Change factors in depression and substance use treatment: a longitudinal integrative model

Abstract

Background: While psychotherapy efficacy has been largely supported, patient change processes across time are still less clear. This is specially true for complex clinical populations, such as dually diagnosed individuals.

Aims: A Latent Growth Model was tested to describe change in depression and substance use in patients from two Randomized Clinical Trials (RTCs). The model tested the impact of the alliance and four evidence-based change principles as predictors of symptomatic change during treatment and follow-up, in addition to advancing the premise that initial treatment matching to non-diagnostic patient characteristics improved the quality of the therapeutic alliance.

Method: Patients (N = 190) completed one of five 20-session treatments, to which they were randomly assigned.

Results: A good fit was found for a quadratic change function for both outcome measures, that is, symptoms decreased in a curve shape followed by a small increase. Full predictive models yielded acceptable fit. Patients high on subjective distress at treatment-entry benefited from affective-focused interventions in terms of their substance use. Matching patient coping style (externalizing vs. internalizing) with treatment focus (symptom change vs. insight-oriented) predicted a better alliance.

Conclusions: Effects of both relationship and interventions were found, as well as their interplay.

Keywords:

• psychotherapy research
• dual diagnosis
• treatment matching
• patient characteristics
• change processes
• latent growth modeling

Acknowledgements

The first author would like to thank the STS Research Lab Team members and collaborators at the University of California, Santa Barbara, too numerous to name, but whose hard work contributed to this study.