Kemani, M. (2015) Efficacy and Processes of Change in Acceptance and Commitment Therapy for Chronic Pain. (Dissertation) Karolinska Institutet, Stockholm, Sweden.

Background:

Approximately one out of five adults in Sweden suffer from longstanding pain and many experience pain-related disability and reduced quality of life. Long-standing pain is also associated with substantial societal costs and psychological factors are central to its development and maintenance. Studies imply the efficacy of cognitive behavior therapy (CBT) and Acceptance and Commitment Therapy (ACT) for long-standing pain in adults, but there is a need for: (1) Randomized controlled trials evaluating the efficacy of ACT compared to established treatments; (2) adequate measures for use in treatment-evaluations that assess the impact of pain on behavior; (3) further evaluations of processes of change in ACT, and specifically for studies that more adequately model change over time and investigate the temporal precedence of change in the mediator in relation to the outcome; and (4) studies evaluating the cost-effectiveness of ACT.

Aims:

The overarching aims of the doctoral thesis were to evaluate the efficacy and processes of change in ACT for longstanding pain. Specifically, the aims were to evaluate: (1) The efficacy of ACT delivered in a group setting (Study I and Study II); (2) the cost-effectiveness of ACT (Study II); (3) the psychometric properties of a brief measure assessing pain interference, the Pain Interference Index (PII) (Study III); and (4) if changes in psychological inflexibility mediated changes in pain disability (Study I) and pain interference (Study IV).

Methods:

The efficacy of ACT was tested in two randomized controlled trials. The first trial utilized a wait-list control condition and included women with fibromyalgia(Study I). Assessments were done pre-and post-treatment and at 3-month follow-up. In Study II adults with non-specific longstanding pain were randomized to ACT or applied relaxation (AR). Data was collected pre-, mid- and post-treatment, and at 3- and 6- month follow-up. Pain disability was the primary outcome measure in both studies. Data was analyzed using hierarchical linear modeling and latent growth curve modeling. Cost-effectiveness was evaluated by calculating the quotient of the difference in average changes in costs and pain disability in ACT and AR (Study II).

The psychometric properties of the PII were evaluated using cross-sectional data from adults with non-specific longstanding pain (Study III). Analyses comprised a principal component analysis, analysis of item statistics, corrected item-total correlations and inter-item correlations. Concurrent criteria validity was evaluated using zero-order correlations and ordinary least squares regression analyses including pain intensity, pain disability, health-related quality of life and depression. In Study I mediation analyses were based on change scores in psychological inflexibility during treatment and change scores between pre- to follow-up assessment in outcomes (e.g. pain disability). Study IV incorporated the specified timeline between mediator (e.g. psychological inflexibility) and outcome (pain interference) based on session-to-session assessments from participants in Study II and used multilevel regression analyses to model change.