Improving the arthritis component of the SLEDAI-2K

To propose a new definition for SLEDAI arthritis informed by imaging.

We performed a planned secondary analysis of observational data from a multicentre study evaluating SLE patients with inflammatory joint pain (swelling not required) using various clinical instruments, laboratory tests and ultrasound. For SLEDAI arthritis, assessors (blinded to ultrasound) were asked which of the glossary terms for arthritis in any version of the SLEDAI drove their decision to score for arthritis. These definitions were tested against ultrasound and other clinical variables. ROC analysis was used to test optimal joint count thresholds.

A total of 78/133 patients had arthritis on SLEDAI-2K. In 21/78, clinician-observed swelling was not a reason for scoring (tenderness: 16/21; reported swelling between visits: 4/21; both of these: 1/21). No patient was scored for warmth or erythema alone. In total, 57 (73.1%) patients were scored for SLEDAI arthritis due to observed swelling, 90% had abnormal ultrasound (PPV = 90%, 95%CI: 79, 94). Of 21 patients with SLEDAI arthritis without observed swelling, 48% had abnormal ultrasound (PPV: 48%, 95%CI: 31, 67). Patients with SLEDAI arthritis with swelling had higher ESR, physician MSK-VAS, patient early morning stiffness (EMS)-VAS and IgG compared with other patients. Optimal sensitivity and specificity for ultrasound synovitis was 1 swollen joint using Youden's criteria.

Our data suggest that the definition of arthritis in SLEDAI be modified to: 'Inflammatory musculoskeletal pain (symmetrical small joint distribution), with one or more clinically swollen joints witnessed on examination, that is not explained by another arthropathy'.