The site selection process for the National Exposure Registry.

Exposure assessment

Estimation of the potential for adverse health outcomes associated with environmental contamination is not only complicated but also full of uncertainties and assumptions (Gochfeld, 1991). It is composed of two primary activities: 1) exposure assessment and 2) effects assessment (Sexton et al., 1992). The process of selecting sites for inclusion in the NER focuses on exposure assessment.

Exposure to a contaminant is defined as an event that occurs when there is contact at a boundary between a human and the contaminant or a mixture of contaminants located within environmental media (Sexton et al., 1992; Ott, 1982), as shown in Figure 2. Exposure can be defined mathematically as:

|Mathematical Expression Omitted~

where E is exposure; C(t) is a concentration which varies as a function of time; and dt is an increment of time from |t.sub.1~ to |t.sub.2~ (Lioy, 1990). Although this expression represents the optimum, it should be noted that there is no uniform, well-established procedure for expressing exposure. Magnitude, usually expressed as concentration, is perhaps the most commonly reported parameter (Sexton et al., 1992). Other possibilities for expressing magnitude include quantities available for absorption, rates of intake, concentrations in body tissues, body burden and organ dose (Brown, 1987).

According to Michaud et al. (1991), the objective of an exposure assessment is to measure or estimate the intensity, frequency and duration of human exposure to an agent, such as a hazardous substance, present in the environment. In its most complete form, an exposure assessment should describe the source, magnitude, duration, schedule and route of exposure; the populations exposed; and the uncertainties in assessments. The site selection process for the NER essentially parallels the goals of exposure assessments. These goals include:

* Identifying the sources of a specific TABULAR DATA OMITTED substance and quantifying releases into the environment.

* Determining the movement of the substance from its sources to the point of contact with the population and identifying the pathways of exposure.

* Identifying important compounds and determining the concentration of the substance in the media at the point of contact with the population; in other words, determining exposure.

* Determining the applied dose of the substance or its metabolite(s) within the individual (via dose reconstruction) (Akland, 1991).

The actual site selection process for the NER is outlined in Figure 3 and discussed in the following section.

Phase I -- Identification of Potential Sites

The first step in Phase I of site identification is to locate sites that are reported to have the primary substance present in the environmental media of concern. Identification of potential sites is initiated by reviewing data from available national databases. It should be noted that, despite their obvious importance, human exposure data typically are not collected in a systematic or comprehensive manner (Sexton et al., 1992; Burke et al., 1992; Stevens and Swackhamer, 1989); therefore, the mere presence of the primary contaminant in a medium of concern qualifies a site for initial review.

ATSDR has developed a database, HazDat (ATSDR, 1993), in which all the information available on NPL sites, ATSDR public health assessment sites, emergency response sites, sites for which a health consultation has been conducted, and federal facilities sites is stored. This database, as well as such others as the Federal Reporting Data System (FRDS) (USEPA, 1991a) and the Storage and Retrieval of U.S. Waterways Parametric Data (STORET) (USEPA, 1991b), is queried for a list of sites with the substance of concern in the media of concern. Finally, notices are sent to ATSDR regional representatives and placed on the Public Health Network electronic bulletin board requesting information for additional non-NPL sites at which potential exposure to the primary contaminant has been indicated by quantitative laboratory analysis of the media of concern. It should be noted that all potential sites, whether NPL or non-NPL, are given equal consideration for inclusion in the NER.

After a list of potential sites has been developed, the applicable ATSDR public health assessment, available for each NPL site and ATSDR-petitioned site, is reviewed. (Note: An ATSDR public health assessment is an evaluation of data and information on the release of hazardous substances into the environment designed to assess any current or future effects on public health, develop public health advisories or other health recommendations and identify studies or actions needed to evaluate and mitigate or prevent human health effects (ATSDR, 1992). The following criteria, in alphabetical order, are used to assess the candidacy of each site.

* Completed exposure pathways. It can be confirmed that the population was exposed to the environmental media contaminated with the primary contaminant.

* Contaminant concentration(s). The concentration(s) at which the primary contaminant was found in the environmental media during the period of exposure must be documented.

* Documented contamination of media. Environmental samples indicate that the medium of concern was contaminated with the primary contaminant; samples were taken from representative locations; and samples were collected and transported appropriately to be analyzed by a certified laboratory.

* Duration of exposure. The duration of the exposure; the current exposure status of the population; and the length of time since the exposure ceased must be determined.

* Size of the potentially exposed population. The number of people who may have been exposed to the substance from the site must be determined (ATSDR, 1988).

All sites that are reviewed are logged into an in-house database so that disposition of the site considered can be tracked in an orderly fashion. Sites not meeting the preceding criteria are removed from further consideration. All information collected and retained for a site is entered into a tracking system, which is maintained for historical purposes.

Phase II -- Data Collection

In the second screening of candidate sites, all ATSDR site files are reviewed. These files, compiled and maintained by ATSDR, typically contain the remedial investigation/feasibility study reports, the hazard review scoring packages, the remedial action master plans, and risk assessments (for NPL sites), as well as sampling information and reports of other health studies that might have been conducted at the site. In addition to the criteria listed previously, the site files are reviewed for information pertaining to the following secondary criteria (ATSDR, 1988).

* Assessment of participation. Information, such as news articles or records of public meetings, exists regarding community interest and/or prior community participation in site-related activities. The levels(s) of support from state and local officials for the registry can be determined.

* Existing biomonitoring data. Information is available indicating the body burden of the primary contaminant in the population associated with this site.

* Number of secondary and/or potential confounding contaminants. The presence or absence of other contaminants has been analytically confirmed in the environmental media of concern at this site. If contaminants are present, the levels have been determined. Evidence exists to indicate whether exposure to this substance(s) is likely to result in health effects that would mask those potentially attributable to the primary contaminant.

