Patient-Reported Outcomes of Depression and Fibromyalgia Symptoms Do Not Predict Non-Inflammatory versus Inflammatory Diagnoses at Initial Rheumatology Consultation

Abstract

Objective: The objective of this study was to assess the potential value of patient-reported outcomes (PROs) of depression, fibromyalgia symptoms, and pain in predicting non-inflammatory vs. inflammatory diagnoses in rheumatology patients. Methods: This retrospective, single-center study evaluated electronic health record (EHR) data from adults who were seen for their first rheumatology consultation and subsequently received a diagnosis of an inflammatory (e.g., rheumatoid arthritis or spondyloarthritis) or non-inflammatory (e.g., osteoarthritis or fibromyalgia) condition. The PROs evaluated included depressive symptoms (Patient Health Questionnaire-2 [PHQ-2]), fibromyalgia symptom severity (FM SS), and pain. Results: A total of 3669 patients were evaluated, including patients with (n = 984; 26.82%) and without (n = 2685; 73.18%) inflammatory rheumatologic disease, of whom 141 (3.8%) had fibromyalgia. The non-inflammatory subgroup reported higher FM SS scores, and the inflammatory subgroup had higher pain and inflammatory markers. Bivariate models based on PHQ-2 and FM SS had a very low specificity (0.3%) for predicting non-inflammatory conditions, resulting in the misclassification of >99% of inflammatory cases. Adding pain, inflammatory markers, and other relevant EHR variables increased specificity but still resulted in a high level of misclassification. Conclusions: The PROs evaluated in this study are not suitable for predicting non-inflammatory vs. inflammatory rheumatologic disease, even when combined with other EHR variables.