RESUMO
BACKGROUND: To date, no data have been available concerning the psychometric characteristics of the Internalized Stigma of Mental Illness scale (ISMI-29) in Latin American countries. The aim of this study was to validate a Latin American version of the ISMI in people with schizophrenia. METHODS: The study included 253 stabilized outpatients with schizophrenia from 3 Mental Health Services in three Latin American countries: Bolivia (N = 83), Chile (N = 85) and Peru (N = 85). We analyzed the psychometric properties using item response and classical test theories. An item reduction was then performed to improve the psychometric properties of the ISMI-29. The final version of the ISMI was tested for construct validity, reliability, external validity and differential item functioning (DIF). RESULTS: The five-factor structure of the ISMI-29 was not confirmed using confirmatory factor analysis (RMSEA = 0.12, CFI = 0.77, and WRMR = 2.20). Seventeen items were discarded to obtain a satisfactory psychometric version. The ISMI-12 evaluates 3 dimensions: social stigma (4 items), stigma experience (4), and self-stigma (3). The factor structure accounted for 68% of the total variance. Internal consistency was satisfactory. The scalability was satisfactory, with INFIT statistics within an acceptable range. In addition, the results confirmed the absence of DIF and supported the invariance of the item calibrations between countries. CONCLUSION: The ISMI-29 is not valid in our sample and should not be used in Latin American countries. The ISMI-12 is the first internalized stigma questionnaire with satisfactory psychometric properties available in Latin American countries. Its brevity could facilitate its dissemination and use in clinical settings.