Understanding internet-supported self-management for low back pain in primary care: a qualitative process evaluation of the SupportBack 2 randomised controlled trial

Objective: The SupportBack 2 randomised controlled trial (RCT) compared the clinical and cost-effectiveness of an internet intervention supporting self-management versus usual primary care in reducing low back pain (LBP)-related disability. In this study, we aimed to identify and understand key processes and potential mechanisms underlying the impact of the intervention. Design: This was a nested qualitative process evaluation of the SupportBack 2 RCT (ISRCTN: 14736486 pre-results). Setting: Primary care in the UK (England). Participants: 46 trial participants experiencing LBP without indicators of serious spinal pathologies (eg, fractures, infection) took part in telephone interviews at either 3 (n=15), 6 (n=14) or 12 months (n=17) post randomisation. Five physiotherapists who provided telephone support for the internet intervention also took part in telephone interviews. Intervention: An internet intervention ‘SupportBack’ supporting self-management of LBP primarily through physical activity and exercise delivered in addition to usual care, with and without physiotherapist telephone support. Analysis: Data were analysed thematically, applying a realist logic to develop context-mechanism-outcome configurations. Results: Four explanatory themes were developed, with five context-mechanism-outcome configurations. Where benefit was reported, SupportBack appeared to work by facilitating a central associative process where participants linked increases in physical activity or exercise with improvements in LBP, then continued to use physical activity or exercise as key regulatory strategies. Participants who reported little or no benefit from the intervention appeared to experience several barriers to this associative process, including negative expectations, prohibitive beliefs about the cause of LBP or functional limitations preventing engagement. Physiotherapists appeared to provide accountability and validation for some; however, the remote telephone support that lacked physical assessment was viewed as limiting its potential value. Conclusions: Digital interventions targeting physical activity and exercise to support LBP self-management may rely on mechanisms that are easily inhibited in complex, heterogeneous populations. Future research should focus on identifying and removing barriers that may limit the effectiveness of digital self-management support for LBP.