1. SUMMARY

Site-specific health risk assessment provides an appraisal of the nature and magnitude of the risks arising from chemical contamination of a site. The assessment takes into account factors relevant to the site such as the proposed use, physico-chemical and bioavailability characteristics of the particular contaminant(s), and the depth and distribution of the contamination. Health risk assessment complements the process of ecological risk assessment.

Site-specific health risk assessment is intended "to provide complete information to risk managers, specifically policymakers and regulators, so that the best possible decisions are made" (Paustenbach, 1989, p28). Good risk assessment is dependent upon a high degree of objectivity and scientific skill and should be distinguished from the risk management process which selects options in response to health risk assessments and which incorporates "scientific, social, economic and political information" and which "requires value judgements eg on the tolerability of risk and reasonableness of costs" (ANZECC/NHMRC 1992, piii).

A preliminary site-specific appraisal risk assessment can be undertaken by comparing site results with the Health-based Investigation Levels appropriate to the site and to its current or proposed use. These are derived using risk assessment techniques and can be applied generically to a range of exposure settings. Where there are exceedances of the Health-based Investigation Levels, site-specific health risk assessments may be used to determine whether further action is needed for a site. The action may range from informing residents or owners of the site of the contamination, to requiring large scale remediation.

The process of risk assessment is intended to achieve the following objectives when assessing site contamination (US EPA 1989):

There are several models of risk assessments and various sets of definitions for the relevant terms. This document uses a model comprising:

•          data collection and evaluation of the chemical condition of the site;

These four stages are closely linked and highly dependent on the preceding stages.

Data collection entails the acquisition and analysis of information about chemicals on a site that may affect human health and which will be the focus for the particular risk assessment (US EPA, 1989).

Toxicity assessment considers:

•          the nature of adverse effects related to the exposure;

•          the dose-response relationship for various effects;

•          the weight of evidence for effects such as carcinogenicity; and

•          the relevance of animal data to humans.

Both qualitative and quantitative toxicity information is evaluated to determine "the incidence of adverse effects occurring in humans at different exposure levels" (US EPA, 1989, p 1.6).

Exposure assessment involves the determination of the frequency, extent and duration of exposures in the past, currently, and in the future. There is also the identification of exposed populations and particularly sensitive subpopulations, and potential exposure pathways. Environmental monitoring and predictive models can be used to determine the levels of exposure at particular points on the exposure pathways. The contaminant intakes from the various pathways under a range of scenarios can then be estimated (US EPA, 1989).

Given this information, risk characterisation details the nature and potential incidence of effects for the exposure conditions described in the exposure assessment. An integral part of this stage is to evaluate the uncertainties and assumptions in the risk assessment process (Langley and El Saadi, 1991). The uncertainties should be "taken into account in planning the management of a site" (ANZECC/NHMRC 1992, p34). The uncertainties may be addressed by gathering further information, the incorporation of safety factors (eg in the development of criteria) and conservatism, and professional judgement.

"The process of risk assessment should enable consistent decisions to be made by the specialists undertaking the process. Expert professional judgement is an integral part of the process. Site-specific risk assessments should not lead to significant variations in the management of similar sites" (ibid, p34)

In many instances site-specific health risk assessments will not be necessary as problems will be 'obvious' and the significant resources required for an adequate site-specific risk assessment or the generation of site-specific soil criteria should be directed to the management of the site. For some sites health risk assessment may be unnecessary as "there may be no population at risk, or decisions may be made on other grounds" (ibid, p20). Site-specific risk assessments may be required as part of the Occupational Health and Safety procedures relevant to site assessment activities.

Numerical estimates of risk will rarely be feasible because of "limitations in toxicological and exposure data" (ibid, p34) which will be reflected in the uncertainty assessment, but a degree of quantification may be possible for some components such as data collection and exposure assessment.

It should be recognised that, as a consequence of data limitations (for example, from the rates of sampling and the analytes chosen), site-specific health risk assessment is a screening process where there may be low rates of false negatives and false positives. "Risk assessment is based on probabilities rather than absolutes and this should be reflected in decision-making" (ibid, p34).

This document provides an approach to site-specific health risk assessment. Due to the complexity and scale of the health risk assessment process a concise 'cookbook' is not practicable. Similarly, the site-specific issues are often sufficiently complex and 'site-specific' for a particular site that a manageable and complete algorithm for decision-making cannot be drafted: the document provides a series of guidelines (and prescriptions) to assist the decision-making process. Where possible, the document is prescriptive about certain aspects of risk assessment. Having specific requirements for the content of investigations and having them presented in uniform, coherent and logically developed reports will enable more efficient, accurate, timely and transparent decision-making and a greater consistency of decision-making across Australia. The principles and guidelines in this document are intended to assist that process and the qualitative process of determining whether remediation is required or not for the proposed use.

The site-specific process is a multi-disciplinary task and requires considerable expertise. People involved in specific components of the health risk assessment process should be adequately qualified and experienced and have a broad understanding of health risk assessment and management and the practical realities of contaminated sites. Professional skills that may be used include: soil science, engineering, geology, history, chemistry, planning, statistics, occupational hygiene, occupational and public health medicine, environmental health, toxicology and health science and epidemiology. While it is unlikely that one person will have the breadth of skill to undertake all components of the health risk assessment, there must be a single person coordinating and taking responsibility for the assessment.

2. ABBREVIATIONS

3. GLOSSARIES

3.1 Soil Criteria Terms

There are two prerequisites for comparison of soil and water test results with defined criteria. The first prerequisite is a standardised soil sampling methodology which provides an appropriate amount of information about the distribution and level of contaminants on a piece of land. The second is a standardised approach to data analysis to enable a meaningful interpretation of sampling results.

3.1.1 Investigation Levels:

An investigation level is the concentration of a contaminant above which further appropriate investigation and evaluation will be required. The investigation and evaluation is to ascertain:

•         the typical and extreme concentrations of the contaminant(s) on the site;

•         the horizontal and vertical distribution(s) of the contaminant(s) on the site;

•         the physico-chemical form(s) of the contaminant(s);

•         the bioavailability of the contaminant(s).

(Langley and El Saadi 1991)

Health-based Soil Investigation Levels are not intended to be clean up levels.

Levels slightly in excess of the investigation levels do not imply unacceptability or levels likely to pose a significant health risk (See Figure 3-I).

Once the further investigation(s) is (are) completed, a site-specific health risk assessment will be required to determine the presence of health risk and, if present, its nature and degree. Final assessment of the degree of contamination should take into account any uncertainties arising from the sampling and analytical methodologies.

When dealing with substances which are considered to have possible effects at very low doses (eg. some carcinogens), a specific approach will need to be established to derive the investigation and response levels. The NHMRC Technical Working Party

on the Carcinogenic Risk Assessment for Soil Contaminants has established Guidelines for Cancer Risk Assessment (NHMRC, 1999).

The application of Investigation Levels and Response Levels to site management will be guided by the risk management process which will be driven by scientific, technological, social, political and economic factors.

Investigation levels provide a trigger to assist in judging whether a detailed investigation of a site is necessary.

When assessing the environmental/human health significance of levels of contamination above an investigation level, the following factors should be considered: potential ground water contamination; land use; the history and nature of the contamination; evidence of potential contamination from site inspection; the local background levels; the problems of the presence of multiple contaminants; and the size of the site. Exposure pathways will be more diverse for a larger site.

The principal limitations of health-based soil investigation levels are that they:

3.1.2 Response Levels:

Response Levels are the concentrations of contaminants at a specific site assessment for which some form of response is required to provide an adequate margin of safety to protect public health and/or the environment.

(adapted from Langley and El Saadi 1991)

Different Response Levels are intended to be used for different exposure situations (eg. residential, recreational, or commercial/industrial land uses).

Figure 3-I

The relationship of soil criteria levels for Substance X.

Proposed Land Uses:

1.       Residential

2.       Recreational

3.       Residential (minimal exposure)

4.       Commercial/Industrial

(Figure not to scale, sequence of '1234' will vary from substance to substance. For example, for another substance, the sequence may be 2134).

(adapted from ANZECC/NHMRC, 1992, p36)

Site-specific evaluation of the available data and proposed land use will be required to determine whether single, occasional or typical values in excess of the investigation level will prompt the further investigation.

Overt health effects would not be expected to occur until contamination is present at levels well in excess of response levels.

The nature of the response required to protect human health will depend on the assessment of risk associated with a given level of contamination. Where the risk is assessed as being relatively low, the response may simply involve informing occupants of the site so that they are aware of risks arising from, certain activities such as, pica behaviour in children (see Schedule B(7A), section 12.2). In cases where there is a relatively high risk, complex soil treatment may be required.

More specifically, the nature of the response will be modulated by factors including:

•     Land use eg. residential, recreational or commercial/ industrial.

