Photo Credit: Cinefootage Visuals
The following is a summary of “Identifying Patient Subgroups in the Heterogeneous Chronic Pain Population Using Cluster Analysis.,” published in the January 2025 issue of Pain by Rijsdijk et al.
Chronic pain is a complex condition with biopsychosocial factors, and treatment failure may stem from a limited understanding of diverse patient subgroups.
Researchers conducted a retrospective study to identify subgroups using psychological variables for more tailored interventions.
They extracted patient-reported data from 2 Dutch tertiary multidisciplinary outpatient pain clinics (2018-2023) for unsupervised hierarchical clustering. Clusters were based on anxiety, depression, pain catastrophizing, and kinesiophobia. Sociodemographics, pain characteristics, diagnosis, lifestyle, health-related QoL, and treatment efficacy were compared among clusters. A prediction model was developed using a minimal set of questions to reliably assess cluster allocation.
The results showed 3 clusters emerged from 5,466 patients with chronic pain. Cluster 1 (n=750) had a high psychological burden, low health-related QoL, lower education and employment rates, and higher smoking rates. Cluster 2 (n=1,795) exhibited low psychological burden, intermediate health-related QoL, higher education and employment levels, and more alcohol use. Cluster 3 (n=2,909) displayed intermediate characteristics. Pain reduction was lowest in cluster 1 (28.6% with capsaicin patch, 18.2% with multidisciplinary treatment), while clusters 2 and 3 experienced reductions greater than 50% with both treatments. A model using 15 psychometric questions reliably predicted cluster allocation.
Investigators concluded the identified distinct patients with chronic pain clusters based on psychological factors, with 1 cluster demonstrating significantly poorer treatment outcomes, and developed a predictive model integrated into a web-based tool to potentially guide clinicians towards more effective, subgroup-specific treatment approaches.
Source: jpain.org/article/S1526-5900(25)00018-5/fulltext
