Routine monitoring of therapeutic alliance to predict treatment engagement in a Veterans Affairs substance use disorders clinic.

Measurement-based care (MBC) can improve mental health treatment outcomes and is a priority within the Department of Veterans Affairs (VA). However, to date, MBC efforts within the VA have focused on assessment of psychological symptoms to the exclusion of psychotherapy process variables such as the therapeutic alliance that may predict treatment response. This quality improvement project involved the implementation of routine monitoring of alliance within a VA substance use disorder (SUD) clinic predominantly serving veterans with serious mental illness. Alliance ratings were provided by 98 veterans following group therapy sessions. Low alliance ratings were used by the clinicians (n = 4) leading the groups (n = 9) as opportunities to discuss veterans' treatment experience and increase engagement. Using multilevel models that accounted for the nested nature of the data and veteran demographics, alliance ratings showed a small increase over time (B = 0.075, p < .001, f2 = 0.033). In addition, maximum alliance rating (i.e., patients' highest rating of alliance across all observations) was significantly but modestly associated with attendance at both MBC group sessions and all SUD-related visits in the 3 months following the initial alliance rating (Bs = 0.96 and 1.79; ps = .006 and .004; f2s = 0.079 and 0.088, respectively). Average alliance rating, however, was not associated with treatment attendance (ps > .050). Findings suggest that assessment of alliance is feasible within a VA SUD clinic and may provide information signaling risk for disengagement that could be used for increasing treatment engagement. (PsycInfo Database Record (c) 2020 APA, all rights reserved).