•     Potential child occupancy.

•     Potential environmental effects including leaching into groundwater.

•     Single or multiple contaminants.

•     Depth of contamination.

•     Level and distribution of contamination.

•     Bioavailability of the contaminant(s) eg. related to speciation, route of exposure.

•     Physico-chemical properties of the contaminant(s).

•     State of the site surface eg. paved, grassed or exposed.

•          Potential exposure pathways.

•          Uncertainties with the sampling methodology and toxicological assessment.

Where a site specific assessment is being carried out with a view to defining response levels, consideration should also be given to the possible risk associated with mixtures of contaminants, since in some circumstances such risks may necessitate a more or less extensive response than would be required to deal with a single contaminant." (Langley and El Saadi 1991, Imray and Langley 1996)

3.2 General terms

(Adapted from NHMRC 1997)

ADI is the Acceptable Daily Intake. The daily intake of a chemical which, during a lifetime, appears to be without appreciable risk, on the basis of all the facts known at the time. It is expressed in milligrams per kilogram of body weight per day (mg/kg/day) (WHO, 1989a)

Adverse Effect is the change in morphology, physiology, growth, development or life span of an organism which results in impairment of functional capacity or impairment of capacity to compensate for additional stress or increase in susceptibility to the harmful effects of other environmental influences.

Adduct is a chemical moiety which is covalently bound to a large molecule such as DNA or protein (DOH, 1991)

Agent is any chemical, physical, biological or social substance or factor being assessed, unless otherwise noted

BMD (Benchmark Dose) is the dose associated with a given incidence (eg. 1%, 5% or 10% incidence) of effect, the Benchmark Risk, based on the best-fitting dose-response curve.

BMR (Benchmark Risk) is a predetermined incidence of adverse response which determines the Benchmark dose.

Background Concentration is the naturally occurring, ambient concentrations of substances in the local area of a site.

Bioavailability is a unitless measure of the ratio of the amount of chemical exposure (applied dose) and the amount of chemical that enters the tissues of exposed biota (absorbed dose).

Biological Monitoring is the measurement of a contaminant or metabolite in body tissue or fluid. It is usually used as a marker or indicator of exposure to environmental chemicals.

Biota are plants, animals, including humans, fungi or bacteria.

Cancer is a disease of heritable, somatic mutations affecting cell growth and differentiation. That is, genetic alterations incurred in the first damaged cells are acquired in subsequent cells after cell division within the same individual.

Carcinogen is a cancer-causing agent.

Carcinogenesis is the origin, causation and development of tumours. The term applies to all forms of tumours (eg. benign and malignant).

Confidence is the weight assigned by the evaluator to the quality of the information available (high, medium or low confidence) to indicate that a chemical possesses certain toxicological properties.

Confidence Limits are the range of values determined by the degree of presumed random variability in a set of data, within which the value of a parameter, eg. the mean, lies, with a specified level of confidence or probability (eg. 95%). The confidence limit refers to the upper or lower value of the range (DOH, 1991).

Confounding Factor is a factor that distorts the apparent effect or magnitude of the effect of a study factor or risk. Such factors must be controlled for in order to obtain an undistorted estimate of a given effect (DOH, 1991).

Contamination is the condition of land or water where any chemical substance or waste has been added at above background and represents, or potentially represents, an adverse health or environmental impact.

Critical Effect(s) is the adverse effect(s) judged to be most appropriate for determining the tolerable intake.

Default Value is a pragmatic, fixed or standard value used in the absence of relevant data.

Dermal is of the skin, through or by the skin.

DNA is the carrier of genetic information for all living organisms except for the group of RNA viruses.

Dose is the total amount of a chemical that enters or interacts with a receptor (biota including humans). The applied dose is the amount of chemical in contact with the primary absorption boundaries (eg. skin, lungs, gastrointestinal tract) and available for absorption. The absorbed dose is the amount crossing a specific absorption barrier (eg. the exchange boundaries of skin, lung, and digestive tract) through uptake processes. The amount of the chemical available for interaction by any particular organ or cell is termed the delivered dose of that organ or cell (US EPA 1992, p 22933).

Dose-response is the relationship between the dose of a chemical and the extent of the toxic effect produced by the chemical in a biological system.

Endpoint is an observable or measurable biological event used as an indicator of the effect of a chemical on a biological system (cell, organism, organ etc.).

Epidemiology is the study of the distribution and determinants of disease frequency in human populations.

Exposure is contact of a chemical, physical or biological agent with the outer boundary of an organism, such as by inhalation, ingestion or dermal contact.

Exposure Assessment is the estimation (qualitative or quantitative) of the magnitude, frequency, duration, route and extent (for example, number of organisms) of exposure to one or more contaminated media.

Exposure Pathway is the course a chemical or physical agent takes from a source to an exposed organism. An exposure pathway describes a unique mechanism by which an individual or population is exposed to chemicals or physical agents at or originating from a site. Each exposure pathway includes a source or release from a source, an exposure point, and an exposure route. If the exposure point differs from the source, a transport/exposure medium (eg. air) or media (in cases of intermedia transfer) also is indicated. (US EPA, 1989, p. 62)

Exposure Route is the way a chemical enters an organism after contact eg. by ingestion, inhalation, or dermal absorption (US EPA 1992).

Extrapolation means for dose-response curves, an estimate of the response at a point outside the range of the experimental data. Also refers to the estimation of a response in different species or by different routes than that used in the experimental study of interest.

Factor means a single factor or product of several single factors by which the modified-benchmark dose is divided to derive an acceptable intake. These factors account for adequacy of the study, interspecies extrapolation, inter-individual variability in humans, adequacy of the overall data base, nature and extent of toxicity, public health regulatory concern and scientific uncertainty.

Gene means the functional unit of inheritance.

Genotoxic means the chemical agents for which the primary biological activity is the alteration of the information encoded in genetic material (Butterworth, 1990).

Guidance Values are the values such as concentrations in air or water, which are derived after appropriate allocation of Tolerable Intake (TI) among the possible different media of exposure. Combined exposure from all media at the guidance values over a lifetime would be expected to be without appreciable health risk. The aim of a guidance value is to provide quantitative information from risk assessment for risk managers to enable them to make decisions concerning the protection of human health." (WHO, 1994, p16)

Guideline Dose means the average daily intake of a chemical which, for a life-time, should not result in cancer, based on a comprehensive expert assessment of the best information available at the time. The guideline dose is derived by regulatory authorities using cancer risk assessment according to guidelines developed by national health advisory bodies.

Hazard is the capacity of an agent to produce a particular type of adverse health or environmental effect eg. One hazard associated with benzene is that it can cause leukemia; one hazard associated with DDT is that it can cause the thinning of eggshells of some predatory birds.

Health Investigation Level is the concentration of a contaminant above which further appropriate investigation and evaluation will be required.

Health Risk Assessment is the process of estimating the potential impact of a chemical, biological, physical or social agent on a specified human population system under a specific set of conditions and over a certain timeframe.

Health Risk Management is the process of evaluating alternative actions, selecting options and implementing them in response to health risk assessments. The decision making will incorporate scientific, technological, social, economic and political information. The process requires value judgements eg. on the tolerability and reasonableness of costs.

IRIS (Integrated Risk Information System) is the computerised database of the US EPA which provides the Agency’s adopted hazard and dose-response assessment for chemical and radiological agents. Used as guidance and to provide consistency in the Agency’s regulatory decisions designed to reduce risk related to environmental exposures (see abbreviations).

Life-time covers the average life span of an organism (eg. 70 years for humans).

LED10 (Lowest Effective Dose) means the lower 95% confidence limit on a dose associated with an estimated 10% increased tumour or relevant non-tumour response (US EPA, 1996).

LOAEL (Lowest Observed Adverse Effect Level) is the lowest concentration or amount of a substance, found by experiment or observation, that causes adverse alterations of morphology, functional capacity, growth, development or life span of target organisms.

Metabolite is a substance which is the product of biochemical alteration of the parent compound in an organism.

Metastasis is the transfer of abnormal cells or pathogenic microorganisms from one organ to another in the body.

Model is a mathematical representation of a biological system intended to mimic the behaviour of the real system, allowing description about empirical data and predictions about untested states of the system.

NOAEL (No Observed Adverse Effect Level) is the greatest concentration or amount of a substance, found by experiment or observation, which causes no detectable adverse alteration of morphology, functional capacity, growth, development or life span of the target organism under defined conditions of exposure. Alterations of morphology, functional capacity, growth, development or life span of the target organism may be detected which are judged not to be adverse.

Public Health is the science and art of preventing disease, prolonging life and promoting health through organised efforts of society.

