Objectives: Existing research on the alliance between professional and client suggests that a stronger alliance is predictive of better treatment outcomes (Duncan et al., 2009; McCleod, 2011). As most of these studies have... [ view full abstract ]
Objectives: Existing research on the alliance between professional and client suggests that a stronger alliance is predictive of better treatment outcomes (Duncan et al., 2009; McCleod, 2011). As most of these studies have been conducted on adult clients, we know little about the role of alliance in youth care or family care. Another important limitation of existing studies is that they are either cross-sectional in nature, or have not optimally used their longitudinal assessments of alliance during treatment. The latter is crucial because several studies reported that there are differential effects of the alliance, depending on when the alliance has been assessed during treatment (e.g., at the beginning or the end of treatment; Sharf et al., 2010). Thus, to really gain an understanding the role of the alliance in youth or family care, we need studies that include adolescents, have multiple assessments of the alliance over time, and preferably have multiple follow-up outcome assessments to be able to investigate long-term effects of care.
In the presentation, we describe such an empirical study on almost 2,000 adolescents and their families. We will examine how the alliance unfolds during treatment, and how it is related to outcomes assessed at 6, 12, and 18 months follow-up. More specifically, we will use three different models to test the alliance-outcome link (see Owen et al., 2016): (a) mono-method model (i.e., 1 rating of working alliance correlated with outcomes), (b) aggregate-assessment model (i.e., multiple sessions aggregated and correlated with outcomes), and (c) change-based model (i.e., changes in working alliance scores correlated with outcomes).
Method: Alliance was assessed monthly as part of the regular quality assessment of Multisystemic Therapy (MST). The participants consisted of roughly 2,000 adolescents with antisocial and/or behavioural problems, and their families. Latent class growth analyses were conducted to distinguish subgroups characterized by similar developmental trajectories regarding the outcome measures during follow-up. Subgroup membership was regressed on the mono-method, aggregated, and change-based measures of alliance.
Results: The analyses are currently conducted. In line with Owen et al. (2016) we hypothesise that trajectories of care that are characterised by a continuous development of the alliance (change-based model), rather the a good alliance during a specific phase of treatment (mono-method or aggregate-asssessment model), will be most predictive of positive outcomes.
