Physical therapist-delivered pain coping skills training and exercise for knee osteoarthritis: Randomized controlled trial

Kim L Bennell, Yasmin Ahamed, G Jull, C Bryant, MA Hunt, Andrew B. Forbes, Jessica Eleonore Kasza, Muhammad Akram, Ben R Metcalf, Anthony Harris, Thorlene Egerton, Justin A Kenardy, M K Nicholas, FJ Keefe

Research output: Contribution to journalArticleResearchpeer-review

111 Citations (Scopus)


To investigate whether a 12-week physical therapist-delivered combined pain coping skills training (PCST) and exercise (PCST/exercise) is more efficacious and cost effective than either treatment alone for knee osteoarthritis (OA).
This was an assessor-blinded, 3-arm randomized controlled trial in 222 people (73 PCST/exercise, 75 exercise, and 74 PCST) ages ≥50 years with knee OA. All participants received 10 treatments over 12 weeks plus a home program. PCST covered pain education and training in cognitive and behavioral pain coping skills, exercise comprised strengthening exercises, and PCST/exercise integrated both. Primary outcomes were self-reported average knee pain (visual analog scale, range 0-100 mm) and physical function (Western Ontario and McMaster Universities Osteoarthritis Index, range 0-68) at week 12. Secondary outcomes included other pain measures, global change, physical performance, psychological health, physical activity, quality of life, and cost effectiveness. Analyses were by intent-to-treat methodology with multiple imputation for missing data.
A total of 201 participants (91%), 181 participants (82%), and 186 participants (84%) completed week 12, 32, and 52 measurements, respectively. At week 12, there were no significant between-group differences for reductions in pain comparing PCST/exercise versus exercise (mean difference 5.8 mm [95% confidence interval (95% CI) -1.4, 13.0]) and PCST/exercise versus PCST (6.7 mm [95% CI -0.6, 14.1]). Significantly greater improvements in function were found for PCST/exercise versus exercise (3.7 units [95% CI 0.4, 7.0]) and PCST/exercise versus PCST (7.9 units [95% CI 4.7, 11.2]). These differences persisted at weeks 32 (both) and 52 (PCST). Benefits favoring PCST/exercise were seen on several secondary outcomes. Cost effectiveness of PCST/exercise was not demonstrated.
This model of care could improve access to psychological treatment and augment patient outcomes from exercise in knee OA, although it did not appear to be cost effective.
Original languageEnglish
Pages (from-to)590-602
Number of pages13
JournalArthritis Care & Research
Issue number5
Publication statusPublished - May 2016

Cite this