A video-based observation method to assess musculoskeletal load in kitchen work.

This paper describes a new video-based observation method aimed to assess musculoskeletal load in kitchen work, aspects of its repeatability and validity, and problems confronted by the observers. Two pairs of researchers observed individually 117 video clips recorded in kitchens. Interobserver repeatability was assessed by computing the proportion of agreement and weighted kappa values (kappa(w)). Validity was analyzed by studying the distribution of the assessments over the rating scales and the ratings before and after the interventions, which were compared with expert assessments made from the same intervention targets. The proportion of agreement ranged from 57 to 88%. Interobserver repeatability based on weighted kappa values was mainly good to moderate. The method detected the changes in physical load due to the interventions. Direction of the changes corresponded with the expert assessments. Further development of the method is needed to assess the load on the hands and wrists.