EULAR/ACR risk stratification criteria for development of rheumatoid arthritis in the risk stage of arthralgia

The field of rheumatoid arthritis (RA) is moving towards identification of and intervention in people at risk of RA, but a validated risk stratification method is lacking. This work was undertaken to develop a risk stratification method for persons presenting with arthralgia considered to be at risk of RA. A joint European Alliance of Associations for Rheumatology (EULAR)/American College of Rheumatology (ACR) expert committee was established. Risk factor and outcome data from 10 arthralgia cohorts (including clinically suspect arthralgia and autoantibody-positive arthralgia) were studied. The work focused on differentiating the risk of progression to clinically apparent inflammatory arthritis (IA) within 1 year, using clinical and serologic variables, without and with subclinical joint inflammation detected by ultrasound (US) or magnetic resonance imaging (MRI). Developing RA according to the 2010 EULAR/ACR criteria within 1 year was a secondary outcome. A set of validated risk stratification criteria was developed. Using data from 2293 symptomatic at-risk individuals, a stratification method was derived consisting of 6 clinical and serologic variables (morning stiffness, patient-reported joint swelling, difficulty making a fist, C-reactive protein, rheumatoid factor, and anti-citrullinated peptide antibody) yielding an area under the curve (AUC) of 0.80 (95% CI, 0.77-0.83) for IA development. The inclusion of US variables did not increase the discriminative ability. When MRI-detected subclinical inflammation variables were included, the AUC was 0.87 (95% CI, 0.82-0.90). In the presence of clinical, serologic, and MRI variables, a sensitivity and specificity of >75% was achieved. For RA development, the AUC of the criteria with MRI was 0.93 (95% CI, 0.90-0.97). EULAR/ACR risk stratification criteria have been developed for people with arthralgia in secondary care who are considered at risk for RA. They can be applied in the absence or presence of imaging data and have been developed to define homogeneous risk groups for future prevention trials.