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.


Level of Agreement of 1987 ACR and 2010 ACR/EULAR Classification Criteria: An Analysis Based on the ESPOIR Cohort Data.

Fautrel3,  Bruno, Rincheval4,  Nathalie, Combe1,  Bernard G., Dougados2,  Maxime, ESPOIR Scientific Committee,  

Hopital Lapeyronie, Montpellier, France
Hospital Cochin, Paris, France
Pitie Salpetriere Hospital, Paris, France
UIRC Montpellier

New rheumatoid arthritis (RA) classification criteria have been developed conjointly by the ACR and EULAR in 2010 to replace the 1987 ACR ones. However, information is lacking to determine in what extent the "old" and "new" sets identify similar patients as having RA.

Objective:

To assess the agreement between 1987 ACR and 2010 ACR/EULAR criteria and the source of potential disagreement of the 2 sets, based on the ESPOIR cohort data.

Patients and Methods:

813 patients with early arthritis (EA) were included into the cohort between 2002 and 2005 and followed every 6 months during the first 2 years. Inclusion criteria were: 18 to 70 years old; more than 2 swollen joints for more than 6 weeks and less than 6 months; with suspected RA diagnosis.

The agreement between the 2 criteria sets was based on Kappa coefficient calculation.

Results:

Data on 811 patients were available with main characteristics as follows: female 77%; mean age 48 ±13 years; mean HAQ 1.0 ± 0.7; swollen joint count 7.2 ±5; tender joint count 8.4 ±7; DAS28 5.2 ±1.5; ESR 29 ±25 mm/1 hr; CRP 22 ±34 mg/L, IgM RF+ 47 %, anti-CCP+ 39 %; structural damage on X-ray 22 %. 579 (71.4%) patients met the 1987 ACR criteria and 641 (79.0 %) met the 2010 ACR/EULAR ones at baseline. 168 patients were discordant, 115 only satisfying the 2010 criteria and 53 only the 1987 ones. The concordance between the 2 sets was fair with a Kappa coefficient of 0.45 [95% CI: 0.38; 0.52]. Main sources of disagreement are indicated in the table below.

Patients meeting:Both sets (n=526)Only 2010 set (n=115)Only 1987 set (n=53)
Tender joints, mean ± sd10.4 ± 7.27.2 ± 6.23.8 ± 2.7
4 to 10 involved small joints, n(%)91 (17.3)31 (27.0)38 (71.7)
>= 10 involved joints, n (%)421 (80.0)61 (53.1)1 (1.9)
Swollen joints, mean ± sd9.1 ± 5.53.9 ± 3.25.2 ± 2.6
>=3 synovitis, n (%)503 (95.6)52 (45.2)51 (96.2)
Symmetrical arthritis, n (%)486 (92.4)42 (36.5)52 (98.1)
Morning stiffness >=60 min, n (%)497 (94.5)74 (64.4)53 (100)
RF +, n (%)309 (58.7)58 (50.4)3 (5.7)
ACPA +, n (%)274 (52.1)39 (33.9)2 (3.8)
RA diagnosis certainty at inclusion on a 0–10 VAS7.9 ± 1.85.0 ± 2.46.2 ± 2.1
RA diagnosis certainty at 2 years on a 0–10 VAS8.1 ± 2.76.3 ± 3.66.4 ± 3.5

The concordance of the 2010 set with the 1987 ACR criteria at 2 years was also tested since 1987 criteria perform better at that time point (Saraux et al. Arthritis Rheum 2001;44:2485). Kappa coefficient remained fair: 0.42 [95% CI: 0.33; 0.51]

Conclusion:

2010 criteria identify more patients than 1987 criteria as having RA. However, the 2010 set failed to identify as RA patients with symmetrical sero-negative arthritis with limited joint involvement.

To cite this abstract, please use the following information:
Fautrel, Bruno, Rincheval, Nathalie, Combe, Bernard G., Dougados, Maxime, ESPOIR Scientific Committee, ; Level of Agreement of 1987 ACR and 2010 ACR/EULAR Classification Criteria: An Analysis Based on the ESPOIR Cohort Data. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :1760
DOI: 10.1002/art.29525

Abstract Supplement

Meeting Menu

2010 ACR/ARHP