Cross-Ethnic Measurement Invariance of the Brief Symptom Inventory for Individuals with Mental Illness

Abstract

The purpose of this study was to test the measurement invariance of the Brief Symptom Inventory's (BSI) secondary factor model across African, white, and Latino Americans using multiple-group confirmatory factor analyses. This study provides an examination of the BSI's validity for use in mental health service research for people with severe and persistent mental illness (SPMI). The sample consisted of 1,166 individuals receiving treatment for SPMI in a publicly funded mental health system in Los Angeles, California. The results showed that the secondary factor model fit the data very well. The findings suggest that the BSI can be used with African, white, and Latino Americans diagnosed with SPMI to assess their psychological distress and that it can be useful in comparing systematic group differences in observed means and covariance matrices across the three ethnic groups. It is recommended that the nine subscale scores and the Global Severity Index score of the BSI be used in practice and research with African, white, and Latino Americans with SPMI. KEY WORDS: Brief Symptom Inventory; confirmatory factor analysis; ethnic disparities; measurement invariance; severe and persistent mental illness ********** Valid measurement tools are critical for assessing a wide range of phenomena in mental health. Measurement instruments are used in clinical practice and in research to identify patients in a general population, to assess psychological symptoms, and to evaluate individual changes in symptoms resulting from treatment. A critical issue in the use of psychological measures is the possible variance resulting from differences across ethnic groups. The surgeon general's report (U.S. Department of Health and Human Services, 2001) emphasized that conceptual and scale equivalence, also called measurement invariance, is important to mental health services and research. For example, community-based psychosocial treatment programs for people with severe and persistent mental illness (SPMI) are limited in their ability to demonstrate program effectiveness in the case of ethnic minority groups because of the lack of standardized treatment outcome measures that are equivalent across different ethnic groups (National Institute of Mental Health, 2003). Plausible cultural differences can prohibit measures of psychological symptoms from being equivalent across ethnic groups, which results in the lack of validity in making clinical assessments and evaluating treatment outcomes. Recent discussions about ethnic disparities in the access, utilization, and outcome of health and mental health services (Ramirez, Ford, Stewart, & Teresi, 2005; Snowden, 2003) have highlighted the need for assessment tools that are valid across ethnic groups so that these disparities can be adequately understood and ameliorated. Thus, it is imperative to test whether symptom measures of interest hold measurement invariance across ethnic groups. The Brief Symptom Inventory (BSI) is a self-report mental health symptom scale (Derogatis & Melisaratos, 1983). It is a 53-item version of Symptom Checklist-90 (SCL-90-R) (Derogatis, 1992). According to the BSI manual (Derogatis, 1993), it is designed to assess general psychiatric distress using nine subscales. A Global Severity Index (GSI) can be derived, which is the average score of all 53 items. The BSI has been used with a wide range of adult populations. For example, an Ovid Medline search yielded more than 260 publications since 2000 that used the BSI, and PsycINFO listed more than 1,700 publications since 2000 that have used the BSI. The BSI is also widely used in mental health research on people with SPMI as a psychiatric symptom measure. For example, the BSI has been used in studies of comparative culture (Coelho, Strauss, & Jenkins, 1998), pharmacology (Salokangas et al., 1996), quality of life (Ritsner, 2003), family issues (Jenkins & Schumacher, 1999), involuntary outpatients (Swartz, Swanson, Wagner, Burns, & Hiday, 2001), program effectiveness (Kingsep, Nathan, & Castle, 2003), service systems (Mares, Young, McGuire, & Rosenheck, 2002), and measurement studies (Preston & Harrison, 2003). …