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.

A Simplified Disease Activity Score (DAS): Validation in the BeSt Trial.

Koevoets6,  Rosanne, Goekoop-Ruyterman2,  Yvonne, Van Der Heijde5,  Desiree M., Han7,  K. H., Ronday1,  H. K., Huizinga6,  T. W. J., Kerstens3,  P. J. S. M

Haga Hospital, Den Hague
Haga Hospital, Den Hague, The Netherlands
JBI, Amsterdam
Leiden Univ Med Ctr, Leiden, The Netherlands
Leiden University Medical Center, Meerssen, The Netherlands
Leiden University Medical Centre, Leiden, The Netherlands
Maasstad Hospital, Rotterdam
VU Medical Centre, Amsterdam, The Netherlands


Implementation of DAS steered treatment is challenging current daily practice. Some feel that using the Ritchie Articular Index for pain in the original DAS is impractical.


To assess the validity of a simplified version of the original DAS with different versions of the pain component.


Within the first year of the BeSt trial, a 4 strategies trial in recent RA patients aiming at a DAS<=2.4, the percentage agreement and kappa statistics were calculated for 3 different versions of a simplified DAS compared to the original DAS: DAS with the Ritchie score reduced to a 0–1 score ('DAS 0–1'), DAS with a 53 joint 0–1 count for pain (DAS-TJC53), and a DAS with a 44 joint 0–1 count for pain (DAS-TJC44). At the same time for all alternative DAS versions and the DAS28 the DAS was calculated with the VAS general health (GH) and the VAS for patient's global assessment of disease activity (PGA). To display the difference for the continuous scores Bland-Altman plots were created with the mean versus the mean difference. Percentages agreement and kappa statistics were calculated for classification into remission, low disease activity (LDA) or moderate (MDA) or high (HDA) disease activity for all indices both with PGA and GH.


DAS scores with a separate joint count are in general higher since there are no single counting joint groups created. As by definition the 'DAS 0-1' was lower in few cases; in the majority of the patients the DAS scores were equal (see figure1).

Mean difference between DAS and 'DAS 0–1' was -0.03; (limits of agreement (-0.24;0.18) between DAS and 'DAS-TJC53' 0.18 (-0.33;0.66) and between DAS and 'DAS-TJC44' 0.11 (-0.42–0.64) (figure 1). Compared with the original DAS, the 'DAS 0–1', 'DAS TJC53' and the 'DAS TJC44' agreed on the classification of patients in LDA, MDA or HDA in 99% (k=0.98), 86% (k=0.81) and 86% (k=0.80), which would have resulted in the same treatment decisions in this trial. Although separate VAS scores are not highly correlated (r=0.5–0.8) all indices with PGA showed high chance corrected agreement (k=0.87–0.94) compared to the corresponding GH score.

Figure 1. Bland-Altman plots representing the mean of the original DAS and alternative DAS in relation to the difference between both.


There are no important differences between all alternative DAS versions, both with the VAS GH and PGA, and all can therefore be used reliably as alternatives to the original DAS.

To cite this abstract, please use the following information:
Koevoets, Rosanne, Goekoop-Ruyterman, Yvonne, Van Der Heijde, Desiree M., Han, K. H., Ronday, H. K., Huizinga, T. W. J., et al; A Simplified Disease Activity Score (DAS): Validation in the BeSt Trial. [abstract]. Arthritis Rheum 2010;62 Suppl 10 :295
DOI: 10.1002/art.28064

Abstract Supplement

Meeting Menu