For each component of individual belief, the health belief model16 was adapted to include relevant race-specific contexts for US black women in health care settings. For example, susceptibility to or risk of developing a symptom among black women may be associated with a race-specific context, such as the US health care environment, in which the severity of physical ailments among black women has been reported to be minimized.18-21 D&C indicates dilation and curettage; EB, endometrial biopsy; and TVS, transvaginal ultrasound.
eTable. Questions Added to Test Fidelity of Themes in Final Interviews
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Doll KM, Hempstead B, Alson J, Sage L, Lavallee D. Assessment of Prediagnostic Experiences of Black Women With Endometrial Cancer in the United States. JAMA Netw Open. 2020;3(5):e204954. doi:10.1001/jamanetworkopen.2020.4954
What are the beliefs, interpretations, and experiences of black women with endometrial cancer with regard to vaginal bleeding and the disclosure of this symptom in health care settings?
In this qualitative study of the interview responses of 15 black women with endometrial cancer in the United States, participants described knowledge gaps and silence about menopause, misinterpretation of vaginal bleeding, and responses from first-line health care practitioners that were not aligned with the risk of endometrial cancer among black women in the United States.
Among participants in this study, several aspects of the prediagnostic experience suggested novel modifiable factors that may be useful in the development of targeted interventions to improve the rates of early diagnosis among black women with endometrial cancer.
Black women with endometrial cancer have a 90% higher mortality rate than white women with endometrial cancer. The advanced disease stage at which black women receive a diagnosis of endometrial cancer is a major factor in this disparity and is not explained by differences in health care access.
To describe the prediagnostic experiences of symptoms and symptom disclosure among black women with endometrial cancer.
Design, Setting, and Participants
This community-engaged qualitative study developed an interview guide to collect data during semistructured interviews among a sample of 15 black women with endometrial cancer in the United States. Interviews were conducted in person or via a secure conferencing platform. An exploratory and descriptive content analysis was performed using iterative rounds of inductive coding, case summaries, and coanalysis with community input to identify emergent themes. Data were collected from October 3, 2017, to April 15, 2019, and the descriptive content analysis was performed from October 11, 2017, to May 6, 2019.
Main Outcomes and Measures
Beliefs, interpretations, and experiences of black women with endometrial cancer from symptom onset to diagnostic confirmation of cancer.
Participants included 15 women who self-identified as black or African American and ranged in age from 31 to 72 years. Eight participants lived in the Puget Sound region of Washington, 2 participants lived in California, and 1 participant each lived in Alabama, Michigan, Louisiana, Georgia, and New York. Twelve participants were receiving adjuvant therapy during the study, which indicated that they were either in a high-risk group and/or had advanced-stage disease. Thirteen participants had health insurance at the time of symptom onset, and all participants had elected to receive cancer treatment. Participants described knowledge gaps and silence about menopause, misinterpretation of vaginal bleeding, and responses by first-line health care practitioners that were not aligned with the risk of endometrial cancer among black women in the United States.
Conclusions and Relevance
The responses of interviewed black women with endometrial cancer suggest that several mechanisms may be associated with a delay in care before diagnosis among this high-risk population and represent modifiable factors that may be useful in the development of targeted interventions to improve the rates of early diagnosis among black women with endometrial cancer.
Racial disparity in endometrial cancer survival rates is substantial and has increased during the past decade.1 Endometrial cancer is the most common gynecologic cancer and is diagnosed in 1 in 37 women in the United States.1 Black women with endometrial cancer have a 90% higher 5-year mortality risk compared with white women, with a 5-year mortality rate of 39% among black women compared with 20% among white women.2 This higher mortality rate is associated in part with disparities in cancer stage at diagnosis; only 53% of black women receive an early diagnosis.1
Most cases of endometrial cancer are diagnosed in early stages owing to the early onset of clear symptoms, such as postmenopausal bleeding or markedly abnormal premenopausal bleeding. Although national guidelines for diagnostic examinations of postmenopausal bleeding exist,3,4 the guidelines assume that patients promptly report symptoms and that health care practitioners take immediate action. Delays in medical care in response to these symptoms may result in potentially untreated cancer. Yet, to date, little is known about the ways in which black women recognize, interpret, or experience postmenopausal bleeding.5 Although cancer stage at diagnosis has been reported as a major factor in the mortality gap between black and white women,6 few studies have focused on identifying modifiable factors.7-9 To address this gap, we investigated the beliefs, interpretations, and experiences of symptom onset and diagnosis among black women aged 31 to 72 years with endometrial cancer.
