Test-retest reliability to evaluate programs for preventing mercury exposure from gold production

Although gold mining activity has decreased in Brazil during the last two decades, mercury (Hg) emissions from some gold-mining areas, deforestation operations and the presence of Hg residues in the Amazon have maintained humans exposed to this metal in Brazil. In the city of Poconé, a previous research project showed  a four fold increase in average urinary levels of mercury (HgU) among non-occupationally exposed residents, when compared to a proper control group (4.35mg/L vs. 1.25mg/L; n=365). In addition, 14 subjects that reported gold mining activities in their own backyards, and used to burn Hg amalgams inside their houses, presented extremely high levels of HgU (7.0 to 102.4 mg/L). In a subsequent study aiming to prevent indoor burning of mercury-gold amalgams, an educational program conducted with 8^th grade students helped to decrease HgU among residents from 2.90mg/L to 1.49mg/L. Accordingly, in the present study, the reliability of a questionnaire used to evaluate the impact of preventive programs that aim to reduce exposure to mercury used in gold production was examined. Test-retest reliability study was performed at a 21-day interval between the two sessions. Sample comprised 128 students from 8^th grade of elementary school. Reliability was evaluated by Kappa coefficient and McNemar’s chi-square test. Kappa coefficient (IC95%) varied from “moderate” to “almost perfect” agreement (k = 0.551 to 0.913). Our results showed that the questionnaire presents good reliability. Its use to evaluate preventive programs in different productive process may require further reliability assessment.