Extrapolation

understood. In the case of contaminants for which the mode of action is less

clear, elicitation of a distribution through expert judgment can be used to

quantify uncertainty (Evans et al. 1994).

Potential noncarcinogens. The toxicity of a potential noncarcinogen is

described by a reference dose (RfD) for the ingestion route of exposure and a

reference concentration (RfC) for the inhalation route of exposure. The reference

dose is defined as an estimate (with uncertainty spanning perhaps an order of

magnitude) of a daily exposure to the human population (including sensitive

subgroups) that is likely to be without appreciable risk of adverse effects.

Noncarcinogenic toxicity factors assume that a concentration or dose threshold

exists, below which no adverse effects are expected.

The no-threshold assumption suggests there is a range of exposures from zero

to a finite value that can be tolerated by individuals with essentially no chance of

expression of the toxic effect. The goal is to determine the lowest of these

exposures in an effort to protect the most sensitive members of the population.

RfDs and RfCs are derived from an NOAEL from an appropriate animal study.

The highest NOAEL is used if a range of studies are available. A number of

uncertainty factors are applied to NOAELs to derive RfDs and RfCs to account

for animal to human extrapolation, variation in human sensitivity, and varying

toxicity data quality.

Extrapolation

In ecological risk assessment, uncertainty factors are used to compensate for

differences among measurement end points in available test species and

assessment end points in species of concern at a dredging or disposal site.

Typically, these fractional factors vary inversely with the quantity and type of

data available (USEPA 1998a).

In human health risk assessment, factors are used to reduce NOAELs to

account for scientific uncertainty inherent in toxicity databases. These factors are

applied to a Low Observed Adverse Effect Level (LOAEL) when an NOAEL is

not available. Dourson and Stara (1983) provide some empirical justification for

the use of these uncertainty factors, which are typically an order of magnitude for

each type of uncertainty considered.

Intraspecies. For human noncarcinogenic toxicity values, a factor of 10 is

applied to an NOAEL or LOAEL to account for the variation in sensitivity

among members of the population.

Although natural variation contributes to uncertainty in estimates of

ecological risk, uncertainties in toxicity end points (i.e., confidence intervals on

LC50, EC50, or LOAEL) are generally not reported in the literature. In some

cases, adaptation of individuals may result in differences in sensitivity between

animals reared in the laboratory or wild animals that are exposed in the field. For

example, recent work has documented populations of fish in the field that are

resistant to the acute toxic effects of PAHs (van Veld 1997) and PCBs (Nacci

et al. 1997). Overall, the magnitude of uncertainty associated with differences

among individuals in a species is expected to be low.

Chapter 5 Uncertainty in Tier IV Risk Assessments