Standardized initial head position in cervical range-of-motion assessment: reliability and error analysis

To assess the clinical reliability and precision of the OSI CA-6000 Spinal Motion Analyzer for measurement of range of motion in cervical spines of pain-free subjects by using a novel procedure designed to minimize variability and quantitatively evaluate sources of errors.

Twenty asymptomatic volunteer subjects were evaluated twice by each of two trained examiners in one session. Subject position was carefully standardized. Rotation, lateral bending, and flexion-extension were evaluated in repeated movements (cycles) from extreme to extreme.

Descriptive statistics and reliability coefficients (interclass correlation coefficients [ICCs]) were calculated for all full- and half-cycle motions. Possible sources of systematic errors were evaluated, and random errors were estimated.

ICCs indicate that the instrument performs very reliably for rotation and lateral bending (0.93-0.97) and acceptably for flexion-extension (0.75-0.93) measurements. Differences in instrument placement, subject posture, or both in different trials correlate neither with differences in measured values nor with variances. Within-trial errors did not correlate with ranges of motion. Standardizing head position resulted in increases in reliability of from 3% to 15% for axial rotation and lateral bending but actually decreased the ICCs for flexion-extension (up to 14%) compared with data collected under a less-stringent protocol. Errors in clinical use are estimated at 4.5 degrees.

By using our modifications to the accessories and standardization of subject position, the CA-6000 is a highly precise and reliable instrument for measuring active cervical motion about the 3 Cartesian axes. Individuals can repeat the same patterns of motion in sequential trials on the same day with very little variation. Ease of repetitious measurement without examiner intervention contributes to the instrument's ability to obtain highly reliable data. Changes in instrument placement or subject body posture between trials do not give rise to systematic errors. Design of the instrument for flexion-extension could be improved.