Arthritis & Rheumatism, Volume 60,
October 2009 Abstract Supplement
The 2009 ACR/ARHP Annual Scientific Meeting
Philadelphia October 16-21, 2009.
A Cross Sectional Study Comparing the CDA to the VDI for the Assessment of Damage in Vasculitis
Suppiah1, R., Flossmann2, O., Mukhtyar1, C., Judge1, A., Brown1, D., Baslund3, B., Jayne2, D.
Nuffield Orthopaedic Centre, Oxford, United Kingdom
Addenbrooke's Hospital, Cambridge, United Kingdom
The National University Hospital, Copenhagen, Denmark
University of Birmingham, Birmingham, United Kingdom
University of SchleswigHolstein, Bad Bramstedt, Germany
The Vasculitis Damage Index (VDI) measures damage related to vasculitis, but it may not capture all relevant damage, especially in ANCA associated vasculitis, compared to a more comprehensive assessment measure: the Combined Disease Assessment index (CDA). The aims of the study are to compare the CDA and VDI to measure damage in vasculitis and to validate the CDA for truth, discrimination and feasibility.
In a cross sectional study, patients with vasculitis were assessed for disease activity using the Birmingham Vasculitis Activity Score 3 (BVAS) and for damage using the VDI and CDA. Inter and intra-observer differences were measured in a sub group and by using paper cases. Feasibility questionnaires for the VDI and CDA were applied.
We evaluated 288 patients from 11 European centres (disease duration 0480 months). Wegener's granulomatosis (58%) and microscopic polyangiitis (10.4%) were the most common diagnoses. The CDA has already satisfied face and content validity during initial development. The degree of agreement between the VDI and CDA scores (measured using Spearman's rank correlation coefficient) was 0.89 (95%CI 0.870.91) confirming their convergent validity. There was good correlation between individual systems (0.710.94) except for skin and mucous membranes which was 0.53 (95%CI 0.450.60), mainly attributed to the inclusion of easy bruising (18.5% of patients) and cutaneous scarring (10.2%) in the CDA (items not present in the VDI). Inter observer reliability using Lin's concordance correlation coefficient was 0.90 (95% CI 0.830.96) for VDI, and 0.70 (95%CI 0.550.85) for CDA. Intra observer reliability was 0.89 (95% CI 0.771.0) for VDI and 0.86 (95%CI 0.721.0) for CDA. Assessing the use of items in the cohort, 9 items were never recorded in the VDI compared to 26 in the CDA. Neither VDI nor CDA correlated with BVAS, CRP or ANCA. Most observers reported that the CDA covered the full spectrum of damage attributable to vasculitis and that it could be used to record the natural history of vasculitis. Only 66% thought that it was useful for measuring damage in clinical trials compared to 100% for the VDI. The CDA took longer to complete (mean 17mins vs. 8mins); it was difficult to complete (67%); and not practical for routine clinical use (83%). By contrast, VDI was easy to complete (100%) and practical for daily clinical use (83%). Overall, 83% preferred to use the VDI when considering all factors.
The CDA measures damage in more detail but the increased complexities result in inferior intra and inter observer reliability and more redundant items than in VDI; CDA is not feasible for daily clinical practice and therefore does not fulfill the requirements of the OMERACT filter. We propose an improvement in VDI, by removing the 9 redundant items identified and adding easy skin bruising and scarring.
To cite this abstract, please use the following information:
Suppiah, R., Flossmann, O., Mukhtyar, C., Judge, A., Brown, D., Baslund, B., et al; A Cross Sectional Study Comparing the CDA to the VDI for the Assessment of Damage in Vasculitis [abstract]. Arthritis Rheum 2009;60 Suppl 10 :644