This qualitative community-engaged study was performed as a partnership between an institutional research team (led by K.M.D.) and a community partner (B.H.) who is the founder and chief executive officer of Cierra Sisters, a cancer support and advocacy organization for African Americans in the Puget Sound region of Washington. The process of community engagement and the previous experience and positionality of the research team are detailed in a previous publication.10 In brief, we used the tools of storytelling, goal setting, and iterative collaboration to achieve 3 goals: (1) building an equitable researcher and community member (K.M.D. and B.H., respectively) partnership, (2) codeveloping the qualitative study, and (3) successfully securing research funding. This study adhered to the Consolidated Criteria for Reporting Qualitative Research (COREQ) reporting guideline.11 The study was reviewed by the institutional review board of the University of Washington and was deemed exempt based on the University of Washington Human Subjects Division criteria for including interactions involving only interviews with adequate provisions to protect the privacy of participants and maintain the confidentiality of the data. All participants provided verbal consent to be recorded and to have their data and responses published. Each participant received a $50 gift card for participation.
Community involvement occurred actively, with collaboration throughout the design, recruitment, data collection, and data interpretation phases. From September 6, 2017, to April 2, 2019, we recruited a purposive sample of 15 black women who had been diagnosed with endometrial cancer. A sample of 15 women was our a priori target given the likelihood of thematic saturation at this sample size12 and the expectation that recruitment would be resource intensive given the lack of a visible community of black women with endometrial cancer. Recruitment occurred via several mechanisms: direct contact with the gynecology-oncology clinics in the University of Washington medical system, outreach through the social networks of our community liaison and the Cierra Sisters organization, dissemination of study information on social media and on the newly formed Endometrial Cancer Action Network for African-Americans website,13 and posts on a general research recruitment website developed by the Institute of Translational Health Sciences at UW Medicine.14 Individuals were eligible for participation if they were 18 years or older, had a current or previous diagnosis of endometrial cancer (assessed through a series of confirmatory questions), self-identified as black or African American, were willing to discuss personal experiences, and had sufficient English language proficiency (assessed during initial contact).
The interview guide (Table 1) was developed with 3 organizing frameworks in mind and thus focused on the defined interval from symptoms to diagnosis,15 the ways in which beliefs were used to make health decisions,16 and the association of previous racialized experiences with actions and beliefs.17 The interviews focused on 3 events: menopause (to provide a knowledge base for participants’ interpretations and actions), onset of vaginal bleeding, and diagnostic experience after symptom disclosure. Questions were predominantly open ended (eg, “Tell me what menopause was like for you”), with optional probes (eg, “How did your menstrual periods change?”) included to ensure consistent and complete data collection. A few directed questions (eg, “Did your medical practitioner ever mention your race during this discussion?”) were also included. A glossary of medical terms was developed for interviewer and participant reference as needed.
The interview guide was developed through several iterations within the community partnership, and both the development process and the final interview guide are available in a previous publication.10 The Figure illustrates our conceptual framework, which was adapted from the health belief model16; this model suggests that 6 factors are associated with an individual’s health behavior: risk susceptibility, risk severity, benefits to action, barriers to action, self-efficacy, and cues to action. Two authors (K.M.D. and B.H.) conducted and recorded interviews in person (n = 7) or via a secure conferencing platform (n = 8). Interviews were transcribed verbatim, verified against the recordings, and deidentified. The participants were deidentified to the research team by stripping the transcripts of any names, uploading the transcripts to the coding software by participant number, and storing the name-to-patient number linkage document in a separate file.
