Patient-Centered Management of Heavy Menstrual Bleeding while on Anticoagulation: A Survey Study

Abstract

Background

Nearly three-quarters of women experience heavy menstrual bleeding (HMB) during anticoagulation treatment for venous thromboembolism (VTE). HMB can lead to medical interventions, impact school or work and diminish quality of life (QOL). While there are several management options for HMB in patients with VTE, we lack information on how patients approach HMB treatment decisions.

Objectives

We examined patients’ experience with, perceptions of, and preferences for treatment of HMB while on anticoagulation.

Methods

Clinicians and patient advocates developed a web-based survey that was distributed via patient advocacy organizations and social media between May and June 2024, targeting participants 18 years or older who menstruate with a history of VTE and self-reported HMB.

Results

A total of 387 participants completed the survey. All participants (100%) reported a history of HMB, and half (50%) experienced HMB while on anticoagulation for greater than twelve months. The majority (86%) reported a negative effect on their QOL. Only 43% of patients reported receiving treatment for HMB, with 10% reporting temporarily holding anticoagulation due to HMB. Approximately a fifth of respondents associated low-risk treatment modalities with thrombosis, including progesterone-only oral contraceptives (21%), implants (18%), and intrauterine devices (progesterone 18%, copper 17%). The majority (83%) of patients identified thrombotic risk as the most important factor when considering treatment options.

Conclusions

HMB is under-reported and undertreated in patients on anticoagulation. Patients have important misconceptions about treatment options; prioritized recurrent thrombosis risk when weighing treatment options; and preferred a multidisciplinary, multimodal approach. These findings should inform interventions to improve patient education and decision support for HMB on anticoagulants.

Introduction

Heavy menstrual bleeding (HMB) is defined as “excessive menstrual blood loss which interferes with a woman’s physical, social, emotional or material quality of life” (1). It affects up to 30% of patients who menstruate, and the prevalence significantly increases to upwards of 70% among patients who take anticoagulation (2,3,4,5). The prevalence is likely underreported due to many factors including difficulty with diagnosis, social stigma, normalization of symptoms, lack of awareness of treatment options, and potential fear of invasive procedures (6).

HMB can lead to limitations in daily activities, missed school or work, psychological distress, and mood disorders in addition to physical symptoms such as dysmenorrhea, symptoms from iron deficiency, and sexual dysfunction. Patients with HMB score lower than the 25th percentile on quality of life (QOL) measures in comparison to age-controlled counterparts in several domains (7). Furthermore, patients who experience new-onset HMB after initiation of anticoagulation have a larger decrease in their QOL as assessed by the Menstrual Bleeding Questionnaire in comparison to patients who had HMB prior to anticoagulation therapy (8).

Despite the prevalence of HMB, it is commonly undiagnosed or untreated. Approximately 50% of patients who experience HMB discuss their symptoms with a physician and a similar proportion receive treatment (9). Treatment of HMB in patients with a history of venous thromboembolism (VTE) can be safe and effective, although it requires careful consideration of family planning, fertility preservation, and thrombosis risk. Options include hormonal therapies like levonorgestrel intrauterine devices (IUDs); combined hormonal contraceptives (CHCs); progesterone implants, pills, and depot injections; adjustment of anticoagulant agents; procedures such as ablations, uterine artery embolization, or hysterectomy; and anti-fibrinolytic agents (10, 11).

Although there is ample evidence on the prevalence of HMB and its impact on QOL, there remains a paucity of data describing patients’ understanding of treatment options including potential side effects and preferences for treatment. The goal of this survey was to generate patient-derived data on experiences with treatment of HMB in patients on anticoagulants, their perception of management options including side effects, and their goals and preferences for treatment to facilitate individualized care.

Methods

An web-based survey was designed by a multidisciplinary team of clinicians and patient advocates. The survey consisted of 34 questions covering several domains, including VTE and treatment characteristics, patients’ experiences with current and past treatments for HMB, their perceptions of management options and associated side effects, as well as their goals and preferences for treatment (see Appendix 1 for survey). The questions included a mix of multiple-choice and Likert scale formats. Multi-select non-mutually exclusive choices were allowed for certain questions.

