Ultrasound Measurement Error and Its Implications for Carpal Tunnel Syndrome Diagnosis.

Abstract

Background: Ultrasound (US) measurement of the cross-sectional area (CSA) of peripheral nerves has been validated as a method for diagnosing peripheral nerve compression; however, the accuracy of US measurements and variation among examiners have not been well established. The purpose of this study was to determine the percent error from a known measurement among multiple US examiners using a gel phantom model. Methods: A US phantom gel, with 2 embedded, premeasured cylindrical objects, was constructed, and 9 residents, fellows, and attending physicians with upper-extremity US experience were recruited to use US to measure the CSA of each object. Percent error was calculated based on the premeasured value of each object, which was measured using digital calipers, and was calculated using the formula for CSA of a cylinder (A = pi × r2) prior to insertion into the gel phantom. Results: The mean percent error for examiners' CSA measurements was 9.0% ± 4.7% (range, 1.6%-17.0%). There was a greater mean percent error for the smaller measured object compared with the larger one. There was a very minimal decline in percent error with increase in years of training experience (slope = -0.029, R2 = 0.0002). Conclusion: This study demonstrates that US examiners were able to measure CSAs of premeasured objects with approximately 10% error. When considering the use of US for the diagnosis of carpal tunnel syndrome, a cutoff value of 10 mm2 is often used, and therefore in cases where measurements are within 10% of the diagnostic threshold, caution should be exercised in interpreting the results.