Occupational injury trends derived from trauma registry and hospital discharge records: lessons for surveillance and research.

OBJECTIVE: The suitability of the Washington State Trauma Registry (WTR) for occupational injury surveillance was assessed via comparing estimated rates and trends with those derived from state hospital discharge data. METHODS: Descriptive methods and negative binomial regression were used to model occupational injury trends (1998 to 2009). RESULTS: Nonlinear trends based on WTR data closely tracked those based on hospital discharge data, beginning about 2002. Rate estimates differed somewhat by data source and were most similar when a severity threshold was applied. Conclusions regarding temporal trends in work-related injury rates were the same using either data source. CONCLUSIONS: This study found substantial similarity between occupational injury trends estimated using either WTR or hospital discharge data. We conclude that a mature state trauma registry with mandatory reporting requirements can be used for surveillance of severe work-related traumatic injuries.

Linkage and concordance of Trauma Registry and hospital discharge records: lessons for occupational injury surveillance and research.

OBJECTIVE: Occupational injury researchers typically use payer to identify work-related injuries in hospital discharge records. Many trauma registries contain a work-related field, independent of payer. Linked trauma registry and hospital discharge records were used to assess data field concordance and to assess the validity of using payer or external cause of injury (E-codes) to identify work-related injuries. METHODS: Washington State Trauma Registry records were linked to hospital discharges (year 2009). RESULTS: There was substantial agreement between Washington State Trauma Registry and hospital discharge records for workers' compensation as primary payer. E-code based methods of identifying occupational injuries had high specificity (more than 99%) but low sensitivity (less than 14%). Payer was 76% sensitive and 98% specific. CONCLUSIONS: This study found substantial agreement for data fields key to occupational injury surveillance and research. Nevertheless, many work-related injuries could not be identified using hospital discharge records.