Barriers to Psychosocial Services Use For Latina Versus Non-Latina White Breast Cancer Survivors

Individuals with breast cancer commonly experience poor quality of life and higher rates of depression than people without cancer (1, 2). Latinos diagnosed with cancer have poorer psychological outcomes than other cultural groups (3–5). Although Latinos with cancer have poorer mental health outcomes, they are less likely to receive psychological, psychiatric, and other psychosocial interventions than non-Latino whites with cancer (5, 6). Moreover, Latinos often report unique concerns, such as socioeconomic, cultural, and linguistic issues, that affect their access to psychosocial and mental health services (7–9). Furthermore, a cancer diagnosis has been associated with worsened mental health caused by reduced income, and Latinos experience worse outcomes compared with non-Latino whites (10).

Many reasons may explain the lower rates of psychosocial services use among Latinos diagnosed with cancer compared with white counterparts. A prior study conducted with the same sample of Latina breast cancer survivors demonstrated varying patterns of access to psychosocial services. Although one-third (N=87, 33%) of the survivors reported needing mental health or psychosocial services after their cancer diagnosis, only 40% (N=35) of the participants received a referral (6). Additionally, non-Latina whites were significantly more likely than Latinas to have had contact with a social worker and to have received psychotropic medication (6), and only 34% of Latinas discussed their emotional problems or needs postdiagnosis with their oncologist or cancer care provider. Latinos may face unique patient-, provider-, and system-related barriers and challenges to care, such as lack of accurate identification of individuals with psychosocial needs, low availability of culturally and linguistically appropriate services, lack of insurance coverage or underinsurance, and individuals’ beliefs and attitudes toward help-seeking behavior (11). Several studies with nationally representative samples have shown that individuals from ethnic minority groups are less likely to receive mental health services than non-Latino whites (11, 12). Alegría and colleagues (11) have reported that probable causes of this disparity in mental health service utilization include underdetection of depression by health care providers; stigma surrounding use of mental health services; financial and employment constraints; previous experiences of discrimination; receipt of informal care from family, friends, or informal providers; and limited access to providers. However, research in this area has been restricted to community, psychiatric, and population-based samples (11–13).

Multiple systematic reviews and meta-analyses published in the past decade have demonstrated that receipt of psychosocial services, psychotherapy, and/or mental health care has at least moderate effects on psychosocial outcomes (e.g., distress, quality of life) of cancer patients and survivors (14–24). Many psychotherapy studies have shown promise in improving the psychological functioning of cancer patients with varying disease types and at varying stages and of cancer survivors (14–24). Most systematic reviews, however, state that recommendations are made with caution because of the suboptimal quality of most interventional studies (15, 16, 18–20, 23).

Barriers to use of psychosocial services among Latinos diagnosed as having cancer—as well as among the general population with cancer—have been understudied. In a systematic review, Dilworth et al. (25) identified 10 quantitative studies reporting barriers to the use of psychosocial services among the general population. The most frequently reported barrier was “no need for psychosocial services and support,” followed by lack of information or wanting more information about services (25). According to that review (25), health professionals reported that lack of time, insurance coverage, knowledge regarding the benefits, identification of patient needs, referral systems, affordable cost, and resources were the most commonly reported barriers to distress management (25). In the present study, we examined differences in self-reported barriers to use of psychosocial services by Latina versus non-Latina white breast cancer survivors treated at a New York City comprehensive cancer center. The secondary objective was to examine associations between these barriers and the women’s use of psychosocial services following a cancer diagnosis.

Methods

Sample and Setting

This retrospective cross-sectional study was conducted with a sample from a single comprehensive cancer center in New York City. Information about the study methods and assessments has been published (6). Eligible participants were adults (age 21 or older) in remission from a breast cancer diagnosis (identified through their medical records), who had received cancer care at the center. After reviewing the medical records, we identified 409 Latinas and 5,146 non-Latina whites treated between 2009 and 2014. All 409 Latinas and a random sample of 10% of the non-Latina whites were selected for this study in order to have a comparable number of patients in each group. Potential participants were mailed a questionnaire packet and a reminder letter 2 weeks later. All materials in the survey were anonymous. The participants provided informed consent by reviewing a form with information about the study, their rights, and potential benefits and by completing the survey. Of the possible participants, 280 women completed and returned the questionnaire (see flow diagram in the online supplement), for a response rate of 30%. In total, 265 participants completed the portions of the questionnaire about psychosocial services use and barriers. This study was reviewed by Memorial Sloan Kettering Cancer Center’s institutional review and privacy board and was determined to be exempt research.