We conducted an exploratory descriptive content analysis22 by first developing a codebook through an iterative process; using menopause, symptom onset, and diagnostic experience as a loose structure, we performed open inductive coding of the first 3 interviews, reviewing the data for ideas and organizing categories to generate the preliminary codebook. We then reviewed previous literature on the menstrual and menopausal experiences of black women23-27 and solicited feedback from our community partner (B.H.) to revise and update the codebook, which was used for all subsequent interviews.
For each transcript, a primary coder (either K.M.D., L.S., or J.A., on a rotating basis) completed the first analysis and created a case summary with relevant quotes. A secondary coder (either K.M.D., L.S., or J.A., on a rotating basis) then reviewed the transcript, the coding, and the case summary to add additional codes or points of clarification and ensure intercoder reliability. Coding was completed using Dedoose software, version 8 (SocioCultural Research Consultants). We then completed group reviews, reporting each case summary and identifying common themes until we reached saturation (ie, the point at which no new data was being obtained) at interview 12. Our community partner (B.H.) reviewed all case summaries, provided notes from the interviews themselves, and shared feedback and interpretations during group review meetings. Content areas with minimal or conflicting data were identified.
In lieu of performing individual member checking (which is not recommended to improve rigor),12 we incorporated participant feedback (which is important to ensure the trustworthiness of a community-engaged process)28 by adding targeted questions to the final 3 interviews. These questions were aligned with specific themes to clarify conflicting data and test the fidelity of our emergent themes (eTable in the Supplement). Data were collected from October 3, 2017, to April 15, 2019, and the descriptive content analysis was performed from October 11, 2017, to May 6, 2019.
Among 19 potential participants screened, 17 women were eligible; of those, 15 women enrolled in the study. Participants were aged 31 to 72 years and lived in the Puget Sound region of Washington (n = 8), California (n = 2), Alabama (n = 1), Michigan (n = 1), Louisiana (n = 1), Georgia (n = 1), and New York (n = 1). Twelve participants were receiving adjuvant therapy, indicating that they were either in a high-risk group and/or had advanced-stage disease. Thirteen participants had health insurance at the time of symptom onset, and all participants had elected to receive cancer treatment (Table 2).
The qualitative content analysis revealed the following themes: (1) unclear definition of normal vs abnormal menopause, (2) misinterpretation of bleeding symptoms in the context of previous personal or familial reproductive health events, (3) disclosure of vaginal bleeding cued by a waiting period or worsening symptoms, (4) vague responses from health care professionals that did not communicate risk, and (5) shock and surprise at the eventual diagnosis (Table 3).
Women reported a lack of clarity about whether symptoms were normal or abnormal during and after menopause. One woman stated, “Anything can happen in menopause” (patient 1), and another said, “I don't know that I had any real expectations what menopause would be like” (patient 3). Uncertainty in knowledge about this life transition was situated in the context of stigma and silence from peers, friends, and family regarding bleeding patterns and an absence of familial experiences with natural menopause owing to the prevalence of hysterectomy in the black community29,30 (Table 4).
Although many women noted that friends and family were their sources of knowledge, these statements were frequently accompanied by the judgment that such information was inadequate. Women directly communicated the inadequacy of their knowledge by expressing the desire to have known more about menopause. One participant said, “And we never…it was not enough education in the community to alert women, especially African American women, of what to expect. And what some of the changes would be. So I was very uneducational [sic] when it come to that” (patient 2). In addition, women described silence about bleeding symptoms during and after menopause that occurred in conversations throughout their lives, which also featured elements of stigma and shame. One woman relegated discussion of bleeding as inappropriate, stating, “It wasn’t a normal topic of conversation” (patient 10), while another associated silence with the larger experience of stigma and shame among black women (Table 4). Another participant elaborated, stating, “I’ve never really talked to a lot of people about menopause. Most people I know don’t want to discuss it. It’s as simple as that” (patient 6).
Another factor associated with the lack of knowledge regarding normal vs abnormal symptoms was the prevalence of hysterectomy in participants’ social networks, which was reported by 2 participants. Mothers, sisters, and friends were all perceived as insufficient sources of knowledge after they had undergone this procedure (Table 4).
