Postmenopausal hormone replacement therapy (HRT) must be continued long term for it to reduce the risks of osteoporosis1 and cardiovascular disease.2 However, few women continue this therapy for more than a few years.3 The three major reasons women give for discontinuing are that treatment is no longer necessary, that it produces unacceptable side effects, and that it might increase the risk of breast cancer.4 Troublesome menstrual bleeding is a frequently cited side effect of HRT, which appears to contribute to the high rate of discontinuation in women with an intact uterus.4,5
By causing monthly menses, cyclic addition of progesterone to estrogen (CYC-PERT) has been proved to prevent the high rate of endometrial hyperplasia6 associated with unopposed estrogen. To avoid endometrial stimulation without producing cyclic menstrual bleeding is the goal of continuous combined estrogen progestin therapy (CC-PERT). Reports of this regimen were first published in the early 1980 s.7,8 Typically, two thirds of CC-PERT users have unscheduled vaginal bleeding for the first 6-12 months of use, but within a year, most who continue become amenorrheic. The CC-PERT regimen has shown excellent protection against endometrial hyperplasia6,9 and salutary effects on cardiovascular risk factors.6,10 As a result of a good safety profile and presumed ease of use, the daily administration of CC-PERT has become quite popular in the United States. In the Kaiser Permanente Medical Care Program in Northern California, we find that among women treated with estrogen/progestin, 60% are currently using CC-PERT. In 1995, the Food and Drug Administration approved the marketing and promotion of CC-PERT, and most believe that this regimen will continue to gain in popularity.
Considerable effort on the part of health care providers and patients may be wasted if women find HRT unacceptable and do not continue it. Based on double-blind, randomized clinical trials of HRT, women assigned to CYC-PERT6,9,11,12 and CC-PERT6,9,12 have high rates of continuation. In the 3-year postmenopausal estrogen/progestin interventions trial, cumulative discontinuation was <20% in both groups.6 However, no large-scale studies of continuation of these HRT regimens in the context of usual clinical practice have been done.
We compared continuation rates of CYC-PERT and CC-PERT in postmenopausal women who were initiating HRT in the context of usual clinical practice. We hypothesized that because of unacceptable irregular bleeding, women using CC-PERT might discontinue treatment more than women using CYC-PERT in the early months of treatment but that, ultimately, CC-PERT users would discontinue less frequently because they would find amenorrhea induced by the regimen more acceptable than cyclic monthly bleeding.
We chose subjects ≥45 years old who were members of the Kaiser Foundation Health Plan (KFHP); who lived in the area of San Jose, California; and who received care at the KFHP Santa Teresa medical offices. All subjects were required to have had continuous membership from 1991 through 1995 and to have health plan coverage that provided estrogen prescriptions at costs ranging from nothing to $21 for a 100-day supply. The study protocol was approved by the Kaiser Permanente Medical Care Program, Northern California Region, Institutional Review Board.
Beginning in March 1991, pharmacies in the KFHP began recording all prescriptions in a computer database (Pharmacy Information Monitoring Service, or PIMS). Included in the P1MS database are patient's identification, dispensing date, and drug information including name, strength, number of pills dispensed, and directions for use.
Criteria for type of estrogen use
We first identified all women who filled a prescription for 100 tablets of 0.625 mg conjugated estrogens (CE) (Premarin; Wyeth Ayerst Laboratories, Philadelphia, PA, U.S.A.). Further selection as a first-time user was on the basis of a search of the PIMS database showing no prescription for any estrogen or progestin for 9 months before the index prescription. Women who filled this first-time prescription between 1 January 1992 and 30 June 1994 were included. Excluded were women whose CE prescription directions indicated a regimen other than daily use or use as directed (i.e., every other day).
Next, we selected from the CE-using cohort those women who also received a prescription for medroxyprogesterone acetate (MPA), either 2.5 mg or 10 mg, on the same day as the CE prescription. We considered coprescription of 2.5 mg MPA, in units of 100 pills, an indication of CC-PERT use because a 2.5-mg MPA dose is not recommended for cyclic HRT. Because 10 mg MPA has never been recommended for continuous combined regimens, we reasoned that its coprescription with estrogen would be limited to the cyclic regimen. Because 5 mg MPA is occasionally used in continuous combined regimens as well as cyclic regimens, we avoided any ambiguity by excluding women who received 5 mg MPA for their initial prescription.
Continuation of HRT use
The PIMS database from all Northern California KFHP facilities was searched for all CE and MPA prescriptions filled by women in the cohort to identify those women who could have filled prescriptions at facilities other than Santa Teresa. Continuation of CC-PERT was indicated if a woman initially categorized as a CC-PERT user obtained refills of CE within 60 days of the date her previous refill would have been consumed. Because most physicians instruct CC-PERT users to take this combination every day, a >60-day hiatus between refills would indicate a <63% compliance rate. In addition, refills of 2.5 mg MPA were required to have occurred within the same use period. Patterns of discontinuation were divided into two types: continuation of CE but switching of MPA to 5.0 mg or 10.0 mg, indicating a probable change to cyclic HRT, or discontinuation of CE. Continuation of CYC-PERT was indicated if a woman initially categorized as a CYC-PERT user obtained a refill of CE tablets within 60 days of >1.2 times the number of pills dispensed. Because physicians usually instruct CYC-PERT users to take CE for either 25 of 30 days or 30 of 30 days, a 180-day interval for 100 dispensed tablets could indicate either 56% compliance (if use 30/30 days) or 67% compliance (if use 25/30 days). Patterns of discontinuation were again divided into two types: continuation of CE but switching of MPA to 2.5 mg or discontinuation of CE.
