Physical therapy is moving into living rooms, guided by phone cameras that track joint angles and software that counts repetitions and flags poor form. The convenience is real, and so is the trade. What a patient gives up is the part of rehabilitation that requires a clinician's hands and eyes in the same room.
A Bayesian network meta-analysis published this summer in the Journal of Medical Internet Research sharpened that picture for one of the most common reasons Americans see a physical therapist. Pooling 20 randomized trials and 1,854 participants with chronic nonspecific low back pain, researchers ranked four delivery models against each other: telerehabilitation, in-person rehabilitation, usual care, and, as an exploratory category, AI-assisted telerehabilitation.