All participants experienced abnormal or postmenopausal bleeding before receiving a diagnosis of endometrial cancer. Underlying their disparate experiences was an interpretation of bleeding symptoms in the context of their own previous reproductive health events (Table 4).
Participants expressed 1 of 3 initial interpretations of vaginal bleeding: (1) it was a sign of resumed menstrual cycles or continued menopause; (2) it represented an unknown event that was not associated with cancer; and (3) it indicated a serious health issue, including cancer. The first interpretation was the most common. Women directly compared the onset of new bleeding with their previous experiences of menopause or menstruation. One participant described thinking, “Hmm, I’m just not done with this yet” (patient 2) when referring to new vaginal bleeding that occurred several years after menopause. Another participant, also several years past menopause, interpreted her vaginal bleeding as a menstrual period, stating, “Every time my period would go away, it would come back” (patient 8). Along with this interpretation came a resulting lack of concern.
Women who expressed initial nonspecific worry understood that their symptoms were not normal, but they lacked direction regarding the potential association of these symptoms with cancer or other serious illnesses. One woman stated, “It was something wrong, but I didn’t know what.” Previous reproductive health experiences were important factors. As another woman stated, “'I had a very, very light pinkish bleeding, and that's what started me to think that something was off because I had already gone through the menopause, and I had totally stopped any periods at all” (patient 2). Another woman said that her previous experiences with heavy menstrual periods mitigated her worry (Table 4).
When women expressed an immediate interpretation of bleeding as a potentially serious health issue, they did so with reference to their previous reproductive health experiences. The increased severity of initial symptoms in contrast with previous menstrual experiences met the threshold of alarm. One woman woke up with a soaked mattress from new-onset bleeding. Another described a public experience of having an instant menstrual period (Table 4). Although few women experienced acute vaginal bleeding, our findings were supported by another participant who did not experience severe bleeding but remarked that increased bleeding severity would have met her personal threshold for taking action, saying, “And had it been heavier, I might've would've called the doctor” (patient 11).
Not all study participants experienced a natural onset of menopause before receiving a diagnosis of endometrial cancer. Among the 3 participants who had not yet experienced menopausal symptoms, 2 women noted marked changes in their menstrual patterns (eg, spotting between cycles, more severe bleeding during cycles, or both), and 1 woman, who did not note specific changes, had a history of irregular and severely heavy menstrual cycles that required the receipt of blood transfusions, which were normalized by health care practitioners, beginning at age 19 years. In response to her bleeding symptoms, she stated that she “got to the point where I quit talking about it. I just felt that, that was normal for my body…and that this is what was gonna go on. You know, I was kinda scared, but I still thought, well, none of the doctors are saying that it's anything, so I just need to, to deal with it” (patient 14).
We defined a diagnostic experience as the interval from the first disclosure of symptoms to a health care practitioner to the diagnostic confirmation of endometrial cancer. This interval was characterized by 3 events: a waiting period or worsening symptoms that prompted symptom disclosure, a mismatch between the participant’s risk of cancer and the practitioner’s response, and an experience of shock or surprise at diagnosis.
We found that a common strategy among women was to endure the symptoms before reaching a personal threshold, at which point they decided that taking action was the next appropriate step (ie, the cue to action).16 One participant described her endurance of symptoms by stating, “Well this has got to be over soon, one day” (patient 1), while another remarked that she could “just live with it” (patient 4). Endurance of symptoms emerged across the cohort as a lack of reported action beyond the maintenance of hygiene (eg, purchasing sanitary pads).
The cues to action for symptom disclosure included a personally defined waiting period, the onset of heavier vaginal bleeding or other symptoms, or both. Regarding the waiting period, 1 individual decided to disclose her bleeding symptoms “because it had been happening for just about a year” (patient 13). Many women said that their concern increased as bleeding became progressively heavier (Table 4). Other symptoms that prompted disclosure included cramping, bloating, fatigue, and nausea, all potential signs of cancer progression. One woman noted her concern about fatigue after bleeding for more than 6 months, stating, “The fatigue was just totally not me. I mean, it was like somebody drugged me. Totally uncharacteristic of my normal self” (patient 11). Another participant described the cramping symptoms as more concerning than the bleeding (Table 4).
