Quantitative measurement of erosion growth and joint space loss in rheumatoid arthritis hand radiographs.

OBJECTIVE To evaluate the performance of simple, inexpensive quantitative techniques for measuring erosion growth and joint space loss in serial hand radiographs of patients with rheumatoid arthritis (RA). METHODS Erosions were measured using a plastic overlay template of sample erosion sizes. Joint spaces were measured with a hand held measuring micrometer. In the first phase of testing, a spectrum of individual erosions and joint spaces was measured to determine intra and interobserver correlations and variability. In the 2nd phase, the tools were used to measure serial changes in RA hand radiographs. Observer correlations and the ability to discriminate serial changes were determined and compared to the scoring method of Sharp. RESULTS Measurements of individual erosion areas and joint spaces were highly reproducible. Intra and interobserver correlations were significant (p < 0.05) for serial erosion growth and joint space loss measurements, as well as changes in Sharp scores. Quantitative measurements correlated highly with the corresponding Sharp score changes. Of all measurements or scores recorded, quantitative joint space loss measurements were statistically superior in discriminating serial change in RA hand radiographs. Quantitatively measured joint space loss correlated well with both erosion growth measurements and serial change in total Sharp scores. CONCLUSION Quantitative measurement of erosion growth and joint space loss is possible with simple inexpensive techniques. Further study is needed to confirm our data, which suggest that quantitative measurement of joint space narrowing may be the most useful discriminator of serial changes in RA hand radiographs.