The COPES ExTRA trial found a modest advantage in pain interference for self-directed cognitive behavioural therapy supported by daily reporting and personalized weekly coach feedback. “Self-directed” described how treatment was delivered; it did not mean participants were left with an app and no human support. The original JAMA results abstract
That distinction is useful for teams comparing digital care models, because the support around the technology is part of the intervention.
Two ways of delivering therapy
The pragmatic, open-label trial randomized 764 adults with chronic musculoskeletal pain across nine US Veterans Health Administration systems. The comparison was clinician-delivered CBT under usual practice conditions, rather than no treatment.
The self-directed programme lasted 11 weeks and used interactive voice-response reporting with weekly personalized audio feedback. The Pain Management Collaboratory project account identifies the programme and its delivery model; it is a study description, not a separate replication.
Yale’s same-trial account describes handbooks and trained-coach messages. The technology and the feedback should therefore remain together when describing what was tested.
The outcome was interference with daily life
The primary measure concerned how pain interfered with activities, on a zero-to-ten scale. At four months, the abstract reports a between-group difference of 0.98 points favouring the self-directed programme, with a 95% confidence interval from 0.65 to 1.31 points.
The abstract identifies one point as the minimum clinically important difference. The reported interval spans differences below and above that value; the result should retain the authors’ description of a modest improvement rather than becoming a claim of dramatic pain relief.
An interference measure is also different from eliminating pain. Neither the outcome’s name nor its scale supports describing participants as pain-free.
Completion and setting limit the interpretation
The abstract reports four-month assessments for 76% of participants and 12-month assessments for 68%. It reports a sustained advantage and higher treatment completion, but those observations do not make the programme interchangeable with every self-guided app.
This account uses the original results abstract and the programme’s institutional descriptions; the full results paper was not appraised. It reports a defined coached intervention in a veterans’ health system, without recommending a personal treatment, prescribing exercises or claiming the same effect for an unsupported digital product.