Study Overview
Objective. To examine whether discontinuation of nonsteroidal anti-inflammatory drug (NSAID) therapy and initiation of telephone-based cognitive behavioral therapy (CBT) is not worse than continuation of NSAIDs in the management of arthritis pain.
Design. Randomized controlled trial with noninferiority design.
Setting and participants. This study was a multicenter trial conducted across 4 Veterans Affairs health care systems in Boston, Providence, Connecticut, and North Florida/South Georgia that started September 2013 and ended September 2018. Eligibility criteria included being age 20 years or older, radiographic evidence of knee osteoarthritis, and use of an NSAID for knee pain on most days of the month for at least the past 3 months. Exclusion criteria included significant hearing impairments that may impede the conduct of the trial, current opioid prescriptions excluding tramadol, contraindications to NSAID use, recent or scheduled intra-articular injections or surgery, comorbid conditions other than knee pain that limited walking, and bilateral knee replacements or pain only in the replaced knee. Concurrent use of tramadol and other non-NSAID analgesics was permitted.
A total of 490 participants took part in the 2-week run-in period where their NSAID regimen was discontinued and they were started on a standardized dose of the NSAID meloxicam 15 mg daily. During the run-in period, 126 participants were excluded for several reasons, including worsening pain and patient withdrawal, yielding 364 participants who were randomized to continue meloxicam treatment or placebo for 4 weeks with blinding.
Intervention. Subsequent to the 4-week phase 1 placebo controlled trial, participants in the placebo group were given CBT via telephone (unblinded) for 10 weeks, and the meloxicam group continued treatment with meloxicam for phase 2. The CBT group received 10 modules over 10 weeks in 30- to 45-minute telephone contacts with a psychologist using a treatment manual modified for knee osteoarthritis. These modules consisted of 1 introductory module, 8 pain coping skills modules (eg, deep breathing and visual imagery, progressive muscle relaxation, physical activity and bodily mechanics, identifying unhealthy thoughts, balancing unhealthy thoughts, managing stress, time-based pacing, and sleep hygiene), and a final module emphasizing skill consolidation and relapse prevention. Outcomes were assessed at the end of the phase 1 and phase 2 periods.
Main outcome measures. Main study outcome measures included pain as measured with the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) at 4 weeks. Secondary outcomes included the WOMAC pain score, disability score, and global impression of change after treatment at 14 weeks. The WOMAC pain scale ranges from 0 to 20, and consists of 5 questions regarding severity of pain during walking, stair use, lying in bed at night, sitting, and standing, with 0 indicating no pain; 1, mild pain; 2, moderate pain; 3, severe pain; and 4, very severe pain for each item. The WOMAC disability scale measures self-reported difficulty in performing tasks that reflect lower-extremity physical function, including climbing stairs, rising from a chair, walking, and other activities of daily living. The global impression of change after treatment was measured on a 5-point scale (where 1 indicates much better and 5 indicates much worse). The minimum clinically important difference of the WOMAC pain scale is 2, based on prior literature. With the noninferiority design, the margin was set as a score of 1.
Main results. The placebo group consisted of 180 participants, with an average age of 58.2 years (SD, 11.8 years); 89% of them were male. The meloxicam group consisted of 184 participants, with an average age of 58.6 years (SD, 10 years); 84% of them were male. The average body mass index was 33.9 and 33.4 in each group, respectively. For the primary outcome, the placebo group had a worse pain score than the meloxicam group at 4 weeks (difference of 1.4; 95% confidence interval, 0.8- 2.0). At 14 weeks, the placebo group (with CBT) had a worse pain score than the meloxicam group (difference of 0.8; 95% CI, 0.2-1.4). There was no statistically significant difference in the disability score or global impression of change after treatment score between the 2 groups. The observed difference in pain score did not, however, exceed the minimum clinically important difference.