Opportunities to Improve Women's Health: Engaging Racial/Ethnic Diverse Women to Provide Biospecimens for Research

Abstract

Objective: Inclusion of women in biomedical cancer research have the potential to close gaps in cancer health disparities and improve adjuvant therapies for women; yet samples needed to advance this area of science are lacking. We developed low-cost educational recruitment strategies to increase our collection of biospecimens from women.

Materials and Methods: Women diagnosed with hormone receptor positive (HR+) breast cancer that initiated hormonal therapy were recruited from three integrated health systems. The analytical sample (n = 144) consisted of women who consented but did not return a saliva sample within 1 year of the initial assessment (baseline). Brief informational recruitment materials were developed via published literature and preliminary data. Women received recruitment materials, which included a personalized information letter, a colorful low-literacy instruction sheet, a postage-paid envelope, and collection kits. We evaluated intervention materials and performed descriptive and bivariate statistics to describe factors associated with biospecimen donation.

Results: Of the total sample, 61% were white and 34% were black. Overall, 29 surveys (20%) and 25 (17%) saliva kits were returned. Women found the materials helpful and easy to read and understand. Women with higher levels of functional well-being and lower ratings of religiosity were more likely to return biospecimens (p < 0.005) after receiving enhanced materials.

Conclusion: This article provides recruitment strategies to enhance biospecimen samples among women. Receipt of brief informational print materials inclusive of personalized messages enhanced our outreach strategies and increased our overall biospecimen provision rate by 17%. The inclusion of messages with a focus on spirituality and other cultural messages may further increase biospecimen provision in racial/ethnic diverse groups of women; however, further study is needed to support this claim.

Clinical Trail Registration Number: NCT02992730.

Introduction

Breast cancer is the leading cancer diagnosed in women in the United States.^1,2 Furthermore, women encounter an array of severe side effects and health complications after systemic breast cancer (BC) treatments (e.g., lymphedema).^3,4 Genomic research contributes to the development of effective modalities (e.g., precision medicine) throughout the BC continuum of care.^5,6 One opportunity to support women's health is improving participation in precision medicine such as biospecimen research.^7–9 Biospecimen collection (e.g., saliva sample) is a quality data-driven methodology to understand genetic markers in the advancement of personalized medicine.^10,11 This method is supported by the Implementation Science framework, which promotes the translation of practices to address barriers to genomic research.^12,13 Historically, it has been challenging to recruit women in biomedical research until the National Institutes of Health Revitalization Act was signed in 1993 mandating including women in human subject research.¹⁴ 27 years later since this law, and inclusion of women when compared with men in biomedical research programs are still limited, and are further exasperated in black women because of medical mistrust concerns,^15–18 due to the mistreatment and exploitation of these populations in research. Improvement in recruitment strategies for participation in biospecimen research is a critical effort to increase samples of women and to advance precision medicine in women's health.

Black women are less likely to donate biospecimen samples,^19,20 when compared with white women. A prior study found that African Americans did not participate in biospecimen collection or biobanking programs because of privacy concerns, they were not approached by medical researchers, or they felt that researchers lacked transparency with regard to study information.^21,22 The first large cohort study of black women identified predictors of DNA donation such as older age, a physician visit within the past 2 years of data collection, and history of breast, colon, and cervical cancer screening.²³More recently we investigated a diverse sample of BC survivors, and found that women with high ratings of religiosity were less likely to provide a biospecimen and women with lower-stage cancer (vs. higher stage) were more likely to provide a biospecimen.²⁴ Although few studies have explored context, most within the context of clinical trials participation, results appear mixed with regard to the role of religiosity. One study of terminally ill cancer patients found a positive correlation between religiosity and clinical trials participation, whereas another study of Hispanic Americans reported a lack of support for clinical trial participation among individuals with higher ratings of religiosity.^10,11 Although prior research has highlighted the diversity needs among biospecimen donation in women of color, information is lacking about recruitment strategies to improve biospecimen collection among racial/ethnic diverse groups of women. Inclusion of racial/ethnic women from diverse backgrounds in biomedical research for cancer and exploring the role of psychosocial factors are necessary to address long-standing BC survival disparities.^1,2

The objective of this pilot study was focused on identifying and delivering low-cost recruitment strategies to enhance biospecimen donation among BC survivors that did not return specimens. The aims were (1) to assess acceptability of brief informational print materials and (2) to identify demographic and psychosocial factors related to biospecimen donation among women. Improving the use of educational materials in biomedical recruitment may maximize the National Cancer Institute (NCI) priorities related to biospecimen usability, collection, handling, and processing.^10,11 This study aligns with NCI's goals to develop educational materials around biospecimen donation and suggest recruitment strategies to address challenges with inclusion of women in genomic-related research.

