The aim of this study was to assess the reliability of a technique to measure the strength of the intrinsic hand muscles. Intraclass Correlation Coefficients showed an excellent level of reliability for the comparison of muscle strength between groups of patients. However, for the results of individual patients, the calculated Standard Error of Measurements (10-16%) and the Smallest Detectable Differences for intraobserver (31-36%) and interobserver (37-52%) values indicate that only relatively large changes in strength can be confidently detected with this technique. The results of the present study were compared with those of four previous grip strength studies.
Copyright 2000 The British Society for Surgery of the Hand.