Psychiatrist-patient verbal and nonverbal communications during split-treatment appointments

Objective: This study characterized psychiatrist and patient communication behaviors and affective voice tones during pharmacotherapy appointments with depressed patients at four community-based mental health clinics where psychiatrists provided medication management and other mental health professionals provided therapy ("split treatment"). Methods: Audiorecordings of 84 unique pairs of psychiatrists and patients with a depressive disorder were analyzed with the Roter Interaction Analysis System, which identifies 41 discrete speech categories that can be grouped into composites representing broad conceptual communication domains. Cluster analysis identified psychiatrist communication patterns. T test and chi square analyses compared the clusters for verbal dominance, affective voice tone, and characteristics of psychiatrist and patients. Results: On average, 53% of psychiatrist talk was devoted to partnering and relationship building, and 67% of patient talk was about biomedical subjects, such as depression symptoms, and psychosocial information giving. Psychiatrist communication patterns were characterized by two clusters, a biomedical-centered cluster that emphasized biomedical questions (η²=.22, df=82, p<.001) and education or counseling (η²=.20, df=82, p<.001) and a patient-centered cluster focused on psychosocial and lifestyle questions (η²=.24, df=82, p<.001) and information giving (η²=.17, df=82, p<.001). The patient-centered cluster was associated with patients'expression of distress, anger, or other negative affects (t=3.22, df= 82, p=.002). Conclusions: Psychiatrists devoted much of their talk to partnering and relationship building while maintaining a focus on symptoms or psychosocial issues. However, patient behaviors did not reflect a similar level of partnering. Future studies should identify psychiatrist communication behaviors that activate collaborative patient communications or improve treatment outcomes.