PTWI (Provisional Tolerable Weekly Intake) is the intake of a chemical deemed to be tolerable expressed as a weekly amount. The term was established by WHO (1972) for several heavy metals which 'are able to accumulate within the body at a rate and to an extent determined by the level of intake and by the chemical form of the heavy metal present in food.' (WHO, 1989)

QRA (Quantitative Risk Assessment) is a risk assessment procedure which yields numerical descriptors of risk.

Response Level is the concentration of a contaminant at a specific site, based on a site assessment, for which some form of response is required to provide an adequate margin of safety to protect public health and/or the environment.

RfD (Reference Dose) is an estimate (with uncertainty factors spanning perhaps an order of magnitude) of the daily exposure (mg/kg/day) to the general human population (including sensitive sub-groups) that is likely to be without an appreciable risk of deleterious effects during a life time of exposure. It is derived from the NOAEL or the LOAEL by application of uncertainty factors that reflect various types of data used to estimate RfD and an additional modifying factor, which is based on professional judgement of the entire data base of the chemical (IRIS, 1996).

Risk is the probability in a certain timeframe that an adverse outcome will occur in a person, a group, or an ecological system that is exposed to a particular dose or concentration of a hazardous agent, ie. it depends on both the level of toxicity of the hazardous agent and the level of exposure.

Safety factors are the numerical values used to divide the LOAEL or NOAEL when deriving acceptable intakes and account for the adequacy of the study, interspecies extrapolation, inter-individual variability in humans, adequacy of the overall data base, nature and extent of toxicity, public health regulatory concern and scientific uncertainty. Safety factors usually refer to health-related concerns.

Site means the parcel of land being assessed for contamination.

Structure Activity Relationship is the relationship between the biological activity of chemicals or series of chemicals and their structure. The relationships can be described qualitatively and quantitatively.

Threshold Dose is the lowest dose which produces an effect and below which no biological effect is known to occur.

TDI (Tolerable Daily Intake) is an estimate of the intake of a substance which can occur over a lifetime without appreciable health risk. It is the TI expressed as a daily amount. (Imray and Langley, 1996, p18) It may have different units depending on the route of administration (WHO, 1994).

TI (Tolerable Intake) is an estimate of the intake of a substance that over a lifetime is without appreciable health risk. (WHO, 1994)

Toxicity is the quality or degree of being poisonous or harmful to plant, animal or human life.

Transformation is the process by which a normal cell acquires the capacity for neoplastic or carcinogenic growth. It is thought to occur in several stages.

Tumour is a mass of abnormal, disorganised cells, arising from pre-existing tissue, which is characterised by excessive and uncoordinated cell proliferation or growth and by abnormal differentiation (specialisation). There are two types of tumours, benign and malignant. Benign tumours morphologically resemble their tissue of origin, grow slowly (may also stop growing) and form encapsulated masses; they do not infiltrate other tissues, they do not metastasise and are rarely fatal. Malignant tumours resemble their parent tissue less closely and are composed of increasingly abnormal cells genetically, morphologically and functionally. Most grow rapidly, spread progressively through adjacent tissues and metastasise to distant tissues.

Tumour Initiation is the first step in carcinogenesis whereby a small number of cells (or one cell) are irreversibly changed due to genetic damage.

Tumour Progression is the stage in carcinogenesis when tumours acquire the features of malignant growth.

Tumour Promotion is the process by which initiated cells undergo clonal expansion to form overt tumours.

Uncertainty is the lack of knowledge about the correct value, eg a specific exposure measure or estimate.

Variability relates to measurable factors that differ eg height is variable across populations. The major types of variability are temporal, spatial and interindividual. They may be discrete (eg albinism) or continuous (eg body weight). It may be readily identifiable (eg presence of albinism) or difficult to identify (eg ability to detoxify a particular chemical metabolite).

4. FRAMEWORK FOR RISK ASSESSMENT

There are various models available for the health risk assessment of contaminated sites. The principal forms are those of the US EPA (1989) and the National Academy of Sciences (1983).

Risk assessment is intended "to provide complete information to risk managers, specifically policymakers and regulators, so that the best possible decisions are made" (Paustenbach, 1989, p28).

The process of risk assessment is intended to achieve the following objectives when assessing contaminated sites (US EPA 1989):

The framework of risk assessment involves four stages which are described by US EPA, (1989)1 as follows, and these stages are used to describe the process of health risk assessment in this document:

•          data collection and evaluation of the chemical condition of the site;

The relationships between these stages are detailed in Figure 4-I.

Data collection entails the acquisition and analysis of information about chemicals on a site that may affect human health and which will be the focus for the particular risk assessment.

Toxicity assessment considers:

•          the nature of adverse effects related to the exposure;

1 The National Academy of Sciences model (1993) has four stages generally similar to those of the US EPA: hazard identification, dose-response, exposure assessment and risk assessment.

Both qualitative and quantitative toxicity information is evaluated to estimate the incidence of adverse effects occurring in humans at different exposure levels.

Exposure assessment involves estimating the frequency, extent and duration of exposures in the past, present, and in the future. It also identifies exposed populations and particularly sensitive sub-populations, and potential exposure pathways. Environmental monitoring and predictive fate and transport models can be used to determine the levels of exposure at particular points on the exposure pathways. The contaminant intakes from the various pathways under a range of scenarios can then be estimated.

Given this information, risk characterisation details the nature and potential incidence of effects for the exposure conditions described in the exposure assessment. An integral part of this stage is to evaluate the uncertainties and assumptions in the risk assessment process.

Figure 4-I

Risk assessment model

adapted from US EPA, 1989.

The role of site-specific health risk assessment in the process of the assessment of contaminated sites is outlined in Figure 4-II.

In conducting risk assessments two guiding principles are recommended (US EPA 1995, p2):

The assessors should:

•       generate a credible, objective, realistic, and scientifically balanced analysis;

•       present information on the separate components of the risk assessment; and

The risk assessors should do this without considering issues such as cost, feasibility, or how the scientific analysis might influence the regulatory or site-specific decision.

Figure 4-II

Site-specific Health Risk Assessment in the Assessment of Contaminated Sites

Note:

It is important that the basis of the decision-making is clearly documented.

5. DATA COLLECTION

5.1 Introduction

Data collection entails the acquisition and analysis of information about chemicals on a site that may affect human health and which will be the focus for the particular risk assessment (US EPA, 1989). The purpose of data collection is to gather data that will enable a useful risk assessment to be undertaken eg data related to contaminant types and distributions and the potential environmental behaviour of the contaminants.

Data Collection is detailed in Schedule B(2).

Adequate data collection is the foundation to an acceptable health risk assessment.

The data collection phase will comprise the following components as described in Schedule B(2).

1.       Setting Data Quality Objectives (see Section 2.3)

2.       Establishing a Site History (see Section 2.4)

3.       Detailing the Proposed Use (see Section 2.2)

4.       Establishing a Sampling strategy and sampling pattern (see Sections 3.1 – 3.5)

5.       Ensuring appropriate analysis (Choice of Analytes) (see Sections 3.6)

Before any sampling is undertaken, the objectives of the task should be defined

The greatest concern, in collecting soil and water samples, is to ensure that the samples taken represent all the waters, and the soils in all strata present on the site. Consequently it is essential to be fully apprised of the conditions at the site locations and what analytes will be tested in each sample, before sampling commences. (Lock 1996, p2)

6. EXPOSURE ASSESSMENT

6.1 Introduction

Exposure assessment involves the estimation of the magnitude, frequency, extent and duration of exposures in the past, currently, and in the future. It also identifies exposed populations, and particularly sensitive sub-populations, and exposure pathways.

The process involves:

•          Analysis of contaminant releases;

•          Identification of exposed populations;

•          Identification of potential exposure pathways;

•          Estimation of exposure concentrations for each pathway; and

•          Estimation of contaminant intakes for each pathway for a range of scenarios.

Direct measurement of the exposures of the (potentially) affected population provides the best exposure data but this is not always available or practicable and default exposure factor data is often required.

(Langley 1993, p90)

The following issues are relevant to exposure assessment

•          Information to assist in interpretation of results (Section 6.4 below);

•          Exposure comparisons to values such as ADIs (Section 6.5 below);

•          Health monitoring (Section 6.9 below);

•          Default values to be used in the absence of site specific data (Section 6.14 below);

•          Further sources of exposure data (Section 6.15 below).

Accurate and useful exposure assessment requires a detailed understanding both of the strengths and weaknesses of the exposure assessment techniques, and the specific exposure factors used in the assessment.

6.2 Key Points in Exposure Assessments of Contaminated Sites

3.       One exposure route will normally predominate;

(ECETOC, 1990; Langley, 1991)

6.3 Use of Point Estimates and Probability Distributions

6.3.1 Introduction

Point estimates are most commonly used in Australia for exposure assessments. A point estimate is a single value chosen to represent a population eg 70kg as the weight of an adult. Point estimates are usually typical values for a population or an estimate of an upper end of the population's value eg 70 years as the duration of residence on a property. An upper end value may be chosen for reasons of conservatism and/or to provide a "worse case" scenario.

