Expert consensus ratings of job categories from the Third National Health and Nutrition Examination Survey (NHANES III).

BACKGROUND A method of occupational physical exposure assessment is needed to improve analyses using large data sets (e.g., national surveys) that provide only job title/category information as a proxy for exposure. METHODS Five ergonomic experts rated and arrived at consensus ratings for job categories used in the Third National Health and Nutrition Examination Survey. Interrater agreement was examined for initial (pre-consensus) ratings. Correlation between consensus ratings and an independent source of ratings (US Department of Labor (DOL)) was used as a basis of comparison. RESULTS Interrater agreements for the initial ratings were weak. Highest interrater agreement was for sitting (weighted kappa (kappa(w)) = 0.56). Lowest agreement was for standing (kappa(w) = 0.07). Consensus ratings were well correlated with DOL ratings (correlation coefficients >or=0.6). CONCLUSIONS The correlation between consensus ratings and DOL ratings support the use of expert consensus to assess physical exposures in national data sets. However, validation of expert consensus ratings is required.