Objective: To compare in daily clinical practice the reliability of matrices that forecast rapid radiologic progression (RRP) at year one, at year two, and over 2 years in patients with early rheumatoid arthritis (RA).
Methods: Overall, 74 early RA patients with X-rays of hands and feet at baseline, year one, and year two were included. Initial DMARD combination therapy with steroids (ICTS) or DMARD monotherapy (IMT) was initiated according to patients' RA severity, based on rheumatologist opinion. The images were scored via the modified Sharp/van der Heijde (SvH) method. A total Sharp score progression of equal or higher than five per year was considered RRP. Six matrices were tested: ASPIRE CRP/ESR matrices, the BEST matrix, two SWEFOT matrices, and the ESPOIR matrix. Patients were placed in each of them yielding a RRP probability. The performance was tested by Area Under the Curve analysis reflecting the predictive value.
Results: Four patients developed RRP in year one, five in year two, and four over 2 years. With regard to face validity, the predicted probability did not correspond to the risk in reality: the one ICTS patient who developed RRP over 2 years was always found in the lowest RRP categories of all matrices. The ASPIRE CRP matrix yielded at least a moderate predicting value for the three time points. The other matrices showed moderate to no predicting value.
Conclusion: The performance of all matrices was disappointing and it is impossible to fully rely on the existing matrices in daily clinical practice.
Keywords: Daily clinical practice; Early rheumatoid arthritis; Prediction models; Rapid radiologic progression.
© 2013 Published by Elsevier Inc.