Estimating the Prevalence of Toxic Waste Sites in Low- and Middle-Income Countries.

BACKGROUND

Exposure to heavy metals at contaminated industrial and mining sites, known also as hot spots, is a significant source of toxic exposure and adverse health outcomes in countries around the world. The Toxic Sites Identification Program (TSIP) developed by Pure Earth, a New York-based nongovernmental organization, is the only systematic effort to catalogue contaminated sites globally. To date, TSIP has identified and catalogued 3282 sites in low- and middle-income countries. The TSIP methodology is not designed to survey all contaminated sites in a country. Rather sites are prioritized based on their perceived impact on human health, and only a limited number of the most highly hazardous sites are surveyed. The total number of contaminated sites globally and the fraction of contaminated sites captured by TSIP is not known.

OBJECTIVE

To determine the TSIP site capture rate, the fraction of contaminated sites in a country catalogued by TSIP.

METHODS

Ghana was selected for this analysis because it is a rapidly industrializing lower middle income country with a heterogeneous industrial base, a highly urban population (51%), and good public records systems. To develop an estimate of the fraction of sites in Ghana captured by TSIP, assessors targeted randomly selected geographic quadrats for comprehensive assessment using area and population statistics from the Ghana Statistical Service. Investigators physically walked all accessible streets in each quadrat to visually identify all sites. Visual identification was supplemented by field-based confirmation with portable x-ray fluorescence instruments to test soils for metals. To extrapolate from survey findings to develop a range of estimates for the entire country, the investigators used 2 methodologies: a "bottom-up" approach that first estimated the number of waste sites in each region and then summed these regional subtotals to develop a total national estimate; and a "top-down" method that estimated the total number of sites in Ghana and then allocated these sites to each region. Both methods used cluster random sampling principles.

FINDINGS

The investigators identified 72 sites in the sampled quadrats. Extrapolating from these findings to the entire country, the first methodology estimated that there are 1561 sites contaminated by heavy metals in Ghana (confidence interval [CI]: 1134-1987), whereas the second estimated 1944 sites (CI: 812-3075). The estimated total number of contaminated sites in Ghana is thus 7-9 times the number of sites captured through TSIP. On a population basis, it was estimated that there are between 31 and 115 contaminated sites per million inhabitants in Ghana.

CONCLUSIONS

The findings of this study indicate that the TSIP methodology provides a sound statistical basis for policy formulation. The statistical approaches used in this study can be replicated in other countries to improve estimates of the prevalence of contaminated sites. This information provides important input to calculations of the global burden of disease attributable to hazardous exposures at contaminated sites.