Ethnic differences in social support after initial receipt of an abnormal mammogram

Abstract

Objectives

We examine access to and type of social support after initial receipt of an abnormal mammogram across non-Latina White (NLW), African American, and Latina women.

Methods

This cross-sectional study used a mixed method sequential design, with quantitative and qualitative measures. Women were recruited through two community advocates and three breast health-related care organizations.

Results

With regard to access, African American women were less likely to access social support relative to NLW counterparts. Similar non-significant differences were found for Latinas. Women did not discuss results with family and friends to avoid burdening social networks and negative reactions. Networks' geographic constraints and medical mistrust influenced Latina and African American women's decisions to discuss results. With regard to type of social support, women reported emotional support across ethnicity. Latina and African American women reported more instrumental support, whereas NLW women reported more informational support in the context of their well-being.

Conclusions

There are shared and culturally unique aspects of women's experiences with social support after initially receiving an abnormal mammogram. Latina and African American women may particularly benefit from informational support from healthcare professionals. Community-wide efforts to mitigate mistrust and encourage active communication about cancer may improve ethnic disparities in emotional well-being and diagnostic resolution during initial receipt of an abnormal mammogram.

Well-documented ethnic disparities exist along the breast cancer continuum (Li, Malone, & Daling, 2003; Ooi, Martinez, & Li, 2011), including after initial receipt of an abnormal mammogram result (Adams et al., 2009; Arnsberger et al., 2006; Press, Carrasquillo, Sciacca, & Giardina, 2008; Ramirez et al., 2013). Ethnic differences in social factors and functioning during treatment and survivorship are well-documented, as indicated by multiple reviews (Russell, Voh Ah, Giesler, Storniolo, & Haase, 2008; Yanez, Thompson, & Stanton, 2011). Few studies have quantified differences in access to social support during diagnosis between non-Latina Whites (NLWs) and minorities (Molina, Beresford, Espinoza, & Thompson, 2014). Extant literature has not clarified which factors contribute to decisions to not disclose across ethnicity. Some factors may include family separation (Shelton, Goldman, Emmons, Sorenson, & Allen, 2011; Tejeda et al., 2013); gendered cultural values emphasizing the placement of loved ones over one's needs (Austin, Ahmad, McNally, & Stewart, 2002; Misha, DeForge, Barnet, Ntiri, & Grant, 2012; Rao, Debb, Blitz, Choi, & Cella, 2008), and high community levels of fatalism and mistrust (Deshpande, Thompson, Vaugh, & Kreuter, 2009; Glen, Chawla, & Bastani, 2013; Soskolne, Marie, & Manor, 2007; Zapka, Stoddard, Costanza, & Greene, 1989). Finally, little is known about ethnic differences in the types of support that are received, which may be dependent on the types of resources available to social networks (e.g., ability to provide transportation; advice about next steps).

The current study used a sequential mixed-methods approach to address these gaps. We collected quantitative data to determine ethnic differences in access to social support disclosure and withdrawal after initial receipt of an abnormal mammogram result. We collected qualitative data to explore ethnic differences in what factors influenced women's decisions about accessing social support and the types of social support women received.

Methods

Setting and Participants

This study combines data from two projects conducted between February 2012 and July 2013 (Mujeres Poderosas/Powerful Women (MPPW) - Molina, Hohl et al., 2014, Molina, Beresford, et al., 2014; Sisters-Healing-Advocating-Praising-Empowering-Sisters (SHAPES) - Molina, Hempstead et al., 2015). Staff recruited women from the Puget Sound area, wherein African American and Latina residents represent respectively 8% and 6% of the population (WA Department of Public Health, 2013) and exhibit breast cancer disparities that mirror those observed at a national level (Adams et al., 2009; Arnsberger et al., 2006; Press et al., 2008; Ramirez et al., 2013). Sixty-nine women participated in these studies, 22 identified as NLW, 19 as Latina, and 28 as African American. Their range of ages was 37 to 74 years old, the median yearly household income was <$20,000, and approximately 10% were ultimately diagnosed with breast cancer.

Procedures

Projects' materials and procedures were approved by the Institutional Review Board. Studies were identical in recruitment materials and sites with the exception of eligibility criteria listed; MPPW recruited only Latina and NLW participants, SHAPES recruited only African American women; and images posted on the recruitment materials. For both projects, representatives from the same two mobile mammography services, the local affiliate of the National Breast and Cervical Cancer Early Detection Program (NBCCEDP), and two NBCCEDP community advocates distributed fliers. Across projects, women were eligible if they: 1) self-identified as being Hispanic/Latina/Chicana, White/Caucasian, or African American/Black/of African descent; 2) were between 30-74 years old; 3) received an abnormal mammogram within the past four years; and 4) had no previous personal history of breast cancer. Interested women contacted one of the co-authors who screened for eligibility and scheduled interviews in the participant's preferred language (Spanish, English) and location. Participants signed an informed consent and completed a quantitative survey assessing demographics, social withdrawal, and emotional distress. Using a semi-structured guide, Latina and African American research team members conducted audio-recorded 30-60 minute-interviews to explore stressors, coping strategies, and social network dynamics during the abnormal mammogram experience. Participants were compensated $25-35.

