Utilisation of hormone replacement therapy by women doctors

BMJ 1995; 311 doi: (Published 25 November 1995) Cite this as: BMJ 1995;311:1399
  1. A J Isaacs, visiting research fellowa,
  2. A R Britton, research assistanta,
  3. Klim McPherson, professor of public health epidemiologya
  1. Department of Public Health and Policy, London School of Hygiene and Tropical Medicine, London WC1E 7HT
  1. Correspondence to: Professor McPherson.
  • Accepted 28 September 1995


Objectives: To ascertain the prevalence and duration of use of hormone replacement therapy by menopausal women doctors.

Design: Postal questionnaire.

Setting: General practices in the United Kingdom.

Subjects: Randomised stratified sample of women doctors who obtained full registration between 1952 and 1976, taken from the current principal list of the Medical Register.

Main outcome measures: Prevalence and duration of use of hormone replacement therapy; menopausal status.

Results: Overall, 45.7% (436/954) of women doctors aged between 45 and 65 years had ever used hormone replacement therapy. When the results from women still menstruating regularly were excluded, 55.2% (428) were ever users and 41.2% (319) current users. The cumulative probability of remaining on hormone replacement therapy was 0.707 at five years and 0.576 at 10 years.

Conclusions: Women doctors have a higher prevalence of use of hormone replacement therapy than has been reported for other women in the United Kingdom, and most users seem to be taking hormone replacement therapy for more than five years. The results may become generalisable to the wider population as information on the potential benefits of hormone replacement therapy is disseminated.

Key messages

  • Key messages

  • Over half of women starting hormone replacement therapy may be expected to continue the treatment for 10 years or more

  • A considerable number of women will be unlikely to take up hormone replacement therapy in the absence of full evaluation of benefit and risk

  • Full evaluation will require randomised controlled trials


Long term hormone replacement therapy is increasingly accepted as instrumental in reducing the risk of osteoporosis and cardiovascular disease in postmenopausal women.1 2 Little is known for certain, however, of the overall uptake of hormone replacement therapy in the United Kingdom and the use in the longer term which may produce a considerable impact on public health.

The attitudes of general practitioners3 may be important determinants of women's choices on whether to use hormone replacement therapy and for how long, particularly as many women consider that the menopause should be viewed as a medical condition.4 We investigated the use of hormone replacement therapy by women doctors as they are a well informed group in a strong position to influence the behaviour of other women.


Sampling--The sampling frame consisted of all women doctors who obtained full registration with the General Medical Council between 1952 and 1976 inclusive and whose names appeared on the principal list of the Medical Register in 1993. A randomised sample of 1550 was taken, stratified by five year bands, to obtain an approximately even age distribution across the age range 40 to 65 years. Of these, 36 living abroad were excluded, giving a final sample size of 1514.

Questionnaire--A postal questionnaire, explanatory letter, and reply paid envelope were sent to all doctors in the sample in June 1993. Initial non-responders were sent a reminder letter, and finally a further letter with a second copy of the questionnaire and another reply paid envelope was sent in July. All responses received by the end of 1993 were coded and the data entered onto a computerised database. Statistical analyses were carried out with Epi-Info.5 When answers to specific questions were missing these respondents were omitted from the relevant analyses. The questionnaire covered various demographic and behavioural factors which will be reported fully elsewhere, the current report being restricted to the prevalence and duration of use of hormone replacement therapy.


Response rate--A total of 1211 completed responses were received. Four were returned as having failed to reach the intended recipient. The total valid response rate was therefore 80.2%. Table I shows the numbers responding by year of registration.


Response rate to questionnaire on use of hormone replacement therapy in women doctors by registration group

View this table:

Menopausal status--Periods had ceased completely in 771 women; 45 were perimenopausal; 93 had started hormone replacement therapy before the menopause; and 302 were still menstruating regularly. A total of 186 women had ceased menstruating as a result of surgery.

Prevalence of hormone replacement therapy use--Overall, 480 of the 1211 respondents (39.6%) had ever used hormone replacement therapy, of whom 344 (28.4% of the whole group) were still using it (table II). Of the 954 women aged between 45 and 65, 436 (45.7%) had ever used hormone replacement therapy. Prevalence of current use was highest in the 50-59 year age band (199; 44.2%); this declined after age 60. When we excluded data on women still menstruating regularly, 472 out of 909 (51.9%) had ever used hormone replacement therapy. Between the ages of 45 and 65 years, 55.2% (428) were ever users and 41.2% (319) current users. The prevalence of ever use was highest below age 60 (322/519; 62%) and declined thereafter. Current use was stable up to 55 years (143/266; 53.8%) and then declined. Use of hormone replacement therapy was higher in women with a surgical menopause, particularly after bilateral oophorectomy. Of 50 women in the latter group, 42 had received hormone replacement therapy at some time and 29 were current users.


