A nested case-control study of fatal work related injuries among Brazilian steel workers

To estimate the relative risk of death from work related injury in a steelworks, associated with exposure to various occupational hazards, sociodemographic factors, and medical history.

The study was a nested case-control design. It was based on a cohort of men employed in the steel plant of USIMINAS, Brazil between January 1977 and August 1990, who were followed up to November 1992. The cases were defined as all workers in the cohort who died from injury in the study period and whose death had been notified to the Brazilian Ministry of Labour as being related to work. Four controls per case, matched to cases on year of birth, were randomly selected from among workers employed in the plant at the time of death of the matching case. Data on potential risk factors for occupational injury were extracted from company records; for the controls these data were abstracted for the period preceding the death of the matching case.

There were 37 deaths related to work injuries during the study period. Four surviving workers were selected as controls for each case, but for eight the personnel records were incomplete, leaving 140 controls in all. Significantly increased risk of fatal injury related to work was associated with exposure to noise, heat, dust and fumes, gases and vapours, rotating shift work, being a manual worker, and working in the steel mill, coke ovens, blast furnaces, and energy and water supply areas. Risk of fatal injury related to work increased with intensity of exposure to noise (P (trend) = 0.004) and heat (P < 0.001), and increased greatly with a hazard score that combined information on noise, heat, dust, and gas exposure (P < 0.001). Number of years of schooling (P = 0.03) and salary level (P = 0.03) were both negatively associated with risk. In a multivariate analysis including all these significant factors, only hazard score and area of work remained associated with death from injury related to work. The highest risks were for men exposed to all four environmental hazards (odds ratio (OR) 19.4; 95% confidence interval (95% CI) 1.1 to 352.1) and those working in the energy supply area (OR 18.0; 1.6 to 198.1).

The study identified parts of the steelworks and types of hazard associated with greatly increased risk of fatal accident. Research and measures to prevent accidents need to concentrate on these areas and the people working in them. The use of a hazard score was successful in identifying high risk, and similar scoring might prove useful in other industrial situations.