We validated our initial categorization of women to CYC-PERT and CC-PERT by reviewing 320 randomly selected medical records from among those women we had assigned to be CC-PERT (n = 147) or CYC-PERT (n = 173) users on the basis of their PIMS data. The type of HRT indicated in the medical record agreed with the type indicated in the PIMS data in 99% of both CC-PERT and CYC-PERT users. Whereas all subjects were considered first-time users on the basis of PIMS data, medical record notations indicated that 8.8% of CC-PERT users and 8.1% of CYC-PERT users were prescribed HRT in the year before the index prescription. Using the same medical records, we confirmed the following prescribing patterns. Of CYC-PERT users, 63.6% were prescribed estrogen for 25 days a month and 35.8% daily estrogen; neither subgroup was prescribed 2.5 mg MPA. Of CC-PERT users, all received only 2.5 mg MPA, and both estrogen and MPA were prescribed daily.
Cox proportional hazards regression was used to estimate the rate ratio (RR) and its confidence interval (CI) in evaluating the difference in continuation rate between CC-PERT users and CYC-PERT users after adjustment for age, entry year, health plan prescription coverage, and type of physician prescriber (gynecologist or not).13 Kaplan-Meier estimates were also obtained to visually compare maintenance of continuation of the two regimens.
CC-PERT users were older and more likely to have started the regimen after 1992 than CYC-PERT users (Table 1). The median cost for a 100-day supply of both conjugated estrogens and medroxyprogesterone was $2.00 (range, 0-$42.00); 10.7% paid nothing, 84.9% paid $2-10, and 4.4% paid $20-42. Prescription coverage was similarly distributed in the two cohorts. Prescribing was by gynecologists for 77.4% of CYC-PERT users and 74.6% of CC-PERT users.
The probability and pattern of continuing HRT are shown in Table 2 and Fig. 1. About 35% of users on either regimen stopped using HRT after their first prescription, and 76-81% of the women had stopped by the end of the third year. CC-PERT users were more likely to discontinue the regimen than CYC-PERT users. After adjustment for age, entry year, copayment plan, and gynecologist prescriber. CC-PERT users had, on average, an 18% higher discontinuation rate than CYC-PERT users (RR = 1.18, 95% CI = 1.04-1.35). Differences between CC-PERT and CYC-PERT discontinuation rates were greatest in the first year of use and diminished over time, although the trend was not statistically significant. Most discontinuation for both regimens was because the patient had stopped HRT entirely; few switched MPA regimens (23.0% switched from CYC-PERT to CC-PERT, and 13.7% switched from CC-PERT to CYC-PERT). When subjects switched from 0.625 mg CE to other CE doses (0.3, 0.9, 1.25 mg) and we considered continuation unaltered, the probability of continuation of CC-PERT at the end of year 1 was 0.45, for year 2 it was 0.29, and for year 3 it was 0.19. For CYC-PERT at end of year 1 the probability of continuation was 0.54, for year 2 it was 0.38, and for year 3 it was 0.24. This lack of difference with the more liberal usage criterion was probably because only 7% of users switched their CE dose once or more.
The age at initiation of HRT regimens and the amount of copayment were not associated with discontinuation of HRT regimens. The calendar year in which treatment began was not a factor in overall discontinuation and did not explain the higher discontinuation among CC-PERT users compared with CYC-PERT users; the rate ratio of discontinuation associated with CC-PERT versus CYC-PERT use was 1.2 for women who started HRT regimens in 1992, 1.3 for 1993, and 1.1 for 1994.
We found discontinuation of HRT in clinical practice to be much higher than that reported in randomized, double-blind clinical trials,6,9 Furthermore, women initiating CC-PERT were more likely to discontinue than those initiating CYC-PERT. Although some women switched between regimens, most women stopped HRT altogether.
We studied the records of women from a large, relatively homogeneous population in similar geographic locations who had equal access to physicians and pharmacies. The PIMS database provided a fairly reliable method of tracking use of prescribed medication but did not indicate whether women took the medication once they filled the order. We consider it unlikely that women would frequently visit pharmacies outside the health plan because the prescription cost for HRT would be many times greater. Moreover, prescription underreporting or failure to use an obtained prescription seemed unlikely to be associated with either HRT regimen.
The setting of our study could limit the generalizability of our findings. Women within an HMO have few barriers to receiving health care, and nearly all were receiving their HRT treatment from obstetrician-gynecologists instead of general practitioners. Together with the low cost of prescriptions, these factors may have caused our subjects to continue treatment longer than would be the case among women not enrolled in HMOs.
