NCSAB Risk Assessment Guidelines

Secretary's Science Advisory Board on Toxic Air Pollutants:
Internal guidelines for toxicological evaluation of chemicals released to the air

This document outlines the risk assessment guidelines the NCSAB uses to formulate recommendations for acceptable ambient concentrations for toxic air pollutants for the North Carolina Department of Environment and Natural Resources (DENR). Risk assessment is the evaluation of scientific information on the hazardous properties of environmental agents, the doses of the agents that may cause adverse effects, and the extent of human and environmental exposure to those agents. The product of a risk assessment in general is a statement regarding the probability that populations exposed to a given agent will be harmed at a given exposure level, and to what degree. The NCSAB has agreed to forward acceptable ambient level (AAL) air concentration recommendations in a form that includes a range of values with the associated risks at each value. In this manner, recommendations based on the strongest science will be accompanied by alternate values that may be considered by risk managers during the adoption of AAL guidelines. NCSAB health-based recommendations will be forwarded to risk managers in the context of a summary document that outlines the NCSAB deliberation process and gives a brief overview of the toxicology of the compound and the risks associated with exposure. These summaries will be submitted to the Air Quality Committee (AQC) and the Environmental Management Commission (EMC) to aid the risk management process. The EMC makes the final decision on changes to AALs.

Risk Assessment
The NCSAB uses risk assessment methodology to arrive at health-based recommendations. Risk assessment is a methodical process involving characterization of toxicity, likelihood for exposure and likelihood for adverse health effects as a consequence of exposure. The four widely acknowledged components of risk assessment are hazard identification, dose-response assessment, exposure assessment and risk characterization. The NCSAB follows this risk assessment framework to determine acceptable ambient exposure levels for toxic pollutants released to the environment. This section outlines the general process involved in determining safe levels of exposure.

Hazard Identification: Hazard identification entails the review of relevant biological and chemical information to determine whether or not a chemical or agent may cause adverse health effects. The physical and chemical properties of each agent are considered in relation to the potential for inhalation and/or dermal exposures. At the outset, the toxicological literature is collected and reviewed. The relevant scientific literature for a given compound may include epidemiological or workplace studies, controlled human studies, laboratory animal bioassays and/or other laboratory studies. In general, the NCSAB prefers to base their recommendations on carefully conducted human epidemiological or workplace monitoring studies. However, in many cases quality studies on human health effects are lacking, and animal or laboratory studies are used to assess risk. The literature review provides the NCSAB with the information necessary to classify toxicological effects as carcinogenic and/or noncarcinogenic.

Response Assessment: The dose-response relationship describes the correlation between extent of exposure to an agent (dose) and the resulting effects (response). A dose-response curve can then be developed for a given chemical to the extent that this relationship is described in the literature. This type of exercise is necessary to correlate high-level experimental type exposures to the types of exposures that are more likely to take place in the environment. Often toxic agents will give rise to a variety of adverse health effects. In these cases, an effort is made to establish a dose-response relationship for each effect so that a comparison of potency can be undertaken. Generally the NCSAB aims to protect against the most sensitive effect since this level of protection will also be protective for other, less sensitive endpoints. The dose-response data in the literature may come from human or animal studies, and may provide information regarding mechanism of action of the toxicant. In general, the NCSAB uses different risk assessment methods when considering noncarcinogenic and carcinogenic effects of compounds.

Exposure Assessment: Exposure assessment provides an estimate of actual or potential population exposures and an estimate of exposure levels that may result in an adverse health effect. NCSAB exposure assessments generally take into account potential inhalation exposures only, although dermal exposure may also be considered. Consideration of exposure due to deposition of airborne compounds onto soil or water may be important for some compounds, and may be considered when available data indicate deposition may lead to significant exposure. Each compound is considered individually, though consideration of mixtures of compounds with similar mechanisms of action may be necessary, as the DENR requests.