We found a pattern of health care practitioner responses after symptom disclosure that lacked explicit mention of cancer as a potential explanation for vaginal bleeding despite the fact that women reported symptoms of concern. This mismatch was noted during the women’s recounting of reassuring remarks from their practitioners (Table 4). One participant could not recall when she reported symptoms, as doing so had prompted normalization of the symptoms by her practitioner. She recalled the practitioner saying, “Well, sometimes that happens, you know?” (patient 4).
All but 1 of the women received recommendations for further diagnostic testing in the form of a transvaginal ultrasound, endometrial biopsy, or dilation and curettage procedure. The theme of practitioner response mismatch with women’s health risks was also observed in the vague rationale given for the need for diagnostic tests. Women received vague explanations, such as, “I just wanna make sure there's nothing going on in there” (patient 1). After ultrasound results indicated enlarged measurements of the endometrium (as reported by participants) that suggested cancer risk, the vagueness in descriptions continued, with one practitioner saying, “Oh, it looks like there's heavy lining in your uterine walls, so I'm just gonna send you to this doctor and have it checked out” (patient 1). No women reported receiving information that a transvaginal ultrasound was a screening and not a confirmatory test.
The combination of reassurance and the lack of explicit discussion of cancer risk led to shock and surprise when the endometrial cancer diagnosis was received. One woman recalled the following discussion with her practitioner: “I said to her, ‘What do you think this is?’ And she said to me, ‘If I was a betting woman, we say you have cancer.’ And I was in the middle of [censored] street on a sunny day carrying my lunch that I had just picked up. And I dropped it” (patient 13).
Participants also expressed surprise that cancer could have been associated with their symptoms at all. As 1 woman expressed, “Cancer never came to my mind, never entered my spirit” (patient 12).
To our knowledge, this is the first study to explore the prediagnostic experiences of black women with endometrial cancer. The women in the study described knowledge gaps and silence about menopause, misinterpretation of vaginal bleeding in the context of these knowledge gaps, and responses by first-line health care practitioners that were not aligned with the high-risk status of the women. In the following paragraphs, we place these findings in the larger context of US black women’s gynecologic health and consider the ways in which, within these experiences, multiple novel factors associated with care delays may exist both before and immediately after symptom disclosure to health care professionals.
Both the nature and interpretation of symptoms are associated with the length of time from symptom onset to the perception of individual risk. The interpretation of symptoms as not new, not bothersome, or not painful has been associated with care delay.15 These interpretations, however, are subjective and dependent on the individual’s previous experiences and norms. The women in our study did not have a clear expectation or standard definition of menopause or an awareness of their menopause knowledge gaps, which they associated with silence regarding vaginal bleeding among family and friends and the occurrence of hysterectomy among other family members. We hypothesize that the commonality of hysterectomy among US black women (who are 3-4 times more likely to have a premenopausal hysterectomy compared with US white women)29,30 may be associated with a lack of familial and community communication about menopausal transition. Unclear definitions of menopause and expectations about this transition may leave black women vulnerable to labeling any experience—bleeding, pain, bloating, or fatigue—as normal.
In addition, US black women have a high prevalence of fibroids or leiomyomas,31-33 which are benign tumors of the uterus. The presence of fibroids has been associated with abnormally heavy and irregular menstrual cycles.34 In a previous mixed-methods study, reproductive-aged black women were reported to have a concerning level of normalization of severe bleeding symptoms.35,36 We hypothesize that a phenomenon may exist in which postmenopausal black women do not regard the onset of vaginal bleeding as new because they have a history of irregular menstrual cycles or as bothersome because the new bleeding and pain are less severe than they were during previous experiences in their youth. Given that pain among black people in US health care settings has been reported to be discounted and undertreated,18-21 it is worth considering the possibility that black women may be discounting their own pain in adaptation to this environment.