Our study was not designed to find reasons for HRT discontinuation. Perhaps scheduled and unscheduled vaginal bleeding are important factors in a woman's decision to stop treatment. We are now conducting a study of the frequency of bleeding episodes and the resultant use of medical resources by women on CC-PERT and CYC-PERT regimens. We also did not examine the degree to which patients were counseled and educated about the risks and benefits of HRT. We assume that gynecologists are well informed and are likely to convey their knowledge and enthusiasm for HRT when beginning treatment for women. We did test the hypothesis that physicians, by progressively becoming more knowledgeable and enthusiastic about CC-PERT, would have been more likely to encourage continuation of CC-PERT in 1994 than in previous years. Our analyses did not provide evidence to support this hypothesis.
The percentages of patients who discontinue HRT in large, randomized clinical trials6,9,11,12 are substantially smaller than the rates we observed. In such clinical trials, the annual discontinuation rates are usually 5-10% and do not differ between women assigned to cyclic or to continuous combined regimens. In open, nonrandomized trials, annual discontinuation rates are ∼20-30%,7,14,15 but these rates are still about half the rates we observed in our study.
We hypothesize that differences in continuation among different types of clinical trials as well as between clinical trials and clinical practice may be related to the degree to which women are informed, encouraged, and supported by those who prescribe the medication. The influences of the physician's knowledge, enthusiasm, and interaction with the patient deserve further study.
We conclude that the likelihood of women continuing HRT beyond 3 years is low. Furthermore, compared with CYC-PERT users, those receiving CC-PERT have a higher probability of discontinuation, and this probability does not diminish after the first 6-12 months, after which troublesome irregular bleeding is likely to resolve. Efforts should be made to understand why three quarters of women beginning HRT will stop it long before it can provide major long-term benefits.
This research was supported in part by a grant-in-aid from Wyeth Ayerst Laboratories, Radnor, Pennsylvania. Lisa J Herrinton, Ph.D., contributed to the study design. The Medical Editing Department, Kaiser Foundation Research Institute, provided editorial assistance.
1. Cauley JA, Seeley DG, Ensrud K, Ettinger B, Black D, Cummings SR. Estrogen replacement therapy
and fractures in older women. Study of Osteoporotic Fractures Research Group. Ann Intern Med
2. Stampfer MJ, Willett WC, Colditz GA, Rosner B, Speizer FE, Hennekens CH. A prospective study of postmenopausal estrogen therapy and coronary heart disease. N Engl J Med
3. Ravnikar VA. Compliance with hormone therapy. Am J Obstet Gynecol
4. Utian WH, Schiff I. NAMS-Gallup survey on women's knowledge, information sources, and attitudes to menopause
and hormone replacement therapy. Menopause
5. Hahn RG. Compliance considerations with estrogen replacement: withdrawal bleeding and other factors. Am J Obstet Gynecol
1989; 161 (6 Pt 2):1854–1858.
6. The Writing Group for the PEPI Trial. Effects of estrogen or estrogen progestin regimens on heart disease risk factors in postmenopausal women: the Postmenopausal Estrogen/Progestin Interventions (PEPI) Trial. JAMA
7. Staland B. Continuous treatment with natural oestrogens and progestogens: a method to avoid endometrial stimulation. Maturitas
1981; 3 2:145–156.
8. Mattsson L-A, Cullberg G, Samsioe G. Evaluation of a continuous oestrogen-progestogen regimen for climacteric complaints. Maturitas
1982; 4 2:95–102.
9. Archer DF, Pickar JH, Bottiglioni F. Bleeding patterns in postmenopausal women taking continuous combined or sequential regimens of conjugated estrogens with medroxyprogesterone acetate. Menopause
Study Group. Obstet Gynecol
1994; 83 (5 Pt 1):686–692.
10. Lobo RA, Pickar JH, Wild RA, Walsh B, Hirvonen E. Metabolic impact of adding medroxyprogesterone acetate to conjugated estrogen therapy in postmenopausal women. The Menopause
Study Group. Obstet Gynecol
11. Marslew U, Munk-Nielsen N, Nilas L, Riis BJ, Christiansen C. Bleeding pattern and climacteric symptoms during different sequential combined HRT regimens in current use. Maturitas
1994; 19 3:225–237.
12. Eiken P, Kolthoff N. Compliance with long-term oral hormonal replacement therapy. Maturitas
1995; 22 2:97–103.
13. Cox DR, Oakes D. Analysis of survival data. 1984; New York: Chapman and Hall, (Monographs on Statistics and Applied Probability).
14. Magos AL, Brincat M, Studd JWW, Wardle P, Schlesinger P, O’Dowd T. Amenorrhea and endometrial atrophy with continuous oral estrogen and progestogen therapy in postmenopausal women. Obstet Gynecol
15. Weinstein L, Bewtra C, Gallagher JC. Evaluation of a continuous combined low-dose regimen of estrogen-progestin for treatment of the menopausal patient. Am J Obstet Gynecol
Keywords:© 2018 by The North American Menopause Society.
Combination therapy; Estrogen replacement therapy; Menopause; Patient compliance; Postmenopause