Send to

Choose Destination
Pain. 2016 Sep;157(9):2115-23. doi: 10.1097/j.pain.0000000000000626.

Cognitive behavioral therapy for chronic pain is effective, but for whom?

Author information

aCenter for Self-Report Science bCenter for Economic & Social Research, University of Southern California, Los Angeles, CA, USA cDepartment of Psychiatry and Behavioral Sciences, Duke University Medical Center, Durham, NC, USA dDepartment of Medicine, Duke University Medical Center, Durham, NC, USA eDepartment of Psychiatry and Behavioral Science fSchool of Nursing gDepartment of Medicine, Rheumatology Division hDepartment of Radiology, Stony Brook University, Stony Brook, NY, USA.


Moderator analyses are reported for posttreatment outcomes in a large, randomized, controlled effectiveness trial for chronic pain for hip and knee osteoarthritis (N = 256). Pain Coping Skills Training, a form of cognitive behavioral therapy, was compared to usual care. Treatment was delivered by nurse practitioners in patients' community doctors' offices. Consistent with meta-analyses of pain cognitive behavioral therapy efficacy, treatment effects in this trial were significant for several primary and secondary outcomes, but tended to be small. This study was designed to examine differential response to treatment for patient subgroups to guide clinical decision-making for treatment. Based on existing literature, demographic (age, sex, race/ethnicity, and education) and clinical variables (disease severity, body mass index, patient treatment expectations, depression, and patient pain coping style) were specified a priori as potential moderators. Trial outcome variables (N = 15) included pain, fatigue, self-efficacy, quality of life, catastrophizing, and use of pain medication. Results yielded 5 significant moderators for outcomes at posttreatment: pain coping style, patient expectation for treatment response, radiographically assessed disease severity, age, and education. Thus, sex, race/ethnicity, body mass index, and depression at baseline were not associated with level of treatment response. In contrast, patients with interpersonal problems associated with pain coping did not benefit much from the treatment. Although most patients projected positive expectations for the treatment prior to randomization, only those with moderate to high expectations benefited. Patients with moderate to high osteoarthritis disease severity showed stronger treatment effects. Finally, the oldest and most educated patients showed strong treatment effects, while younger and less educated did not.

[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Wolters Kluwer
Loading ...
Support Center