Risk Characterization: Risk characterization integrates the hazard identification, dose-response assessment and exposure assessment in order to describe the nature and magnitude of health risk. Thorough analysis of the toxicological literature aids in determining the critical (or most sensitive) effect for risk assessment purposes. The critical effect is the adverse effect that occurs at the lowest dose, taking into account biological and statistical significance of effects, reversible versus irreversible endpoints, and effects on sensitive subpopulations. The NCSAB often develops preliminary conclusions based on more than one endpoint for comparison to insure the final recommendation will protect against all endpoints of concern. The NCSAB presents multiple risk estimates, where possible, for cancer and non-cancer effects. For cancer estimates, this might include an upper bound estimate on risk as well as a most likely estimate of risk. For non-cancer estimates, the NCSAB may arrive at several risk estimates based on different methods for estimating risks including the use of animal vs. human data or various toxic endpoints. While all NCSAB recommendations forwarded to risk managers will have a scientific foundation, the uncertainties associated with each estimate will be explained in each case, and the NCSAB will indicate the estimate(s) which have the strongest science base and which endpoints are of the greatest public health concern. In this fashion, the NCSAB will be providing a "range of risks" to the risk managers responsible for making final decisions concerning AALs and the potential impact on public health.

Other Considerations
The NCSAB strives to use up-to-date risk assessment methods whenever appropriate. Examples or this include using computer modeling to estimate carcinogenic risk and physiologically based pharmacokinetic models to estimate non-carcinogenic risk. The updated risk assessment methods used by the NCSAB are widely accepted by risk assessment bodies and can be found in such publications as the EPA Guidelines for Carcinogen Risk Assessment, which represents EPA carcinogen risk assessment policy. In those cases where the NCSAB adopts a recommendation using a "non-traditional" or recently developed method for risk assessment, the accompanying narrative will indicate this and attempt to provide an explanation of the methodology chosen. For comparison, EPA estimates of risk - when available - may be included in the narratives of the NCSAB risk assessments.

The NCSAB utilizes uncertainty factors when gaps in the toxicological databases necessitate their use. This is necessary to provide a degree of conservatism when toxicological information is lacking for a chemical. The NCSAB tends to use uncertainty factors tailored to the chemical being considered. An example of this is the range of values available for LOAEL to NOAEL conversion; smaller values (normally 3) are used when the dose-response pattern seems to call for reduced uncertainty factors and larger values (normally 10) are used when the dose-response curve appears to be broad or dose-response information is lacking. As the science of risk assessment changes, the NCSAB risk assessment methodology will change accordingly. However, the approaches outlined here and in the table below describe the risk assessment methods currently used by the NCSAB.

The NCSAB issues health-based AAL recommendations for toxic air pollutants but does not recommend averaging times. The Division of Air Quality, through its Toxics Protection Branch, is responsible for recommending appropriate averaging times based on the guidelines found in the 1986 North Carolina Academy of Sciences report entitled Report and Recommendations of the Air Toxics Panel of the North Carolina Academy of Sciences. In general: irritants and acute systemic toxicants are usually assigned 1 hour averaging times, chronic systemic toxicants are usually assigned 24 hour averaging times, and carcinogens are assigned annual averaging times.

NCSAB recommendations with attached averaging times will be forwarded to DENR through Division of Air Quality (DAQ). The Liaison to the NCSAB will prepare a narrative for each compound reviewed that will describe the decision process adopted by the NCSAB. These narratives are not in-depth examinations of the toxicology of each reviewed compound. Rather, they provide a narrative of the NCSAB recommendations and bases for those recommendations in plain terms, with adequate explanation of all uncertainty factors used, points of uncertainty or disagreement, a discussion on alternate recommendations and an estimate of the resulting risks to public health (i.e. the "range of risks"), and a mathematical representation of the final risk assessment methodology chosen. These summaries will be attached to all NCSAB recommendations when they are forwarded to the Air Quality Committee and the Environmental Management Commission for rule consideration.

Uncertainty factors developed by the NCSAB for risk assessment of noncarcinogens
Noncarcinogenic effects, conversions Uncertainty Factors Circumstances for use
LOAEL to NOAEL 10 Dose: response curve is broad, effect is severe
5 Dose: response curve is broad
2 Dose: response curve is steep
Animal to Human 3 some data available
10 no data available to relate animal to human
Inter-individual Variation 10 account for variability in human sensitivity
Acute to Chronic exposure 4 data from acute studies used to estimate chronic effects