Usefulness of occupation and industry information in mortality data in South Africa from 2006 to 2015

BACKGROUND: There is no population based occupational health surveillance system in South Africa, thus mortality data may be a cost effective means of monitoring trends and possible associations with occupation. The aim of this study was to use deaths due to pneumoconiosis (a known occupational disease) to determine if the South African mortality data are a valid data source for occupational health surveillance in South Africa. METHODS: Proportions of complete occupation and industry information for the years 2006–2015 were calculated for working age and retired adults. Deaths due to pneumoconiosis were identified in the data set and mortality odds ratios calculated for specific occupations and industry in reference to those who reported being unemployed using logistic regression. RESULTS: Only 16.1% of death notifications provided a usual occupation despite 43.1% of the population being employed in the year. The MORs for occupation provided significant increased odds of pneumoconiosis for miners (9.04), those involved in manufacturing (4.77), engineers and machinery mechanics (6.85). Along with these jobs the Mining (9.8), Manufacture (2.2) and Maintenance and repair industries (6.0) have significantly increased odds of pneumoconiosis deaths. The data can be said to provide a useful source of occupational disease information for surveillance where active surveillance systems do not exist. CONCLUSION: The findings indicate valid associations were found between occupational disease and expected jobs and industry. The most useful data are from 2013 onwards due to more detailed coding of occupation and industry.