Where a risk assessment uses a series of upper end estimates, the result can be a worse than worse case scenario due to the compounding effects of the estimates eg the person with the upper end value for weight is unlikely to have; the upper end value for water consumption that has the upper end value for contamination; the upper end value for duration of residence; the upper end value for soil ingestion, etc.

In recent years there has been increasing attention paid to the use of Monte Carlo-type exposure assessments and such methods have been acknowledged by the US EPA and the UK Department of the Environment (US EPA, 1992a; Ferguson et al, 1994).

These methods are 'more informative and inherently more representative' (Ruffle et al, 1994, p 403) than point estimates. If applied appropriately point estimates still have a major role in exposure assessment as they are readily understood and

applied, and may incorporate safety factors that could be lost with Monte Carlo-type exposure assessments.

The Monte Carlo-type exposure assessments rely on the use of probability distribution functions. A distribution of possible values for each of the parameters (is) described …. along with the probability of occurrence of each. Using standard mathematical formulae several thousand iterations of a mock mathematical model are performed. For each iteration, values for each parameter are selected randomly from each distribution based upon the probability of occurrence. The estimated risk values are combined to provide a frequency distribution of possible risk (Alsop et al., 1993). Figure 6-I demonstrates the process of the Monte Carlo method. The end result is a more representative picture of the range of exposures, and hence risks, in a population than occurs with the use of point estimates.

If a Monte Carlo assessment is performed the methodology must be 'transparent' or problems will arise in community consultation. As with any form of risk assessment, the basic principles of the method must be able to be understood by the affected community.

On a small site, the use of Monte Carlo methods is likely to be to complex and/or costly and it may be more appropriate to do direct measurements of exposure. The exposures of 'outliers' must always be acknowledged in risk assessments and ways of identifying and accommodating them must be considered.

This is particularly important in the assessment of an existing site contamination (eg. a site where housing has already been developed), rather than a forecast exposure scenario, where the presence of outliers will severely affect the credibility in any risk communication exercise. (Langley and Sabordo 1996)

Figure 6-I

Principles of the Monte Carlo Method

 (adapated from Ferguson 1994))

6.3.2 Weaknesses With the Monte Carlo Technique

Langley and Sabordo (1996) reviewed the key limitations of the Monte Carlo technique. The limitations include:

6.3.2.1 Complexity

While the Monte Carlo method has a very general applicability changing one variable may mean large amounts of re-calculation because of the extent of the iterative process when using this model. The complexity reduces the 'transparency' of the method. This may create difficulties in community consultation and risk communication, it obscures errors, and creates difficulties for checking by both the modellers and administering authorities.

6.3.2.2 Loss of factor distinctions

The method does not indicate which variables are the most important contributors to output uncertainty. (US EPA, 1992).

6.3.2.3 Unrealistic probability assessments

US EPA (1992) notes that simulations such as that found with the Monte Carlo model often "include low probability estimates at the upper end that are higher than those actually experienced in a given population, due to improbability of finding these exposures or doses in a specific population of limited size, or due to nonobvious correlations among parameters at the high ends of their ranges". This results in overestimations of exposure dose or risk. The Science Advisory Board of the US EPA has noted that "for large populations, simulated exposures, doses and risks above the

99.9 percentile may not be meaningful when unbounded lognormal distributions are used as a default". (ibid, p22922)

6.3.2.4 Assessment endpoints

With Monte Carlo-type assessments there is still a need to determine what is an acceptable level of exposure. Smith (1994) considers that 'the level of exposure exceeded by 1 in 20 exposed persons would seem to be an appropriate reasonable maximum'. This would allow 5% of the population not to be included in the exposure assessment.

6.3.2.5 Variability-uncertainty confusion

Smith (1994) highlights the need to distinguish between 'variability' (measurable factors that differ across populations such as height) and 'uncertainty' (unknown, difficult to measure factors such as frequency of trespassing on a site). Currently available software packages do not distinguish between variability and uncertainty. An administrator reviewing a Monte Carlo risk assessment will, however, need to appreciate the differences between variability and uncertainty and the nature and extent of both.

6.3.2.6 Limited exposure data

Limited information is available about many variables for the exposure assessments. As a consequence of this, many input variables are described as triangular distributions. Smith (1994) stresses the need 'to collect and verify distributions from many currently undescribed input assumptions' to improve accuracy.

The use of Monte Carlo methods may be inappropriate where the predictions of exposure are so dominated by uncertainties. McKone (1994) gives the example of benzo(a)pyrene, where information on benzo(a)pyrene exposure is 'not readily available' so that the use of Monte Carlo methods to assess variability in population exposures is somewhat redundant.

6.3.2.7 Simplification of complex situations

Exposure assessments are comprised of combinations of modelling, sampling, and modelling/sampling combinations. Even the use of complex models still provides a static picture of a dynamic world albeit a more elaborate representation of reality (McKone, 1994) and such a picture must be placed within a sound theoretical framework.

6.3.2.8 Misleading precision

The use of more complex models 'does not necessarily increase precision'. The costs of collecting and analysing data, and constructing new models 'must be balanced by the value of the information obtained' . There is a need to appraise the value of information along with its uncertainties in 'defining the capabilities and limits of exposure models'. (McKone, 1994)

6.3.3 Characterisation of Extreme Values

The 50th percentile can always be estimated with less uncertainty than the 99th percentile (Finley et al, 1994). Problems in estimating the extreme percentiles can come from limitations in the measurement techniques (eg incorrect and implausible estimates of dietary consumption may be accepted into the survey); the duration over which exposure data was collected (see short term and long term variation, above); and whether there are sub-populations who may have unusual exposures (eg. vegetarians, subsistence fishermen) (Finley et al, 1994)

6.3.4 Selecting Appropriate Data Sets

For describing a probability distribution, the relevant studies and the quality of the data produced may vary considerably. Unless data sets are rigorously scrutinised 'the resulting uncertainty in the range of risk estimates could be greater than obtained using point estimates' (Finley et al, 1994, p 536)

6.3.5 Principles For The Use Of Monte Carlo-Type Techniques

Burmaster and Anderson (1994) stress that any method of exposure assessment must have a 'clearly defined assessment end point' and provide all relevant information so that the assessment can be reproduced and evaluated. They detail fourteen principles for good practice in Monte Carlo assessments. These are:

1.       Detail all formulae.

5.       Provide detailed information about input distributions with the minimum being:

Burmaster and Anderson state that the principles are not mutually exclusive nor collectively exhaustive."

6.3.6 Administrative Requirements for the Use of Monte Carlo Methods

Regulatory authorities in Australia will require assessments using Monte Carlo methods to meet the following criteria:

•          Meeting the 14 principles of Good Practice detailed above.

On a large site divided into housing lots, the results for specific housing lots that may be affected by atypically elevated concentrations should not be obscured by averaging or Monte Carlo techniques applied to the entire site. In many instances, Monte Carlo methods will only be relevant to large sites or sites where direct measurements of exposure are not practicable.

Government authorities will need to define the range of acceptable exposures. Given that the Monte Carlo method loses much of the conservatism usually inherent in point estimates and hence the safety factors, it is proposed that the 99th percentile of exposures for particular groups (eg. by age: young children; children; and adults, and by susceptibility eg. asthmatics) be chosen. Depending on the conservatism of the toxicological assessment, this should result in adequate protection for at least 99% of the population. It should be noted that for a population of 1 million, 10 000 people would exceed the 99th percentile for exposure." (Langley and Sabordo 1996, p141)

6.4 Appraising Exposure Assessments

Factors that tend to result in underestimates of exposure (EPA, 1992):

•          Overlooking a significant exposure or metabolic pathway.

•          Failure to evaluate all contaminants of concern in the mixture.

Factors which can cause overestimates of exposure include (EPA, 1992):

•          The use of unrealistically conservative exposure parameters.

•          Portraying hypothetical potential exposures as existing exposures.

Factors that may cause underestimates or overestimates include (EPA, 1992):

•          Computational errors.

•          Use of inappropriate factors eg. for intake routes.

•          Insufficient uncertainty assessment to put the exposure assessment in perspective.

6.5 Exposure Durations and Exceedances of Acceptable Daily Intakes (ADIs)

Appropriate durations of exposure need to be assessed so that transient (short term) and important exposures are not obscured by the use, for example, of average lifetime exposures. This is important in the Australian context where Acceptable Daily Intake values from WHO have been used to establish Health-based Investigation Levels. The duration and magnitude of exceedances of the ADIs must be obvious in exposure assessments.

