Telehealth and Online Cognitive Behavioral Therapy-Based Treatments for High-Impact Chronic Pain: A Randomized Clinical Trial

Abstract

Importance: Cognitive behavioral therapy (CBT) skills training interventions are recommended first-line nonpharmacologic treatment for chronic pain, yet they are not widely accessible.

Objective: To examine effectiveness of remote, scalable CBT-based chronic pain (CBT-CP) treatments (telehealth and self-completed online) for individuals with high-impact chronic pain, compared with usual care.

Design, setting, and participants: This comparative effectiveness, 3-group, phase 3 randomized clinical trial enrolled 2331 eligible patients with high-impact chronic musculoskeletal pain from 4 geographically diverse health care systems in the US from January 2021 through February 2023. Follow-up concluded in April 2024.

Interventions: Participants were randomized 1:1:1 to 1 of 2 remote, 8-session, CBT-based skills training treatments: health coach-led via telephone/videoconferencing (health coach; n = 778) or online self-completed program (painTRAINER; n = 776); or to usual care plus a resource guide (n = 777).

Main outcomes and measures: The primary outcome was attaining or exceeding the minimal clinically important difference (MCID) in pain severity score (≥30% decrease; score range, 0-10) on the 11-item Brief Pain Inventory-Short Form from baseline to 3 months; 6 and 12 months from baseline were secondary time points. Secondary outcomes at 3, 6, and 12 months included pain intensity, pain-related interference, PROMIS (Patient-Reported Outcomes Measurement Information System) social role and physical functioning; and patient global impression of change.

Results: Among 2331 eligible randomized individuals (mean age, 58.8 [SD, 14.3] years; 1712 [74%] women; 1030 [44%] rural/medically underserved), 2210 (94.8%) completed the trial. At 3 months, the adjusted percentage of participants achieving 30% or greater decrease in pain severity score was 32.0 (95% CI, 29.3-35.0) in the health coach group, 26.6 (95% CI, 23.4-30.2) in the painTRAINER group, and 20.8 (95% CI, 18.0-24.0) in the usual care group. Both intervention groups were significantly more likely to attain an MCID in pain severity compared with control (health coach vs usual care: relative risk [RR], 1.54 [95% CI, 1.30-1.82]; painTRAINER vs usual care: RR, 1.28 [95% CI, 1.06-1.55]), and the health coach program was more effective than the online self-completed painTRAINER program (health coach vs painTRAINER: RR, 1.20 [95% CI, 1.03-1.40]). Statistically significant benefits were observed for both intervention groups vs usual care at 6 and 12 months after randomization for the pain severity outcomes and for other secondary pain and functioning outcomes.

Conclusions and relevance: Remote, scalable CBT-CP treatments (delivered either via telehealth or self-completed modules online) resulted in modest improvements in pain and related functional/quality-of-life outcomes compared with usual care among individuals with high-impact chronic pain. These lower-resource CBT-CP treatments could improve availability of evidence-based nonpharmacologic pain treatments within health care systems.

Trial registration: ClinicalTrials.gov Identifier: NCT04523714.

Figure 1.. Flow of Participants Through Trial

ACPA indicates American Chronic Pain Association. ^aPain, Enjoyment of Life, General Activity (PEG) scale assessing pain intensity and interference, calculated as sum of 3 items (each 0-10; total scale range, 0-30; higher score worse). Eligibility screening was tiered. Individuals without high-impact chronic pain were not asked the PEG questions. Those without PEG score ≥12 were not asked the other questions about behavioral treatment and planned surgery. ^bRandomization stratified by sex (male vs female), clinical site, baseline pain severity score (<7 vs ≥7), and rural/medically underserved residency (yes vs no). ^cOf 249 (painTRAINER): 13 withdrew, 1 became incarcerated, and 2 died; 233 remained in study but 212 could not be reached during the treatment period, 8 no longer interested in intervention, 5 had other health/life issues, 3 did not like intervention services, 3 gave no reason, and 2 did not have time. Of 118 (health coach): 16 withdrew and 1 died; 101 remained in study but 74 could not be reached, 16 did not have time, 3 had other health/life issues, 3 gave no reason, 2 did not like intervention services, 2 no longer interested, and 1 had privacy concerns. ^dTwelve treatment providers delivering intervention; median number of patients treated by each, 69 (IQR, 52-76) (range, 24-110). ^eDid not meet the electronic health records–based eligibility criteria. ^fPatients were assessed at follow-up regardless of number of sessions completed. ^gMay have completed follow-up assessment(s) prior to the event and would consequently be counted as having completed any follow-up.

Figure 2.. Adjusted Percentage With 30% or Greater Reduction in Pain Severity (Primary Outcome)

The primary outcome occurred at 3 months. Whiskers indicate 95% CIs. Adjustment of outcomes is explained in footnote b of Table 2.