Comparing Digital to Conventional Physical Therapy for Chronic Shoulder Pain: Randomized Controlled Trial.
Featured on Physle Daily · Sunday, 9 August 2026
This randomized controlled trial compared a fully remote digital physical therapy program to conventional in-person physical therapy for chronic shoulder pain, a condition with causes like rotator cuff disorders, adhesive capsulitis, or arthritis. Of 90 participants randomized, 82 received their allocated 8-week intervention, either home exercise, education and cognitive behavioral therapy with device-based biofeedback and remote therapist monitoring, or in-person exercise, manual therapy, education and CBT. The primary outcome was change in arm/shoulder function and symptoms. Both groups improved significantly on the function questionnaire, with no statistically significant difference between them.
Secondary outcomes showed no differences in surgery intent, analgesic use, mental health, or worst pain, though the conventional group had somewhat greater reductions in average and least pain, described by the authors as unlikely to be clinically meaningful given small effect sizes. High adherence and satisfaction occurred in both groups, with no adverse events reported. The single-center design is a notable limitation for generalizing these findings.
This study looked at whether doing physical therapy for chronic shoulder pain remotely, through a digital program with exercises, education, and video-based coaching support, works as well as going to in-person sessions with a therapist. It included 82 people who completed one of the two 8-week programs, out of 90 who were randomized. Both groups saw meaningful improvement in shoulder function and symptoms, and there was no statistically significant difference between the digital and in-person approaches.
People in the in-person group had slightly better improvements in some pain measures, but the researchers noted these differences were small enough that they probably don't matter much in real life. Both groups stuck with their programs well and reported being satisfied, and no adverse events were reported. Because this was a single-center study, the results may not apply to every setting or population.
The findings suggest digital and in-person programs led to similar outcomes in this trial, though this is one study and not proof that either approach works better overall.
This is a summary of research, not medical advice. Talk to your own healthcare provider about anything affecting your care.
This single-center RCT (N=90 randomized, 82 receiving allocated treatment) compared 8 weeks of fully remote digital physical therapy, incorporating home exercise, education, CBT, and asynchronous therapist monitoring via biofeedback device, against in-person conventional therapy for chronic shoulder pain. Primary outcome was change in QuickDASH score from baseline to 8 weeks. Both arms showed significant within-group improvement in function, with no statistically significant between-group difference (-1.8, 95% CI -13.5 to 9.8; P=.75).
Secondary outcomes including surgery intent, analgesic use, and mental health showed no group differences. The conventional group had greater reductions in average and least pain, but effect sizes were small (0.15 and 0.16) and judged unlikely to be clinically meaningful by the authors.
Adherence and satisfaction were high across both arms, and no adverse events were reported. The single-center design and reliance on self-reported, electronically-delivered outcome measures limit generalizability and introduce potential for measurement bias. As a single trial, these findings support further investigation of digital delivery as a scalable alternative but do not establish superiority or equivalence definitively given the sample size and confidence interval width.
In this RCT of 82 completers, digital and in-person physical therapy produced similar improvements in shoulder function with no statistically significant between-group difference (-1.8, 95% CI -13.5 to 9.8; P=.75).
Conventional therapy showed slightly greater reductions in average and least pain, but small effect sizes (0.15-0.16) suggest these differences are unlikely to be clinically meaningful.
This is a single-center trial with a modest sample size, limiting generalizability despite high adherence, satisfaction, and no adverse events reported in both groups.
Source
Journal of medical Internet research
Pak SS, Janela D, Freitas N et al. · 2023
doi: 10.2196/49236