Introduction

The underutlization of mental health services among Latino immigrants appears to be a growing concern given that risk factors such as poverty and unemployment, the acculturation process, low level of education, and separation from children and family have been linked to depression [1–6]. However, Latinos are less likely to seek treatment due to a variety of barriers such as lack of health insurance and accessibility of mental health facilities [4]. However, when Latinos do seek treatment they tend to prematurely drop out [7]. To address this concern there have been efforts to integrate culturally responsive psychological services into community-based comprehensive health care clinics that offer not only adult, pediatric and dental services, but also an array of social services, all with the goal of increasing access and keeping Latinos in treatment [8–11]. Such models include Spanish speaking health and mental health care providers who are knowledgeable about the Latino culture, and the unique life circumstances and challenges that many Latino immigrants face. This direction is supported by earlier research in which Latino clients who received mental health services in community-based agencies had lower drop out rates than those who received such services in agencies that did not have Spanish speaking staff and were not located in Latino communities [12]. This is especially important for recent immigrants who often rely on their physicians for mental health care. While such findings are noteworthy and seen as a potential solution to the problem of poor treatment engagement and and retention, there is limited knowledge of specific factors that may be related to staying in treatment among Latinos who are receiving mental health services in community-based agencies that offer culturally responsive services. This is particularly relevant because Latinos tend to prefer psychological treatment over pharmacological treatment [13].

The agency where this study was conducted is located in a predominately Latino community, offering comprehensive health care services including behavioral health. Specifically, the clinic offers outpatient mental health services, is staffed by eighteen mental health service providers of which the majority (14) are Spanish/English language bilinguals, and all have expertise in working with this population.

As previously stated, the argument has been that this type of setting may reduce drop out rates; however there is limited research in identifying specific factors that contribute to staying in treatment or dropping out. Therefore, the purpose of this initial study was to investigate specific psychosocial and cultural variables that may affect the treatment retention of Latinos diagnosed with depression in a community-based clinic that provides culturally appropriate mental health care.

A retrospective design was used to compare individuals classified as premature treatment termintors (PTs) with others classified as regular attendees (RAs) on a number of demographic and related variables potentially related to treatment outcome. Based on the literature, these included a number of factors hypothesized as barriers to completing treatment such as lack of transportation, child care issues, insurance, medication usage, employment status and income. We examined the importance of family and religiosity—cultural values among Latinos that have been noted in the literature, as playing an important role in coping with stressful life situations and life altering illness [14, 15]. Finally, we examined acculturation and acculturative stress, two related variables that have been linked to depression in the Latino population. In essence, we explored whether these factors played a role in differentiating clients who stayed in treatment from those who dropped out.

Method

Participants

Latino adult clients who either were currently receiving or had previously received therapy at a comprehensive community-based clinic in the center of a Latino community in a large Midwest city were recruited for the study. Since this was a pilot study, our goal was to recruit 20 participants who were regular attendees (RAs) and 20 participants who were premature terminators (PTs). RAs were defined as individuals who attended six or more sessions, attended 70% or more of scheduled sessions, and had no breaks of more than 3 months between sessions. PTs were defined as individuals who had been discharged after attending three or fewer therapy sessions and the reason for discontinuing treatment was unknown. In other words, no specific reason was noted on the chart. All participants had a primary diagnosis of Major Depressive Disorder as measured by the PRIME-MD [16] at intake according to chart records. No co-morbidities were noted in the chart records.

Table 1 presents demographic variables for RAs, PTs and the total sample. Overall, the sample consisted of 30 females and 6 males, with an average of 38 years of age, household income of approximately $11, 315, and a little over 8 years of education. Regarding country of birth, the sample of those born in Mexico and Puerto Rico were almost equal in numbers. Interestingly, half (18 of 36) were currently married or in a common law marriage and the other 18 were separated, divorced, widowed or never married. Participants, on average, had three children. Furthermore, the average number of years living in the United States post-migration was a little over 12 years. RAs and PTs did not significantly differ by gender, age, number of children, country of birth, education, number of years in the United States, or income.

