Ethnic and gender differences in mental health service utilization are well documented. Individuals from ethnic and racial minority groups in the United States have been reported to underutilize mental health services when compared to those from the majority group. Once they have accessed services, individuals from racial minority groups have been found to average significantly fewer treatment sessions than white clients and to drop out of therapy at significantly higher rates (Sue, 1977; Vail, 1996; Vernon and Roberts, 1982).
With regard to gender differences in mental health service utilization, the picture is more complex. While data suggest that men underutilize services more than women do, among consumers of mental health treatment, men use costly services such as inpatient treatment more frequently than do women. In 1987, women were 60% more likely than men to use mental health services, but average spending by men who used services was more than 80% higher than average spending by women who use services (Glied and Kofman, 1995). Increased use of mental health services by men for treatment of less severe symptoms might decrease their need for costly inpatient services. These patterns raise the important question of how we might make services more acceptable, accessible and attractive-particularly to men and patients from racial minority groups.
Many of the factors purported to influence accessing mental health services by men and ethnic minorities are systemic in nature, ingrained within our culture, and consequently, difficult to change (e.g., gender differences in attitudes toward help-seeking, ethnic differences in the use of alternative healing resources). However efforts have been made within the mental health system to make services more acceptable to men and minority group members who choose to, or are able to, access the system. For example, most training programs now include instruction in culturally competent counseling. Also, many clinics have implemented special gender-sensitive or ethnicity-sensitive treatments. One practice adopted by many clinics which has been argued to increase the acceptability of treatment services is matching the race and sex of the therapist to that of the client.
There are fairly extensive bodies of research on ethnic matching and gender matching of therapists and clients. Research examining clients' stated preference find that ethnic minorities, particularly African Americans, want to be treated by individuals from their ethnic group (Atkinson, 1983). Research on the effects of race or ethnicity matching of client and therapist on dropout rate and treatment duration has been somewhat inconsistent.
In Atkinson's review of research conducted prior to 1982, there was very little support for the superiority of ethnic matching of client and therapist. It must be noted that the studies reviewed by Atkinson primarily focused on African Americans. In more recent research, ethnic matching has more consistently been related to increased treatment duration and decreased dropout-especially among Asian ethnic groups and less so among African Americans (Flaskerud, 1991; Sue et al., 1991).
Research on the benefits of gender matching has been less clear. Early reviews of this literature reached differing conclusions with one asserting that gender matching results in superior outcomes (Luborsky et al., 1971). Others concluded that there was no clear relationship between patient-therapist gender matching and outcome (Garfield, 1978). Most of the more recent research seems to suggest that gender matching is beneficial only for women and may be relatively detrimental for men-or to state more simply, that both male and female clients prefer women therapists (Jones et al., 1987) suggests that clients treated by female therapists experienced more symptom improvement and reported more satisfaction with treatment than did those treated by male therapists.
It is critical to note, however, that the majority of the clients in the above-mentioned studies were white, middle class clients. Studies using minority clients report contradictory findings. In Vail's study of black clients, more people dropped out of treatment if assigned a therapist of the same sex. Two studies of Asian clients found that gender matching predicted longer treatment duration for males but not females (Wu and Windle, 1980; Flaskerud, 1991). Asian male clients stayed in treatment longer if assigned to a male therapist than if treated by a female therapist.
These studies, across racial groups, on the impact of gender matching on treatment duration, point to the importance of considering both gender and race when making assignments to treatment. However, most of the past empirical studies have typically focused separately on the impact of race or gender in service utilization and have failed to consider the additive effects of race and gender matching.
In a recent study, we examined the treatment records of black patients and white patients, both women and men, who initiated outpatient therapy at a large Midwestern department of psychiatry. The sample consisted of 154 black men, 131 white men, 235 black women and 177 white women. The number of clinical encounters in the six-month period following treatment onset was tabulated using the department's database. Individuals who did not return for appointments following the initial encounter were considered treatment dropouts. Our objective was to determine if these two variables-number of clinical encounters and dropout status-were related to gender and race matching for black and white patients.
Consistent with previous research, black clients were much more likely than whites to drop out after the first session. Closer examination of the data revealed that this effect was accounted for by those black clients who were assigned to a therapist of a different race. Black clients assigned to a black therapist were no more likely than white clients to drop out of treatment. In multivariate tests, controlling for the patients' Medicaid status, diagnosis and age, we found that race matching was associated with decreased dropout and increased treatment duration for black but not for white clients.
Also consistent with previous research, males were more likely than females to drop out after the first session. Interestingly, sex matching had an opposite effect on dropout status and treatment duration for male and female patients. Sex matching was associated with decreased dropout and increased treatment duration for females but with increased dropout and decreased treatment duration for male patients.
What does this all mean? First it is important to point out that we are simply looking at treatment duration, not symptom improvement. While black clients attended more sessions with a black therapist and all clients attended more sessions if seen by women, it is not at all clear that black clients are benefited more from treatment with black therapists, or that all clients were aided more by therapy with women. Previous research with Asian clients did not find improvements in Global Assessment Functioning scores based upon ethnicity matching but did report increased treatment duration if client and therapist were ethnically matched.
While perhaps not related to symptom improvement, increased treatment duration and decreased dropout have been argued to be proxy variables for client satisfaction with treatment. If we accept this argument, what might the current findings mean for clients' satisfaction with treatment? Why might black clients be more satisfied with black therapists than with therapists from other racial groups, while white clients are about equally satisfied with white and non-white therapists? Additionally, why might clients be more satisfied with female than with male therapists? Perhaps as argued by social-role theorists, individuals in positions of lower power and status are relatively expert at reading/perceiving members of higher status groups than are those from higher status groups at reading people from lower status groups. Thus, for example, women may be more adept at attending to men, than men are at attending to the feelings and experiences of women. Similarly white providers may be less familiar with the culture and life experience of minority clients than are minority providers with white clients (Sue and Zane, 1987).