Asbestos Exposure at the Long Beach Naval Shipyard; Reproducibility and Validity of Self Reported Exposure

Objective. To assess the reproducibility and validity of self reported asbestos exposure in a naval shipyard.;Methods. 1,734 workers at the Long Beach Naval Shipyard completed exposure and health effects questionnaires in 1983 and 1985. 209 of these workers completed the questionnaires in both 1983 and 1985. Kappa statistics were calculated to determine the reproducibility of self reported exposure to asbestos and other agents for these 209 workers. Validity of asbestos self report was assessed by comparing the association between self reported asbestos exposure and self reported asbestos disease to the association between asbestos exposure assigned by occupation and self reported asbestos disease. Exposure assignment by occupation was based upon literature review. Logistic regression was used to analyze the associations between exposure and disease outcome.;Summary of findings. Of the 209 workers who completed both the 1983 and 1985 questionnaires, only 14 answered the question regarding past asbestos exposure discordantly. The resulting kappa statistic of 0.65, which indicates good agreement, ranked second only to fiberglass among 21 exposures assessed. The odds ratio for self reported asbestos disease based upon self reported asbestos exposure was 11.1 (3.98-30.8). The odds ratio for self reported asbestos disease based upon exposure determined by occupation was 2.70 (1.76-4.16). This disparity in odds ratios may reflect a reporting bias created by simultaneous assessment of self reported exposure and disease. Exposure misclassification by occupational assignment may have also played a role.