Article Text
Abstract
Objectives Against the background of an epidemic of silicosis and related tuberculosis revealed by field and autopsy studies in recent decades, the South African gold mining industry has committed itself to the elimination of silicosis.
Methods Annual chest radiographs of active miners were taken by mine occupational health services. These were collected in a representative sample of miners enrolled into a baseline survey between 2004 and 2009 as part of a cluster randomised trial of community-wide isoniazid preventive therapy. All radiographs were read for silicosis by an experienced lay reader. All images classified as abnormal and a random sample of normals was re-read by a ‘B’ reader for validation.
Results A total of 14 322 radiographs from 15 goldmining shafts from three companies were read by the lay reader for an overall silicosis prevalence of 3.7% (> ILO 1/1), reaching 6.6% for workers with > 20 years since first employment. Silicosis prevalences adjusted for the readings of the B-reader were much lower: 1.7% and 3.2% respectively. On either reading, these industry-wide radiological prevalences are lower than single workforce surveys dating from 2000/2001. They are also at considerable variance with statutory autopsy data from in-service black miners.
Conclusions In the absence of evidence of lower dust levels in gold mines in the decades prior to this study, one explanation for these findings is that goldminers with silicosis are increasingly being selected out of the industry by ill-health, particularly HIV and tuberculosis, and/or by stricter hiring practices in a declining industry. High silicosis prevalence is thus likely still to be found among former goldminers. Autopsy data from active miners further suggest a shift towards sub-radiological silicosis, which has been shown to confer a substantially increased risk of tuberculosis even in the absence or radiological disease.