Orthopedics · Medical Literature
Comparative Effectiveness of AI-Assisted Telerehabilitation, Telerehabilitation, In-Person Care, and Usual Care for Chronic Nonspecific Low Back Pain: Bayesian Network Meta-Analysis.
Abstract
Guided exercise is central to rehabilitation for chronic nonspecific low back pain. Telerehabilitation enables remote delivery of guided exercise, but its effectiveness vs other rehabilitation modalities remains uncertain. This review systematically assessed the comparative efficacy of telerehabilitation, in-person rehabilitation (IPR), and usual care (UC) for improving pain, disability, kinesiophobia, and health-related quality of life in patients with chronic nonspecific low back pain. Telerehabilitation combined with artificial intelligence (TLRH-AI) was evaluated as an exploratory intervention because available evidence was limited. Following PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 guidelines, we searched randomized controlled trials in PubMed, Cochrane Library, Web of Science, and Embase from inception to April 30, 2026. A Bayesian network meta-analysis was conducted using R (version 4.4.1). Interventions were ranked using surface under the cumulative ranking curve (SUCRA) values. Evidence certainty was assessed using the GRADE (Grading of Recommendations Assessment, Development, and Evaluation) framework. Findings were interpreted considering heterogeneity, risk of bias, inconsistency, and estimated prediction intervals. Among 2491 records, 20 randomized controlled trials involving 1854 participants were included. For pain intensity, IPR showed the greatest benefit at 4 weeks (low-certainty evidence), telerehabilitation at 8 weeks (moderate-certainty evidence), and telerehabilitation ranked highest at 12 weeks (SUCRA 87.2%; moderate-certainty evidence). For the Oswestry Disability Index-based disability, IPR ranked highest at 4 weeks (SUCRA 98.2%; low-certainty evidence) and 12 weeks (SUCRA 86.7%; low-certainty evidence), whereas telerehabilitation ranked highest at 8 weeks (SUCRA 90.4%; high-certainty evidence). For the Roland-Morris Disability Questionnaire-based disability, IPR was among the more effective interventions (SUCRA 67.3%; low-certainty evidence). For kinesiophobia, IPR ranked highest (SUCRA 99%; low-certainty evidence). For health-related quality of life, telerehabilitation significantly improved the physical component summary score (mean difference 6.05, 95% credible interval [CrI] 2.89-9.22; moderate-certainty evidence), whereas IPR showed a nonsignificant trend toward an improved mental component summary score (mean difference 2.79, 95% CrI -1.61 to 7.17; low-certainty evidence). Evidence for TLRH-AI remained limited and descriptive, suggesting possible short-term benefits with low to very low certainty. No significant small-study effects or global inconsistency were detected, although potentially important local inconsistency was observed in the 4-week Oswestry Disability Index comparison between UC and IPR. This review uniquely compared telerehabilitation, IPR, UC, and exploratory TLRH-AI within a Bayesian network meta-analysis. Unlike previous reviews focused mainly on telerehabilitation vs conventional care, it provides a comparative hierarchy across delivery models, follow-up windows, and outcomes while incorporating evidence certainty and heterogeneity. The findings support individualized rehabilitation selection. In practice, telerehabilitation may offer a scalable option for longer-term pain relief and physical function improvement, whereas IPR may remain important for supervised functional recovery and psychological support. TLRH-AI remains exploratory and should not guide clinical decision-making until adequately powered trials are available. PROSPERO CRD420251146712; https://www.crd.york.ac.uk/PROSPERO/view/CRD420251146712.
