Perceptions of Endocrine Therapy in African-American Breast Cancer Survivors: Mixed Methods Study

Background Although the incidence of breast cancer is lower in African-American women than in White women, African-American women have a decreased survival rate. The difference in survival rate may stem from poor endocrine therapy adherence, which increases breast cancer recurrence. Therefore, accessible and culturally sensitive interventions to increase endocrine therapy adherence are necessary. Objective The purpose of this concurrent convergent mixed methods study was to provide further data to guide the development of the proposed culturally sensitive mHealth app, STORY+ for African-American women with breast cancer. Methods We recruited 20 African-American women diagnosed with estrogen-positive breast cancer and currently prescribed endocrine therapy. We used a concurrent convergent data collection method to (1) assess the use of smartphones and computers related to health care and (2) identify foundational aspects to support endocrine therapy adherence for incorporation in a mobile health app. Results Overwhelmingly, the participants preferred using smartphones to using computers for health care. Communicating with health care providers and pharmacies was the most frequent health care use of smartphones, followed by exercise tracking, and accessing the patient portal. We identified 4 aspects of adherence to endocrine therapy and smartphone use for incorporation in app development. The factors that emerged from the integrated qualitative and quantitative data were (1) willingness to use, (2) side effects, (3) social connection, and (4) beliefs about endocrine therapy. Conclusions Further research is needed to develop a culturally sensitive app for African-American women with breast cancer to improve adherence to endocrine therapy. Our work strongly suggests that this population would use the app to connect with other African-American breast cancer survivors and manage endocrine therapy.


Introduction
Breast cancer is the second-most common cancer among women in the United States and results in the second-highest cancer death rates [1]. Earlier detection through screening and advances in treatments have contributed to an increase in the 5-and 10-year breast cancer survival rates [2]; however, African-American women have decreased survival rates and higher mortality rates compared to White women despite lower incidence rates [3]. One potential explanation for the differences in survival rates and mortality among women with estrogen-positive breast cancer is adherence to long-term endocrine therapy, which may last for 5 to 10 years [4]. Endocrine therapy, which blocks estrogen receptors in breast cancer cells, can reduce recurrence by 40% and lower the risk of dying by one-third [5]. Overall, African-American women have low long-term adherence rates to endocrine therapy [6,7]. Therefore, interventions to assist this vulnerable population with endocrine therapy adherence are essential [4,8].
We conducted a literature review on interventions that improve endocrine therapy adherence, which identified 2 consequential gaps in existing intervention research: use of education only interventions and a lack of cultural adaptations [9][10][11][12][13][14][15][16]. Next, we examined commercially available cancer and medication adherence mobile health (mHealth) apps as a possible option to address these gaps and assist African-American women with endocrine therapy adherence [17][18][19][20][21]. Unfortunately, available cancer and medication adherence apps also have several areas that cause concern: (1) a lack of adequate development or testing in clinical practice [17,22,23], (2) a lack of input from patients or providers resulting in significant usability problems [24][25][26], and (3) a lack of scientifically valid information, with the majority of cancer-related apps created to promote a pharmacy or organization and not to assist African-American women with endocrine therapy adherence [20,26]. Importantly, over 90% of medication adherence mHealth apps were simple reminders and not effective in improving adherence [20]. Cancer mHealth apps research has explored functionality and acceptability, but only one evaluated effectiveness or clinical outcomes [17][18][19].
To address these gaps in endocrine therapy adherence and mHealth apps for medication adherence, we propose to extend our earlier teleconferencing work, Sisters Tell Others and Revive Yourself (STORY), a culturally tailored intervention that connected African-American women with breast cancer to support and educated them during initial diagnosis and treatment [27]. We wanted to explore how STORY components could assist endocrine therapy adherence in a more accessible platform such as an mHealth app. The purpose of this concurrent convergent mixed methods study [28] was to provide further data to guide the development of the proposed culturally sensitive mHealth app, STORY+ for African-American women with breast cancer. Our initial work has been reported [14,[29][30][31][32][33][34]. The aims of this study were to (1) assess the use of smartphones and computers related to health care in African-American women with breast cancer, and (2) identify foundational aspects to support endocrine therapy adherence for incorporation in an mHealth app.

