Using multiple data sets for public health tracking of work-related injuries and illnesses in California

Source: Lauren Joe, Rachel Roisman, Stella Beckman, Martha Jones, John Beckman, Matt Frederick and Robert Harrison, American Journal of Industrial Medicine, Early View, August 5, 2014
(subscription required)

From the abstract:
Background: Research suggests the U.S. Bureau of Labor Statistics Survey of Occupational Injuries and Illnesses underestimates the magnitude of workplace injuries and illnesses. Enumerating workplace injuries and illnesses may be improved by utilizing multiple state-based data sources.

Methods: Using California-based datasets (workers’ compensation claims, health care facility data, and physician reports), we enumerated unique cases of amputations and carpal tunnel syndrome (2007–2008), and evaluated the datasets for usefulness in occupational health tracking by performing record linkage across all datasets and calculating match rates between them.

Results: 6,862 amputation and 39,589 carpal tunnel syndrome (CTS) cases were identified. Match rates between the datasets ranged from 34.0% to 45.6% (amputations) and 3.0% to 43.5% (CTS). Enumerated amputation and CTS cases from state-based sources were about five and ten times greater than the BLS SOII estimates (1,390 and 3,720).

Conclusions: Successful demonstration of this state level approach has broad implications for improving federal and state efforts to track and prevent work-related injuries and illnesses. …