An investigation of exposure assessment methods for selected physical demands in hand -intensive work.

Work-related upper extremity musculoskeletal disorders are still prevalent in many hand-intensive industries. Current ergonomic exposure assessment methods do not adequately address the needs of ergonomists and epidemiologists. The purpose of this work was to evaluate and improve the capability and reliability of subjective exposure assessment methods for selected upper extremity ergonomic risk factors in manual assembly work. Specifically, the work examined the reliability when using a reduced number of analysts, compared worker assessments with observer assessments and electromyography, and examined the relationship between work element and overall job ratings. The observational analyses and worker self-assessments were applied at both job and work element levels. This work also proposed and tested a tool to predict Hand Activity Level from work methods information and compared results to observed data. Measures of inter-after reliability and statistical correlation were used to assess reliability of the above comparisons. Results from this work indicated that for selected evaluations, measurement reliability was maintained while using pairs of analysts because interclass correlation coefficients were generally greater than 0.60 after discussion between analysts. In the case of Hand Activity Level and peak hand force, a single analyst could be used because the interclass correlation coefficients were greater than 0.60 before discussion. Additionally, the inter-rater reliability of observational measurements was higher for all risk factors during on-site job-level evaluations. Inter-rater reliability (0.21--0.59 vs. 0.73--0.78) and worker-observer agreement (r2 = 0.04 vs. r2 = 0.36) also improved significantly by assessing work elements and combining them analytically to create an overall job score. These combined elemental ratings were significantly correlated to overall job ratings, but were higher in magnitude than job-level ratings. This work demonstrated additional uses for information gained from element-level analysis. When comparing worker self-assessments of hand force to electromyography, muscle activity explained 25% of variance in worker ratings while other job and personal factors accounted for an additional 48%. Prediction of Hand Activity Level from work methods information showed a strong relationship (r 2 = 0.64) between calculated and observed scores suggesting that this tool could be used to predict and track exposure to repetitive motion/exertion.PhDApplied SciencesHealth and Environmental SciencesIndustrial engineeringOccupational safetyPublic healthUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/125349/2/3192626.pd