Self-Directed vs Clinician-Delivered Cognitive Behavioral Therapy for Chronic Pain
Alicia A. Heapy, Mary A. Driscoll, Sara Edmond, Kathryn M. LaChappelle, Eugenia Buta, Allison Halat, Sarah L. Krein, Steve Martino, Lynn DeBar, Jennifer L. Murphy, Kristin Mattocks, Brian C. Coleman, Christopher Ruser, John D. Piette, R. Ross MacLean, Brett Ankawi, Dana Cervone, Todd H. Wagner, Diana M. HigginsImportance
Cognitive behavioral therapy for chronic pain (CBT-CP) is a first-line nonpharmacological treatment, but uptake remains low due to multiple access barriers.
Objective
To assess the effectiveness of self-directed CBT-CP with asynchronous, personalized feedback relative to clinician-delivered CBT-CP under usual clinical practice conditions.
Design, Setting, and Participants
This randomized, open-label pragmatic superiority trial enrolled 764 patients with chronic musculoskeletal pain from 9 US Veterans Health Administration (VHA) health care systems from December 20, 2019, to February 20, 2024; follow-up was completed in February 2025.
Interventions
Participants were allocated 1:1 to self-directed CBT-CP (n = 384) or clinician-delivered CBT-CP (n = 380). The self-directed group included 11 weeks of treatment with weekly personalized audio-recorded feedback provided by coaches. Feedback was based on participants’ daily reports of pain coping skill practice, physical activity, and pain-relevant ratings collected by the interactive voice response system. Clinician-delivered CBT-CP included 4 to 11 weekly sessions provided under usual practice conditions.
Main Outcomes and Measures
The primary outcome was patient-reported pain interference measured by the 7-item Brief Pain Inventory-Interference (BPI-I) subscale at 4 months (score range, 0-10; higher scores indicate worse function; minimum clinically important difference, 1). Secondary outcomes included BPI-I at 6 and 12 months; pain intensity, pain impact, catastrophizing, self-efficacy, sleep, global impression of change, depressive symptoms (all at 4 months); and treatment dose.
Results
Of the 764 randomized participants (mean age, 52.8 [SD, 12.3] years; 299 women [39.1%]; 289 Black [39.1%]; 399 White [54.0%]; and 107 Hispanic ethnicity [14.0%] and 186 participants [24%] who resided in a rural area), 583 participants (76%) completed the 4-month assessment and 523 (68%) the 12-month assessment. At 4 months, the self-directed CBT-CP was superior to clinician-delivered CBT-CP for reduction in pain interference (mean score, 5.26 vs 6.23, respectively; mean difference, −0.98; 95% CI, –1.31 to −0.65, P < .001) and maintained superiority at 6 and 12 months. Self-directed CBT-CP was superior to clinician-delivered CBT-CP for all other outcomes at 4 months ( P ≤ .001). Participants in the self-directed CBT-CP group completed more expected treatment sessions than patients in the clinician-delivered CBT-CP group.
Conclusions and Relevance
Self-directed CBT-CP was associated with modest improvements in pain interference at 4 months that were sustained to 12 months, small to moderate improvements in all secondary outcomes at 4 months, and higher session-completion rates, relative to clinician-delivered CBT-CP. Scalable, convenient self-directed CBT may be an effective alternative to clinician-delivered CBT that could increase uptake of CBT-CP.
Trial Registration
ClinicalTrials.gov Identifier: