A meta-model of low back pain to examine collective expert knowledge of the effects of treatments and their mechanisms

Purpose Low back pain (LBP) is a complex, multifactorial condition with diverse contributors across biopsychosocial domains. Although personalized treatment is advocated, clear guidance on tailoring interventions is lacking. To help address this gap, we synthesized expert knowledge on treatment effectiveness and underlying mechanisms using a systems-based, collaborative modeling approach. Methods Twenty-nine experts from diverse disciplines created individual fuzzy cognitive maps (FCMs) to represent their understanding of factors affecting pain, disability, and quality of life (QoL), along with treatment mechanisms. These maps were aggregated into a meta-model comprising 142 Components and 1,161 weighted Connections. Centrality was used to identify the most central domains of the meta-model. Simulations with the meta-model based on expert knowledge 1) estimated the relative effectiveness of treatments on pain, disability, and QoL and 2) identified key Mediators and mediating Domains based on their relative contribution to mediating treatment effects. Results Psychological, biomechanical, and social/contextual Domains were central to expert conceptualizations of LBP. Simulation indicated cognitive behavioral therapy was considered the most effective among all interventions. Most interventions were mediated by Components across multiple Domains, with psychological factors frequently serving as mediators. The conceptual meta-model underscored the complexity of LBP, reflecting both its multifactorial nature and the diversity of expert perspectives on factors related to treatment effectiveness. Conclusion The developed meta-model provides a novel, systems-based representation of expert knowledge about LBP, enabling quantitative exploration of treatment effects and underlying mechanisms. This conceptual framework also offers a foundation for advancing research on multi-modal, personalized care.