Methods and Variables

The questionnaire packet captured demographic information and assessed respondents’ attitudes toward and history of using psychosocial services and perceived barriers to use of these services. The section on demographic characteristics included questions about age, education, religion, marital status, socioeconomic status, family composition, living situation, employment status, ethnicity, race, language preference, birthplace, and years living in the United States.

The psychosocial services section included questions about the participants’ psychosocial services usage before and after the cancer diagnosis. These questions also assessed patients’ preferences for services, including type of counseling, delivery format, and type of mental health professional. Types of mental health professional were defined according to the National Comprehensive Cancer Network (NCCN) distress management guidelines and included social workers, psychologists, psychiatrists, and religious counselors (i.e., priests, pastoral counselors, or ministers). Types of psychosocial services also were defined according to the NCCN guidelines and included individual counseling, family therapy, couples therapy, support groups, group therapy, and pharmacotherapy.

The questions about barriers to access asked about attitudes toward mental health and counseling services and patient-, physician-, and system-level barriers to use of such services. We categorized 14 potential barriers (Cronbach’s alpha=0.83) as patient-level barriers or attitudes (e.g., self-reliance, lack of knowledge, stigma, preference for informal care). The remaining six (Cronbach’s alpha=0.70) were categorized as physician- or system-level barriers (e.g., insurance and cost issues, lack of available providers, and lack of recommendation or referral). The questions were based on prior literature and the authors’ clinical experience with Latino populations in mental health and cancer settings (26–29). Respondents were asked whether they faced each barrier (“yes,” “maybe,” or “no”). To ensure an adequate number of respondents to conduct the analyses, “yes” and “maybe” were collapsed into a single category, resulting in two categories (“yes and/or maybe” and “no”).

We used SPSS, version 19, to analyze the statistical data. We first described the demographic characteristics, self-reported use of psychosocial services after the cancer diagnosis, and barriers to use of psychosocial services by using descriptive statistics. We then assessed the self-reported barriers for the full sample and for the two groups. We further used descriptive statistics to characterize the sample by ethnicity. Chi-square analyses were used to evaluate differences in self-reported barriers between the two groups, with a two-sided significance level of p<0.05.

We used bivariate logistic regression models to determine associations between the self-reported barriers to contact with mental health professionals and receipt of mental health services after a cancer diagnosis for the total sample. Because of the small sample size, we were unable to conduct these analyses by racial-ethnic subgroup. The evaluable subsamples for the unadjusted models ranged from 240 to 261 participants because of missing data. We analyzed data on contact with mental health professionals (social worker, psychiatrist, or psychologist) and receipt of mental health services (individual, family, couple, or group counseling) after the cancer diagnosis separately. We reduced these classifications to binomial outcomes for analysis (i.e., contact with mental health professionals, yes or no).

We conducted adjusted logistic regression models to examine the association between self-reported barriers and contact with psychosocial services professionals and use of psychosocial services after the cancer diagnosis. Because of the small sample size, we could not conduct these analyses by ethnic subgroup. We adjusted the models for sociodemographic and medical factors, such as age, marital status, ethnicity, education level, and time since diagnosis. Age, time since diagnosis, marital status, and education level were dummy coded and entered into the models. The subsamples for evaluating the adjusted models ranged from 213 to 220 participants because of missing data. We calculated adjusted ORs with 95% confidence intervals to assess the relationships between the barriers and both the use of services and contact with professionals. A two-sided significance level of p<0.05 was considered statistically significant.

Results

Table 1 shows the demographic profile of the participants. About two-thirds of the sample was married or partnered (63%). Latinas had lower levels of education: 30% had attained a high school education or less, compared with 17% of the non-Latina whites. About one-third of Latinas (32%) preferred their care delivered in Spanish. Thirty percent of Latinas and 91% of non-Latina whites were born in the continental United States. Many Latinas were born in Puerto Rico (22%) or South America (22%). One-quarter of the sample (25%) had been diagnosed as having breast cancer less than 2 years prior. In recalling the time after the cancer diagnosis, 43% (N=116) of participants (49% [N=47] of Latinas and 40% [N=69] of non-Latina whites) expressed that they wanted or needed psychosocial services after the diagnosis. Of the participants who reported needing services after the cancer diagnosis, two-thirds (N=76; 57% [N=27] of Latinas and 71% [N=49] of non-Latina whites) reported having contact with psychosocial services providers, and 61% (N=71) reported using psychosocial services after the diagnosis (53% [N=25] of Latinas and 67% [N=46] of non-Latina whites).

