Supporting self-management of low back pain with an internet intervention with and without telephone support in primary care (SupportBack 2): a randomised controlled trial of clinical and cost-effectiveness

Background: Low back pain is prevalent and a leading cause of disability. We aimed to determine the clinical and cost-effectiveness of an accessible, scalable internet intervention for supporting behavioural self-management (SupportBack).

Methods: Participants in UK primary care with low back pain without serious spinal pathology were randomly assigned 1:1:1 using computer algorithms stratified by disability level and telephone-support centre to usual care, usual care and SupportBack, or usual care and SupportBack with physiotherapist telephone-support (three brief calls). The primary outcome was low back pain-related disability (Roland Morris Disability Questionnaire [RMDQ] score) at 6 weeks, 3 months, 6 months, and 12 months using a repeated measures model, analysed by intention to treat using 97·5% CIs. A parallel economic evaluation from a health services perspective was used to estimate cost-effectiveness. People with lived experience of low back pain were involved in this trial from the outset. This completed trial was registered with ISRCTN, ISRCTN14736486.

Findings: Between Nov 29, 2018, and Jan 12, 2021, 825 participants were randomly assigned (274 to usual care, 275 to SupportBack only, 276 to SupportBack with telephone-support). Participants had a mean age of 54 (SD 15), 479 (58%) of 821 were women and 342 (42%) were men, and 591 (92%) of 641 were White. Follow-up rates were 687 (83%) at 6 weeks, 598 (73%) at 3 months, 589 (72%) at 6 months, and 652 (79%) at 12 months. For the primary analysis, 736 participants were analysed (249 usual care, 245 SupportBack, and 242 SupportBack with telephone support). At a significance level of 0·025, there was no difference in RMDQ over 12 months with SupportBack versus usual care (adjusted mean difference -0·5 [97·5% CI -1·2 to 0·2]; p=0·085) or SupportBack with telephone-support versus usual care (-0·6 [-1·2 to 0·1]; p=0·048). There were no treatment-related serious adverse events. The economic evaluation showed that the SupportBack group dominated usual care, being both more effective and less costly. Both interventions were likely to be cost-effective at a threshold of £20 000 per quality adjusted life year compared with usual care.

Interpretation: The SupportBack internet interventions did not significantly reduce low back pain-related disability over 12 months compared with usual care. They were likely to be cost-effective and safe. Clinical effectiveness, cost-effectiveness, and safety should be considered together when determining whether to apply these interventions in clinical practice.

Funding: National Institute for Health and Care Research Health Technology Assessment (16/111/78).