between groups. Comparison of patient characteristics between concordant and discordant groups showed no significant differences in age, sex, or provider type (Table 1). The discordant group reported higher pain by VAS than the concordant group (median 60.0 vs. 23.5; P < .001). Negative results for RF (P = .01) and anti-CCP antibodies (P = .045), lack of radiographic joint erosions (P = .002), and presence of ≥ 2 tender joints (P = .02) were significantly associated with patient-provider discordance. No significant differences were observed between the groups in swollen joint counts or inflammatory markers. Among the comorbidities, fibromyalgia (P < .001) and depression (P = .02) showed significant association with discordance. However, excluding patients with fibromyalgia did not affect the association of pain VAS with positive discordance (mean pain VAS of 56 in the discordant group vs. 31 in the concordant group; P < .001). Current use of opioids (32% vs. 19%, P = .009), fibromyalgia medications (17% vs. 7%, P = .005), antidepressants or anxiolytics (34% vs. 19%, P = .003), and sleep aids (17% vs. 8%, P = .007) was significantly higher in the discordant group. Both groups had similar treatment with conventional and biologic DMARDs, without significant differences in DMARD modification at the index visit. Of 140 patients approached for the prospective substudy, 70 patients agreed to participate, including 50 patients from the discordant group and 20 from the concordant group. Four patients in the discordant group were among the 11 in the study population with lower discordance who were excluded, leaving 46 patients in the discordant group for analysis. In the discordant group, the disease of two patients was in remission, and 26 had LDA, 13 had MDA, and 5 had HDA. The discordant group reported higher median scores for pain VAS (49.0 vs. 25.0, P = .002), SF-MPQ-2 continuous pain (3.3 vs. 2.2, P = .03), neuropathic pain (2.0 vs. 1.2, P = .045), Fibromyalgia Survey WPI (7.0 vs. 4.5, P = .008), and polysymptomatic distress (11.5 vs. 7.5, P = .007) than the concordant group (Table 2). Of note, fatigue measures were not significantly different between the groups, either by Bristol Rheumatoid Arthritis Fatigue or PROMIS Fatigue Survey. Five patients (10.8%) in the discordant group had PHQ-9 scores ≥ 10, indicating moderate to severe clinical depression. An investigator (JMD) screened these five patients, and one patient consented to psychiatric evaluation on a nonurgent basis; the other patients chose to follow-up with their primary rheumatologist or primary care provider. None expressed suicidal ideation or required urgent evaluation. Comparison of patient-reported outcomes between the concordant and discordant groups in the prospective substudy To address the possibility of confounding of patient-reported outcome analyses by disease activity states, the next analysis separately compared individual patient-reported outcomes between discordant and concordant groups among patients with LDA versus MHDA (Table 3). Irrespective of disease activity category, the discordant group had higher median scores for pain VAS (for LDA, 37.0 vs. 18.5, P = .006; for MHDA, 66.0 vs. 36.0, P = .02) and higher median HAQ-II disability scores (for LDA, 0.9 vs. 0.2, P = .001; for MHDA, 1.2 vs. 0.9, P = .10). In the LDA category, the discordant group had impaired ability to function in activities and social roles according to the PROMIS ability to participate instrument (median, 3.5 vs. 4.3; P = .02) and