Three barriers or attitudes were endorsed by more than 50% of the sample (Table 2): “I feel I can take care of my problems myself” (63%), “I prefer to go back to my normal routine rather than to see a therapist or counselor” (60%), and “I prefer to seek counseling and support from my friends and family” (56%).

Latinas were more likely than non-Latina whites to endorse the following statements: “Mental health services are too expensive” (43% versus 30%, χ²=3.93, df=1, p=0.05); “Using an interpreter for counseling or therapy would make me uncomfortable” (22% versus 1%, χ²=14.11, df=1, p<0.001); “There are no counselors or therapists who understand my cultural values or background” (21% versus 7%, χ²=11.04, df=1, p=0.001); and “There are no counselors or therapists who speak my language” (9% versus 0%, χ²=6.61, df=1, p=0.01). Non-Latina whites more frequently endorsed the statement, “My doctor did not tell me about counseling or therapy” (24% versus 36%, χ²=3.56, df=1, p=0.06).

Analyses adjusted for demographic characteristics (ethnicity, age, education, marital status, and time since diagnosis) showed that wanting to return to normalcy, lack of information about where to seek services or how counseling can help, believing it is normal to be distressed after a cancer diagnosis, feeling embarrassment about seeking services, and believing that counselors will not understand the individual’s cultural background were significant predictors of lack of contact with counseling professionals and/or psychosocial services (see Table 3).

Discussion

This study evaluated barriers to psychosocial services use faced by Latina and non-Latina white breast cancer survivors at a comprehensive cancer center. Among the women who reported needing psychosocial services, Latinas were more likely to report that there were no counselors who could speak their language or understand their cultural values and background and that mental health services were too expensive. Our prior study of the same sample demonstrated differences in perception of psychosocial services availability, acceptability, and help-seeking behaviors by Latina and non-Latina white breast cancer survivors (6). This prior study showed a third of the women reported needing mental health or psychosocial services postdiagnosis (6). However, ethnic differences in type of psychosocial service use were shown: non-Latina whites were significantly more likely than Latinas to have contact with a social worker and to receive psychotropic medication, and Latinas were significantly more likely to receive spiritual counseling than non-Latina whites (11% [N=5] versus 3% [N=2], respectively). The present study adds to our previous finding by illustrating and comparing the barriers Latina and non-Latina white cancer survivors face in accessing psychosocial services.

Research examining barriers to psychosocial care among people with cancer has been scarce. In our sample of breast cancer survivors, the most commonly cited barriers to psychosocial care or reasons for declining services included beliefs about self-reliance, desire to return to normalcy, and preference for receiving care from family and friends. However, the barriers that predicted lack of use of psychosocial services and/or contact with professionals among the women who reported having emotional problems or needs were the desire to return to normalcy, not understanding the overall or direct benefit of therapy, not knowing where to access services, stigma around the need for mental health services, the belief that it is normal to feel distressed after a cancer diagnosis, the belief that counselors might not understand their cultural values or backgrounds, and discomfort with possibly needing an interpreter. A 2014 systematic review (25) including 10 quantitative studies reporting psychosocial services barriers found that the barriers most frequently reported were lack of perceived need for psychosocial services and support and lack of information or wanting more information about such services. Other studies have reported the main barriers to be uncertainty about the need for mental health care and the cost of such treatment (30), receipt of support from sources outside the cancer clinic, lack of awareness of the service, and lack of provider referral (31). Our findings support that commonly reported barriers, such as lack of information or wanting more information about services, lack of awareness of the service, and uncertainty about the need for mental health care, have an association with patient use of psychosocial services and/or contact with professionals. However, it is unclear whether other frequently cited barriers, such as having no perceived need for psychosocial services, the cost of treatment, receipt of support from sources outside the cancer clinic, and lack of provider referral, also have associations with use of services and access to professionals. In terms of perception of need of services, patients who endorsed, “I feel I can take care of my problems myself,” were as likely to use services as their counterparts. Furthermore, perceiving that mental health services were too expensive; seeking counseling and support from friends, family, or a priest, pastor, or minister; and not receiving a recommendation for counseling or therapy from their doctors were not associated with receipt of psychosocial care.

The NCCN guidelines (32) for psychosocial care of individuals with cancer have established a set of quality measures for screening and algorithms for managing distress and psychiatric disorders (e.g., adjustment disorders, depression, suicide and suicide risk, and cognitive disorders). Despite patients from ethnic and racial minority groups having less access to psychosocial services and facing more barriers to access, the NCCN guidelines fail to include specialized recommendations for the distress screening and treatment management of culturally and linguistically diverse patients. Our results highlight the need to study and address this gap in national clinical guidelines. Patients, cancer care and mental health providers, administrators, and institutional leaders need education about patients’ psychosocial issues, and interventions that are culturally and linguistically sensitive need to be created.