6.6 Biological Monitoring

Biological monitoring (based on Langley, 1991a) is a measuring procedure whereby validated indicators of the uptake of contaminants, or their metabolites, and people's individual responses are determined and interpreted. Whereas environmental monitoring measures the composition of the external environment around a person, biological monitoring measures the amount of contaminant absorbed into the body.

If biological monitoring is practicable it will be more valuable than environmental monitoring in determining the level of risk from an environment as it will measure whether exposure is occurring and the level of exposure.

The biological samples used for monitoring include: blood, urine, fat, hair, and expired air.

Biological monitoring should not be commenced before:

•          The objective of the biological monitoring is defined clearly.

Results should always be available to participants in biological monitoring with an explanation of the results.

Several aspects must be considered:

6.7 Choice of a Test

Optimally, a biological monitoring test (based on Langley, 1991) would give a result which reflected the exposure, the concentration of the substance in the target organ and the risks of adverse effects (Friberg, 1985). Few tests are available which approach this ideal (Langley et al 1998).

In Australia, exposures from contaminated soil will be generally low, creating problems in accurate measurement at low levels and the possibility of results being overwhelmingly influenced by other sources of exposure (eg. the influence of cadmium in food, tobacco smoke and the occupational environment will generally be far greater than the influence of cadmium contamination of soils).

For many substances, biological monitoring is impracticable because:

Hair is an inappropriate tissue for biological monitoring on or near contaminated sites. External contamination of the hair cannot be adequately removed during sample preparation and an accurate measure of excretion via hair cannot be performed. Hair analysis may be useful for assessing intake from purely dietary sources when there is no general environmental contamination.

Substances for which biological monitoring of general environmental exposures is practicable are detailed in Table 6-A.

Table 6-A

Substances likely to be suitable for biological monitoring

Adapted from Langley (1991a)

Most organic contaminants are not amenable to biological monitoring in general environmental situations because of the low levels of exposure and the lack of comparison data compared to occupational situations. Specialised studies may make biological monitoring of some inorganic substances practicable (eg. manganese, radioactive isotopes).

A good knowledge of the toxicokinetics of a substance is required for the correct choice of method and interpretation of results eg. individual results may be distorted if there is not constant exposure or equilibrium within the body.

Under the NOHSC National Model Regulations for the Control of Workplace Hazardous Substances (adopted by the States and Territories), health surveillance is required for specified substances. Biological monitoring methods developed for some of these methods are detailed in the NOHSC Guidelines for Health Surveillance.

6.8 Biomarkers

The term 'biomarker' has been introduced recently and refers broadly to "almost any measurement reflecting an interaction between a biological system and an environmental agent, which may be chemical, physical or biological" (WHO, 1993). Three classes of biomarker are identified by WHO (1993, p12):

Some examples of commonly used biomarkers are serum cholinesterase for organophosphate exposure and serum enzymes for liver damage.

6.9 Health Monitoring

From Australian and international experience, health effects are likely to be found in only a very limited number of situations of extreme soil contamination. Subtle effects may only be able to be determined on a group basis rather than on an individual basis (eg. subtle neurodevelopmental effects determined by sophisticated testing in groups of children with different lead exposures). Similar problems of causation relating to individual findings rather than group findings arise if the putative effects are common in the general population eg. headache, fatigue. Health effects are rarely as specific to an exposure as chloracne with PCB or dioxin exposure, mesothelioma and asbestos exposure, and vinyl chloride monomer and haemangiosarcoma.

Health monitoring for specific health effects is warranted where environmental or biological monitoring has indicated a significant risk of effects eg. specific tests of renal function if urinary cadmium levels above the levels of concern are detected in biological monitoring.

When health monitoring is done it should rarely be done in isolation from environmental and/or biological monitoring. Clearly defined health effects should be sought with specific case-definition criteria. Records of other symptoms and clinical findings should also be kept to enable epidemiological assessment of other potential health effects. (Langley 1991a, p195)

6.10 Default Exposure Settings

Taylor and Langley (1998) Exposure scenarios and exposure setting details the derivation of default exposure settings and the qualifications for their use and these are provided in Schedule B(7B). It includes 'default exposure ratios' which are standard multiplication factors which can be applied to investigation levels for each setting to take into account expected differences in levels of exposure.

Table 6-B

Exposure Settings and Default Exposure Ratios for establishment of soil investigation criteria

(from Taylor and Langley 1998, p14)

Notes to Table 6-A:

7.     National Occupational Exposure Standards have been developed with an undefined career duration.

"…Day-care centres and preschools (and primary schools to a lesser extent) potentially provide situations which are comparable to residential dwellings in terms of soil access by young children, and can be placed in the 'base case' residential setting.

Whilst inclusion of primary school sites in a 'residential' category may be seen as overly conservative in view of diminished mouthing behaviour and soil ingestion expected in this age group compared with infants and toddlers, primary school sites have been included in this category because some contain preschool or child-care centres; some contain special education units where children may be at increased risk of hand-mouth or pica behaviours; and social and community considerations about 'acceptable risk' have been taken into account in the regulatory framework. It is acknowledged that exposures in primary schools may be similar to exposures in secondary schools. If well-maintained barriers to soil access exist (eg in the form of paving such as cobblestones, gravel, or a substantial pine bark ground covering) then a primary school setting may not be comparable to a standard low-density 'residential with backyard garden' setting but more akin to high-density residential land use with reduced opportunities for soil access.

Similarly, a residence where the yard space is fully and permanently paved (eg concrete), or the contaminated soil is fully and permanently contained, affords minimal opportunities for contaminated soil access and investigation levels may be more appropriately considered in the context of a separate, lower-risk category.

A residential setting with accessible soil but minimal or negligible home food production has usually formed the baseline case for development of investigation levels to date in this country, but this approach has not explicitly quantified the home food production pathway.

It is advisable to distinguish those households with free-range poultry as special cases since this pathway may significantly influence exposure levels (Cross and Taylor 1994,1996). The great majority of urban local governments in recent times either prohibit poultry-keeping altogether, or require poultry to be kept on concrete pads where they remain out of contact with soil. If free-range poultry are being kept on contaminated soil then site-specific sampling of produce is likely to be the best means of determining exposure and the level of risk." (Taylor and Langley, 1998, p11)

6.11 Variation from Default Exposure Settings

Taylor and Langley (1998, p16) state: " For default assumptions not to be used, realistic and appropriately inclusive exposure opportunities for the proposed land use would need to be detailed, with sufficient safeguards for other potential future exposures. This may require annotations on the title documents or elsewhere stating the constraints on other possible land uses. The alternative exposure scenarios would have to differ markedly (possibly by an order of magnitude) from the defaults proposed here, in order for them to be used in preference to the defaults in establishing site-specific soil criteria."

A degree of conservatism is built into the default exposure settings as these relate to generic Health-based Investigation Levels that must provide for a wide range of scenarios within each default setting. To deal with uncertainties, some conservatism should remain when setting site-specific Response Levels, although it is anticipated that site-specific Response Levels will more closely reflect the site-specific exposure assessment.

6.12 Exposure Assessment of Volatile Contaminants

Volatile contaminants require specialised sampling techniques to ensure that the contaminants are not lost during and after sampling so that analytical results accurately represent the concentrations present on a site. The inhalation route will be more important than for non-volatile contaminants. It is often impractical to undertake environmental (ie air) sampling because of the constant variations over time of the concentrations as a result of fluctuation in temperature, wind speed and direction. Other factors that will have a significant effect are: soil disturbance; the physico-chemical properties of the soil and contaminants; and whether there is a renewable source or whether the contamination will dissipate over time. Exposure assessment will often depend on modelling. Models relevant to Australia are being developed by CSIRO in conjunction with the Environment Protection Authority (NSW) and the Public & Environmental Health Service in Adelaide. Preliminary details are provided in: Anderssen and Markey (1996); Anderssen and Markey (1997); Anderssen, de Hoog and Markey (1997); and, Anderssen and Markey (1998). Another methodology, “Guidelines for the Management of Hydrocarbon Impacted Land” is being developed by the Australian Institute of Petroleum. A process for the appraisal of the methodologies and determination of soil criteria may be considered as part of a future work plan that may arise from the Measure.

6.13 Exposure Assessment Reports

The following checklist details matters that should be appropriately addressed in an exposure assessment. Some material may be omitted, if justification can be provided. It is adapted from US EPA (1995):

1. What are the most significant sources of environmental exposure?

•          What are the most-significant environmental pathways for exposure?

4.       What are the key descriptors of exposure?

•          How was the high-end estimate developed?

6.       Summarise exposure conclusions and discuss the following-,

•          limitations of each, and the range-of most reasonable values; and

•          confidence in the results obtained, and the limitations to the results."

6.14 Default Values for Exposure Assessments

The following default values have been used in exposure models since 1991 to derive Health-based Soil Investigation Levels. They are adapted from Langley and Sabordo (1996, p184). These values should be used unless values more pertinent to the relevant population can be provided and justified. Factors should be relevant to the population about whom the exposure assessment is being done.