Materials and Methods

Sample

This was a secondary analysis of baseline data from the Women's Hormonal Initiation and Persistence (WHIP) prospective study (2012–2017). Participants were recruited from three integrated health networks and community outreach activities. Further details about WHIP data collection procedures are previously described.¹² Women were eligible if they were diagnosed with hormone receptor (HR)-positive BC within 12 months, ≥21 years old, and initiated adjuvant endocrine therapy within 1-year postdiagnosis. For the purpose of this study, self-identified black/African American and white women were only included if they did not provide a biospecimen during baseline (n = 144). Institutional Review Board (IRB) approval was obtained by all participating institutions. Figure 1 illustrates the study schema.

FIG. 1.

Biospecimen study schema.

Biospecimen collection process

Women diagnosed with HR-positive BC included in the WHIP study were sent a study flyer/letter with the option to provide a saliva sample. A trained clinical research assistant led the verbal consent process; in this process women were informed in detail about the biospecimen collection procedure and invited to ask questions about the process. Specifically, women were told that their samples would be safely secured in a laboratory with other samples from women in the study, and that only the investigator and laboratory technicians would have access to the samples. Participants were also informed that investigators on the study would determine which tests were to be used on the samples, to better understand how adjuvant endocrine therapy medications function in women bodies. There were no tests or results distributed back to study participants. Consented women who enrolled in the study completed a baseline survey and received sample kits, if they agreed to provide a saliva sample (n = 575). Saliva was collected using the sample kits that were mailed to each participant, along with a $15 gift card in the same envelope. “Each kit included a 1.22-fl oz. bottle of Crest Pro-Health mouthwash, a screw top biospecimen collection cup, a ziplock plastic bag with an absorbent cloth inside, and a stamped return bubble-wrapped envelope.”²⁴

Sample kits were given to women along with low-literacy education materials written at a sixth-grade reading level that provided instructions on how to provide the sample. In these instructions, women were asked to complete the following steps: (1) rinse the mouth with plain water; (2) swish three quarters of the mouthwash in their mouths for 45 seconds; (3) deposit the sample in the biospecimen cup; (4) place the sample inside the plastic Ziploc bag, and seal securely inside the bubble wrap envelope; (5) seal the envelope and place in the addressed postage page envelope. “Once received, research staff logged samples and sent them directly to the Molecular Epidemiological Laboratory at Georgetown University Medical Center for processing. Cell pellets were harvested from mouthwashes by centrifugation, followed by washing with phosphate-buffered saline solution twice for long-term storage in a −80°C freezer that is connected to online monitoring. Cell pellet sizes were roughly estimated as none, very small, small, medium, and large. DNA was purified using Qiagen's QIAamp DNA mini kit. DNA concentrations were determined by ultraviolet spectrophotometry.”²⁴

Brief information materials

In an effort to increase biospecimen provision among our cohort, we developed brief informational print materials. Using clear communication strategies and health literacy techniques, we developed a tailored letter from the PI that included personalized messages about biospecimen donation. Along with the letter we created a four-step instruction sheet for saliva collection (Fig. 2).

FIG. 2.

Saliva collection instructions.

Enhanced aforementioned print materials were sent to women who did not provide a biospecimen sample after baseline data collection. They also received another biospecimen kit²⁴ and a stamped self-addressed envelope in which to return the biospecimen sample. Envelopes were sent with prepaid postages for convenience.

Data collection

In addition to the enhanced print materials and saliva collection kit, women were asked to complete an assessment survey about the materials and provision of biospecimen data for clinical research.