Overview
The concurrent convergent mixed methods design of our study incorporated the simultaneous collection, analysis, and interpretation of qualitative and quantitative data to inform the design of an mHealth app to increase adherence to endocrine therapy in African-American women with breast cancer. We recruited 20 African-American women with a diagnosis of estrogen-positive breast cancer and prescribed endocrine therapy. We conducted an ex-ante study (ie, before the design and construction of the app). Prior to the start of the study, the University of South Carolina Institutional Review Board approved the study (Pro00085557).

Sampling Strategy and Inclusion Criteria
We combined 2 sampling schemes (ie, criterion and convenience) to select our sample. These strategies followed the works of Onwuegbuzie and Collins [35], who established typologies for sampling designs in social science research and suggested appropriate sample sizes. We also used design science research criteria (ie, innovation and evolution) to establish the legitimacy of our sample size. A sample of 20 or fewer for formative research is often employed [36,37]. We needed women who met our enrollment criteria, but we wanted heterogeneity in the sample. We used a random number generator to select potential participants for recruitment from a pool of 1577 patients. We reviewed the potential participants' medical records and selected the first 50 patients who met the inclusion criteria. This process created our first batch of potential participants. From previous recruitment efforts, we had estimated that 50 potential participants would supply our recruitment goal of 20 participants. We planned to recruit until we reached informational redundancy [38,39] (ie, both qualitative and quantitative data provided no new information). If the initial batch did not supply sufficient participants, the process was to be repeated, and data collection would be continued.
Inclusion criteria were African-American women who were 18 years of age or older, had been diagnosed with estrogen-positive breast cancer in the past 10 years, were currently prescribed endocrine therapy, and were able to speak and understand English. Reading English was not a requirement because the researcher read the informed consent and questions to the participants. Exclusion criteria were individuals with a diagnosis of psychosis, with significant cognitive impairment, or undergoing current treatment for another cancer excluding squamous cell (any type).

Recruitment and Retention
Women who were eligible based on the inclusion and exclusion criteria were recruited through a comprehensive oncology outpatient practice in South Carolina (500 new breast cancer patients annually). Our recruitment plan was covered by a Health Information Portability and Accountability Act waiver with the oncology outpatient cancer physician practice. We used the Heiney-Adams Recruitment Framework to guide our recruitment efforts [40]. The principal investigator mailed a personal letter to potential participants with a colorful and readable frequently asked questions flyer using the STORY logo (designed by focus group participants in STORY). Within 5 days of the mailing, a racially concordant researcher contacted participants by telephone. The researcher followed a script that asked if the patient had a diagnosis of estrogen-positive breast cancer and had been prescribed tamoxifen or an aromatase inhibitor. To improve recruitment and retention in the study, participants received a gift card for their time and effort. We successfully recruited 20 patients. From the original list of 50 potential participants, 19 could not be reached by phone or mail, 7 declined to participate, 2 did not meet the criteria, and 2 were not contacted because accrual was reached ( Figure 1).

Data Collection
The research team, which included nurses and software computer engineers, developed a data collection tool using Research Electronic Data Capture (REDCap [41]; Vanderbilt University). The data collection tool consisted of a qualitative interview guide (Multimedia Appendix 1) and quantitative (open-and closed-ended) questions (Multimedia Appendix 2). Data collection, which we digitally recorded, began with structured questions and ended with semistructured questions. The researcher pilot tested the REDCap data collection tool with 1 African-American breast cancer survivor before the data collection process, resulting in minor adjustments of the tool based upon the feedback. This pilot testing established the ease of data collection using REDCap during an interview, not reliability or validity.
Once the researcher received verbal consent from the participant during the telephone conversations, she scheduled the assessment appointment at the participant's choice of location, usually in the home or a meeting room in the local library. During this assessment appointment, the researcher (1) read all interview questions to the participant, (2) recorded the responses directly in the REDCap data collection tool via a tablet for the quantitative questions, and (3) digitally recorded the qualitative interview questions and participants' responses. A professional transcriptionist transcribed the deidentified recordings verbatim.

Quantitative Data
We exported the data from REDCap into Excel (Microsoft Inc) for analysis. For continuous variables, we calculated means, standard deviations, and ranges. For categorical variables, we calculated frequencies and percentages.

Qualitative Data
We used thematic analysis [42]. No software was used to analyze the qualitative data. Briefly, data analysis began by listening to the digital recordings of participant responses to the semistructured interview questions. Next, 2 researchers read and reread the transcripts and began open line by line coding of the data. We coded significant and salient phrases and words within each transcript. We identified and discussed themes from these codes until consensus was reached.

