the population, some caution is required in extrapolating these findings. The estimated hysterectomy rate of about 20% by age 55 is in line with other similar estimates for the United Kingdom.