6.14.1 Dermal absorption factors

•          The child will wash once each day.

•          Australian washing/bathing values are to be used where available.

6.14.2 Inhalation factors

•          Australian dust values are to be used where available.

Inhalation factors for particulates are dealt with in the NOHSC Exposure Standard Guidelines which are required by law for assessing occupational exposures.

6.14.3 Ingestion factors

•          Soil ingestion rates will be:

*conservative estimates from ANZECC/NHMRC(1992)

6.15 Sources of Exposure Assessment Data

Data must be pertinent to the relevant population. Where available, data from Australian populations are preferred.

Sources of information and data include:

Langley AJ. (1993). 'Refining exposure assessment.'

Langley AJ and Sabordo L (1996) Exposure Factors in Risk Assessment

Langley AJ, Taylor A and Dal Grande E (1998) 1996 Australian Exposure Factors

The Australian Bureau of Statistics can provide a range of Australian data.

The American Industrial Health Council's 'Exposure Factors Sourcebook' (1994) provides examples of probability distributions for a range of exposure factors. These largely relate to the US population. These, and similar US-based data, should only be used if they can be demonstrated to be relevant to the Australian population.

7. TOXICITY ASSESSMENT

7.1 Introduction

Toxicity assessment considers:

• the nature of adverse effects related to the exposure;

•          the dose-response relationship for various effects;

•          the weight of evidence for effects such as carcinogenicity; and

•          the relevance of animal data to humans.

Both qualitative and quantitative toxicity information is evaluated in assessing the incidence of adverse effects occurring in humans at different exposure levels (US EPA, 1989).

There are two elements to the toxicological assessment: hazard identification and dose-response assessment.

Hazard identification examines the capacity of an agent to cause adverse health effects in humans and other animals. It is a essentially qualitative description based on the type and quality of the data, complementary information (eg structure-activity analysis, genetic toxicity, pharmacokinetic), and the weight of evidence from these various sources (US EPA, 1995). Key issues include (ibid):

•          nature, reliability and consistency of human and animal studies

•          the available information on the mechanisms of toxic effect; and

•          the relevance of the animal studies to humans.

The dose-response assessment examines the quantitative relationships between exposure and the effects of concern. The determination of whether there is a hazard is often dependent on whether a dose-response relationship is present. Key issues include (US EPA, 1995):

•          the potential for differing susceptibilities in population subgroups.

The Toxicity Assessment phase is similar to the Hazard Identification and Dose Response stages in the National Academy of Sciences (1983) Risk Assessment model and other models.

The matters covered in the toxicity assessment phase of this document are:

•          the components of a toxicological appraisal (Section 7.2 below)

•          the sources and ranking of toxicological assessment data (Section 7.4 below)

7.2 Toxicological Appraisals

The following checklist is adapted with slight modification from US EPA (1995) and should be the basis of toxicological appraisals. A summarised version can be used if Tolerable Intake data from WHO or NHMRC are used as these matters should have been considered when the Tolerable Intake was set.

7.2.1 Hazard Identification

•          How good is the key study?

•          Describe other studies that support this finding.

•          Discuss any valid studies which conflict with this finding.

As many relevant studies as possible should be collated and rigorously assessed as to their strengths and weaknesses to determine the key studies. This is particularly important where quantitative risk estimates will be undertaken or where there are apparently contradictory studies; in the latter case, the studies that are considered to be adequate in their design and interpretation will need to be appraised to determine the overall weight-of-evidence.

• What are the significant data gaps?

3. Discuss available epidemiological or clinical data. For epidemiological studies:

•          What types of studies were used, ie, ecologic, case-control, cohort?

•          Describe the degree to which exposures were adequately described.

•          Describe the degree to which other causal factors were excluded.

In assessing the relationship between a possible cause and an outcome, there must be a careful appraisal of whether the results could be explained by selection or measurement bias, confounding or chance (Beaglehole et al 1993). Particularly rigorous scrutiny should be given to studies giving a positive but not statistically significant result. A systematic and sequential approach to determining the nature of an association is detailed in ‘Guidelines for causation’ (Beaglehole et al 1993, p 76):

•          Discuss relevant studies of mechanisms- of- action or metabolism.

•          Does this information aid in the interpretation of the toxicity data?

•          What are the implications for potential health effects?

6.       Summarise the hazard identification and discuss the significance of the following:

•          confidence in conclusions;

•          alternative conclusions that are also supported by the data;

•          significant data gaps; and

•          highlights of major assumptions.

7.2.2 Characterisation of Dose-Response

If animal data were used:

•          Which species were used: most sensitive, average of all species, or other?

•          Were any studies excluded? Why?

•          If epidemiological data were used:

For non-carcinogenic hazards:

•          How was the Tolerable Intake (or the acceptable range) calculated?;

•          What assumptions or uncertainty factors were used?

•          What is the confidence in the estimates?

•          For carcinogenic hazards:

7.3 Mixtures

Currently there is no agreed Australian approach to assessing mixtures of contaminants. Where data (including mechanistic data) is available on the interaction of contaminants this should be taken into account in the assessment of a site. For more discussion on mixtures, see Schedule B(7A), Section 14.

A process for the appraisal of mixtures has been proposed as part of the future work plan that will arise from the Measure.

7.4 Sources of Toxicological Assessment Data

The following categories are given in order of preference. All documents, particularly those in the second and third categories require rigorous appraisal for relevance, validity and accuracy.

7.4.1 Principal Sources:

•          International Agency for Research on Cancer (IARC) documents

•          NICNAS Priority Existing Chemical (PEC) reports

•          IPCS Concise Information Chemical Assessment Documents (CICAD)

•          OECD Standard Information Data Sets (SIDS)

7.4.2 Secondary Sources

7.4.2.1 Peer-reviewed journals

These may not provide opinions that meet general scientific agreement. With justification, and acceptance by the local jurisdiction, they may be suitable for use. Examples are:

• Risk Analysis

•          Regulatory Toxicology and Pharmacology

•          Human and Ecological Risk Assessment

•          Food and Chemical Toxicology

•          Carcinogenesis

•          Environmental Health Perspectives

•          Scandinavian Journal of Work, Environment and Health

•          Journal of Occupational and Environmental Medicine

7.4.2.2 Industry Publications

With justification, and acceptance by the local jurisdiction, they may be suitable for use:

•          Chemical Industry Institute of Toxicology (CIIT) reports

7.4.2.3 Occupational Health & Safety Sources

These may be a useful source for toxicological data and reviews but occupational exposure criteria must not be used in a general public health context without appropriate adjustment for the different durations of exposure, the inclusion of susceptible subpopulation in the general community (eg children) and the methodological differences in the setting of criteria.

7.4.3 Tertiary Sources

The use of this information requires justification that no other sources are available and an appraisal of the methodology detailing the level of conservatism and range of uncertainties inherent in the approach.

7.5 Methods for Risk Assessment of Carcinogens

The material in 7.5.1 and 7.5.2 is drawn, with amendment, from the NHMRC 'Draft Cancer Risk Assessment for Environmental Contaminants' (1997, pp1-16) which was prepared by a Technical Working Party (TWP). It has been reviewed and changed following a process of public consultation. The redrafted version is titled 'Toxicity Assessment Guidelines for Carcinogenic Soil Contaminants'. It was endorsed by NHMRC in September 1998.

The methodology was presented at the Fourth National Workshop on Health Risk Assessment and Management of Contaminated Sites held in Brisbane, October 1996 (Langley et al, 1998) and was considered by the Joint ANZECC/NHMRC Contaminated Sites Technical Review Committee also in October 1996. Both the

Workshop and the Joint ANZECC/NHMRC Contaminated Sites Technical Review Committee endorsed the approach taken by the TWP and supported the further development of the methodology.

This approach is consistent with other international risk assessment methodologies. The development and use of an agent-specific Guideline Dose is consistent with current risk assessment practice in Australia as well as with international practice. For example, the Guideline Dose and its use in risk assessment is analogous to the Acceptable Daily Intake (ADI) and the US Reference Dose (RfD). In Australia, due to variations in the circumstances of use and the nature of the chemicals being regulated, a range of different methodologies are used by different government agencies

7.5.1 Background

A variety of risk assessment methods has been used elsewhere, for example by the United States Environmental Protection Agency (US EPA, 1986), and the World Health Organisation (WHO, 1993).

As biological advances, mechanistic data, pharmacokinetic data and other relevant data are increasingly being taken into account in classifying and assessing the risks of carcinogens.

The TWP believed that existing methodologies had difficulties in conveying the health implications of exposure to environmental pollutants. The result in many cases has been an inequity of regulatory, political and public attention between cancer and other-than-cancer health effects.

