A comparison of two portable dynamometers in the assessment of shoulder and elbow strength

Objectives To compare the practical applicability and measurement properties of a hand-held dynamometer (MicroFET2®) and a fixed dynamometer (Isobex2.1®) in determining isometric strength of the shoulder and elbow.

Design Muscle strength in four directions (glenohumeral abduction, external rotation and elevation and elbow flexion) was measured using both instruments by two examiners. The assessments were repeated by one of the examiners 3 days later.

Setting Leiden University Medical Center.

Participants Twenty healthy volunteers.

Main outcome measures Time to complete a set of measurements and discomfort were recorded. To determine intra- and inter-observer reliability, intra-class correlation coefficients (ICCs), limits of agreement and smallest detectable difference were computed.

Results The time to complete a set of measurements was significantly shorter for the hand-held dynamometer than for the fixed dynamometer in both examiners. The number of subjects reporting discomfort was similar with the two dynamometers. Except for glenohumeral abduction, the forces measured using the hand-held dynamometer were significantly higher than those when using the fixed dynamometer in both examiners. The intra- and inter-observer ICCs for the four directions ranged from 0.82 to 0.98 for both dynamometers. However, the mean differences between replications and the wide limits of agreement suggest substantial bias and variability. For example, for the measurement of shoulder abduction with the fixed dynamometer by one tester (190N), the results suggest that on 95% of occasions the second tester's measurement would be between 158 and 275N.

Conclusions Although time taken and discomfort should be considered in the selection of dynamometers, due consideration should be given to the significant differences in absolute results. Neither the dynamometers nor the testers can be considered interchangeable. Both the intra- and inter-observer reliability of the two dynamometers were similar, yet both demonstrated systematic bias and variability in the measurements obtained.