Future research should be refocused from documenting disparities in access to reducing and addressing the disparities (33). Interventions need to address not only individual-level barriers, but also provider- and institutional-level barriers. Given the emergence of widespread use of technology in the recent years, psychosocial care will need to draw increasingly on the use of the Internet, social media, text messaging, and other technologies to engage patients and providers (33). The key barriers found in this study can be addressed with interventions that consider individual-, provider-, institutional-, and system-level barriers by building and/or supporting educational programs for patients and cancer care providers, specialized training for mental health providers, institutional programs and policies to improve cultural competency and access to such programs, and system-level interventions to promote screening and management. Educational interventions for patients could increase awareness among patients about available psychosocial services and providers, normalize the use of such services, provide information about the benefits of such services and where to receive them, and confirm the availability of culturally congruent counselors and/or interpreters. Because of the limited information about effective stigma-reducing interventions for patients who have cancer (34, 35), these interventions should be based on a theoretical framework informed by pilot studies and targeted to an oncological population. Modalities using the Internet, text messaging, social media, patient portals (health care–related online applications), and patient navigation programs can be used to deliver these interventions (36).

Educational interventions are also needed for oncology and mental health providers. Cancer care providers could benefit from culturally sensitive training on communications skills to teach how to elicit patients’ psychosocial needs and knowledge gaps and to address patients’ concerns and questions. These interventions can be embedded in routine provider training and can be delivered through remote or hybrid formats. Psychosocial service providers also can benefit from receiving training to provide culturally responsive services and to address concerns about language interpretation in the context of mental health services delivery. Virtual courses and seminars and professional networks or organizations, such as the International Psycho-Oncology Society (https://www.ipos-society.org) and the Latin American Psychosocial Oncology Research Training network (www.fipol.info), can provide alternatives to facilitate specialized training.

Health care institutions can also create policies, processes, and programs to improve institutional cultural competency (37). These actions can include developing and supporting culturally sensitive training for providers and support staff, maintaining a diverse workforce of bilingual and bicultural cancer care and psychosocial providers and support staff; offering no-cost, widely advertised language assistance to patients and family members with limited English proficiency; translating all materials to patients’ preferred languages; conducting ongoing assessment of policies regarding culturally and linguistically appropriate care; collecting and maintaining reliable demographic data; partnering with the community to ensure culturally and linguistically appropriate psychosocial services; partnering with community clinics and programs to ensure effective transition of psychosocial care; providing case management and patient navigation to address logistical and psychosocioeconomic needs or partnering with community programs to ensure effective delivery of these services; and maintaining cultural and community advisory boards to plan and monitor programs and policies to improve the institution’s cultural competency.

This study was limited by its sample population of a cohort of breast cancer survivors who retrospectively answered questions about barriers to psychosocial services use after receiving a cancer diagnosis. Because patients responded during remission, their responses may have been influenced by time and memory. Also, the institutional policies regarding psychosocial care may have changed from the time of receipt of therapy to the time of data collection. Second, the non-Latina white response rate was higher than that of the Latina response rate. Because the Latina participants had lower educational attainment, their lower response rate may have been due to lower literacy levels, which may have prevented participation in a self-administered survey, or due to lack of familiarity with completing or responding to surveys. Third, the sample was small and was recruited from a comprehensive cancer center in New York City, thus limiting its generalizability, especially to participants receiving care in safety-net hospitals that may have more uninsured or underinsured patients and/or greater barriers to psychosocial services. Fourth, these results may not reflect the experiences of uninsured patients, because all our participants were insured. Financial costs or lack of insurance may be important barriers to patients who receive care in public hospitals. Fifth, in the absence of a validated instrument to measure barriers to psychosocial services, the main study instrument was based on the literature and on the research team’s clinical and research experience. Further study of the psychometric properties of the type of instrument used in the current study may be helpful for future studies in other clinical settings.

Conclusions

This study illustrated how cultural and linguistic issues can become barriers to psychosocial service use. Our findings should help inform future strategies to make psychosocial services more accessible to non-Latina white and Latina people with breast cancer and to address their barriers to care. It is crucial to study and address the multilevel barriers in patient access to and use of psychosocial care that contribute to health disparities and discrepancies among racial-ethnic minority groups in the context of cancer treatment. Finally, recommendations to address key barriers at the patient, provider (both cancer and mental health care providers), and system levels have been provided.