Arthritis & Rheumatism, Volume 62,
November 2010 Abstract Supplement

Abstracts of the American College of
Rheumatology/Association of Rheumatology Health Professionals
Annual Scientific Meeting
Atlanta, Georgia November 6-11, 2010.


The 2010 ACR/EULAR Criteria for Rheumatoid Arthritis Perform Well in Prediction of Clinical RA Diagnosis at 6 Months in Very Early Arthritis Patients: Longitudinal Data from the NOR-VEAC Cohort.

Mjaavatten3,  Maria D., Van Der Heijde8,  Desiree M., Uhlig5,  Till, Haugen1,  Anne J., Nygaard6,  Halvor, Bjorneboe9,  Olav, Arvidson7,  Nils G.

Østfold Hospital, Moss, Norway
Betanien Hospital, Skien, Norway
Diakonhjemmet Hospital, Oslo, Norway
Diakonhjemmet Hospital, Oslo, Norway
Diakonhjemmet Hospital, Oslo, Norway
Hospital for Rheumatic Diseases, Lillehammer, Norway
Innlandet Hospital, Kongsvinger, Norway
Leiden University Medical Center, Meerssen, The Netherlands
Martina Hansen's Hospital, Sandvika, Norway

Background:

The new ACR/EULAR criteria for rheumatoid arthritis (RA) were designed to predict risk of RA development at an early stage. The algorithm yields a score from 0–10 points where a patient needs >= 6 points to be classified as RA. The criteria need validation in cohorts of early arthritis patients.

Methods:

Patients in the Norwegian Very Early Arthritis Clinic (NOR-VEAC) (swelling in>=1 joint with duration <=16 weeks, 18–75 years) were followed for 6 months and assessed for development of RA (clinical diagnosis). Only patients that did not contribute with data to Phase 1 of the criteria development were included to minimize overfitting of data. Sensitivity, specificity and likelihood ratios (LR) for different cut-offs were calculated and overall discriminatory ability was assessed with a receiver operating characteristics (ROC) curve.

Results:

298 patients (50 % females, median age 46.6 years, median duration of joint swelling 39 days, anti-CCP/IgM rheumatoid factor positives 18.5 %/14.2 %) were included in the analyses. After 6 months, 72 (24.2 %) patients had RA by the judgement of a rheumatologist (seropositive RA 45 (62.5 %)). Median (range) ACR/EULAR score at inclusion in RA/non-RA patients was 7(3–10)/3(0–9). Area under the ROC curve (95% CI) was 0.93 (0.90–0.96). The proposed cut-off of 6 points for RA classification showed a good combination of sensitivity, specificity and LRs (table). All seropositive RA patients had a score >=6, while 19/27 (70.4 %) seronegative clinical RA patients fulfilled the new criteria.

Conclusion:

The new ACR/EULAR RA classification criteria performed well for prediction of clinical RA diagnosis in a cohort of very early arthritis patients. This validation supports the use of these criteria in clinical practice.

Table. Sensitivity, specificity and likelihood ratios for clinical RA diagnosis for different cut-offs for the 2010 ACR/EULAR RA criteria.

Cut-off (points)Sensitivity (%)Specificity (%)LR+LR-
310036.31.570.00
498.658.82.390.02
591.776.53.900.11
688.983.25.290.13
763.993.49.680.39
847.299.478.670.53
LR+: positive likelihood ratio; LR-:negative likelihood ratio.

Figure. Predicted probability of RA development for increasing number of points in the algorithm

To cite this abstract, please use the following information:
Mjaavatten, Maria D., Van Der Heijde, Desiree M., Uhlig, Till, Haugen, Anne J., Nygaard, Halvor, Bjorneboe, Olav, et al; The 2010 ACR/EULAR Criteria for Rheumatoid Arthritis Perform Well in Prediction of Clinical RA Diagnosis at 6 Months in Very Early Arthritis Patients: Longitudinal Data from the NOR-VEAC Cohort. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1774
DOI: 10.1002/art.29539

Abstract Supplement

Meeting Menu

2010 ACR/ARHP