Menopause hormone therapy prescribing in ambulatory care visits among midlife and older U.S. women from 2018–2019

Abstract

The rates of prescription for menopause hormone therapy have been low in the U.S. since the 2002 Women’s Health Initiative study, but no recent studies have assessed the prescribing of hormone therapy in the U.S. Using the National Ambulatory Medical Care Survey data from 2018 to 2019, we found that hormone therapy was prescribed in 3.8% of U.S. visits by midlife and older women, with 60% of these visits including estradiol-only prescriptions. Older age and Hispanic/Latina ethnicity were associated with decreased odds of prescribing, while White race and depression were associated with increased odds, indicating possible disparities in menopause care.

Introduction

Menopause hormone therapy (HT) is indicated for the treatment of vasomotor symptoms of menopause, premature menopause, genitourinary syndrome of menopause, and prevention of osteoporosis [1]. After the 2002 Women’s Health Initiative (WHI) study results were published, HT prescribing rates plummeted 66% amidst fears of cardiovascular disease, venous thromboembolism, and breast cancer [2]. In 2009–2010, the prevalence of HT use was 4.7%, decreased from 22.4% in 1999–2000 [3]. Almost no recent studies have reported HT prescribing rates in representative patient samples in the U.S. This report examines HT prescribing in ambulatory U.S. care settings in 2018–2019 (the most recent available National Ambulatory Medical Care Survey [NAMCS] data and representing the last year before the COVID-19 pandemic), including commonly prescribed HT types and HT prescribing characteristics.

Methods

The NAMCS 2018–2019 used multistage probability sampling to obtain a representative sample of U.S. ambulatory visits in which non-federally employed office-based clinicians engaged in patient care in private practices, academic medical centers, local government clinics, and health maintenance organizations or their office staff abstracted visit data for a randomly assigned 1-week reporting period [4]. Physicians from the specialties of anesthesiology, pathology, and radiology were excluded from the NAMCS dataset [4]. Descriptive statistics were used to summarize documented new or continued HT prescriptions (either estrogen-only or estrogen-progestin therapies) in visits involving women aged 50 years and older. Medication codes specific to hormone preparations typically used as contraceptives and vaginal-only estrogens were excluded [1]. Chi-square tests were used to examine differences in HT prescribing by patient demographics (age, race, Hispanic ethnicity, insurance status) and selected clinical characteristics/diagnoses with the potential to be influenced by HT use (obesity based on body mass index [BMI], dementia, cardiovascular disease, cancer, cerebrovascular disease, depression, diabetes, hyperlipidemia, osteoporosis, and venous thromboembolism). Multivariable logistic regression models were fit to examine independent associations between all the above patient characteristics and HT use. All analyses incorporated NAMCS sampling weights.

Results

Among the 5,959 ambulatory visits involving women aged 50 years and older analyzed (extrapolating to approximately 633 million visits nationally), HT prescribing was documented in 248 (3.8%) visits (extrapolating to approximately 24 million visits nationally). In visits where HT was documented, 59.7% of visits included estradiol-only prescriptions, 25.2% included conjugated estrogen-only prescriptions, 12.0% included estrogen plus progestin prescriptions, 1.6% included combination estrogen and non-progestin (ex: bazedoxifene) prescriptions, and 1.9% included duplicate estrogen prescriptions.

Only 2.5% of visits involving HT were with Hispanic/Latina women, compared to 11.7% of visits not involving HT (p<0.001). Women with elevated BMI (≥25 kg/m²) represented 56.4% of visits involving HT, compared to 70.8% of visits not involving HT (p=0.02 for heterogeneity in BMI by HT status). A diagnosis of depression was more common in visits involving HT compared to visits not involving HT (22.7% vs 14%, p=0.03), but no significant differences in documented diagnoses of Alzheimer’s disease/dementia, coronary artery disease, congestive heart failure, cancer, cerebrovascular disease, diabetes mellitus, hyperlipidemia, osteoporosis, or venous thromboembolism were detected (Table 1).

Table 1.

