The process for selecting observational methods for evaluating work-related upper limb musucloskeletal disorders (WRULMSDs) is complex, but essential to risk management. In a Portuguese automotive industry plant (Setúbal, 2004-05), where all jobs had been previously analyzed, OCRA checklist was reapplied (final assembling and painting) in workstations (n=152) classified as moderate/high risk (score > 16.5 OCRA). In all risk confirmed cases (n=71) three other methods were applied: RULA, SI, and HAL. Sequences of working activities were recorded on video to estimate the predictive validity of the methods. The results show differences in the final scores of the methods in the same workstations: OCRA reveals 34 workplaces with high risk; SI only ranked 31 positions with high risk; HAL has 35 high risk workplaces, and RULA only 7. Analysis of the effect of each independent variable (risk factor) in the final scores of the methods, based on linear regression analysis, reveals different contributions and, consequently, different ways of assessing WRULMSD risk. The video analysis also highlights different predictive validities for the risk factors in each method. These results indicate the need to select the appropriate WRULMSD risk assessment instrument adequate to each real work situation.
WRULMSD risk assessment; Occupational Repetitive Actions checklist - OCRA; Rapid Upper Limb Assessment - RULA; Strain Index - SI; Hand Activity Level - HAL