The traditional benchmark dose methodology (traditional BMD) has been developed over the last two decades and is now being given serious consideration as a useful tool in risk assessment (Dourson, 1984; Barnes et al., 1995; US EPA, 1995a; 1996). More recently, the approach has also been proposed for cancer risk assessment (eg US EPA, 1996)

The TWP sought to develop a methodology for use in Australia which avoided some of the limitations inherent in existing cancer risk assessment methods. In particular, the methodology optimises the advantage of using all relevant scientific data in the decision-making process and provides for a clear separation and justification of the major components of the process: public health policy, professional judgement and scientific principles and data.

The methodology is a two step process. Firstly, the modified-BMD is derived from the experimental data. Secondly, the modified-BMD is divided by cumulative factors to derive a Guideline Dose for human exposure. The steps are outlined in Figure 7-I.

The modified-BMD is set using 5% extra risk determined from animal or epidemiological studies. After consideration of all the available toxicological data, this extra risk is then divided by a series of modifying factors (potentially up to 50,000) according to a specified decision tree to derive an agent specific Guideline

Dose protective of public health. These factors relate to inter- and intraspecies variation, quality of the data base and other factors for the seriousness of the carcinogenic response. The factors are derived using a decision tree which takes into account all of the available data; Scientific judgement is used to address a number of the uncertainties in the risk assessment process and in the development of safety factors.

The TWP supported the use of all available, relevant information in the risk assessment process. In cases where there are few or inadequate data, conservatism may be justified and the use of conservative (default) assumptions was supported. Recommendations on default assumptions are provided for cases where the data are incomplete to bridge data gaps and allow the risk assessment to proceed. All choices, both those based on scientific data and those based on default assumptions, must be supported by reasoned and critical analytical arguments.

The Guideline Dose is established by regulatory authorities and is defined as the daily intake of a chemical agent which, during a life time, is unlikely to result in cancer, based on a comprehensive expert assessment of the best information available at the time. It is considered that the Guideline Dose is protective of public health.

The Guideline Dose may be used in the development of health investigation levels, response levels and risk characterisation of human exposures to contaminants in soil.

The Guideline Dose does not attempt to model or predict a response incidence at low environmental exposure. It is an estimate of the dose which is considered protective of public health (the compounding use of factors assures a high level of safety). This places the focus of regulation on the control of exposure to environmental contaminants rather than calculation or discussions of risk. This approach has the added benefit of allowing comparisons with guidance values based on non-cancer health effects for chemical agents.

The TWP did not recommend a numerical value which would constitute an acceptable level of risk for low-level environmental exposure to carcinogens. Whilst there has been considerable debate over the last twenty years about what constitutes an acceptable risk, there is no agreed position internationally on this issue (see Department of the Environment, 1993).

Key points about the methodology are:

modified-BMD which are determined are not greatly influenced by the mathematical model chosen. Therefore, different models can be fitted to the data with similar goodness of fit. In contrast, extrapolation well below the experimental range by other quantitative risk assessment methods is very much model dependent and results are highly variable with different models (Maynard et al., 1995).

The methodology can be compared to non-threshold models currently in use which assume low dose linearity (eg the US EPA methodology). The non-threshold models are inflexible and generally do not take account of the complexities of the events between exposure to an agent and the induction of a neoplasm. Risks estimated at doses below the range of experimental data can vary considerably depending on the model used, even though the various mathematical models used generally fit the experimental data equally well (Crump, 1984; Paustenbach, 1995). The numerical expression of the calculated level of risk falsely gives the impression that it represents an exact measure of actual risk. This numerical expression provides little or no information on the uncertainties related to the calculated level of risk, nor does it allow comparison with values for non-cancer health effects. Low-dose linearity assumes a positive slope of the dose-response curve at zero dose and implies that a single, irreversible genetic event at the initiation stage of carcinogenesis leading to transformation of a cell, is sufficient by itself to lead to the development of cancer. The major difficulty in this debate is the impossibility of testing experimentally the shape of the dose-response curve at extremely low doses (Purchase and Auton, 1995).

7.5.2 General principles of the methodology

1.       Identify the relevant soil contaminants.

If no ADI, PTWI or TDI is available, proceed as follows: (Note: Appendices refer to NHMRC (1999))

8.       Use route to route extrapolation where appropriate (Appendix E).

12.    Write the report (Appendix G).

Figure 7-I

Decision Tree for Cancer Risk Assessment

from NHMRC (1999)

Note: Appendices refer to NHMRC (1999)

7.5.3 Further actions

Guidelines are being developed because of deficiencies in current methodologies. Efforts are being made to develop and instigate guidelines and to provide Guideline Doses for specific chemicals. As an interim measure, advice on specific chemicals should be sought from the relevant regulatory body. When probabilistic estimates of risk are the only guidance available, there needs to be a full appreciation of the differences between ‘real’, ‘calculated’ and ‘perceived’ risk. The ‘real risk’ is the actual risk. The objective of risk assessment is to calculate as closely as possible the ‘real risk’. All participants in the process (different experts, different members of the community) will have different perceptions of the nature and magnitude of risk and these perceptions will change as circumstances change.

8. RISK CHARACTERISATION

8.1 Introduction

Risk characterisation is the final step in the risk assessment process that:

The final risk characterisation is rarely accurately quantitative because of the limitations of the data which will be reflected in the uncertainty assessment. The process requires considerable expertise. If data are collected and analysed according to the principles and guidelines in this document the process will become more transparent and consistent. Some parts of the risk assessment process such as 'data collection' and 'exposure assessment' will be, at least in part, quantifiable. These guidelines are intended to assist the qualitative process of determining whether remediation is required or not for the proposed use. Due to the complexities of the matter, the risk characterisation process cannot be reduced to a 'cookbook'.

8.2 Key Principles in Health Risk Characterisation

There are a number of principles which form the basis for a risk characterisation:

Risk assessments should be transparent, in that the conclusions drawn from the science are identified separately from policy judgements, and the use of default values or-methods and the use of assumptions in the risk assessment are clearly articulated.

describe the likelihood of harm. The summary should include a description of the overall strengths and the limitations (including uncertainties) of the assessment and conclusions.

(dotpoints 5-10 adapted from EPA NSW, 1998)

US EPA (1995) provides a common format to assist risk managers in evaluating and using risk characterisation.

The outline has two parts. The first part requires summaries of the major components of the risk assessment leading up to the risk characterisation.

The second part draws all of the information together to characterise risk. The outline described below represents the expected findings for a typical complete chemical assessment for a single chemical. Exceptions due to the circumstances of individual assessments such as particular statutory requirements, resource limitations, and other specific factors should be explained as part of the risk characterisation.

Minor variations in its application from one instance to another are appropriate and expected and are not a legitimate basis for delaying or complicating action on otherwise satisfactory scientific, technical, and regulatory products. (US EPA, 1995)

8.4 Risk Conclusions

A modified version of the US EPA (1995) framework for risk conclusions is a useful model to be used in Australia

8.4.1 Risk Conclusions

3.       What are the major limitations and uncertainties in the three main analyses?

4.       What are the science policy options in each of the three major analyses?

5.       What alternative approaches have been evaluated?

6.       What are the reasons for the choices made?

8.4.2 Risk Context

2.       What are the alternatives to this hazard? How do the risks compare?

3. How does this risk compare to other risks?

•          Describe the limitations of making these comparisons.

8.4.3 Existing Risk Information

Comment on other risk assessments that have been done on this chemical by State, Territory or federal agencies, or other organisations. Are there significantly different conclusions that merit discussion?

8.4.4 Other Information

Is there other information that would be useful to the risk manager, or the public in this situation that has not been described above?

8.5 Uncertainty

Uncertainty analysis must be addressed for each step of the risk assessment and for its cumulative effect from all of the steps.

The assessment of uncertainty is a critical part of the risk assessment process. Uncertainty characterisation is an essentially qualitative process relating to the selection and rejection of specific data, estimates, scenarios, etc (US EPA, 1992). Uncertainty assessment can be more quantitative and it may be represented by more simple measures such as ranges, simple analytical methods such as sensitivity analysis and may progress to complex measures and techniques (Langley, 1993).

Uncertainty (ie. the lack of knowledge about the correct value, for example a specific exposure measure or estimate) must be distinguished from variability (ie. different levels of exposure experienced by different individuals).

Uncertainty may need to be addressed by the collection of further data.

Further detail on uncertainty analysis is available in Vic EPA (1997a).

Table 8-A gives an example of an uncertainty table.

Table 8-A

Example of an uncertainty table for exposure assessment

2 As a general guideline, assumptions marked as "low", may affect estimates of exposure by less than one order of magnitude; assumptions marked "moderate" may affect estimates of exposure by between one and two orders of magnitude; and assumptions marked "high" may affect estimates of exposure by more than two orders of magnitude. (adapted from US EPA, 1989a, p6-51)

Reasons for addressing uncertainties in assessments include (US EPA, 1992 with slight amendment):

•          The combination of uncertain information from various sources.