The survey underwent pilot testing prior to distribution. The finalized survey was openly distributed online over a two-month period (May–June 2024) to allow for timely data collection while minimizing recall bias and maintaining feasibility. To enable recruitment of a broad and geographically diverse sample, participants were recruited through online platforms including social media and email distribution lists. Recruitment efforts targeted adults aged 18 years or older with a history of VTE and self-reported HMB while on therapeutic anticoagulation therapy. Patients were excluded if they were under the age of 18, did not report a history of HMB, or did not report a history of anticoagulation. Survey participation was voluntary, and an optional financial incentive (a drawing for a $250 gift card) was provided in order to increase participation. The survey was anonymous, and no protected health information was collected. Participants provided informed consent prior to accessing the survey, and the institutional review board at the University of Utah deemed this study as exempt.

Responses were collected using REDcap software (12). Summary statistics were reported as numbers, frequencies, and proportions.

Results

Sociodemographics and VTE Treatment Characteristics

The survey yielded 387 responses from 13 different countries (Table 1). Due to the nature of the openly distributed survey, a response rate could not be calculated. Although the majority (77%) were from the United States, responses additionally came from Europe, Oceania, Asia, and Africa. Nearly three-fourths of respondents were aged 30-49 years and most (87%) identified as White. The population was highly educated with approximately 90% of respondents having completed at least 1 year of college.

Table 1.

Sociodemographic

Race and Ethnicity*	Count (N=387) (%)
White	337 (87)
Hispanic	19 (4.9)
Black or African American	19 (4.9)
American Indian	1 (0.3)
Asian	8 (2.1)
Native Hawaiian	2 (0.5)
Multiracial / other / not specified	23 (6.0)
Country	Count (N=387) (%)
United States	298 (77)
Other †	89 (23)
Current age	Count (N=387) (%)
18-29	31 (8)
30-39	120 (31)
40-49	166 (43)
50-59	66 (17)
60 years or older	4 (1)
Age at VTE Diagnosis	Count (N=387) (%)
18-29	104 (27)
30-39	131 (34)
40-49	127 (33)
50-59	22 (6)
60 years or older	3 (1)
Level of Education	Count (N=387) (%)
Did not finish high school	5 (1.3)
High school diploma, GED or alternative credentials	26 (6.7)
Some college credit	49 (12.7)
Associates degree	35 (9)
Bachelor’s degree	114 (29.5)
Graduate degree	91 (23.5)
Professional degree	12 (3.1)
I prefer not to respond	55 (14.2)

^*Respondents were able to select more than one option.

^†Canada, UK, Germany, Spain, New Zealand, Australia, Hong Kong, Croatia, Turkey, UAE, Philippines, Ghana

Table 2 describes the type and number of VTE events respondents experienced, as well as their most recent anticoagulant therapy. Respondents were able to select more than one type of VTE. Sixty percent of respondents experienced one VTE event, most commonly pulmonary embolism. Direct oral anticoagulants (DOACs) were the most common anticoagulant most recently taken by respondents, with a frequency of 90% (apixaban 51%, rivaroxaban 35%).

Table 2.

VTE and Treatment Characteristics

Type of VTE*	Count (N=387) (%)
Pulmonary embolism	139 (36)
Deep vein thrombosis	93 (24)
Both	112 (29)
Other	42 (11)
Multiple VTE	116 (30)
Total Number of Thrombotic events	Count (N=387) (%)
1	232 (60)
2	96 (25)
3	34 (9)
4	4 (1)
5	7 (2)
>5	14 (3)
Most Recent Anticoagulant	Count (N=387) (%)
Apixaban	197 (50.9)
Rivaroxaban	135 (34.9)
Warfarin	31 (8)
Enoxaparin	9 (2.3)
Dabigatran	5 (1.3)
Edoxaban	5 (1.3)
Other	5 (1.3)

^*Respondents were able to select more than one option.

