Objective: To compare fibromyalgia (FM) characteristics among patients identified in a community-based chronic pain cohort based on traditional International Classification of Diagnoses 9th revision (ICD-9) diagnostic coding, with that of patients identified using a novel predictive model.
Methods: This retrospective study used data collected from July 1999 to February 17, 2015, in multiple chronic pain clinics in the United States. Patients were assigned to the FM case group based on specific inclusion criteria using ICD-9 codes or, separately, from results of a novel FM predictive model that was developed using random forest and logistic regression techniques. Propensity scoring (1:1) matched FM patients (cases) to nonmalignant chronic pain patients without FM (controls). Patient-reported measures (eg, pain, fatigue, quality of sleep) and clinical characteristics (ie, comorbidities, procedures, and regions of pain) were outcomes for analysis.
Results: Nine ICD-9 clinical modification diagnoses had odds ratios with large effect sizes (Cohen's d > 0.8), demonstrating the magnitude of the difference between the FM and matched non-FM cohorts: chronic pain syndrome, latex allergy, muscle spasm, fasciitis, cervicalgia, thoracic pain, shoulder pain, arthritis, and cervical disorders (all P < 0.0001). Six diagnoses were found to have a moderate effect size (Cohen's 0.5 < d > 0.8): cystitis, cervical degeneration, anxiety, joint pain, lumbago, and cervical radiculitis.
Conclusions: The identification of multiple comorbidities, diagnoses, and musculoskeletal procedures that were significantly associated with FM may facilitate differentiation of FM patients from other conditions characterized by chronic widespread pain. Predictive modeling may enhance identification of FM patients who may otherwise go undiagnosed.
Keywords: arthritis; cervical pain; depression; fibromyalgia; myelitis; myofacial pain syndromes; pain; shoulder pain.
© 2017 World Institute of Pain.