Correlation of the Lateral Wrist Radiograph to Ulnar Variance: A Cadaveric Study.

ABSTRACT

PURPOSE: Both positive and negative ulnar variance have been implicated in a variety of wrist disorders. Surgery aims to correct the variance in these pathologic conditions. This necessitates accurate and reproducible measuring tools; however, the most accurate radiographic measurement technique remains unclear. The purposes of this study were to evaluate 3 methods for determining ulnar variance and to compare each with direct anatomic measurement in a cadaver model. METHODS: We fixed 10 fresh above-elbow cadaver specimens in neutral rotation and obtained standardized fluoroscopic posteroanterior and lateral wrist images. A dorsal approach was performed and two independent investigators directly measured ulnar variance using digital calipers with the cartilage both intact and denuded. Ulnar variance was measured radiographically using the lateral, perpendicular, and central reference point methods. The reliability of each set of measurements (within a 1-mm cutoff) was assessed by the intraclass coefficient; agreement between radiographic and direct measurements was evaluated by the Bland-Altman method. RESULTS: Each method of determining ulnar variance demonstrated near perfect agreement by the intraclass coefficient. The lateral radiograph method correlated highly with the directly measured ulnar variance with the cartilage denuded with an average measurement difference of 0.06 mm. No radiographic measurement technique demonstrated consistent agreement within 1 mm of the measured ulnar variance with the cartilage intact. CONCLUSIONS: Ulnar variance measured by the lateral wrist radiograph technique correlates highly with the directly measured osseous ulnar variance. The remaining measurement techniques did not correlate reliably to within 1 mm of the directly measured ulnar variance with 95% confidence. No method was able to account accurately for the articular cartilage thicknesses at the lunate facet of the radius or the distal ulnar head, which we found to vary in an unpredictable manner. Whereas the lateral radiograph has been shown to allow for more reliable standardization of wrist position compared with the posteroanterior view, this study also highlights the inherent limitations of using static radiographic images in evaluating ulnar variance. CLINICAL RELEVANCE The results of the current study demonstrate the utility of the lateral wrist radiograph for assessing bony ulnar variance.