Measures

Outcome. The outcome of interest was provision of saliva biospecimen kits (yes vs. no). Sociodemographic factors included race, age, education, and marital status. Clinical factors. Clinical data were abstracted from hospital records and pathology records. Those factors included BC stage and surgery type (lumpectomy or mastectomy). Health care process factors. PSQ-18 assessed women's satisfaction with their health care.²⁵ PSQ-18 includes seven domains: financial concern (Cronbach's α = 0.73), interpersonal manner (Cronbach's α = 0.66) access convenience (Cronbach's α = 0.75), technical quality (Cronbach's α = 0.74), time spent with doctor (Cronbach's α = 0.77), and communication (Cronbach's α = 0.64). Higher scores represented higher levels of satisfaction. Psychosocial factors. Religiosity was assessed using Lukwago (2001) 9-item scale.²⁶ Women responded to items using a 4-point Likert scale. Possible scores ranged from 9 to 36, where higher scores indicated higher religiosity (Cronbach's α = 0.99). Medical mistrust was measured using a 7-item scale²⁷ where higher scores indicated more mistrust (scale range = 7–35) (Cronbach's α = 0.76). Finally, health-related quality of life was measured using the functional well-being subscale from the Functional Assessment of Cancer Therapy—Breast Cancer scale (Cronbach's α = 0.90).²⁸

Statistical analysis

We used descriptive statistics to characterize sociodemographic, clinical, health care process, and psychosocial factors. Bivariate analyses were performed to examine associations with women's compliance to biospecimen donation (yes vs. no). Independent sample t-test and chi square tests were used to obtain p-values for categorical (e.g., education) and continuous independent variables (e.g., medical mistrust), respectively. Frequencies were also used to describe the sample of women who returned their saliva sample and survey assessment on knowledge and personal concerns with biospecimen donation. All statistical analyses were conducted using SAS Version 9.4 using two-tailed alpha level of 0.05.

Results

The analytic sample consisted of n = 144 women. Overall, 29 surveys (20%) and 25 (17%) saliva kits were returned (Table 1). Ages of women ranged from 25 to 89 (M = 57 years SD = 11.8). Majority of the sample included white women (61%) vs. black women (34%). Most women were college educated (77%) and married (65%). Of the 144 women who received the brief informational materials and saliva kits, 21 (15%) white women returned biospecimen samples and only a few returned samples were from black women 3 (2%).

Table 1.

Sample Characteristics and Predictors of Returned Saliva Biospecimens (n = 144)

Variables	Returned biospecimen		p
	No (%)	Yes (%)
	119 (82)	25 (17)
Age, M ± SD	57 (11.8)	62 (12.3)	0.075
Race
Black	46 (32)	3 (2)	0.087
White	66 (46)	21 (15)
Education
<HS diploma/GED	27 (19)	5 (4)	0.722
College or higher	89 (63)	20 (14)
Bachelors and above
Marital status
Married	78 (54)	16 (11)	0.882
Single	41 (28)	9 (6)
Breast cancer stage
Stage I	53 (42)	18 (14)
Stage II	36 (28)	6 (5)	0.055
Stage III	14 (11)
Breast cancer surgery
Mastectomy	40 (28)	5 (3)	0.121
Lumpectomy	34 (24)	9 (6)
Psychosocial factors
Religiosity, M ± SD	29.0 (6.4)	23.8 (8)	0.0007**
Well-being function, M ± SD	19.9 (5.4)	22.7 (4.6)	0.018*
Discrimination, M ± SD	14.8 (1.8)	14.6 (1.3)	0.533
Medical mistrust, M ± SD	20.7 (5.4)	19.4 (4.0)	0.25
Health literacy, M ± SD	1.1 (1.8)	0.5 (1.4)	0.112
Health care process factors
Technical quality, M ± SD	4.0 (0.6)	4.1 (0.5)	0.334
Interpersonal manner, M ± SD	4.2 (0.6)	4.3 (0.8)	0.889
Financial aspects, M ± SD	3.6 (1.0)	4.0 (0.69)	0.077
Communication	4.0 (0.6)	4.2 (0.5)	0.189
General satisfaction	3.7 (0.7)	3.9 (0.7)	0.501
Access convenience	4.0 (0.6)	3.9 (0.7)	0.437
Time spent with doctor	3.69 (3.82)	3.82 (4.1)	0.431

^* p < 0.05; ^**p < 0.01.

M, mean; SD, standard deviation.

Demographic factors (age, race, marital status, and education) and clinical factors (BC stage or surgery) were not significantly associated with saliva biospecimen provision (p > 0.05). Among psychosocial factors, women with higher levels of functional well-being and lower ratings of religiosity were more likely to return biospecimens (p < 0.005) after receiving the enhanced information when compared with women who did not provide a biospecimen (Table 1).