Data Integration
For clarity, we organized the findings by aim. Most findings are reported with quantitative data and supporting qualitative quotes [39]. In some instances, qualitative themes emerged that were not explored in the quantitative questions and vice versa. In these cases, the results are displayed separately.

Smartphone and Computer Use for Health Care
Overwhelmingly, the participants preferred using a smartphone for health care over using a computer.

I pretty much do everything on my cell phone because it's always with me. [Participant 13]
With respect to health care use, participants reported communicating with health care providers and pharmacies most frequently, followed by exercise tracking, and accessing the patient portal (Table 2). "Well, I went through the program with a dieting thing. It was to help you lose weight and it sets you up. They send you a scale. You weigh every day. They give you a coach. She logs in every day with different ideas and stuff for you to do meals to plan and it also helps, but my insurance will come down instead of paying a surcharge." [Participant 20] 3 (15) 1 (5) Diet tracking, coaching, or management a 2 (10) 1 (5) Personal health records a A representative quotation is not available.

Overview
We identified 4 aspects of endocrine therapy adherence and smartphone use that may guide app development. The factors that emerged from the integrated qualitative and quantitative data included (1) willingness to use, (2) side effects, (3) social connection, and (4) beliefs about endocrine therapy.

Willingness to Use an App for African-American Breast Cancer Survivors
The majority of participants (17/20, 85%) stated that they would use an mHealth app to assist with endocrine therapy adherence. They specifically mentioned an online community of other breast cancer survivors and African-American-tailored graphics (ie, emojis, videos). The qualitative interviews provided additional details on the importance of an app for African-American breast cancer survivors:

On the [majority] of apps, I see it is still [for] white [women]. I don't normally see too many blacks that I can reach out to. I would love to share with an African-American female [with breast cancer] what I've been through. Every app that I [see] dealing with breast cancer shows white [women], and I would love to talk to African-American women, women of color.
[

Self-Reported Side Effects
In patients' self-reported experiences (ie, reports were not verified by health care providers) with the side effects of endocrine therapy (Table 3), almost one-third (6/20, 30%) of participants stated that they called their health care provider about the side effects, with 15% (3/20) reporting they made an appointment with the health care provider to discuss the side effects. In addition, 80% (16/20) of participants talked with a family member, friend, or significant other about the side effects.

Beliefs About the Value of Endocrine Therapy
There were many motivators, especially external motivators, for patients to continue endocrine therapy. The most prominent motivators for endocrine therapy adherence included the desire to live longer (16/20); children (16/20); and religion, church, or a higher being (15/20

General
This concurrent convergent mixed methods study provides additional data to guide the development of STORY+. We assessed the use of smartphones and computers related to health care in the targeted population and identified foundational mHealth features to support endocrine therapy adherence. Overall, the participants used smartphones more frequently than they used computers for health care, especially for social interactions such as social media, messaging, and email. This finding suggests that African-American breast cancer survivors are more likely to use an mHealth app than they are to use a webpage via computer to assist with endocrine therapy adherence. Overwhelmingly, African-American breast cancer survivors use smartphones in managing their health through communication with their health care provider and pharmacy to promote their health.
While fewer participants reported using mHealth apps for medication tracking and health care record management, qualitative results indicated that combining social and cultural features with mHealth functions would be desirable. Other researchers have also identified the need for racial and cultural content [43][44][45][46][47][48][49]. Our study verifies that African-American women do not have a culturally sensitive mechanism to track symptoms and discuss symptoms with health care providers. We concluded that foundational STORY+ features should include the following: an online community of other African-American breast cancer survivors, tailored graphics, information about prescribed endocrine therapy and its potential side effects, and a method for tracking side effects, sharing the frequency and severity of side effects, and recording medication adherence. Previous findings [50] have also supported the use of interactive features to promote positive health outcomes-participants overwhelmingly supported the development of an mHealth app to assist with endocrine therapy adherence and social connections with other African-American breast cancer survivors.

Limitations
The study is not generalizable beyond the immediate needs of the research team; however, the results of this study provide a foundation for the mHealth app development, STORY+. The literature suggests that other breast cancer survivors experience similar endocrine therapy side effects [51].

Conclusions
This concurrent convergent mixed methods study established the use of smartphones by African-American breast cancer survivors for health care management. We also identified foundational features for STORY+. This work recognizes that one app does not fit all needs and advances the science of cultural and racial appropriate mHealth apps. Future work will include the development and testing of STORY+ for African-American breast cancer survivors to promote endocrine therapy adherence.