JMIR Publications

ABSTRACT

Background:

Guided exercise is a core component of rehabilitation for chronic nonspecific low back pain (CNSLBP). Although remote rehabilitation has emerged as a novel approach for delivering guided exercise in this condition, its relative effectiveness compared to face-to-face rehabilitation remains unclear.

Objective:

This study aimed to systematically assess the comparative efficacy of remote rehabilitation, AI-assisted remote rehabilitation, face-to-face rehabilitation, and conventional care in improving pain, disability, kinesiophobia, and quality of life (QoL) in CNSLBP patients.

Methods:

Following the PRISMA 2020 guidelines, a systematic search of relevant randomized controlled trials (RCTs) was conducted in databases such as PubMed, Cochrane, and other self-built databases up to September 2025. A Bayesian network meta-analysis was performed using R software (v4.4.1), and the interventions were ranked using the surface under the cumulative ranking curve (SUCRA) values. The Grading of Recommendations, Assessment, Development, and Evaluation (GRADE) framework was applied to assess the certainty of evidence for all comparative outcomes.

Results:

A total of 164 records were retrieved from databases and other sources, of which 20 records met the study criteria, involving 1,854 participants. According to the SUCRA and league matrix results, Tele-rehabilitation combined with Artificial Intelligence（TLRH-AI）demonstrated the greatest effect in reducing pain intensity at 4 and 8 weeks post-intervention (SUCRA >99%), although the certainty of evidence was low. TLRH ranked highest at 12 weeks (SUCRA 87.2%), with moderate certainty of evidence. In terms of improving disability, IPR showed the most significant improvement at 4 and 12 weeks. Furthermore, IPR was most advantageous in reducing kinesiophobia. In improving quality of life, TLRH-AI and IPR both excelled in the physical and mental components summary, respectively.

Conclusions:

TLRH-AI demonstrated a clear advantage in the short-term relief of pain and improvement in quality of life, whereas IPR showed greater effectiveness in addressing disability recovery and providing psychological support. It is recommended to adopt a phased, individualized rehabilitation strategy based on the patient's rehabilitation stage and needs, with TLRH-AI prioritized in the early stages, followed by a combination of IPR in the later stages to consolidate the therapeutic effects. Clinical Trial: Not applicable.