Validation of methods for converting the original Disease Activity Score (DAS) to the DAS28

Abstract

The Disease Activity Score (DAS) is integral in tailoring the clinical management of rheumatoid arthritis (RA) patients and is an important measure in clinical research. Different versions have been developed over the years to improve reliability and ease of use. Combining the original DAS and the newer DAS28 data in both contemporary and historical studies is important for both primary and secondary data analyses. As such, a methodologically robust means of converting the old DAS to the new DAS28 measure would be invaluable. Using data from The Early RA Study (ERAS), a sub-sample of patients with both DAS and DAS28 data were used to develop new regression imputation formulas using the total DAS score (univariate), and using the separate components of the DAS score (multivariate). DAS were transformed to DAS28 using an existing formula quoted in the literature, and the newly developed formulas. Bland and Altman plots were used to compare the transformed DAS with the recorded DAS28 to ascertain levels of agreement. The current transformation formula tended to overestimate the true DAS28 score, particularly at the higher end of the scale. A formula which uses all separate components of the DAS was found to estimate the scores with a higher level of precision. A new formula is proposed that can be used by other early RA cohorts to convert the original DAS to DAS28.