upper-bound estimate of the Carcinogenic Potency Factor is used in 

 carcinogenic risk calculations. Substitution of the mean estimate or the 

 lower bound of the 95 percent confidence interval for the potency 

 factor in the risk calculations is generally not done because of the 

 instability of these estimates (U.S. EPA 1980b, 1986a). 



The U.S. EPA (1986b) guidelines on exposure assessment and Whit- 

 more (1985) summarize the primary methods for characterizing uncer- 

 tainty in exposure estimates m relation to attributes of the exposed 

 population and the exposure data. In many cases, data will be sufficient 

 only to use parametric/bounding analysis, as described above. In any 

 case, a discussion of qualitative uncertainties in the analysis should 

 always accompany presentation of risk assessment results. For ex- 

 ample, limitations of data related to inadequate survey design or 

 insensitive analytical chemistry methods should be described. The 

 extent of chemical data for geographic locations of interest should be 

 summarized. Insufficient information on characteristics of the exposed 

 population should be noted. The level of confidence in data used to 

 develop RfDs, Carcinogenic Potency Factors, and weight-of-evidence 

 classifications based on IRIS Chemical Files should be indicated. 



Additional information to support risk assessment of contaminated 

 fish and shellfish consumption may include: 



• Comparisons of tissue concentrations of contaminants with 

 FDA action (or tolerance) levels 



• Statistical comparisons of mean contaminant concentrations 

 among fishery species and among locations 



• Statistical comparisons of mean contaminant concentrations in 

 fishery species with those in other foods. 



FDA hmits on contaminants in fishery products are shown in Appendix 

 I. Limitations to use of these values for assessing health risk were 

 discussed earher (see above, Overview of Risk Assessment). For com- 

 parison, legal limits on fishery contaminants established by other 

 countries are also provided in Appendix I. 



Some resource management agencies have developed advisories based 

 simply on comparisons between contaminant concentrations in fishery 

 species and those in corresponding species from reference or control 

 areas. For example, the Northeast Shellfish Sanitation Commission has 

 established "alert levels" for metals in shellfish as the concentration 

 equal to one standard deviation above the mean background (refer- 

 ence) concentration. These alert levels are not based on health effects, 

 but assume that the level of concern is related to an elevation above 

 average background conditions. 



Supplementary 

 Information 



75 



