RT Journal Article SR Electronic T1 SAT0063 A prospective study on comparison of composite indices with ultrasound for detecting remission and prediction of flare in 2 years JF Annals of the Rheumatic Diseases JO Ann Rheum Dis FD BMJ Publishing Group Ltd and European League Against Rheumatism SP 791 OP 792 DO 10.1136/annrheumdis-2017-eular.3227 VO 76 IS Suppl 2 A1 Olmez, MO A1 Gunal, EK A1 Ureyen, SB A1 Keskin, H A1 Ozturk, AB A1 Yeter, HG A1 Cobanoglu, E A1 Aydin, SZ YR 2017 UL http://ard.bmj.com/content/76/Suppl_2/791.2.abstract AB Background Treat-to-target (T2T) approach suggests using a composite index when following patients with rheumatoid arthritis (RA) without identifying which one to use.Objectives In this prospective study we aimed to compare the accuracy of different indices for RA patients in remission taking Ultrasound Global Synovitis Score (GLOESS) as a gold standard and their predictive value for flares in 2 years.Methods RA patients who were considered to be in clinical remission according to the clinician were recruited. Disease activity was assessed using DAS28-CRP, CDAI, SDAI and RAPID-3 and 38 joints per patient were scanned by US and scored according to GLOESS. The total GLOESS scores were calculated for 38 joints and also for 28 joints by excluding the MTP joints. The number of joints with ≥2 GLOESS was calculated. Flare data was collected in 3 subsequent visits in the following 2 years, whenever available.Results Ninety-six consecutive patients (80.2% females) were recruited. Patients were more frequently categorized as being in remission using DAS28 (80%) compared to CDAI (50%), SDAI (45.2%) and RAPID 3 (37.5%). Patients that were in remission according to CDAI had lower GLOESS scores on 28 joints (p=0.05) and had less joints with ≥2 signals (p=0.04) (table). For SDAI patients in remission had significantly less number of joints with grade 3 signals (p=0.03) and tend to have lower GLOESS scores on 28 joints as well as lower number of joints with ≥2 signals (p=0.06). None of the US scores were able to differentiate different disease states according to DAS28-CRP or RAPID. Flare data was available in 76 patients, 22 of whom had flares. Patients that had flare had higher GLOESS scores on 28 joints at baseline (p=0.05) and tend to have higher number of joints with grade 3 signals (p=0.06). Although numerically higher, none of the clinical indices were able to predict flares based on remission status (remission vs non remission: CDAI: 22.5% vs 36.1, p=0.2; SDAI: 22.9 vs 36.8, p=0.2; DAS28: 25.4% vs 50%, p=0.1; RAPID3: 20% vs 34.8%, p=0.2).View this table:Table 1. The distribution of US-GLOESS according to different categories by indicesConclusions Our results show that CDAI is superior then other clinical indices to assess remission in RA. US has the superiority over clinical indices to predict flares and 28 joint GLOESS is superior to 38 joints.Disclosure of Interest None declared