Measures

Quantitative

Socio-demographic and clinical variables

All women responded to standard US Census questions concerning socio-demographics and their language(s) of preference. Women also reported their insurance status (0=No insurance, 1=Public, 2=Private), initial review of frequency distributions resulted in dichotomization of the variable (0=Un/underinsured, 1=Privately insured). Finally, women indicated if they had received a mammogram prior to this experience and their age at mammogram.

Access to social support

To assess disclosure, we asked women if they had spoken to their family and friends about results after initial receipt of their abnormal mammograms (yes/no). To assess social withdrawal, we administered the 3-item negative social consequences subscale of the Psychological Consequences Questionnaire (Cockburn, De Luise, Hurley, & Clover, 1992). To minimize recall bias, items were administered only to women who had the abnormal mammogram experience within three months of the interview (n= 39). A sample item includes “As a result of thoughts about your abnormal mammogram result, have you found noticeably withdrawing from those who are close to you?.” Responses ranged from 0=Not at all to 3=Quite a lot of the time. Cronbach's alpha for this sample was 0.70.

Qualitative

The study team developed a semi-structured interview guide to elucidate stressors, coping strategies, and family and friend dynamics following initial receipt of an abnormal mammogram. Questions from the guide were selected based on women's responses to the quantitative survey.

Analysis Plan

Quantitative

We conducted bivariate associations to identify potential socio-demographic and clinical differences by race/ethnicity. Crude and adjusted regression models were conducted to examine differences in disclosure and social withdrawal across race/ethnicity.

Qualitative

Three authors independently read each transcript, transcribed all interviews verbatim, checked for accuracy, and uploaded into ATLAS.ti version 7 (Berlin, Germany). Spanish transcripts were back- and forward-translated before transcription. The team applied an inductive content analysis approach, in which themes from different samples were explored and new themes were identified from raw interview data (Bernard & Ryan, 2010). A preliminary codebook was generated to capture the essence of each idea. When new codes emerged, the team returned to coded transcripts to apply the new codes across all interviews. They compared notes, reviewed data, and clustered similar ideas together into categories representative of each broader emergent theme.

Results

Initial bivariate analyses revealed Latina participants had less education and lower income, odds of being insured, and English proficient relative to NLW participants (Table 1). African American women were comparable to NLWs in socio-demographic and healthcare characteristics. Given this, education, income, and insurance status were included in subsequent quantitative analyses.

Table 1.

Sample characteristics.

		NLW (n = 22)	Latina (n = 19)	African American (n =28)
	Range	M (SD)	M (SD)	M (SD)
Household size	1-9	2.93 (2.25)[REF]	2.93 (1.59)	2.93 (1.93)
Age	37-74	51.77 (5.77)[REF]	49.68 (6.54)	54.71 (8.90)
Emotional distress1	0-15	5.23 (4.50)[REF]	9.92 (3.47)***	7.78 (5.12)
Social withdrawal1	0-9	1.83 (2.15)[REF]	3.62 (2.84)	4.44 (2.92)*
		n (%)	n (%)	n (%)
≤High school degree (Education)2		4 (18%)[REF]	10 (53%)*	5 (18%)
<$20,000 (Median Income)2		11 (50%)[REF]	16 (83%)**	10 (36%)
Un- and underinsured		14 (64%)[REF]	18 (95%)*	19 (79%)
English proficient		22 (100%)[REF]	8 (42%)***	28 (100%)
Discussed results with family/friends		17 (77%)[REF]	9 (47%)*	14 (50%)*
Diagnosed with cancer		1 (5%)[REF]	1 (5%)	5 (18%)

¹Psychological Consequences Questionnaire - emotional and social consequences subscales. Administered to women who received abnormal result within 3 months of interview (15 NLW, 13 Latina, and 9 Black/African American).

²Variable analyzed continuously, but depicted categorically to facilitate interpretation.

^*p <.05,

^**p <.01,

^***p <.0001

Quantitative Findings

Latina and African American women were less likely to report disclosing initial receipt of abnormal results with family and friends relative to NLWs, OR=0.3, 95%CI [0.1, 1.0], p=.05 for all. African American women reported greater social withdrawal relative to NLWs, B=2.6, 95%CI [0.4, 4.8], p=.02. There was a non-significant association between Latinas and NLWs, wherein Latinas reported slightly more withdrawal, B=.8, 95%CI [-0.2, 3.8], p=.08. After adjusting for education, income, and insurance status, differences in disclosure between Latina and NLW women were attenuated (p=.09), but differences in initial disclosure of results (p=.03) and social withdrawal (p=.05) between African American and NLW women remained significant.

