Representativeness of deaths identified through the injury-at-work item on the death certificate: implications for surveillance.

BACKGROUND This research investigated the accuracy of the injury-at-work item on the death certificate for surveillance of occupational injury deaths in Oklahoma during 1985 and 1986. METHODS Representativeness of occupational injury deaths identified by death certificates was assessed by comparing these deaths with all occupational injury deaths identified through death certificates, workers' compensation reports, medical examiner reports, and OSHA records for categories of occupation, industry, and external causes of death. RESULTS Certain external causes of death (e.g., motor vehicle traffic deaths) and certain occupations (e.g., farming) and industries (agriculture and services) are more often underidentified through death certificates. CONCLUSIONS The findings of this study support Baker's observation that no single data source contains all deaths or all the data elements necessary to describe occupational injury deaths. Data sources may be combined to improve representativeness through more complete case ascertainment.