Umeå University's logo

Background: To develop effective injury prevention approaches for work-related musculoskeletal disorders in hand-intensive work, accurate quantitative risk assessments of workers' hand activity and force level are essential. However, to what extent workers’ self-ratings and observers’ ratings of hand activity and force levels correlate to corresponding objective measures is not well known. 

Methods: Fifty-nine workers (29 women and 30 men) were video recorded performing a hand-intensive work task for 15 minutes in their regular work. Workers self-rated hand activity level and hand force level (Borg CR10), both being variables in the method Hand Activity Threshold Limit Value®. Furthermore, four observers, two woman–man pairs, rated each worker's hand activity and force levels from the video recordings. Workers’ wrist flexion angular velocity was measured using inertial movement units (IMUs) on the hand and distal forearm. The 10th, 50th, and 90th percentiles were computed. The muscle activity in the forearm muscles flexor carpi radialis and extensor carpi radialis was measured with electromyography (EMG, Delsys, Trigno).  Root-mean-square values (RMS) were computed and normalized to the RMS during maximal voluntary electrical activation (%MVE). The 50th percentile of %MVE and proportion of time >30%MVE in the work task were computed from the normalized EMG.

Results: Spearman correlations between self-rated hand activity and wrist angular velocity were 0.35, 0.38, and 0.28 at the 10th, 50th, and 90th percentiles, respectively. Corresponding correlations for observers’ ratings were 0.40, 0.56, and 0.45. All correlations were statistically significant.  None of the self-reported force levels were significantly correlated to muscle activity (50th and >30 %MVE).  Correlation between observer-reported force level and flexor carpi radialis activity was 0.40 and 0.56 of 50th and 30%MVE, respectively. Corresponding correlation between observer-reported force level and extensor carpi radialis activity was 0.37 for >30 %MVE. All observer-rated force levels were significantly correlated to muscle activity, except for 50th for extensor carpi radialis.

Conclusion:  Observer ratings correlated more strongly than self-ratings with technical wrist angular flexion velocity and muscular activity measures. If ratings are used, observers’ ratings should be preferred since they seem more valid than the workers’ own ratings. Overall, the correlations between ratings and technical measures were modest. This suggests that technical measures of wrist flexion angular velocity and muscular activity should be preferred for exposure assessment, when possible.