Combined data sources, including union administrative records and workers' compensation claims, were used to construct event histories for a dynamic cohort of union carpenters from Washington State during the period 1989-1992. Person-time at risk and the events of interest were stratified by age, sex, time in the union, and predominant type of carpentry work. Poisson regression techniques were used to identify subgroups at greatest risk of filing claims for a variety of musculoskeletal disorders defined by ANSI codes for body part injured and injury nature.
Distinguishing different kinds of musculoskeletal disorders, even crudely with ANSI codes, led to different conclusions about the effects of the explanatory variables. Among older workers, the rates of fractures of the foot were higher, while rates of contusions of the hand and foot were lower. Women had higher rates of sprain/strains and nerve conditions of the wrist/forearm. Higher rates of injuries to the axial skeleton were seen among carpenters who did predominantly light commercial and drywall work, while piledrivers had lower rates of these injuries. Drywall workers had higher rates of sprains to the ankle/lower leg. Workers who were members of the union as long as four years had lower risks for the vast majority of musculoskeletal disorders studied. Similar patterns were seen for more serious claims that resulted in paid lost time from work. Am.