CHAPTER 6

Mental Health Care for Hispanic Americans

Appropriateness and Outcomes of Mental Health Services

Studies on Treatment Outcomes

Few studies on the response of Latinos to mental health care are available. Only three small studies of depression have been published. They investigated the care for depression given to unmarried Puerto Rican mothers with depressive symptoms (Comas-Diaz, 1981), to Mexican American women (Alonso et al., 1997), and to Puerto Rican adolescents (Rossello & Bernal, 1999). Although all found that those who were treated had favorable results, the sample sizes are far too small to establish the response of Latinos to care for depression.

Another study examined interventions for schizophrenia among Latinos. In this randomized study, members of low-income, Spanish-speaking families were more likely to suffer a significant exacerbation of symptoms in highly structured family therapy than in the less structured case management (Telles et al., 1995). The authors of this study speculated that these individuals may have found this highly structured treatment too intrusive.

Several preventive intervention studies have focused on Latino children and families (Costantino et al., 1986, 1988; Szapocznik et al., 1989; Malgady et al., 1990; Lieberman et al., 1991). In these studies, mental health professionals provided culturally adapted preventive care to immigrant mothers and infants in San Francisco (Lieberman et al., 1991), Puerto Rican children and parents in New York City (Costantino et al., 1986), and families in Miami (Szapocznik et al., 1989). In general, the interventions resulted in short-term gains, but long-term follow-up evaluations to determine whether they actually prevented later mental disorders were not reported.

Two effectiveness studies examined treatment for depression among ethnically mixed samples of primary care patients with significant proportions of Latinos. In the first study, Miranda and Munoz (1994) investigated the effectiveness of group cognitive treatment for minor depression. Although analyses were not run separately for Latinos, who comprised 24 percent of the sample, the findings indicated that patients receiving the cognitive treatment improved significantly more than those who received no intervention or who watched a 40-minute videotape.

The second study was more ambitious. It was carried out in 46 primary care clinics across six managed systems of care (Wells et al., 2000). Two of the cities in the study, San Luis, Colorado, and San Antonio, Texas, have large Mexican American communities. Latinos comprised nearly a third (30 %) of the enrolled sample (N = 1,356). The purpose of the study was to assess the effects of programs to improve the quality of care for depression. Specifically, usual care was compared with two interventions, one for which medication was administered and closely followed for 6 or 12 months and the other for which local psychotherapists provided cognitive-behavior treatment ranging from 4 sessions for minor depression and related problems to 10–16 sessions for major depression. Although results broken down by ethnicity have yet to be published, the initial findings indicate that, relative to usual care, the quality improvement programs had significant effects on treatment process, clinical outcome, and even social outcomes such as employment.

Diagnostic and Testing Issues

Quality care requires valid diagnostic and clinical assessment. Several studies have found that bilingual patients are evaluated differently when interviewed in English as opposed to Spanish (Del Castillo, 1970; Marcos et al., 1973; Price & Cuellar, 1981; Malgady & Costantino, 1998); however, the extent to which these factors result in misdiagnoses is not known. One small study examining records of patients with bipolar disorder (manic depressive illness) found that in the past, both African American and Latino patients were more likely to have been misdiagnosed as schizophrenic than whites (Mukherjee et al., 1983). Further research is needed to clarify how cultural and linguistic factors influence diagnoses (Malgady et al., 1987; Lopez, 1988).

Psychological testing can also be affected by language and cultural factors. Of particular interest is testing that contributes to the diagnosis of mental retardation (e.g., cognitive intelligence tests), dementia (neuropsychological testing), and mental disorders (psychological tests such as the MMPI-2). The two main positions on testing are that (1) tests are biased against minority group members (e.g., Guthrie, 1998), and (2) there is no evidence of ethnic or cultural bias (Gottfredson, 1997). Cole (1981) refers to these positions as those of the reformers and the defenders. Most of the literature involves African Americans (e.g., Helms, 1992), and when Latinos are included, they are mostly English-speaking Latinos (e.g., Sandoval, 1979). However, the literature concerning Latinos and the particular challenge of assessing bilingual persons and those with limited English proficiency is growing (e.g., Jacobs et al., 1997).