Prevalence of use of hormone replacement therapy by age group in women doctors*

View this table:

Duration of hormone replacement therapy--Past users had received hormone replacement therapy for a median (range) period of 0.8 (0.1-20) years, whereas current users had taken hormone replacement therapy for 3.6 (0.1-26.1) years. Life table analysis of the combined group showed that over 70.7% of ever users were still receiving hormone replacement therapy five years after starting and over 57.6% at 10 years (figure). When asked their intentions, 130 (37.8%) current users anticipated taking hormone replacement therapy for between five and 10 years altogether and 166 (48.1%) for more than 10 years.


Probability of continuing to take hormone replacement therapy. Vertical bars represent 95% confidence intervals


There have been relatively few published surveys on the use of hormone replacement therapy in the United Kingdom, and these have differed in respect of methodology, target age group, and geographical location. A survey in Greater London of women aged 45-65 showed 10% of the whole group (and 18% of those no longer having periods) had ever received hormone replacement therapy.6 Doctors in the Medical Research Council's general practice research framework prescribed hormone replacement therapy to an estimated 9% of their female patients aged 40-64 in 1989, with female doctors treating an average of about five more women than male doctors.7 An international survey indicated current rate of use of hormone replacement therapy of 7% in women aged 40-69 in the United Kingdom.8 Another study in Scotland found a prevalence of 9% for current use and 16% for ever use among postmenopausal women aged 33-68.4

The above rates are all considerably lower than those found in the present study, in which current use of hormone replacement therapy was 33% in the 45-64 age group and ever use was 46%, these figures rising to 41% and 55% if data for premenopausal women were excluded. The prevalence of use of hormone replacement therapy among women doctors in the United Kingdom is thus closer to the reported rates in the western United States9 10 11 than to those among other women in the United Kingdom (though unpublished data suggest that rates in the latter are now rising).

In the United States rates of use of surgical procedures by doctors have been shown to reflect or only slightly exceed those in comparable groups of lay people,12 whereas there is greater use of obstetric interventions by women doctors.13 The substantial difference for hormone replacement therapy suggested by our results may in part be attributable to the high socioeconomic status of women doctors, a known determinant of use of hormone replacement therapy in the United Kingdom14 15 as elsewhere. Women doctors may perhaps be pace setters for the wider female population in this respect, as for smoking,16 another health related behaviour.

The benefits of hormone replacement therapy to the individual in terms of reduction of risks of fracture and cardiovascular disease are thought to be related to duration of use. Risk-benefit and cost effectiveness models of the impact on public health, taking into account potential adverse consequences such as an increased risk of breast cancer risk17 as well as benefits, often assume compliance with treatment over 10 years.18 19 Cross sectional surveys in the United Kingdom have not so far studied the issue of length of treatment. Studies from the United States have reported rather low rates of compliance overall,20 21 and it has been suggested that these could be increased by education of both patients and physicians about the value of hormone replacement therapy.22 The current survey indicated that, although 10% of women doctors stopped hormone replacement therapy within six months, the discontinuation rate subsequently declined, and over 50% of those starting would still be expected to be taking hormone replacement therapy at 10 years. This may be because they rely less on the views of their general practitioners than do other women.

Though these data are based on the respondents' unvalidated recollections of their past use of hormone replacement therapy, a longitudinal study showed reasonable consistency of reporting of perimenopausal use of oestrogen, at least up to 10 years since last use.23 Furthermore, the subjects' own predictions of their likely duration of future use gave similar results. It may therefore be reasonable to take 50% compliance at 10 years as a target for other groups of women and to use this figure in cost effectiveness analyses. The possibility of increasing rates of use and compliance in the general population has been shown by the establishment of a dedicated clinic in primary care.24

Overall, the results of the survey are consistent with increasing use of hormone replacement therapy by younger postmenopausal women doctors. This may presage more widespread use in the general population, particularly as more information becomes available on the preventive effects of hormone replacement therapy.25 About 40% of those in the most eligible age group, however, had never tried it, and more general use still may depend on the further elucidation of the benefit:risk ratio, which is likely to have to await the results of large scale randomised controlled trials.26

The authors take sole responsibility for the views expressed. We thank Jeremy Isaacs for assiduous data input and processing.


  • Funding Department of Health.

  • Conflict of interest None.


View Abstract