Comparative validation of clinical disease activity index (CDAI) and simplified disease activity index (SDAI) in rheumatoid arthritis in India

Abstract

BackgroundCDAI and SDAI have been frequently used to categorize disease activity in patients with rheumatoid arthritis (RA), but have not been comparatively validated in Indian population. ObjectiveTo validate CDAI and SDAI in RA, taking DAS-28 as gold standard and to derive new cutoffs for CDAI and SDAI. MethodsPatients fulfilling ACR/EULAR criteria for diagnosis of RA were studied. After complete history, physical examination and biochemical tests, patients were grouped into remission, low, moderate and high activity on the basis of pre-defined cut-offs for DAS-28, CDAI, and SDAI. Spearman’s correlation and group wise inter-rater agreement tests were performed. Using DAS-28 as gold standard, the sensitivity and specificity of CDAI and SDAI cut off were determined for predicting levels of disease activity by area under receiver operator characteristics curves. (AUROC) ResultsWe studied 112 patients with RA, there was excellent correlation between DAS-28 and CDAI (r = 0.96 with 95% C.I. = 0.94−0.97), CDAI and SDAI (r=0.99, 95% C.I. 0.98−1), and DAS-28 and SDAI (r = 0.96, 95% C.I. = 0.94−0.97). There was a good inter-rater agreement between the various levels of disease activity as defined by DAS-28 and CDAI (weighed k = 0.598) and DAS-28 and SDAI (weighed k = 0.699) with excellent agreement between SDAI and CDAI categories (weighed k = 0.816). There was no statistically significant difference between AUROC of CDAI and SDAI and both performed equally well. ConclusionCDAI and SDAI are highly correlated with DAS-28 score hence are good markers of disease activity. The cut-off values for CDAI and SDAI used in western literature can be used with minor modifications in Indian scenario.