Table 1 Demographic information
Full size table

Data Collection

An existing client database was used to recruit participants. With respect to recruitment, 100 of the most recently seen clients at the clinic who met RA criteria, and 100 of the most recently seen clients who met PT criteria were selected based on chart review. Participants were randomly selected from these lists and recruited for the study. The recruitment procedure involved a telephone call by a member of the research team who read a standard script inviting the individual to participate in the study. They were told that we were gathering information to improve psychological services for Latinos suffering from depression. Also, they were told that we wanted to learn about their experiences living in the United States, how they valued their family, and their experiences receiving treatment for depression at the clinic. Initially, 63 PTs and 27 RAs were contacted which yielded 20 PTs and 20 RAs. More PTs were contacted compared to the RAs because a significant number of their telephones were disconnected, they moved, or they did not return the telephone call (approximately 30%).

The interviewers consisted of an advanced doctoral student in counseling psychology and a counseling psychologist. Both were fluent in Spanish and English, had considerable experience working with this population and were skilled at conducting interviews.

Participants met the interviewer at the clinic and were escorted to a private room for the interview and completions of questionnaires. Participants provided informed consent, then completed the questionnaires and a semi-structured interview. Because of the relatively low literacy level among this population all questionnaires were read to each participant to ensure that the items were clearly understood thereby increasing validity in the study. Information was gathered on all 40 participants; however, two were dropped because they were inadvertently mis-identified as eligible for the study, and two were born in the United States. This yielded a final sample size of 36 (18 PTs and 18 RAs). The authors of this article received IRB approval from the University to conduct this study.

Measures

Pan Hispanic Familism Scale (PHF) is a five-item self-report questionnaire that assesses the importance of immediate and extended family (familismo) [17]. Participants rate each item on a 5-point scale with higher scores indicating greater commitment to one’s family (1 = “Strongly Disagree”; 5 = “Strongly Agree”) with possible scores ranging from 5 to 25. Based on our samplethe PHF has good internal consistency (α = .85), and has been validated in both English and Spanish and across country of origin.

Short Acculturation Scale for Hispanics (SASH) is a 12-item self-report questionnaire that assesses degree of acculturation [18]. Participants rate each item on a 5-point scale (1 = “Only Spanish/All Latinos/Hispanic”; 5 = “Only English/All Americans”) with higher scores indicating greater acculturation. The SASH produces a total score ranging from 12 to 60 and three subscales (Language Preference, Media Preference, and Ethnic Social Relations; ranging from 5 to 25, 3 to 15, and 4 to 20, respectively). Based on our sample the SASH’s full scale and subscales have good reliability (α’s ranging from .60 to .84).

Multidimensional Acculturative Stress Inventory (MASI) is a 25-item self-report questionnaire, with English and Spanish equivalence, that assesses stress associated with acculturation [19]. Participants rate each item on a 6-point scale (0 = “No” (i.e., not applicable); 5 = “Extremely Stressful”). The MASI contains four scales (Spanish Competency Pressures, English Competency Pressures, Pressure to Acculturate, and Pressure Against Acculturation), with scores ranging from 0 to 35 except for Pressure Against Acculturation which ranges from 0 to 20. Higher scores indicate greater pressure on all scales. Based on our sample the MASI’s full scale and subscales have good reliability (α’s ranging from .62 to .83).

Interview Survey:Footnote 1 an interview survey was developed by the investigators for this study consisting of forced choice (e.g., yes/no) and open ended questions. This survey was developed to assess demographic characteristics (e.g., income, marital status, education, religion, employment status, medical insurance coverage, number of children, gender, year of migration, length of residency in the United States). We also asked about their transition when they arrived in the United States and whether or not it was difficult. In addition, the survey included questions that were directed only at PTs related to potential barriers to continuing treatment. Finally, the survey included questions about how participants viewed various aspects of agency service provision and agency characteristics. These questions are listed in Table 2 and were responded to on a 4-point scale of 1 (poor), 2 (fair), 3 (good), and 4 (excellent). It was initially piloted with two clients from the clinic, and their feedback was used to modify and clarify the items.

Table 2 Client impression of agency
Full size table

Results

Perception of Clinic

Initial analyses sought to confirm that the clinic was meeting the needs of the participants, to rule out the possibility that RAs and PTs differed in terms of their experiences at the clinic rather than the psychosocial variables of interest. Table 2 presents participants responses to questions about the clinic. No significant differences were found between RAs and PTs on any of these variables. Overall, all but one report that the clinic was helpful and would recommend the clinic to someone else (35/36). Average ratings of “good” or above (3 = good; 4 = excellent) were found for the location of the clinic, the lobby area, reception by office staff, promptness in scheduling appointments, and promptness connecting the participant with ancillary services. The availability of parking was rated slightly less than “fair” (1 = poor; 2 = fair) and safety and security were rated slightly less than “good.”

