Self-Directed CBT-CP Outperformed Clinician-Delivered CBT in VHA Trial

Key Takeaways

• Self-directed CBT-CP was associated with lower pain-interference scores at the primary 4-month assessment, and the between-group difference persisted at 6 and 12 months.
• At 4 months, all other prespecified patient-reported outcomes favored the self-directed program, and the authors characterized those differences as small to moderate.
• Participants assigned to self-directed CBT-CP completed more expected sessions, and the abstract did not describe adverse events or other safety outcomes.

In a JAMA randomized trial involving veterans with chronic musculoskeletal pain, self-directed CBT-CP was associated with nearly a 1-point lower pain-interference score than clinician-delivered CBT-CP at 4 months. The comparison paired an 11-week self-directed program that included coping-skills work and weekly personalized audio-recorded feedback with clinician-delivered CBT-CP provided in 4 to 11 weekly sessions under usual practice conditions. The trial was conducted in a pragmatic Veterans Health Administration setting across multiple health systems, and the primary measure was a patient-reported interference scale reflecting how much pain affected function. The between-group difference on pain interference persisted at 6 and 12 months.

The randomized, open-label pragmatic superiority trial was conducted across 9 US Veterans Health Administration health care systems to assess effectiveness under usual clinical practice conditions. A total of 764 adults with chronic musculoskeletal pain were randomized, with 384 assigned to self-directed CBT-CP and 380 assigned to clinician-delivered CBT-CP. In the self-directed group, participants received 11 weeks of treatment with weekly personalized audio-recorded feedback from coaches, based on interactive voice response reports about coping-skill practice, physical activity, and pain-relevant ratings. Clinician-delivered CBT-CP was provided in 4 to 11 weekly sessions. Mean age was 52.8 years, 39.1% were women, and the primary endpoint was 4-month pain interference on the 7-item Brief Pain Inventory-Interference subscale.

At 4 months, mean BPI-I scores were 5.26 in the self-directed group and 6.23 in the clinician-delivered group, yielding a mean difference of -0.98 (95% CI, -1.31 to -0.65; P<.001). The 7-item scale ranges from 0 to 10, with higher scores indicating worse function, and 1 point was listed as the minimum clinically important difference. The difference on pain interference persisted at 6 and 12 months, and at 4 months pain intensity, pain impact, catastrophizing, self-efficacy, sleep, global impression of change, and depressive symptoms also favored self-directed CBT-CP.

Participants assigned to self-directed CBT-CP completed more expected treatment sessions than those assigned to clinician-delivered CBT-CP. Enrollment ran from December 20, 2019, through February 20, 2024, and follow-up was completed in February 2025. Outcome data were available for 583 participants, or 76%, at 4 months and for 523, or 68%, at 12 months.

The authors described the results as modest improvements in pain interference sustained to 12 months, along with small to moderate secondary differences and higher session completion.