Psychometric Properties of Scale to Assess the Therapeutic Relationship—JapaneseVersion (STAR-J)

Abstract

Background: A good therapeutic relationship between patient and psychiatrist is vital for effective mental health care. However, no instruments to assess this relationship are available in Japan. This study aimed to develop a Japanese version of a Scale To Assess Therapeutic Relationship (STAR-J), which measures such relationships from the viewpoints of both the patient (STAR-J-P) and clinician (STAR-J-C). We examined the tool’s psychometric properties, including factor structure, internal consistency, convergent validity, and test-retest reliability among psychiatric outpatients and psychiatrists. Methods: Study participants comprised 139 outpatients and 10 psychiatrists. Exploratory factor analysis was conducted to investigate factor structure; to confirm cross-validity, confirmatory factor analysis was conducted using a different sample constituting 195 participants in an assertive community treatment program and their 91 case managers. Cronbach’s alpha was used to assess internal consistency. For STAR-J-P only, the intra-class correlation coefficient (ICC) was computed for 17 patients to determine test-retest reliability. Spearman’s correlation coefficients were calculated to examine convergent validity with service satisfaction, empowerment, and medication adherence. Results: We identified a two-factor structure for STAR-J-P and a one-factor structure for STAR-J-C. Cronbach’s alphas for the two STAR-J-P factors were 0.897 and 0.645, and that for the STAR-J-C factor was 0.949. The ICCs for STAR-J-P factors 1 and 2 were 0.765 and 0.630, respectively. STAR-J-P and STAR-J-C were not significantly correlated. STAR-J-P factors 1 and 2 showed significant correlations with service satisfaction (factor 1: ρ = 0.648, p < 0.001; factor 2: ρ = 0.238, p = 0.005) and medication adherence (factor 1: ρ = 0.508, p < 0.001; factor 2: ρ = 0.347, p < 0.001), but only factor 1 showed a significant relationship with empowerment (ρ = 0.283, p = 0.001). STAR-J-C was significantly correlated only with empowerment (ρ = 0.207, p = 0.017). Conclusions: STAR-J appears to be a useful instrument for assessing therapeutic relationships in the Japanese psychiatric outpatient setting. Further studies should test its validity and applicability in different mental health service settings.