“…A primary source of epidemiologic data on CTDUE has been workers' compensation databases. Unfortunately, numerous limitations of workers' compensation data for research purposes have been cited, such as variation in policy by region and over time [Bednar et al, 1998]; variation in compliance with reporting requirements [Korrick et al, 1994]; incomplete coverage of all workers [Yassi et al, 1996]; difficulty in generating direct ''at risk'' denominators [Franklin et al, 1991;Ashbury, 1995;Yassi et al, 1996;Sprout, 1997]; differential treatment of sudden onset vs. gradual onset injuries [Webb et al, 1989;Blessman, 1991]; significant under-reporting by workers and health professionals [Fine et al, 1986;Cummings et al, 1989;Webb et al, 1989;Franklin et al, 1991;Korrick et al, 1994;Beaton, 1995;Maizlish et al, 1995;Yassi et al, 1996;Herbert et al, 1997;Meservy et al, 1997;Silverstein et al, 1997;Biddle, 1998;Homan et al, 1999]; obtaining benefits under other sickness and disability insurance [Maizlish et al, 1995]; poor measurement of exposure due to broad industrial and occupational categories [Franklin et al, 1991;Silverstein et al, 1997]; lack of validated, quantitative case definitions [Cummings et al, 1989;Franklin et al, 1991;Silverstein et al, 1997]; the limited number of coded variables, particularly with respect to clinical data [Korrick et al, Job-related factors . Differences in jobs normally held by men and women Mergler et al,1987;Ashbury, 1995;Brogmus et al,1996;…”