Qualitative Findings

Qualitative findings clarified quantitative associations with regard to factors contributing to access to social support by ethnicity. Qualitative findings also depicted the types of support received after initial receipt of results by ethnicity.

Factors influencing decisions to discuss abnormal mammogram results

When discussing situations in which they chose not to disclose their initial results with family and friends, women noted not wanting to worry members of their social network, not wanting to experience negative reactions, and unwanted pity.

Components of these reasons were shared across race/ethnicity. All women indicated their family and friends had their own problems and not needing additional burdens. A NLW woman reported, “I didn't even tell my husband yet… because there's too much on his plate too.” An African-American respondent similarly shared: “They would have lost it…. They would have been a mess.” Other women stated wishing to avoid family and friends' vicarious fatalism. A Latina affirmed, “I don't want them to say, ‘Oh poor thing, look what she has.’” Similarly, an African American respondent said she felt “isolation because I didn't share it…Sometimes people in their kindness and their concern, it allows you to be weak and I wanted to stay strong, so I didn't share it.”

Unique components of these factors were relevant to Latinas and African American women's broader psychosocial contexts. Immigrant Latinas indicated geographic isolation was influential. Women said this separation would amplify worry and helplessness among family and friends, if they knew about the abnormal results. One Latina respondent reported:

When my father got sick, we had all the things ready…there was nothing so important as the health of my father. But here [United States]?…No! I cannot tell [my daughter]! It wouldn't be a solution. How can I give her a pain? What good would it do to tell my family? …I cannot tell my daughter.

Latina and African American respondents indicated another unique negative reaction they wished to avoid was family and friends' medical mistrust and anticipations of racism. One African American woman reported, “You hear about other people's scary stories…that if you're African American…it's going to be racist…they're not going to pay attention to your mammogram as much.” A Latina stated, “People just mistrust doctors…they want to give you an illness to make money, so they can run more tests.”

Types of support received when discussing abnormal mammogram results

The manifestation of different types of social support after disclosing abnormal results varied across ethnicity. Differences reflected how community norms, perceptions and experiences may influence individuals.

Emotional Support

NLW women largely discussed support as bolstering their emotional well-being during the abnormal mammogram experience, as highlighted by this respondent:

“Friends who um would check…‘How are you doing today?’ ‘I'm fine.[laughs] …Thank you for thinking of me.’…And the other one had actually gone on a trip to Phoenix, so she was texting me, you know, and saying ‘Just checking on you, seeing how you were doing.’… And that's very helpful for me. It's a very good support group.”

Latina and African American women reported feeling emotionally supported when family and friends provided encouragement about adhering to recommended care. A Latina recalled, “They were caring. They remembered I had to go to the doctor. They were telling me to go. ‘What time are you going to go? You're going to go to your appointment?’” Correspondingly, an African American reported, “My girlfriend and sister, they would coach me, and my daughter… ‘Come on, Mom, you can do this. We're behind you.’”

Latina and African American women further discussed emotional support as intersecting with instrumental support. A Latina respondent recalled, “My boyfriend said, ‘You better read that [results letter].’ I said, ‘I need your help a little bit…just to make sure…I read it correctly.’” Similarly, an African American respondent said, “My friend was comforting, you know, he fed me, took me there, took me home.”

Informational Support

NLW women more often discussed the informational support they received from family and friends who had undergone similar diagnostic procedures, including individuals with cancer history. A NLW respondent reported, “Cause my neighbor that I talked to quite a bit, you know, she's my age and she said she's had 2 or 3 or 4 [follow-up appointments] and it was no big deal.” Another NLW woman reported, “My sister has had a bad mammogram come back. She explained to me the calcium deposits and that kind of stuff. She had to have a biopsy too.” Notably, NLW women said that through receipt of informational support, they felt better prepared to appraise their risk of breast cancer and anticipate upcoming procedures.

Fewer African American than NLW women said they were able to access similar sources of informational support among their family and friends. These women noted the need for such support, as described by this African American respondent, who said, “They should have a support group… in the process of giving the patient test results that are abnormal and frightening.”

Although Latina respondents reported knowing breast cancer patients and survivors in their area and their countries of origin, they were reluctant to inquire for informational support from these sources. They said the breast cancer patients and survivors they knew were often too emotionally traumatized to be reliable sources of informational support. For example, a Latina respondent asserted,

“I have a friend…she has breast cancer. She isolated herself. No one knew about her. She wouldn't open the door to us. She wouldn't answer the phone. We didn't know if she ate or not, if she was living there or not. She isolated herself completely, because she was going to die…and because of that, I didn't [contact her].”

