Modelling changes in anxiety and depression during low‐intensity cognitive behavioural therapy: An application of growth mixture models. (7th November 2019)

• Record Type: Journal Article
• Authors: McDevitt‐Petrovic, Orla; Shevlin, Mark; Kirby, Karen
• Is Part Of: British journal of clinical psychology. Volume 59:Number 2(2020)
• Page Start: 169
• Page End: 185
• Publication Date: 2019-11-07
• Subjects: low‐intensity CBT -- anxiety -- depression -- growth mixture modelling -- psychological well‐being practitioner -- treatment response; 616.89
• DOI: 10.1111/bjc.12237
• Languages: English
• ISSNs: 0144-6657

Abstract

Abstract : Objectives: Research largely supports the clinical effectiveness of low‐intensity cognitive behavioural therapy (LICBT) for mild‐to‐moderate anxiety and depression, delivered by psychological well‐being practitioners (PWPs). Knowledge regarding the predictors of treatment response, however, is relatively limited. The primary aim of this study was to model the heterogeneity in longitudinal changes in anxiety and depression throughout LICBT provided by PWPs in Northern Ireland (NI), and to explore associations between pre‐treatment variables and differences in treatment response. Methods: Growth mixture modelling (GMM) techniques were employed to examine changes in psychological status in clients ( N = 253) over the first six sessions of treatment, to identify divergent early response trajectories. A series of pre‐treatment variables were used to predict class membership using chi‐square tests and binary logistic regression models. Results: There was one class representing improvement and one representing no improvement for both anxiety and depression. Class membership was predictive of treatment outcome. Pre‐treatment variables associated with less improvement included unemployment, risk of suicide, neglect of self or others, using medication, receiving previous or concurrent treatments, a longer duration of difficulties, and comorbidities. Conclusions: Findings indicate most of the sample populated an 'improvers' class for both depression and anxiety. Pre‐treatment variables identified as predictive of poor treatment response may need to be considered by practitioners in potential triage referral decision policies, supporting cost‐effective and efficient services. Further research around predictors of clinical outcome is recommended. Practitioner points: Most of the sample belonged to an 'improvers' class. Several pre‐treatment variables predicted poor treatment response (unemployment, suicide risk, neglect, medication, previous or concurrent treatments, longer duration of difficulties, and comorbidities). Few studies have utilized GMM to determine predictors of outcome following LICBT Regarding pre‐treatment variables, the possibility of self‐report bias cannot be excluded. The time period was relatively short, although represented the optimum number of sessions recommended for LICBT. The lack of a control group and random allocation were the main limitations.