MODEL-pain MOdeling Diverse Experiences and Longitudinal trajectories of pain

Emergency departments (EDs) provide ~140 million encounters annually for all populations (i.e., varying demographics, injuries, comorbidities) and serve as a primary safety net in U.S. healthcare infrastructure for individuals often not encountered elsewhere. Moreover, over 80% of these patients are experiencing pain with variable causes, perceptions, and environmental influences challenging our understanding to optimize diagnosis, prognostication, and treatment. This proposal seeks to leverage the ED setting to comprehensively characterize the full range of interindividual differences in acute pain phenotypes, optimal treatment, and trajectories of pain within a biopsychosocial and health services framework. We aim to: 1) Characterize clinically informed pain phenotypes with a data-driven modeling approach, 2) Identify optimal pain treatment strategies using individualized treatment regimen (ITR) estimation method, and 3) Develop predictive models of pain trajectories and predict transition from acute to chronic pain. We will collect the necessary data for robust and comprehensive data-driven modeling in a prospective observational study of (n=2,400) ED patients seeking care for acute pain. In addition to NIH HEAL core domains, we will collect self-reported data, health records, digital physiologic markers, and biologic samples to measure: (i) AHRQ social determinants of health, (ii) character, severity, timing of pain, (iii) perceptions, resiliency, and beliefs, (iv) co-morbidities (medical and mental health), (v) physical function and sleep interruptions, (vi) digital physiologic data, (vii) measures of pain sensitivity (e.g. Von Frey, thermal thresholds), (viii) healthcare utilization, (ix) treatment access/effectiveness, and (x) genome-wide single nucleotide polymorphisms. Our comprehensive modeling approach will combine advanced and cutting-edge methods, informed by an expert panel including persons with lived experience of chronic pain, community stakeholders, and clinicians allowing the development of clinically interpretable predictive models. This innovative and high-impact investigation will provide the missing link in our understanding of pain by embracing the required breadth and complexity. Our remarkably interdisciplinary team is well-positioned for this significant challenge. We will capitalize on (i) ED care for all types of pain occurring in persons with comorbidities and (ii) our proven capacity for a longitudinal study of often challenging populations in episodic care environments without ongoing provider relationships. Our inclusion of psychosocial, cognitive, biologic, and health services measurement is an enormous and necessary advance. The quintessential contribution of this research will be models appropriate to the challenge of pain heterogeneity, health disparities, and pain itself. Without this foundation, we cannot hope to identify intervention targets and develop novel interventions sufficient to fundamentally transform current patterns in pain-related health outcomes and disparities.