Background/Purpose:
Radiographs of the hands and feet are used to evaluate the disease course of rheumatoid arthritis (RA) with several scoring systems previously described. However, most of them are time-intensive, require specialized expertise and are thus not practical for use in routine clinical care. We assessed a novel method of radiographic scoring using the VARA (Veterans Affairs RA) registry. The goal of our study was to compare radiographic scores using the ‘VARA method’ with a well-established scoring method (Sharp score) and to assess the correlations and agreement between these two approaches.
Methods:
The study was done on a multi-center database of veterans with RA. A total of 198 patients were studied. Bilateral hand radiographs were read by an experienced rheumatologist trained in the Sharp scoring method and radiographic scores were calculated independently by the Sharp and VARA methods. The simplified VARA method involves assessments for the presence of erosions and joint space narrowing in 3 joint regions of the hand- including wrist, metacarpophalangeal (MCP) and proximal interphalangeal (PIP) joints. The same individual joints are also evaluated for the Sharp score. For VARA scoring, the presence of erosion or narrowing each gives a score of 1 and absence gives a score of 0. The maximum erosion and narrowing score for each hand is 3 each and minimum score is 0. Thus, the range for the total cumulative VARA score is 0 to 6 while the range in Sharp score is 0 to 314.
Correlations of the two systems were examined by calculating non-parametric Pearson correlations while formal agreement for the two approaches in capturing the presence of erosions and narrowing were assessed through calculation of a kappa coefficient. A sub-analysis was done to estimate time needed to complete the scoring for each individual set of radiographs.
Results:
Of the RA patients examined (n=198), there were moderate and highly statistically significant correlations between the VARA and Sharp scores (r = 0.66, p < 0.001 for erosion score and r = 0.68, p < 0.001 for narrowing score). The average time needed for VARA scoring is 3.3 minutes, whereas time for the Sharp score is 7-9 minutes. Additionally, there is near perfect agreement between the 2 methods with highly significant p-values in the detection of any erosion or any joint space narrowing (kappa = 0.984 for erosion and kappa = 1.00 for narrowing; p-value < 0.001).
Conclusion:
The VARA method demonstrates significant correlations with the standard radiographic Sharp scoring method. The average calculation time decreases by 2/3rd when VARA method is adopted and it appears to provide a meaningful quantitative measure of global joint involvement in RA at a single point of time. Potential disadvantages of this novel scoring method include limitations in sensitivity that may preclude its use in identifying small changes in disease progression over a period of time.
Nevertheless, the VARA method is a very practical method of radiographic scoring and may be particularly useful for observational studies and in clinical practice where quantitative assessments of joint damage at a single time point are needed.
Disclosure:
P. Vashisht,
None;
H. Sayles,
None;
T. R. Mikuls,
None;
A. R. Erickson,
None.
« Back to 2013 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-novel-feasible-and-reliable-method-of-radiographic-scoring-of-rheumatoid-arthritis/