Asbestos-related pleural thickening: effect of threshold criteria on interpretation

Radiographic evidence of pleural thickening was evaluated in 1,216 shipyard workers (high-risk group) and 214 executives (low-risk group) and classified according to two threshold levels: low (any detectable thickening) and high (thickening less than or equal to 2 mm excluded). Results were markedly different, affecting the low-risk group more than the high-risk group. Changing from the high to the low threshold markedly increased pleural findings and scores as well as inter-reader correlation while reducing both differences between risk groups and detection of additional findings on supplementary oblique views. When data on prevalence and extent of pleural thickening were combined, differences between risk groups were maximized; however, when only prevalence was considered, such differences were minimized, particularly with a low threshold. These findings indicate that different or ambiguous threshold criteria produce divergent results in screening surveys for asbestos-related disease. The authors recommend that explicit minimal pleural threshold criteria be adopted.