Background In some countries, including Turkey, reimbursement criteria require a DAS28 score of >5.1 for initiating anti-TNF therapy in patients with rheumatoid arthritis. DAS28 score can be calculated based on C-reactive protein (CRP) levels (DAS28-CRP) or erythrocyte sedimentation rate (DAS28-ESR). The results of a recent mini-survey suggest that Turkish rheumatologists use both methods interchangeably, but with a clear preference of DAS28-CRP. However, levels of agreement between the two definitions of DAS28 at individual level are not known in Turkish RA patients.
Objectives To assess the agreement between the ESR- and CRP-based DAS28 scores in Turkish patients with RA.
Methods A total of 61 RA patients with full data for both DAS28-ESR and DAS28-CRP at baseline were identified in our recently established biological therapy database (TURKBIO). All patients were receiving anti-TNF treatment. We compared their DAS28-CRP and DAS-ESR scores obtained from baseline and following visits. Mean DAS28-ESR and DAS28-CRP values were compared by Spearman correlation and linear regression analysis. Percent agreement, kappa statistic, and weighted kappa statistic weredetermined for the two definitions. Bland–Altman plots were generated for assessment of the variation between DAS28-ESR and DAS28-CRP.
Results Of the 61 patients 54 were female (86%) with a mean age 52.2 (±13.4). Disease duration was 9.83 years (±5.8) RF positivity was 65.5 % and anti-CCP positivity was 69.1%. All patients were on anti-TNF treatment with 82% of them also receiving methotrexate and/or leflunomide. DAS28-CRP and DAS28-ESR scores were available for 209 visits. Mean DAS28-CRP and DAS-ESR values were 3.4 ±1.5 and 4.1 ±1.5, respectively, showing a strong correlation with each other [(Spearman correlation coefficient: 0.940, p<0.001) and linear regression analysis (R²=0.91, p<0.001)]. The agreement rate, kappa and weighted kappa values were69%, 0.53 and 0.77, respectively. The number of visits with high disease activity based on DAS28-ESR was greater than that based on DAS28-CRP (p=0.08). 27 (41.5%) visits with moderate disease activity according to DAS28-CRP were classified as high disease activity according to DAS28-ESR (Table). Similar results were obtained when the same analysis was repeated using the baseline data of individual patients. Bland-Altman plot analysis showed a mean difference (95% CI) of 0.69 (0.63 - 0.75) between the two methods (DAS28-ESR – DAS28-CRP) with slightly larger difference at low mean values. The 95% limits of agreement were between -0.20 and 1.58.
Conclusions These results suggest that DAS28-CRP underestimates disease activity in Turkish RA patients, as compared to DAS28-ESR. A significant proportion of Turkish RA patients, who would not qualify for reimbursement of anti-TNF therapy according to DAS28-CRP, would be classified as having high disease activity according to DAS28-ESR. Therefore the two definitions of DAS28 should not be used interchangeably in daily practice.
Disclosure of Interest None Declared