•          As a means of highlighting biases that may have crept into the process.

8.6 Exceedances of Acceptable Daily Intakes (ADIs)

Renwick and Walker (1993, p 464) provide two excerpts from WHO documents relating to exceedances of ADIs:

'Because in most cases, data are extrapolated from life-time animal studies, the ADI relates to life-time use and provides a margin of safety large enough for toxicologists not to be particularly concerned about short-term use at exposure levels exceeding the ADI, providing the average intake over longer periods does not exceed it.' (WHO, 1987)

Further information was provided in a report in 1989 which added that:

It is impossible to make generalisations concerning the length of time during which intakes in excess of the PTWI (provisional tolerable weekly intake) would be toxicologically detrimental.

Any detrimental effect would depend upon the nature of the toxicity and the biological half-life of the chemical concerned.' (WHO, 1989)

The following discussion on the significance of exceeding the ADI applies equally to other recommended limits of intake or exposure, such as TDI or PTWI.

Renwick and Walker (1993) propose three questions that need to be considered if there are potential exceedances of the ADI.

•          What proportion of the population should be allowed to exceed the ADI?

•          To what extent can the ADI be exceeded without any real concern?

(Renwick and Walker, 1993, p 464)

They stress 'the significance of any minor excursions of intake above the ADI can only be put into context by reference back to the animal data and to the NOEL which gave rise to the ADI' (p 465).

Renwick and Walker describe three parameters governing the precision of the No Observed Effect Level:

•          the group size studied which tends to be less important than;

Renwick and Walker (1993) conclude that the significance of the exceedance must be assessed on a substance-specific basis and by reference to the toxicological (and especially the NOEL) data; the magnitude of the exceedance, and the duration of the exceedance. (Langley and Sabordo 1996, p144)

9. APPRAISAL OF ASSESSMENTS

Langley (1993a) presents methods for appraising assessments.

9.1 Introduction

Accurate and timely site-specific health risk assessments will depend upon coherent and logically developed reports; See Schedule B(2) for standardised formats for data Collection and reporting. Regulatory agencies should immediately reject reports that:

•          are unclear and confusing

•          do not meet appropriate levels of coherence or logic

9.2 General

A person preparing or reviewing an assessment will consider questions such as:

9.2.1 Key Aspects

•          Has the objective of the report been defined clearly?

•

•          Is it clear how results of any sampling plan were analysed and interpreted?

•          Have data been analysed en masse or for the appropriate strata?

•          Were environmental fate and transport mechanisms understood?

•          How were abnormal results or findings managed?

•          Were the uncertainties of the assessment identified and understood?

9.2.2 Interpretation of Data

An appraisal of data must show an understanding of:

•          the site history (and gaps in the history)

•          topography of the site

•          soil structure (eg presence of clay or fill, and the depths of individual strata)

•          the proposed land use

Too often numerical data are considered in isolation from other key parameters such as:

•          soil characteristics (eg from what stratum did the sample come)

•          the levels of detection (and reporting)

•          the geographical relationship of one sample to another

•          the proposed landuse

Other key failings in the analysis of numerical data include:

Given two similar results, the result that can be explained (eg by history, or similarities with results from similar strata) will tend to be of less concern.

9.2.3 Use of Subjective Terms

The use of subjective terms in reports (eg 'heavy/medium/light contamination') or terms that are used in common parlance but may have legalistic definitions (eg 'contamination') is confusing and should be avoided in reports. The use of the term 'hot spot' can be rather misleading to the general public and is best avoided.

9.3 Specific

9.3.1 Human health risk assessment checklist

The following checklist has been adapted from EPA NSW (1998) and Vic EPA (1997a) and should be addressed in human health risk assessments.

9.3.1.1 Data Collection

•          Have the objectives of the risk assessment been stated?

•          Has the background to the events leading to the risk assessment been provided?

•          Have all chemicals of potential concern been identified and appraised?

•          Have the sources of the contaminants been identified?

•          Have the environmental fate and transport of the contaminants been identified?

9.3.1.2 Toxicological Assessment

•          Have all relevant toxicological facts been checked for accuracy and currency?

•          Has the adequacy of the available toxicological database been appraised?

•          Has the critical toxic effect(s) and organ/body system been identified?

•          Have toxicologically sensitive subpopulations been identified?

9.3.1.3 Exposure Assessment

•        Has the potentially exposed population been identified?

9.3.1.4 Equations

•          Have all equations used in the risk assessment been presented in the report?

•          Are all equations consistent?

•          Have all parameters in each equation been clearly defined?

•          Have the correct units been allocated to each parameter?

•          Are all equations dimensionally correct?

9.3.1.5 Data Evaluation

•          Has laboratory QA/QC been reported and analysed?

•          Has field QA/QC been reported and analysed?

9.3.1.6 Assessment and report presentation

•          Have all tables and figures been referred to correctly in the text of the report?

•          Has irrelevant information from other sites been excluded from the report?

•          Have all assumptions and default data been identified and justified?

•          Has the analysis been based on an up-to-date literature appraisal?

•          Have all conclusions been justified?

•          Has a detailed uncertainty discussion been included in the report?

The following method was used for developing Health-based Investigation Levels. Similar principles have been used for determining HILs for contaminants with and without cancer endpoints. The methodology was initially endorsed in ANZECC/NHMRC(1992). This approach can also be used for developing Response Levels by using site-specific data and appropriate safety factors to accommodate uncertainty and variability. Site-specific Response Levels developed by risk assessors will require endorsement by regulators and/or auditors. Guideline Doses for carcinogenic soil contaminants will be established by regulatory authorities.

Investigation levels will be determined taking into account (ANZECC/NHMRC 1992, Imray and Langley 1996):

The total exposure to a substance 'X' can be represented by the equation:

Exposure to substance X = Background Exposures (eg. from food and water)

+

Exposures from contaminated soil by ingestion,

inhalation and skin absorption)

= Background Exposures

+

Amount of substance absorbed from soil.

= BE

+

(Sing x Cing x Bing + Sinh x Cinh x Binh + Sskin x Cskin x Bskin)

= BE + SEsoil

BE = Background Exposures (eg. from food and water).

Sing = Amount of soil ingested.

Sinh = Amount of soil/dust inhaled and retained.

Sskin = Amount of soil on skin.

Cing = Concentration of substance in soil ingested.

Cinh = Concentration of substance in soil/dust inhaled and retained.

Cskin = Concentration of substance in soil on skin.

Bing = Bioavailability, ie. percentage absorbed, of substance when ingested.

Binh = Bioavailability of substance when inhaled.

Bskin = Bioavailability of substance when on skin.

SEsoil = Substance exposure from soil."

(ANZECC/NHMRC, 1992, p37)

Different levels of bioavailability will occur between soil ingested, inhaled or in contact with skin.

National health investigation level guidelines will be set by national health advisory bodies.

A variable percentage of the TI will be allowed for exposure to contaminated soil. The variation will largely relate to contributions of other background factors, especially food. This is consistent with the IPCS approach and that used in the four Australian workshops on the health risk assessment and management of contaminated sites.

When the PTWI/ADI is used for establishing investigation levels for individual contaminants, the basis for the PTWI/ADI should be sought from appropriate World Health Organization documents (eg. WHO 1987, WHO 1989). This information should include target organ(s) and effect(s) (eg. nature, reversibility, severity, LOAEL for most significant toxic effect); bioavailability; and safety factors accounting for variations in human sensitivity and extrapolations from animal studies. Similarly, when a Guideline Dose derived using the NHMRC 'Toxicity Assessment Guidelines for Carcinogenic Soil Contaminants' is used, the basis for the derivation should be fully documented. Guideline Doses for soil contaminants with cancer effects will be determined by national health advisory bodies or their appointees.

If no PTWI, ADI, or GD is available a specific approach acceptable to the relevant health agencies will need to be determined using WHO (1994) for non carcinogens, or National Environmental Health Forum Guidance for Cancer Risk Assessment for substances with cancer effects.

It is considered that these methods for determining investigation levels will protect the entire population with few exceptions. Where a significant proportion of the population demonstrates allergic sensitisation to a substance (eg nickel) this will need to be considered in criteria setting. People who may have unusual sensitivity to contaminants may need to be considered in a site assessment (Imray and Langley, 1996).

Qualifications to setting the Health-based Investigation Levels are:

A decision tree detailing the use of Health Risk Assessment to develop health-based soil criteria is provided in Figure 10-I.

Figure 10-I

Decision Tree for the development of health-based soil criteria

11. REFERENCES

Sites Monograph Series Number 7. South Australian Health Commission, pp 137-190.