The lack of reliable and valid tests normed on con-temporary samples of Latinos, both Spanish-speaking and English-speaking, is a significant obstacle to carrying out the appropriate assessment of Latinos (Bird et al., 1987; Loewenstein et al., 1994; Velasquez et al., 1998). Two of the most widely used tests for diagnostic purposes are the Wechsler scales of intelligence and the MMPI-2. The available Wechsler test for Spanish-speaking adults, Escala Inteligencia de Wechsler para Adultos (EIWA), was published in 1968 and was based on a standardization sample of Puerto Rican islanders (Wechsler, 1968). Since then, two English language versions have been standardized and published (Wechsler, 1981, 1998).

The current Spanish language norms are significantly outdated, and available research has demonstrated their overestimating the level of functioning of some Spanish-speaking adults (e.g., Lopez & Taussig, 1991). The children’s version of the WAIS, however, has been developed and standardized on a more contemporary sample of Puerto Rican island children (Wechsler, 1989). In the restandardization of the MMPI (MMPI–2; Butcher et al., 1989), little consideration was given to Latinos. Of the 2,600 who comprised the standardization sample, only 73, or 2.8 percent, were identified as Hispanic. This percentage reflected only one-third of the actual Hispanic representation in the Nation at that time. Both the EIWA and MMPI-2 demonstrate that some test publishers assign little importance to providing contemporary and representative norms of Latinos in the United States. This statement does not apply to all tests, since recent advances have been made in the development of language skills tests in Spanish and English (e.g., Woodcock & Munoz, 1993) and nonverbal tests (e.g., Bracken & McCallum, 1998, Naglieri & Bardos, 1999). At the very least, tests based on normative samples of U.S. adults or children should include subsamples of Latinos that accurately reflect their representation in the Nation. At best, Latinos should be oversampled so that tests of fairness can be carried out that attend to differences among sub-groups within the Hispanic American population as well as differences between Hispanic Americans and other racial and ethnic groups.

Evidence-Based Treatment

To determine whether there are disparities in mental health care, it is important to discover whether Latinos are as likely as white Americans to receive care that is consistent with guidelines established by recognized psychiatric and psychological organizations. Recent data suggest that Latinos are less likely than whites to receive treatment according to evidence-based guidelines. Evidence from a representative national sample suggests that many individuals with depression and anxiety do not receive appropriate care (Young et al., 2001); fewer Hispanics receive appropriate care (24 %) than do whites (34 %).

Another study examined the use of antidepressants among clients who had visited a general medical doctor (National Ambulatory Medical Care Surveys of 1992–1993 and 1994–1995). During the two time periods in the early 1990s that were evaluated, Latinos were less than half as likely as whites to have received either a diagnosis of depression or antidepressant medication (Sclar et al., 1999).

A few small preliminary studies have examined pharmacologic responses in Latino populations. In the research that does exist, data are often drawn from aggregate samples of several different Hispanic groups in attempts to characterize a typical Hispanic response (Mendoza & Smith, 2000). However, evidence of important genetic variation among subgroups (i.e., Mexican Americans, Puerto Ricans, and Colombians) implies that disaggregated data are needed before any ethnopsychopharmacological findings should be considered conclusive (Hanis et al., 1991; Mendoza & Smith, 2000).

Cultural Competence

Sue and colleagues (1991) studied community mental health centers in Los Angeles in order to examine ethnically matched provider services versus nonmatched provider services. Ethnic match resulted in longer duration of treatment for Mexican Americans, as well as better patient response to treatment based on a global indicator of functioning. This suggests that ethnic match of provider and consumer can be important in providing services for some Latinos.

One limitation of ethnic match research is that there is no direct assessment of clinicians’ cultural under-standing or skills. Therefore, it is not clear if the cultural competence of practitioners is related to the positive findings of ethnic match. Direct study of cultural competence for Latinos is needed. Although there have been efforts to develop specific cultural competence guide-lines for Latinos (Western Interstate Commission for Higher Education, 1996), most models that have been developed apply across ethnic groups.

Cultural competence has received widespread attention across the Nation. Some State and local policymakers now require cultural competence training for their practitioners. Federal agencies are supporting the development and implementation of guidelines (e.g., CMHS, 2000). Despite the several models and the growing interest in cultural competence, much work needs to be done before cultural competence will positively impact mental health service delivery for Latinos and other ethnic groups. Currently, cultural competence is largely a set of guiding principles that lack empirical validation. Thus, an essential step in advancing culturally competent services for Latinos is to carry out research to test the guide-lines, standards, or models proposed by these expert clinicians and administrators. Bernal et al. (1995) and Lopez et al. (in press) discuss multiple strategies to develop culturally informed interventions.