Clinical relevance
Among adults with chronic nonspecific low back pain, how does AI-assisted telerehabilitation (TLRH-AI) compare with standard telerehabilitation, in-person rehabilitation (IPR), and usual care (UC) in improving pain intensity, disability, kinesiophobia, and health-related quality of life? __SNAPSHOT__ {"studyDesign":"Systematic review with Bayesian network meta-analysis (NMA) of randomized controlled trials; PRISMA 2020-compliant","setting":"International; multi-database search (PubMed, Cochrane Library, Web of Science, Embase) from inception to April 30, 2026","sampleSize":"n=1,854 across 20 RCTs (intervention-level allocation not reported at the aggregate level in the available abstract)","population":"Adults with chronic nonspecific low back pain (symptom duration >3 months without identifiable specific pathology); baseline characteristics of included RCTs not reported in available abstract","intervention":"Telerehabilitation (remote guided exercise via digital platforms) and, as an exploratory node, AI-assisted telerehabilitation (TLRH-AI) — defined as telerehabilitation combined with artificial intelligence feedback systems","comparator":"In-person rehabilitation (IPR) and usual care (UC); four-arm network including telerehabilitation, TLRH-AI, IPR, and UC","primaryOutcome":"Pain intensity, disability (Oswestry Disability Index [ODI] and Roland-Morris Disability Questionnaire [RMDQ]), kinesiophobia, and health-related quality of life (physical and mental component summary scores); assessed at 4, 8, and 12 weeks","followUp":"Outcome windows of 4, 8, and 12 weeks; longer-term follow-up not reported","fundingSource":"Not reported in available abstract"}
Key findings
Among 20 RCTs (n=1,854), network meta-analysis ranked interventions using SUCRA values across multiple time points and outcomes. For pain intensity, IPR ranked highest at 4 weeks (low-certainty evidence) and telerehabilitation ranked highest at both 8 weeks (moderate-certainty evidence) and 12 weeks (SUCRA 87.2%; moderate-certainty evidence). For ODI-based disability, IPR ranked highest at 4 weeks (SUCRA 98.2%; low-certainty evidence) and 12 weeks (SUCRA 86.7%; low-certainty evidence), while telerehabilitation ranked highest at 8 weeks (SUCRA 90.4%; high-certainty evidence — the single highest-certainty finding in the analysis). For RMDQ-based disability, IPR was among the more effective options (SUCRA 67.3%; low-certainty evidence). For kinesiophobia, IPR ranked highest (SUCRA 99%; low-certainty evidence). For health-related quality of life, telerehabilitation significantly improved the physical component summary (PCS) score vs. UC (mean difference 6.05, 95% credible interval [CrI] 2.89–9.22; moderate-certainty evidence); IPR showed a non-significant trend toward improved mental component summary (MCS) score vs. UC (mean difference 2.79, 95% CrI −1.61 to 7.17; low-certainty evidence). Evidence for TLRH-AI was limited and descriptive, suggesting possible short-term benefits, but certainty was rated low to very low across all outcomes. No significant global inconsistency or small-study effects were detected; however, potentially important local inconsistency was observed in the 4-week ODI comparison between UC and IPR. NNT/NNH values cannot be calculated from the available abstract data.
Practice implications
This NMA established that telerehabilitation is a viable and in some instances superior alternative to in-person rehabilitation for pain intensity and physical disability at 8–12 weeks, supported by moderate- to high-certainty evidence for select endpoints. IPR demonstrated consistent ranking advantages for kinesiophobia and short-term (4-week) disability, with low-certainty evidence, suggesting it may still be preferable for patients with significant fear-avoidance behavior or when early supervised functional recovery is a priority. The TLRH-AI node did not generate sufficient evidence to support any clinical recommendation — its inclusion was explicitly exploratory, and findings carry only low to very low GRADE certainty. The analysis is limited to the abstract (full text not available for review), so specific effect estimates for pain intensity at each time point, between-treatment differences for most comparisons, and the magnitude of disability score changes cannot be independently verified. The study population characteristics of included RCTs are not described, limiting assessment of how broadly findings apply; populations with significant comorbidities, radicular symptoms, or high chronicity burden may not be adequately represented. The short maximum follow-up of 12 weeks means nothing can be concluded about maintenance of benefit or long-term outcomes. Local inconsistency at 4-week ODI warrants caution in interpreting that specific comparison. __LIMITATIONS__ Only the abstract is available for this summary; full data tables, forest plots, and subgroup analyses cannot be independently assessed.,Maximum follow-up of 12 weeks precludes conclusions about sustained benefit, relapse, or long-term functional outcomes.,The TLRH-AI network node was sparse, relying on a very limited number of trials, rendering all AI-specific estimates descriptive rather than inferential.,Predominance of indirect comparisons in a Bayesian NMA requires the transitivity assumption to hold; patient characteristics and co-interventions across RCTs may differ in ways not fully assessed.,Local inconsistency detected in the 4-week ODI comparison between UC and IPR potentially undermines reliability of that network pathway.,Evidence certainty was low for most outcomes (only the 8-week telerehabilitation ODI comparison reached high certainty), substantially limiting the confidence that ranked SUCRA estimates reflect true clinical hierarchy.,SUCRA rankings indicate probability of superiority, not magnitude of benefit; statistically favored interventions may not exceed minimal clinically important difference thresholds.,Funding source not reported; potential for industry influence on included trial design cannot be excluded.,Baseline severity, chronicity, comorbidities, and cultural context of included RCT populations not described; generalizability to diverse clinical settings is uncertain. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ Telerehabilitation demonstrates moderate-certainty evidence of benefit for pain intensity at 8–12 weeks and high-certainty evidence for ODI-based disability reduction at 8 weeks compared to usual care, supporting its use as a scalable alternative to in-person rehabilitation for chronic nonspecific low back pain. In-person rehabilitation retains an advantage for kinesiophobia and short-term disability, and should be prioritized when fear-avoidance behavior or closely supervised functional recovery is the clinical goal. AI-assisted telerehabilitation remains entirely exploratory with low to very low certainty evidence and should not yet influence clinical decision-making.