Number, Percentage, and Characteristics of U.S. Ambulatory Visits in Women 50 Years or Older, by Prescription or Continuation of Hormone Therapy

Patient Characteristics	Visits with Hormone Therapy (N=248)	Visits without Hormone Therapy (N=5,711)	P-value
Age (years old)
50–60	79 (29.7%)	1551 (29.5%)	0.11
61–70	92 (41.6%)	1818 (32.5%)
>70	77 (28.7%)	2342 (38.1%)
Race
White	225 (90.0%)	4998 (84.0%)	0.16
Black or African American	14 (6.1%)	430 (9.8%)
Other	9 (3.9%)	283 (6.1%)
Ethnicity
Hispanic or Latina	10 (2.5%)	556 (11.7%)	<0.001
Not Hispanic or Latina	238 (97.5%)	5155 (88.3%)
Source of payment
Medicare	104 (46.2%)	2968 (53.5%)	0.61
Medicaid	6 (8.2%)	268 (4.8%)
Self-pay	12 (4.2%)	204 (4.5%)
Private	109 (41.5%)	1869 (37.2%)
Body mass index (kg/m 2 )
<18	2 (0.2%)	41 (1.6%)	0.02
18–24.9	78 (43.4%)	930 (28.6%)
25–29.9	47 (29.1%)	936 (29.0%)
≥ 30	46 (27.3%)	1211 (40.8%)
Clinical conditions
Alzheimer’s disease/dementia	2 (2.2%)	108 (2.0%)	0.87
Coronary artery disease, congestive heart failure	10 (6.3%)	512 (10.1%)	0.21
Cancer	24 (17.2%)	654 (13.0%)	0.36
Cerebrovascular disease (history of CVA/TIA)	7 (2.5%)	170 (2.9%)	0.77
Depression	46 (22.7%)	738 (14.0%)	0.03
Diabetes mellitus	20 (11.7%)	975 (19.8%)	0.07
Hyperlipidemia	61 (28.0%)	1428 (30.1%)	0.69
Osteoporosis	16 (8.9%)	303 (7.2%)	0.55
Venous thromboembolic history (PE, DVT)	3 (1.5%)	60 (1.2%)	0.77

All percentages are column percentages. Percentages are weighted percentages incorporating recommended NAMCS sampling weights, but Ns are unweighted Ns.

Key: CVA= cerebrovascular accident; DVT= deep venous thrombosis; PE= pulmonary embolism; TIA= transient ischemic attack

In multivariable analyses, older patient age was associated with a decreased odds of HT prescribing (OR 0.90, 95% CI 0.82–0.99 per each 5 years in age), as was Hispanic/Latina patient ethnicity (OR 0.13, 95% CI 0.05–0.33). White patient race was associated with an increased odds of HT prescribing (OR 2.0, 95% CI 1.09–3.64, for White versus all other races), as was diagnosed depression (OR 1.9, 95% CI 1.04–3.48). No significant independent associations between HT prescribing and other characteristics or diagnoses were detected.

Discussion

In a nationally representative sample of U.S. ambulatory care visits with midlife or older women in 2018–2019, only 3.8% of visits documented HT prescribing or continuation. HT prescribing was appropriately decreased in visits with older women, but it was less appropriately decreased in visits with non-white or Hispanic/Latina women even after adjustment for other patient characteristics, indicating possible disparities in menopause care. The odds of HT prescribing were 90% higher in visits with patients diagnosed with depression, which is known to overlap with several types of menopause symptoms but did not differ significantly based on concomitant diagnosis of other conditions that could be worsened by HT (ex: coronary artery disease) or improved by HT (ex: osteoporosis).

Most of the visits documenting HT involved estrogen-only therapy rather than estrogen plus progestin therapy. This may reflect greater clinician comfort with prescribing estrogen-only HT to postmenopausal women who have undergone hysterectomy, and thus, do not need concomitant progestin therapy.

Limitations of this research include possible inaccuracy of medication abstraction, potentially leading to the misclassification of local vaginal estrogen therapy as systemic HT. Additionally, the survey data distinguished between healthcare visits, not patients, so the same patient could be represented more than once in the data if they presented to multiple clinicians participating in the survey. Almost no nationally representative studies have been published, though one analysis of insurance claims data including only new users of estrogen and progestin or bazedoxifene therapy reported potential associations with cancer diagnoses. This study found even lower prescribing rates of 0.05%, which may be due to its focus on new hormone therapy prescriptions and exclusion of estrogen-only prescriptions. [5]

Given the high prevalence of menopause symptoms in U.S. women, these findings suggest that many women are not being prescribed HT who may have indications for it. Future research should investigate patient and clinician concerns about HT use, explore variations in clinician background or beliefs associated with HT prescribing, and assess clinician decision-making when prescribing HT in women across different demographic and clinical characteristics.

Highlights.

• The rates of prescription for menopause hormone therapy continue to be low in the U.S.

• Estrogen-only therapy was more often prescribed than estrogen-progestin therapy.

• Hormone therapy was prescribed more in visits with a diagnosis of depression.

• Hormone therapy was prescribed more in visits with women of White race.

• Hormone therapy was prescribed less in visits with older and Hispanic/Latina women.

Data sharing and collaboration

There are no linked research data sets for this paper. Data will be made available on request.

This research and data were presented in a podium presentation in the national scientific meeting of The Menopause Society on September 29, 2023.