Predictors of knowledge and attitudes toward palliative care among patients with cancer: a cross-sectional study.

Chronic pain remains a complex and poorly understood condition, presenting significant challenges for both diagnosis and treatment. A major obstacle is identifying the underlying mechanisms that drive chronic pain, which often vary significantly between individuals. Previous studies often but not exclusively sought to identify subgroups of common underlying pathomechanisms inferred from sensory profiles. Similarly, our group previously aimed at clustering individuals with different chronic pain conditions, i.e., complex regional pain syndrome (CRPS) and low back pain (LBP), based on their quantitative sensory testing (QST) profile assessed in the most painful area. This approach strived to identify sensory patterns that might suggest distinct pathomechanisms as well as disclosing the origin of such pain-generating mechanisms, e.g., peripheral, spinal, or supra-spinal, as proposed previously regardless of the clinical diagnosis. While two distinct subgroups emerged with potentially different underlying pathomechanisms, many individuals with LBP exhibited sensory profiles similar to those of pain-free controls, raising questions about whether QST alone is suitable to detect clinically relevant clusters. Adding psychological and social factors as well as complementary experimental paradigms might improve clustering performance. The biopsychosocial model of pain has gained traction in the understanding and treatment of chronic pain, recognizing that psychological and social factors, in addition to biological ones, play crucial roles in shaping pain perception. Recent work using data from this study explored how psychological and physiological factors contribute to acute pain perception. However, their focus was on acute, experimentally-induced, rather than chronic, spontaneous pain. Another study developed a model that predicted the incidence, severity, and spread of chronic pain based on biopsychosocial factors in individuals with various painful and non-painful medical conditions. Factors like fatigue, stress, and depression predicted the spread of chronic pain. However, the model did not include experimental paradigms indicative of nociceptive processing (i.e., biological factors). Comprehensive pain paradigms can illuminate the balance between anti- and pronociceptive processes relevant for chronic pain. Various approaches, including the investigation of the spinothalamic integrity, temporal summation of pain (TSP), conditioned pain modulation (CPM), and experimental pain habituation offer valuable insights into peripheral, spinal, and supra-spinal sensitization, as well as top-down modulation processes. We aimed to classify individuals with differing chronic pain intensity using three classification models: one that focused solely on sensory profiles (QST model), one integrating a broad set of biological, psychological, and social features (biopsychosocial model), and one including the same data as the latter model but excluding the QST-data (noQST model). The biopsychosocial model incorporated experimental pain paradigms such as TSP, CPM and experimental pain habituation, alongside psychological variables (depression, anxiety, pain catastrophizing) and broader health-related factors such as quality of life, fatigue, and general health. The study population consisted of a diverse chronic pain cohort, including individuals with CRPS, LBP, and neuropathic pain after spinal cord injury (SCI), and pain-free healthy controls (HC). This heterogeneous sample was intended to enhance the study’s ability to identify the most common and influential factors contributing to the perceived severity of chronic pain independent of its etiology. We hypothesized that the biopsychosocial model would significantly outperform the QST model and noQST model in classifying individuals with differing perceived chronic pain intensities. Additionally, and most importantly, we examined the feature importance in all models which potentially infer the underlying mechanism of why some patients experience severe pain regardless of their clinical diagnosis. For the biopsychosocial model we hypothesized that, in addition to factors related to pain perception and modulation, psychological aspects such as pain catastrophizing would play a pivotal role.