Quality of Life and Treatment of HMB

Prior to initiation of anticoagulation, only 18% of respondents had discussed the potential of anticoagulant-related HMB with a provider. After initiation of anticoagulation, this proportion increased to 36%. Half of respondents reported experiencing HMB for more than one year, and 30% reported a duration of greater than two years. Most respondents (86%) reported a negative impact of HMB on their QOL. Forty-three percent of respondents reported receiving treatment for HMB. Figure 1 illustrates the treatment modalities most utilized, and respondents were able to select multiple treatment modalities. The most utilized treatments were progesterone-only therapies (35% IUD, 30% oral). Twenty-one percent of patients altered their anticoagulation regimen, with a notable 10% reporting the need to temporarily hold anticoagulation while experiencing HMB. The survey did not capture whether changes in anticoagulant regimens were made under medical supervision. Twelve percent of the respondents trialed anti-fibrinolytic therapies (12%). Fourteen percent of respondents trialed CHCs including pills, patches, and rings. Nearly one third (29%) of respondents reported undergoing invasive procedures including hysterectomy, uterine artery embolization and others.

Figure 1.

Reported HMB Treatment Modalities

Bar chart showing frequency of reported treatments among study participants (n=387). Progesterone-based therapies were most commonly reported (progesterone IUD 35%, progestin-only pills 30%), followed by anticoagulant management (30%) and surgical procedures (29%). Colors indicate treatment categories: dark blue (progesterone-only), magenta (anticoagulant management), red/orange (procedures), yellow (antifibrinolytic), gray (combined hormonal), green (copper IUD), light blue (other). Percentages represent proportion of respondents reporting each treatment. HMB, heavy menstrual bleeding; IUD, intrauterine device.

(A) IUD with progesterone

(B) Birth control pill without estrogen

(C) Progesterone injection or implant

(D) Change type of blood thinner or dose

(E) Temporary hold of blood thinner

(F) Other procedures

(G) Hysterectomy or uterine artery embolization

(H) Tranexamic acid or aminocaproic acid

(I) Birth control patch or ring

(J) Birth control pill with estrogen

(K) IUD with copper

(L) Other

Treatment Preferences for HMB on Anticoagulants

When asked about their goals of treatment, 69% of respondents wished for their menstrual bleeding to decrease but continue, 41% wished for the menstrual cycles to stop completely and 40% wished for improvement in their pain. The most important factor for patients surveyed when selecting between treatment options was recurrent VTE risk (83%). Fifty-five percent indicated likelihood of decreasing HMB as important. About a fifth of respondents ranked convenience and affordability as important factors (18%), and only 12% of respondents ranked fertility concerns as being important in their decision of treatment. Patients thought it was important to discuss treatment with both their obstetrician/gynecologist (77%) and hematologist (60%). To improve treatment knowledge and decision-making, the majority (61%) preferred written explanations over pictorials, interactive exercises and videos.

Figure 2 depicts responses to which HMB treatment options participants have tried in the past or would be willing to try in the future. Forty-nine percent selected a once daily pill, 20% a pill multiple times per day, and 43% a pill only during their menstruation. Twenty-nine percent selected an injection every few months and 33% selected a procedure. Thirty-six percent selected changing their blood thinner, and 19% selected an intrauterine device. Fourteen percent would rather not receive treatment over the options presented in the survey.

Figure 2.

HMB Treatment Preferences

This bar chart depicts patient treatment preferences for heavy menstrual bleeding management. Daily oral medication was most preferred (49%), followed by cyclical therapy (43%) and anticoagulation modification (36%). Less preferred options included surgical procedures (33%), injections (29%), IUDs (19%), and no treatment (14%). Colors indicate preference categories as labeled. Percentages represent proportion of respondents indicating each preference (n=387). IUD, intrauterine device.

(A) Taking a pill once a day

(B) Taking a pill only during my menstrual cycle

(C) Taking a pill multiple times per day

(D) Making changes to blood thinner

(E) Having a surgery or procedure

(F) Taking a shot every few months

(G) Having an IUD

(H) None

(I) Having an implant in my arm

(J) Placing a vaginal ring monthly

(K) Other

Knowledge of Side Effects of HMB Treatment Options

Most respondents (74%) associated increased risk of thrombosis with estrogen-containing oral contraceptives. A minority also identified this association with estrogen-containing patches or rings (27%) and progesterone injections (22%). A sizable proportion associated thrombosis with lower-risk modalities including progesterone-only oral contraceptives (31%), implants (18%), and IUDs (progesterone 17%, copper 11%). A large portion of respondents had never heard of or been presented with various treatment options: 66% reported being unaware of anti-fibrinolytics, 48% of progesterone implants, 40% of procedures other than hysterectomy, 23% of progesterone pills, and 25% of the progesterone IUD. Overall, 36% were unaware of any hormone-based treatment options.

