Accuracy of the Borg CR10 scale for estimating grip forces associated with hand tool tasks

The gripping of tools is required by many industrial operations, and an important aspect of exposure assessment is determining the grip force output of operators. Ratings of perceived exertion can provide an indirect measure of grip force; however, reports in the literature of the use of Borg CR10 scale ratings as a surrogate measure of grip force have been mixed. During a laboratory study with 16 participants, power grip forces were measured directly during three hand tool task simulations: (1) a screwdriver task, (2) a ratchet task, and (3) a lift and carry task, each performed at four force/load levels. Borg scale ratings reported following each trial were compared with mean, peak, and integrated grip forces for the respective trials. Pearson correlations conducted on an individual basis were greatest for the screwdriver task, r approximately 0.9. Correlations for integrated grip force were generally better than for mean or peak force. Correlations were also performed on data pooled for all participants, simulating a cross-sectional sampling approach. Correlations made with pooled data were weaker than when conducted on an individual basis, ranging from r = 0.26 for peak grip force for the lift and carry task, to r = 0.79 for the screwdriver task. When the pooled data were normalized to individual maximum voluntary grip exertions, correlation generally improved but not to the level of the "individually scaled" data. Based on these findings, a protocol is proposed that could improve the strength of correlations between direct measures of grip force and ratings of perceived exertion. Differences in strength of correlation between task simulations are discussed with respect to differences observed in force distributions about the handle for the three tasks.