Abstract
BACKGROUND: Interest in the assessment of patient satisfaction with in-patient psychiatric treatment has steadily increased and several measurement tools are available for the quantification of patients' experience. However, they are often uni-dimensional or focus mainly on therapeutic relationship and environment, and neglect other important issues such as information about treatment and participation. The BEST questionnaires were developed as comprehensive instruments that include items on all of the mentioned topics. The present study evaluates the psychometric properties of the BEST in a version for adolescents and for parents. Furthermore, the dimensionality of the satisfaction ratings is analyzed. METHOD: Descriptive statistics were applied to data of 1582 adolescents (mean age = 15.0 years, SD = 1.65; 62.4% female) and 1998 parents/guardians assessed in seven in-patient units across Germany. The factorial structure of the BEST questionnaires was determined by exploratory and confirmatory factor analyses, including a bifactor model. RESULTS: The psychometric quality of the scales was strong. Correlations with another assessment instrument of patient satisfaction were good to high, indicating good convergent validity. Exploratory factor analyses revealed three factors in adolescents that were labelled as: Therapeutic relationship, environment, and general satisfaction and treatment success. For parents, the same three factors could be distinguished. Confirmatory bifactor models suggested that the vast majority of variance was accounted for by the general factor; the three specific factors provided some additional information. Agreement between the subscales of adolescents and their parents was only moderate. Parents were usually more satisfied. CONCLUSIONS: The BEST questionnaires can be considered as reliable and valid instruments to not only assess the "classical" aspects of patient satisfaction, but to also assess newer fundamental topics such as children's rights and treatment participation. For scientific usage, the total score seems superior because of the high explained variance by the general factor, but the subscale scores provide further information. The use of single items seems advantageous for quality management purposes.