Variation in measurements of range of motion: a study in reflex sympathetic dystrophy patients

Abstract

Objective: To quantify the amount of variation attributed to different sources of variation in measurement results of upper extremity range of motion, and to estimate the smallest detectable difference (SDD) between measurements in reflex sympathetic dystrophy (RSD) patients. Design: Two observers each measured in two sessions the range of motion of several upper extremity joints of RSD patients participating in an outcome study. Setting: Department of Rehabilitation of a university hospital. Subjects: Twenty-nine upper extremity RSD patients. Main outcome measures: The range of motion of forward flexion and external rotation of the shoulder, extension, flexion and supination of the elbow, palmar, dorsiflexion and ulnar, radial deviation of the wrist of affected and nonaffected sides, using a two-armed goniometer and an inclinometer. The measurement results were analysed using an analysis of variance according to the generalizability theory. Results: The results indicate that observer and patient–observer were important sources of variation. The random error was the most important source of variation. Averaged over all ranges of motion the observer contributed 3.9% to the total variation, patient–observer interactions contributed 5.2% and the random error 20.3%. The SDD was smallest for elbow flexion, 7.1° and 9.6° and was largest for external rotation of the shoulder, 24.8° and 28.7°. The SDD was smaller for the nonaffected side as compared to the affected side for the majority of ranges of motion except elbow extension, wrist dorsiflexion, and radial and ulnar deviation. Conclusion: Clinically, our results indicate that range-of-motion measurements in RSD patients are subject to considerable variation and indicate that results of medical examinations in order to assess disability on the basis of range-of-motion measurements are subject to the same variation.