Work

Comparative effectiveness of digital health interventions for low back pain: systematic review and network meta-analysis of randomised controlled trials

Year: 2026

Type: preprint

Study designs: Systematic Review, Meta-Analysis

Abstract: Summary Background Digital health interventions are increasingly used for low back pain, but their comparative effectiveness across different delivery modalities, clinical outcomes, and follow-up periods remains uncertain. We aimed to compare the effects of major digital health modalities for adults with low back pain. Methods We conducted a systematic review and frequentist random-effects network meta-analysis of randomised controlled trials. PubMed, Embase, CENTRAL, CINAHL, APA PsycINFO, ClinicalTrials.gov, and WHO ICTRP were searched from inception to March 19, 2026, and updated to Aug 8, 2026. Eight treatment nodes were compared: telemedicine, web applications, mobile applications, virtual reality, exergaming, biofeedback devices, face-to-face care, and control. Primary outcomes were pain intensity and physical function; secondary outcomes were pain-related fear avoidance, health-related quality of life, self-efficacy, anxiety, and depression. The review was registered with PROSPERO (CRD420261339841). Findings We included 92 reports from 84 randomised controlled trials involving 11 803 participants. At post-intervention, virtual reality had the most favourable point estimate for pain intensity (SMD −1.43, 95% CI −1.87 to −0.98; moderate confidence). The estimate for physical function also favoured virtual reality (−0.86, −1.45 to −0.26), although confidence was very low. Virtual reality also probably reduced pain-related fear avoidance (−1.14, −1.68 to −0.61) and improved health-related quality of life (0.97, 0.57 to 1.36), both with moderate confidence. Other digital modalities showed favourable effects for selected outcomes, but confidence was generally low or very low. Networks became progressively sparse at later follow-up, and sustained effects were inconsistent. Global inconsistency was detected in the post-intervention pain-intensity network, with local inconsistency for selected exergaming comparisons. Interpretation Digital health interventions for low back pain showed different effects across outcomes and follow-up periods. Virtual reality had the most favourable post-intervention estimates for pain intensity, pain-related fear avoidance, and health-related quality of life compared with control, with moderate confidence in these estimates. Larger head-to-head trials with longer follow-up are needed to determine whether these effects are sustained and to clarify the comparative effectiveness of digital interventions. Funding National Key Research and Development Programme of China (2025YFC3607901). Research in context Evidence before this study We searched PubMed, Embase, the Cochrane Central Register of Controlled Trials, CINAHL, APA PsycINFO, ClinicalTrials.gov, and the WHO International Clinical Trials Registry Platform from database inception to March 19, 2026, and updated the search on Aug 8, 2026. We used terms related to low back pain, digital health technologies, and randomised trials, and screened the reference lists of systematic reviews. Previous reviews examined individual modalities, such as mobile applications, telehealth, or virtual reality, and used pairwise meta-analysis. Existing network meta-analyses either included broader chronic pain populations or compared a narrower range of rehabilitation approaches for low back pain. Although these reviews suggested that some digital interventions might improve pain or physical function, their relative effects across clinical outcomes and follow-up periods remained uncertain. To our knowledge, no previous synthesis had compared digital delivery modalities for low back pain across pain, physical function, and psychosocial outcomes while assessing confidence in the network estimates. Added value of this study This systematic review and network meta-analysis included 92 reports from 84 randomised controlled trials involving 11 803 participants. We compared six digital modalities—telemedicine, web applications, mobile applications, virtual reality, exergaming, and biofeedback devices—with face-to-face care and control across seven outcomes and follow-up periods extending from treatment completion to 12 months or longer. Virtual reality had the broadest favourable estimates at post-intervention, including reductions in pain intensity and pain-related fear avoidance and improvement in health-related quality of life, all supported by moderate-confidence evidence. Other modalities showed favourable effects for selected outcomes, but no stable hierarchy was observed across outcomes and follow-up periods. Confidence in many estimates was low or very low, the post-intervention pain-intensity network showed inconsistency, and longer-term networks were sparse. These findings distinguish favourable post-intervention estimates from evidence of sustained comparative superiority. Implications of all the available evidence Digital health interventions might support the management of low back pain, but the available evidence does not establish that any single delivery modality should be preferred. Although virtual reality produced the broadest favourable post-intervention estimates, evidence of sustained superiority was not established, and the observed effects cannot be separated reliably from differences in therapeutic content, treatment intensity, and clinician support. Selection of a digital intervention should therefore consider the intended treatment outcome, therapeutic components, accessibility, digital literacy, costs, and patient preferences. Adequately powered head-to-head trials with longer follow-up are needed, together with component-based studies that distinguish the contribution of the technological platform from that of the treatment it delivers.

Source: medRxiv

DOI: 10.64898/2026.09.23.26363780