OBJECTIVE: To investigate the construct validity, convergent validity and reliability of the Patient Satisfaction with Mental Health Services Scale (SATIS-BR). METHOD: One hundred and ten psychiatric patients attending five public mental health services located in Minas Gerais, Brazil, participated in this research. The scale was applied in individual structured interviews. The scale has 12 items evaluating patients' satisfaction, with response alternatives disposed in a five points Likert scale. In order to evaluate the convergent validity, another scale was applied, the Perceived Improvement Questionnaire (PIQ). RESULTS: Factorial analysis using the Principal Axis Factoring method revealed a 3 factors structure, evaluating satisfaction regarding the following dimensions: 1. Professional competence and understanding; 2. Help from the professional; 3. Service physical conditions. Reliability results showed adequate internal consistency evaluated by Cronbach analysis (alpha = 0,88). Convergent validity analysis showed a positive significant Pearson Correlation coefficient (r = 0,41; p < 0,001) with the PIQ scale, which evaluates the related construct of patient" perceived improvement. CONCLUSION: The SATIS-BR scale has adequate psychometric properties of construct validity, convergent validity and reliability.
Patient satisfaction; scales; outcome assessment (health care); mental health services
