Effect of a Significant Other on Client Change Talk in Motivational Interviewing

Objective:To examine significant-other (SO) and therapist behaviors as predictors of client change language within motivational interviewing (MI) sessions. Method: Participants from an emergency department received a single session of MI that included SO participation (N = 157). Sessions were coded using therapy process coding systems. Sessions were subdivided into 10 equal deciles to facilitate sequential analyses. Multilevel modeling was used to examine the relationships among the following variables: therapist MI-consistent and MI-inconsistent behaviors; client change talk and sustain talk; and SO Support Change and SO Against Change. Results: Therapist MI-consistent and MI-inconsistent behaviors failed to predict either client change talk or sustain talk at the decile level. Global measures of therapist MI spirit and acceptance were associated with lower levels of client sustain talk (p = .002 and p < .001, respectively). Higher levels of SO Support Change were more likely to be followed by higher levels of client change talk (p < .001) and lower levels of client sustain talk (p < .001). SOs who engaged in behaviors that discouraged the patient's drinking in the 6 months prior to the intervention had higher levels of SO Support Change language (p = .02). When analyzed at the aggregate level, therapist behavior was significantly associated with client change talk, but effect sizes were quite modest. Conclusions: Within-session SO behavior impacts client verbalizations regarding changes in alcohol use. Results raise questions about the role of therapist behavior when an SO is present.