Discussion

This study highlights patients’ preferences for managing HMB experienced while on anticoagulants for VTE and their understanding of potential adverse outcomes associated with treatment. It underscores significant gaps in knowledge, communication, and treatment utilization, highlighting the need for more patient-centered approaches to increase treatment of HMB and improve QOL.

The survey results are in line with available data demonstrating that many patients with HMB on anticoagulants do not receive effective medical treatment despite the availability of evidence-based options (8,13). First, the survey demonstrated high unawareness and low utilization of hormone-based HMB treatment options. Only 14% reported using CHCs, despite being considered first-line treatment options for HMB (10). Concerns about their thrombotic risk may explain their limited use; however, a study by Martinelli et al. suggested no significant increase in recurrent VTE risk among anticoagulated patients using CHCs (14). Progesterone-only therapies were more frequently used, although a significant proportion of respondents (23-48% depending on formulation) were unaware of these options. Additionally, a sizable proportion of respondents associated thrombosis with progesterone-only therapies, while data indicate that low-dose progesterone-only pills (POPs) and the progesterone IUD mitigate the risk of thrombosis (15). For comparison, Table 4 highlights the thrombotic risk associated with these progesterone-only formulations in women of childbearing age in comparison to non-users (15). The survey did not discriminate between low- and high-dose POPs or specify the depot formulation, which may account for some discrepancy, as these progesterone formulations may confer increased thrombosis risk (15). Addressing these knowledge gaps through enhanced patient and provider education may improve treatment uptake.

Table 4.

VTE Risk Associated with Progesterone-Only Contraceptives

Hormonal Therapy	Formulation	Adjusted RR (95%) in comparison to non-users
Progestogen-only Pills	Oral (low dose)	0.9 (0.6-1.5)
Progestogen-only Pills	Oral (high dose)	5.3 (1.5-18.7)
Depot Medroxyprogesterone Acetate	Injection	2.7 (1.3-5.5)
Levonorgestrel	Intrauterine Device	0.6 (0.2-1.5)
Etonogestrel	Subdermal Implant	1.4 (0.6-3.4)

Adapted from: Skeith L, Bates SM. Estrogen, Progestin, and Beyond: Thrombotic Risk and Contraceptive Choices. Hematology Am Soc Hematol Educ Program. 2024 Dec 6;2024(1):644-651. (Reference 15)

Regarding anticoagulation management, 10% reported temporarily interrupting therapy—a practice that poses a significant risk of thrombotic recurrence (16,17). As opposed to interrupting therapy, altering the anticoagulation regimen may be an effective treatment for HMB, as the relative risk of HMB varies by agent with rivaroxaban resulting in nearly twice as much HMB as apixaban and four times as much as dabigatran (10). A large proportion of survey respondents were taking rivaroxaban and only 1% were taking dabigatran, representing a sizable population that may benefit from a change in anticoagulant. However, high quality data to support changing anticoagulants due to HMB is lacking and a nuanced discussion regarding the risks of bleeding and recurrent thrombosis with a medical provider is warranted. Furthermore, insurance barriers, lack of provider awareness of these different drug profiles, or the need for bridging therapy with dabigatran initiation may underly these treatment decisions.

Procedural treatments, including hysterectomy, uterine artery embolization, and other interventions, were reported by 29% of respondents. This aligns with patient preferences, as 33% expressed willingness to undergo a procedure. Although it is challenging to predict patient preferences for conservative versus procedural therapies, a multivariate analysis of patients with HMB in Finland correlated preference for hysterectomy with menstrual pain, menstrual irregularities, completed family size, higher age, lower education level, gynecologist consultations, and unemployment (18). Approximately half of the patients had trialed medical management prior to a procedure. Further, a direct choice experiment in the Netherlands that explored patient preferences between a progesterone IUD versus endometrial ablation found that the most significant factor influencing treatment choice was the presence/absence of hormones (19). Together, this suggests that the significant QOL burden associated with HMB, in the absence of conflicting family planning goals, likely drives interest in definitive procedural solutions. There is likely an element of misunderstanding of the risks and side effects associated with hormonal therapy influencing treatment choices that could be mitigated.