In our assessment of the enhanced recruitment materials most women found it helpful and easy to read and understand (Fig. 3). Majority of women also reported prior participation in genetic research, higher levels of understanding the importance of saliva provision, and noted that spiritual/religious belief had no impact on their sample provision (Fig. 4). There were no frequently cited concerns or problems endorsed by women pertaining to the provision of saliva (Fig. 5).

FIG. 3.

Women's ratings of brief information.

FIG. 4.

Women's ratings of biospecimen knowledge and beliefs.

FIG. 5.

Women's ratings of concerns with providing saliva samples.

Discussion

Information is lacking about enhanced recruitment strategies to increase biospecimen provision among women. To our knowledge, this is among the first attempts to pilot brief low-cost informational tools to enhance biospecimen donation among HR-positive BC survivors. Our brief informational recruitment tools increased our overall biospecimen returned rate by 17%. Women found the enhanced materials helpful and easy to read and understand. Lower ratings of religiosity and higher levels of functional well-being among women were the top drivers in receipt of returned biospecimen samples.

Psychosocial values such as religiosity may influence women's decision-making process to participate in genetic research; however, the role of religiosity is not fully understood in this context. Concordant with our previous study findings that examined predictors of biospecimen provision in HR-positive BC survivors,²⁴ higher ratings of religiosity were negatively associated with biospecimen provision. Religiosity has been observed as a potential barrier with African Americans participation in precision medicine research.^22,29–31 However, this finding offers opportunities to intervene. One of the largest cross-sectional study designs to assess public views on biobanking found higher religiosity as a negative association with biobanking engagement.³² Although these findings demonstrate an inverse relationship between religiosity and participants' decision to participate in genomic research, causation cannot be assumed. Religiosity is multifaceted; thus, additional study is needed to fully understand the interplay of religion and genetic research participation. Moreover, community-engaged study with religious leaders may offer insight to this gap in the literature. This finding supports the role of psychosocial factors in participants' willingness to participate in biospecimen donation and adds the complexity behind religiosity/spiritualty in BC survivors' acceptance of genetic research.^7,31,33

BC survivors' decision to provide biospecimen samples for genetic research were influenced by their overall well-being. In our study, we found women with higher functioning well-being were more likely to donate biospecimen samples. Our finding highlights the role general wellness play in participation in genetic research among BC survivors. BC survivors who report lower physical, emotional, and social well-being may be a key group that lacks informational support and knowledge about participating in biomedical research. In support of this argument, prior study has found black women who practiced preventative health habits (e.g., visiting physician and physical activity) were more likely to donate biospecimen samples.²³ We also observed a trend toward significance for lower stage cancer (vs. higher stage) in our sample. Women with later BC stages and lower well-being may inform future intervention targets for HR+ women.

Although our findings begin to address gaps in the discussion of biospecimen donation among women, there are some limitations to consider. Our study sample size of returned survey and saliva samples was very small, especially among black women. Our study also lacked a comparison group of the brief informational recruitment materials. Although our brief informational tool enhanced provision of biospecimen samples from women, receipt of materials seemed to perform better for white women. Furthermore, knowledge among BC survivors' religiosity and well-being needs (e.g., physical, emotional, and social/family) may continue to inform recruitment strategies in biospecimen research.³⁴

Conclusion

There are limited intervention strategies aimed to increase biospecimen provision among women. Receipt of our enhanced brief informational materials among HR-positive black and white women increased our overall returned biospecimen rate by 17%. Women found the low-cost brief informational recruitment materials helpful, easy to read and understand. These factors are crucial to consider when developing recruitment strategies to optimize implementation science within clinical practice.

Findings from this study highlights that augmented materials that explain the usability, dissemination purposes, and confidentiality of genetic information improves BC survivors biospecimen donation; however, inclusion of cultural messaging may be a better recruitment strategy among black women. To increase representation of minority BC survivor populations in genetic research,^21,29,35,36 strategies may include targeting women at their physician appointments.²⁹ Thus, educating BC survivors from underrepresented groups during their primary care visits about provision of biospecimens may help mitigate missed opportunities to increase access to diverse genetic markers to advance promising treatment modalities.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research was funded by the National Institutes of Health R01CA154848 (Sheppard). It was also supported in part by the NIH-NCI Cancer Center P30 CA016059.