Is self-reported pain an appropriate outcome measure in ergonomic-epidemiologic studies of work-related musculoskeletal disorders?

Of the various overlapping approaches used to define outcome in ergonomic-epidemiologic studies of work-related musculoskeletal disorders, the most widely used are clinical diagnostic entities, whose criteria derive either from clinical examination (consisting of clinical tests in conjunction with symptoms and clinical history) or from findings of special investigations, such as nerve conduction testing. Problems with the use of clinical diagnostic entities as surveillance tools relate to their high definitional variability, to the unknown test attributes and performance in worker populations, and to their lack of field utility. Other approaches to define outcome such as impairment and disability evaluations are seldom used; there are problems with existing instruments in disentangling multiple determinants. The use of subjective measures (self-reported pain) may be the most valuable approach to measuring outcome in population-based surveys. This approach has high capacity (can be used in large populations) and good field utility, has been supported by evidence of construct validity in some ergonomic-epidemiologic studies, and is able to assimilate the diverse and overlapping symptom patterns characteristic of some work-related musculoskeletal disorders. Traditional biases against subjective measures are evident in the literature, but these measures have also been badly used in many studies, with insufficient attention being paid to potential confounders and effect modifiers. Measurement of factors that influence pain perception and reporting need to be incorporated in ergonomic-epidemiologic studies and controlled for in analysis. Outcome definitions should be made more explicit.