The survey highlighted patient priorities when selecting HMB treatments. Risk of recurrent VTE (83%) and reduction in blood loss (55%) were the most important considerations, while fewer respondents prioritized convenience, cost, or fertility preservation which may be related to demographic of respondents as most were 30-49 years of age. Additionally, there was significant interest in the use of as-needed medications to control HMB, yet few patients (12%) reported using anti-fibrinolytics and only 38% were aware of this option. The data regarding anti-fibrinolytic use and VTE risk are conflicting. In patients at high risk for VTE, including those who were postpartum, experienced traumatic hemorrhage, or had undergone an orthopedic procedure, there was not an association with increased VTE risk (20, 21, 22). However, a large Danish cohort study did reveal an increased risk of VTE and arterial thrombosis in patients taking tranexamic acid (23). Despite these limited data, in patients on therapeutic anticoagulation, they could be considered as a potential therapeutic option. A majority of patients favored written educational materials and engaging with both gynecologists and hematologists in decision-making. Multidisciplinary clinics (MDCs) have demonstrated improved outcomes in caring for patients with bleeding disorders who experience HMB, and this model may provide similar benefits for patients who experience HMB while on anticoagulation (24, 25). Further research exploring patient and provider perspectives on HMB management could help identify barriers to care and inform targeted educational interventions.

Our findings support actionable strategies to improve HMB management in patients on anticoagulation. These include the development of patient-facing decision aids to support shared decision-making; routine multi-disciplinary consultation between hematologists, gynecologists, and primary care physicians; and provider-facing educational materials to address misconceptions about hormonal therapies and thrombotic risk. Integrating these elements into care pathways could bridge the knowledge and treatment gap and promote informed, individualized treatment choices.

This study has important limitations. Given it was an online survey, there may be selection bias with those experiencing severe symptoms more likely to participate. In addition, recruiting via online platforms may have preferentially recruited digitally literate participants. The reliance on self-reported data introduces the potential for recall bias, and the survey’s anonymity precluded verification of medical diagnoses, such as inherited thrombophilia, or anticoagulation regimens leading to potential inaccuracies when characterizing VTE events or eligibility for treatment options. Further, the generalizability of these results are limited by the predominately white (87%), highly educated (90% with at least 1 year of college) population, the majority of whom live in the United States (77%). Given that Black and Latina patients are disproportionately affected by HMB, and that highly-educated patients are more likely to have access to medical care, this likely results in an overestimate of HMB treatment rates and an underestimate of the knowledge gaps related to treatment of HMB while on anticoagulation and its associated risks (26, 27, 28). Future steps to expand the study population and increase generalizability could include combining online distribution with in-person recruitment efforts at primary care, hematology, and gynecology clinics. This would have additional benefit of verification of VTE diagnoses and treatment modalities.

Conclusion

In this survey of patients anticoagulated for VTE, we identified opportunities to improve patient-centered management of HMB, for which less than half of patients received treatment. This study highlights the significant impact of heavy menstrual bleeding on the QOL of patients on anticoagulation, underscoring critical gaps in patient education, communication, and treatment. Patients desired accurate information about thrombotic risk and other side effects so they could compare risks and benefits of treatment options and prefer a multi-disciplinary approach to management. When making decisions, they prioritized risk of recurrent thrombosis, reduction in blood loss, and resolution of dysmenorrhea. Future research should expand the study population to include more patients outside of the United States of various races, ethnicities, and educational backgrounds to help further characterize these preferences and knowledge gaps related to treatment of HMB for all patients on anticoagulants. These findings should inform interventions to improve patient and provider education and decision support for management of HMB in patients on anticoagulants.

Table 3.

Unawareness Rates of HMB Treatment Modalities

Treatment Modality	Unawareness Rate
CHC pill	7%
CHC patch or ring	45%
Progesterone only pill	23%
Progesterone injection	42%
Progesterone implant	48%
Progesterone IUD	25%
Antifibrinolytics	62%
Hysterectomy	28%
Other procedures	37%