Comparison of RAs and PTs

We examined specific factors that might differentiate the RAs from the PTs based on our literature review: financial/employment, family, medication usage, religion, and acculturation. Table 3 presents results for PTs, RAs and the total sample for each of these broad areas.

Table 3 Interview and self-report questionnaire results
Full size table

First, no significant differences were found between groups with respect to medical insurance coverage. PTs were significantly more likely to be employed at the time of treatment χ2 = 11.25, P = .001. Although PTs were more likely to be employed, no significant differences were found between groups in mean household income reported.

There were no differences between groups on family variables, including familismo as measured by the PHF, family members they could count on, and family members who were involved in treatment. It is noted, however, that both groups reported very high PHF scores, indicating that family commitment and connectedness were extremely important to both groups. In fact, 19 out of 36 reported that they were comfortable with having family participate in therapy with them; however, only one PT and four RAs actually had family members involved in treatment.

Regarding religion, as expected, participants reported religion as highly important (M = 6.05, SD = 1.67). Interestingly, the RAs rated religion as significantly more important than did the PTs, t(36) = 2.17, P = .04.

PTs and RAs did not significantly differ on use of anti-depressant medications and, in fact, most were on medication (See Table 3). However, significantly more RAs obtained their medication directly from psychiatrists at the clinic where they were also receiving therapy (as opposed to obtaining medication from their primary care physician or from a different treatment center), χ2 = 11.68, P = .001.

There were no differences between groups on any of the acculturation variables. The SASH subscale scores indicate that this sample, as a whole, preferred using Spanish in conversation, viewing and listening to Spanish language media, and socializing with other Latinos. MASI subscale scores indicated that neither group experienced Spanish competency pressure or pressure against acculturation; however, both groups experienced moderate pressure to acculturate and high English competency pressure.

Although there were no differences between RAs and PTs, a significant number of participants reported having a difficult time with the migration process and transition to the life and culture of the U.S. as noted in Table 3. In addition, the interview data indicate that 68% reported having difficulty adjusting to the U.S. of which a significant number of Mexicans said that coming to the U.S. as undocumented made their migration experience even more difficult (e.g., financial hardship, crossing the boarder, and traveling to their destination). In addition, others reported that the demands of adjusting to a new culture and way of life were also difficult.

Situations that Prevented the PT Clients from Continuing Therapy

Additional questions asked just of PTs explored which factors prevented them from continuing treatment. Table 4 shows that the most frequent reason was not being able to take time from work, followed by not having any health insurance, childcare, and transportation. Interesting, three participants dropped out because they had lost their job and were currently unemployed. Other reasons were not being familiar with the counseling process, had family problems and/or had to pick up child/children from school. However, these reasons were reported by just a few of the participants.

Table 4 Situations that prevented clients from staying in treatment (as reported by premature terminators)
Full size table

Discussion

The current study identified specific characteristics of and examined differences between Latino clients, treated for depression at a community clinic, who regularly attended sessions and Latino clients who terminated treatment prematurely. In general, this sample had been living in the United States for over a decade but still reported considerable difficulties with acculturation. Participants, on average, had not acculturated fully into mainstream United States culture, continued to speak Spanish and interacted socially with other Latinos, and viewed religion and family as very important.

The participants, on average, had an annual income below the poverty line, eighth grade education, and predominantly comprised of women (84%). Less than half (44%) of all participants were employed.

Although important to the etiology of depression, socioeconomic status did not distinguish PTs from RAs; however, more PTs were employed compared to RAs. There are several possible explanations for this finding. First, we speculate that perhaps the RAs had a more flexible schedule allowing them to attend therapy sessions on a regular basis since most were unemployed during the time of treatment. Second, PTs may have decided to drop out of treatment because of the fear of losing their jobs for taking time off from work. Furthermore, a significant number reported that they didn’t have health insurance coverage which contributed to their dropping out. In addition, it is possible that PTs left treatment prematurely because they gained employment during treatment which, in turn, limited their time to continue therapy. Finally, the RAs could have been more chronically or severally depressed (a variable not assessed in this study), which might explain attendance at more sessions and unemployment status. Furthermore, it is clear that all but one of the RAs and all of the PTs were satisfied with the agency’s services, reporting that coming to the clinic and seeing a therapist was helpful, and that they would recommend the clinic to someone else. In addition to their general satisfaction with the clinic, 80% (n = 29) of the total sample was on antidepressant medication and most of them (n = 24) obtained their medication from the psychiatrists at the same clinic where they were receiving psychological treatment. Moreover, all but two of the RAs obtained their medications from the clinic’s psychiatrist. One can speculate that this type of treatment setting facilitates better communication between counselors and psychiatrists, allowing for more agreement on treatment goals [20]. It is also possible that offering a variety of services in a single location improves attendance by simply reducing the logistical burden on clients with very little resources (e.g., they only have to learn the bus route to one location; multiple appointments could be scheduled for the same day, minimizing travel and time off work).

