ABSTRACT
STUDY DESIGN: Intraexaminer and interexaminer/procedure reliability and error analysis using a repeated-measures design. OBJECTIVE: To quantify sources of discrepancies in cervical range of motion values between two procedures that use the same potentiometric technology. SUMMARY OF BACKGROUND DATA: Studies using an early version of an electrogoniometer system, which was connected between a helmet worn by the study participant and a chair on which they sat, reported unusually high values for active and passive cervical range of motion, although measurements were reliable. To understand the sources of the discrepancies between that study and later studies (using upgraded software), the current study was designed to quantify possible sources of error contributed by various components of the procedures: helmet, thoracic reference, chair, and software. METHODS: A total of 22 asymptomatic study participants were evaluated in two separate sessions, 1-3 days apart. Components of two procedures were changed systematically in a series of repeated measurements to provide concurrent reliability and to assess sources of error between the two procedures. RESULTS: The reliabilities of both procedures were generally high with no systematic trends, except for lower values for flexion-extension studies with Procedure 2. Procedure 2 also provided systematically greater range of motion values (2-8 degrees ) than Procedure 1, except for flexion (half-cycle). The source of the greatest discrepancy between the two procedures was the software, when comparing the original with the updated versions. With regard to the instrumentation, the greatest source of variability was in the thoracic reference post; next was the helmet, and least significant was the type of seat used. A comparison of overall procedure discrepancies and summation of individual elements of the procedures accounted for virtually all of the observed error. CONCLUSION: The potentiometer-based electrogoniometer is a reliable instrument for determining cervical range of motion. Measurements appear to be more valid when the thoracic reference point is physically attached to the study participant's body. The original software provided with the system appears to have contributed to systematic overestimation of ranges of motion, but current units provide measurements that are both reliable and valid.