Discussion

Our study contributes important nuanced information regarding how available social resources after initial receipt of an abnormal mammogram differs across ethnicity. Our work adds to growing evidence indicating racial/ethnic minorities are less likely to access support resources when undergoing cancer-related care and offers insight into types of support accessed across racial/ethnic backgrounds.

Similar to previous studies (Molina, Beresford, et al., 2014), our quantitative findings suggest minorities may be less likely to disclose and more likely to be socially withdrawn after receiving an abnormal mammogram result. In qualitative interviews, across ethnicity, women did not discuss results so as to avoid concerning their family and friends and to avoid negative reactions. Our work speaks to the engendered norms across many cultures, wherein women are socialized to “suffer with dignity” and thus must maintain an image of strength in order to not burden their loved ones (Fulton, Rakowski, & Jones, 1995; Murphy, Vernon, Diamond, & Tiro, 2014; Otiniario, Carroll-Scott, Toy, & Wallace, 2012; Shelton et al., 2011). Simultaneously, unsolicited reactions from well-meaning friends and family members appeared to trivialize women's self-efficacy to manage implications of an abnormal mammogram. Extant literature has fatalism as a barrier to accessing cancer-related care (Deshpande et al., 2009; Natio, O'Callaghan, & Morissey, 2009; Sheeran, 2002), especially among Latinas and African American women. Our work additionally suggests that fatalism at the interpersonal and community levels can also influence women's experiences when undergoing screening, detection, and treatment.

In addition to these shared experiences, Latina and African American women reported two unique factors underlying differences in discussing results. First, most Latinas in our study were immigrants whose social network largely remained in their country of origin. The stress of this larger context was interwoven in Latinas' unease with burdening family and friends if they discussed results, paralleling previous findings that indicate geographic isolation as a major stressor influencing timely follow-up after receiving an abnormal mammogram (Shelton et al., 2011; Tejeda et al., 2013). Second, networks' medical mistrust may influence women's decisions to disclose their abnormal results, a finding supported in previous work (Deshpande et al., 2009; Molina, Kim, Berrios, & Calhoun, 2015).

Finally, we characterized the types of social support received by women who discussed initial results. There were some ethnic differences in how support emerged. For example, NLW women appreciated sympathy in the form of check-ins and queries about their well-being. Conversely, African American and Latina women preferred motivational verbal support, such as encouragement while interpreting results letters and in attending follow-up appointments. This suggests a complicated relationship of communication for self-efficacy among Latina and African American women, wherein family and friends may reduce self-efficacy through medical mistrust or may provide the necessary collective esteem for women to adhere to recommended care. NLW women were more likely to indicate they discussed results with family and friends who had undergone similar experiences and could provide informational support. Discussions may be impeded, in part, because the uniquely traumatic experiences faced by Latina and African American women diagnosed with cancer and silence about cancer (e.g., late stage detection, unmet healthcare needs)(Deshpande et al., 2009; Orom, Kiviniemi, Shavers, Ross, & Underwood, 2013).

Limitations

This study had several limitations. We did not assess important characteristics of family structures, including marital status and household size. Due to the small sample size, our quantitative findings were likely underpowered. This is particularly the case with regard to our quantitative data, wherein we had relatively few women of color to quantify differences, especially in adjusted models. While we sought to minimize recall bias by limiting quantitative survey measures to women with ‘recent’ experiences (within 3 months), bias may still have occurred. Our study did not quantitatively assess different forms of social support. The themes identified in this convenience-based sample may not be representative of all NLW, African American, and Latina women. Although the two studies described had many similarities, there were differences which may have influenced our analyses, including differences in the timing of studies and age-related eligibility criteria. Larger, population-based studies are needed to confirm differences in access and use of social support after abnormal mammograms among ethnically diverse populations. Longitudinal research is also needed to determine if access to social support after abnormal mammograms can improve timely adherence to follow-up.

Implications for Practice

The current study adds to growing evidence to support the involvement of multiple stakeholders in efforts to mitigate breast cancer disparities, including lay clinicians, community advocates, and researchers. Counseling may be adapted to address ethnic minority women's social isolation, depending on whether it is a result of geographic constraints or a result of community-wide norms concerning silence when undergoing a cancer-related experience. Psychologists and other healthcare professionals may complement Latina and African American women's social networks by supporting women's efforts to self-advocate and offer informational support. Collaborations between healthcare professionals and community advocates may further improve breast cancer outcomes among marginalized populations through efforts that mitigate medical mistrust and promote active, informed, and supportive communication about cancer care within lay social networks.