Several other variables are highlighted as important for future research by the current exploratory study. Religion was rated as important by most participants, but RAs reported religion as significantly more important than did PTs. While this was an unanticipated finding that waits replication, we speculate that participants with strong religious faith may continue treatment because it indirectly enhances their sense of mastery or control over their life circumstances and promotes positive change [15]. Similarly, the family is clearly important to the Latinos we interviewed. Although no statistically significant differences between groups were found, only one PT had family involved in treatment, while 4 RAs did, which may be of practical significance. Incorporating the family into treatment has been a consistent recommendation for Latinos treatment [21]. Therefore, the current research suggests that although it may be beneficial to encourage more family members to become involved in treatment it may not necessarily be easy to engage them.

Current findings are consistent with the notion that the demands of adjusting to different customs and social norms, learning a new language, and becoming familiar with new laws and rules can be a source of stress. In the current study, measures of acculturation and acculturative stress indicated that most participants experienced considerable pressure to learn English and acculturate even though they lived in a predominantly Latino community where their cultural expressions and Spanish language were accepted. Participants preferences for watching Spanish television and socializing with other Latinos may very well be reinforced by living in a predominantly Latino community. We speculate that living in such a community may reinforce the maintenance of the native language, cultural traditions and instill a sense of belonging; at the same time, however, the pressure to learn English from the dominant culture which could be limiting their attempts to integrate fully in a way that would more successfully resolve acculturative stress over time.

In the current sample, this unresolved acculturative stress undoubtedly contributed to depression (and therefore treatment seeking) but did not appear to differentially relate to treatment attendance versus premature termination. Problems with acculturation and acculturative stress may continue many years after migration. In the current sample, despite the number of years living in the United States ranging from 1 to 39, with an average of slightly more than 12 years, the participants, as a whole, were still experiencing acculturative stress. This phenomenon is consistent with the notion that acculturation is gradual and can be a life-long process [22].

In sum, our findings showed that there were very few differences between the two groups on family and acculturation variables; however, employment status and medication were statistically significant. That is, clients who continued treatment were less likely to be employed and were more likely to be on medication prescribed by a psychiatrist at the clinic. This clearly supports the premise that psychological treatment combined with medication treatment in the same setting keeps clients in treatment. What needs further exploration is the role of employment status. For example, we do not know if the severity of depression may have prevented them from gaining employment. Thus, further research is needed with a larger sample to not only explore the variables examined in this study but also to examine the relationship between employment and depression.

Limitations

The primary limitation of the study is that it utilized a retrospective rather than a prospective longitudinal design. That is, interviews were conducted after clients had terminated prematurely or had attended at least six sessions. A future direction is to interview clients during intake and then track their attendance and changes in relevant variables, including increasing the sample size. In addition, the sample is somewhat small, limiting the generalizability of the results. Finally, given that many of the PTs we attempted to contact following chart review were not enrolled, it is possible that our sample of PTs was non-random. That is, it is possible that those who terminated early because they were unhappy with the clinic may have been less likely to return our recruitment phone calls. Nonetheless, it is one of the few studies that investigated specific psychosocial factors and cultural variables that may affect treatment retention among Latino clients in a community-based mental health clinic that provides culturally responsive services.

New Contribution to the Literature

This study is an initial attempt to identify barriers to treatment adherence in a Latino population that significantly underutilizes mental health services compared to Caucasians [8]. Studies that examine perceived and actual barriers encountered by Latino clients are necessary to guide community-based agencies in developing service delivery alternatives that will keep them in treatment. As stated earlier this is particularly important because offering culturally responsive services in community clinics is seen as a way to address poor treatment engagement and retention.