The women’s reported cues to action (ie, increased bleeding severity and pain) to seek health care did not appear to prompt an appropriate level of concern from health care practitioners given the risk of endometrial cancer among black women.3,37 This misalignment between health risk and practitioner response was observed in the lack of communication about cancer risk after the women disclosed their bleeding symptoms. Although it is estimated that only 10% of women with these symptoms will be diagnosed with endometrial cancer,37 in the context of the knowledge gaps and misdirected coping strategies we observed among the black women in our study, such minimization of cancer risk may have exacerbated the delay in diagnosis. Without a clear discussion of cancer risk, women may continue to have assumptions of normality and use coping strategies, such as endurance of symptoms, that may be associated with subsequent delays in scheduling follow-up tests or with not receiving follow-up tests at all.
This study’s strengths include a geographically diverse cohort, a sample that was large enough to reach thematic saturation,12 and a community-engaged design, which allowed for richness, depth, and intimacy in interviews about topics that are often regarded as taboo among black women.38 This study also has limitations. Because this study was qualitative, the findings were not meant to be generalizable but instead aimed to clearly communicate the experiences of the participants. Our sampling strategy may have overrepresented women with health insurance and health care access and likely underrepresented the full breadth of barriers associated with household income and insurance.39 As a community liaison–research group partnership, we purposively did not collect extensive demographic information (eg, income strata and educational level) owing to feedback from our community partner (B.H.), who suggested that these types of questions may deter women from participating in the study and create a barrier to the establishment of a comfortable and trusting environment for the women. By identifying and including women who had already received a diagnosis of endometrial cancer, we risked recall bias. Subsequent research to translate these findings into a large-scale national survey will add quantitative context and provide information about other environmental associations with the themes reported in this study.
Black women with endometrial cancer described multiple experiences that were associated with care delay independent of health care access. Given the importance of cancer stage at diagnosis to overall cancer prognosis and the lack of interventions to address higher mortality rates among black women with endometrial cancer, these findings represent an important first step toward the development of evidenced-based interventions.
Accepted for Publication: March 11, 2020.
Published: May 15, 2020. doi:10.1001/jamanetworkopen.2020.4954
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2020 Doll KM et al. JAMA Network Open.
Corresponding Author: Kemi M. Doll, MD, Department of Obstetrics and Gynecology, University of Washington School of Medicine, 1959 NE Pacific St, PO Box 356460, Seattle, WA 98195-6460 (email@example.com).
Author Contributions: Dr Doll had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Doll, Hempstead, Alson.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Doll, Hempstead.
Critical revision of the manuscript for important intellectual content: All authors.
Statistical analysis: Doll, Sage.
Obtained funding: Doll, Hempstead.
Administrative, technical, or material support: Alson, Sage.
Supervision: Doll, Alson, Lavallee.
Other - Qualitative analysis: Alson.
Conflict of Interest Disclosures: Dr Doll reported receiving grants from the National Comprehensive Cancer Center Network Foundation and the Robert Wood Johnson Foundation during the conduct of the study. Ms Hempstead reported receiving grants from the National Comprehensive Cancer Network Foundation and the Robert Wood Johnson Foundation during the conduct of the study. Ms Alson reported receiving grants from the National Comprehensive Cancer Network Foundation during the conduct of the study. No other disclosures were reported.
Funding/Support: This work was funded by grant 75203 from the Robert Wood Johnson Foundation (Dr Doll, Ms Alson, Ms Sage) and a grant from the National Comprehensive Cancer Network Foundation (Dr Doll, Ms Alson, Ms Sage).
Role of the Funder/Sponsor: Neither funding agency had a role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation and review of the manuscript; and decision to submit the manuscript for publication. The National Comprehensive Cancer Network Foundation reviewed and approved this publication for submission but was not involved in the preparation of the manuscript in any way. Any opinions, findings, and conclusions expressed in this article are solely those of the authors and do not necessarily reflect the opinions of the National Comprehensive Cancer Network or the National Comprehensive Cancer Network Foundation.
Additional Contributions: We thank the interview participants who chose to share their experiences with us.
Create a personal account or sign in to: