The use of different methods for rapid determination of the ESR induces DAS28 misclassification in clinical practice.

OBJECTIVES Monitoring of disease activity using DAS28 is more effective than routine RA care, but the ESR measurement is time consuming. Alternative rapid ESR determination methods can be used but effects on DAS28 classification are unknown. METHODS Alternative rapid ESR methods, including the Starrsed 30-minute mode and Alifax Roller Test-1TH, were compared to the Westergren method. Mean difference, limits of agreement (LoA) and intraclass correlation coefficients (ICC) were calculated. Based on these results, using a longitudinal design the percentage of DAS28 misclassification for the Alifax Roller Test-1TH was measured. RESULTS The Alifax showed acceptable ICCs, but LoA were large. ICC was 0.67 (0.56-0.76), LoA -43;34. The longitudinal study on the Alifax (n=125) showed an ICC of 0.93, a kappa of 0.61, but disease activity was misclassified in 26% of the patients. Use of the ESR from the previous visit resulted in comparable levels of misclassification. CONCLUSIONS ESR measured by automated analysers like Alifax show acceptable ICC but LoA are large compared to the Westergren ESR. The Alifax Roller Test-1TH is very rapid but DAS28 misclassification is considerable and even as large as when using the ESR of the previous visit.