* Particularly sensitive subpopulations. It has been determined whether sensitive subpopulations, such as children or the elderly, have been exposed to the contaminants at the site. The size of the exposed subpopulation(s) has been determined.

TABULAR DATA OMITTED

* Reported health problems. Documentation exists that indicates the population has reported experiencing health problems potentially related to this site (ATSDR, 1988).

As always, personal knowledge of a site is important. For that reason, each ATSDR regional representative is an invaluable source of site-specific information and contacts. At this point in the site selection process, additional documentation of site criteria is solicited from the regional representative for the region in which the site is located. The regional representative is also the liaison for contact with state and local officials who may have specific, in-depth knowledge pertinent to the site that might affect its selection for the subregistry. Other people who might be contacted for site-specific data are U.S. Environmental Protection Agency personnel, including the remedial project manager, and U.S. Geological Survey representatives, who might have additional sampling data or information about the physical characteristics of the site.

After review, sites are either retained for or removed from further consideration. Sites that are retained are ranked using the NER site-scoring protocol. This allows subjective ranking of a site using a scoring system which quantifies the requirements listed in Phases I and II.

Phase III -- Site Visit/Data Validation

Site visits are conducted only for the highest ranking sites. A typical site visit consists of meeting with interested officials from the departments of health and/or environment for the state in order to discuss the NER and how the candidate site might be included in a subregistry. Permission is sought to review all state files for the candidate site. The same process is repeated with the county and local officials. The types of documentation typically reviewed during a site visit are media specific; for example, if the medium of concern is drinking water, the operating records and chemical analysis records for the water system are reviewed. Of particular interest are minutes from previous public meetings at the site and news articles written about the site, which indicate the level of citizen interest in site-related activities.

The neighborhoods affected by the site are inspected to discover any specific characteristics that might affect the site's position in the final ranking of candidate sites. The types of information that might affect a site's rank include previously unreported evidence of the presence of specifically susceptible populations or evidence of a particularly transient population.

Finally, environmental sampling data are reviewed for adherence to quality assurance and quality control (QA/QC) procedures. Data without verifiable QA/QC measures for sampling and analytical procedures are not accepted for consideration. Some procedures that are considered in the review include use of appropriate sample selection and collection procedures; chain of custody; sample storage, packaging and transport; sample identification; split and duplicate sample analysis; analytical procedures (calibration verification, holding time, sample preparation and blanks); and data reporting.

Phase IV -- Final Selection

Upon completion of the site assessment and site visit, a rationale for selecting the site is prepared. The rationale includes information indicating how the site meets the criteria for site selection and which of the criteria, if any, the site does not meet. The rationale paper is reviewed by the administration within ATSDR. Based on the resources available, the site is either approved or disapproved for inclusion in an exposure subregistry.

Following inclusion of a site in the NER, the environmental data used in the selection process then serve as the basis for quantifying each registrant's exposure and accumulated dose. Because the environmental data do not fully characterize exposure throughout the defined exposure period, mathematical models and other methods of estimating exposure and dose are used. Through this process of dose reconstruction, each individual's exposure and/or the concentration of contaminants (or metabolites) in tissues are quantified. It is imperative to recognize, however, the uncertainty of the particular estimation technique when relating exposure/dose to adverse health outcomes.

Problems associated with the site selection process

A common problem when conducting exposure assessments for sites is there is typically little or no information on which to base estimates of exposure. Very often no actual environmental measurements, or far too few to quantify the exposure to a given individual or population, have been collected (Shalat, 1990). Most of the environmental sampling data for a site are data that were collected after the site had been identified and exposures to contaminants had ceased. In addition, the available data tend to be anecdotal, fragmented and focused narrowly on specific pollutants, media and routes of exposure (Sexton et al., 1992). Although valuable for characterization of a site, this information does not provide an adequate basis for determining exposure to receptor populations.

Historical data, which are typically what are available for exposure assessment, can be inadequate for a variety of reasons. Some common problems encountered by ATSDR staff with the use of historical data include:

* No identified source(s) of contamination.

* No information available to confirm individual levels of exposure or potential levels of exposure.

* No information available to determine the duration of the exposure.

* Samples taken in an inappropriate manner or analyses lacking the proper QA/QC measures.

* Gaps in essential records or data; for example, lack of operation records or distribution samples.

Data problems of this type can and frequently do result in the elimination of what could possibly be a viable site for inclusion in the NER.

When data gaps exist, biologic markers can sometimes be used to verify that human contact with hazardous substances has occurred (Gift et al., 1991). A biologic marker of exposure is 1) an exogenous substance or its metabolite(s); 2) the product of an interaction between a xenobiotic agent and some target molecule or cell that is measured within a compartment of an organism; or 3) other events in an organism related to exposure (National Research Council, 1987). As biologic markers become available for field use, they will be incorporated into the NER data collection process, thus overcoming many of the problems listed previously.

Although considerable information exists on NPL sites and some state priority sites that have proceeded through the remedial investigation phase, little or no information exists for thousands of sites that have been reported to local, state or federal authorities. In some cases, although adequate information exists for a site, the site is not listed in any database or on any list of recognized sites. Additionally, access to the data for these sites is often limited, the data may be scattered among multiple files in multiple offices, or the data might even have been lost.

Summary

In summary, through the process of elimination, the number of potential sites for inclusion in a given subregistry is gradually reduced, resulting in the identification of sites with the most complete exposure assessment information for the contaminant of concern. Although these data never fully characterize personal exposure, they serve as the basis for dose reconstruction which is necessary for determining dose/response relationships.

To date, more than 3,000 sites have been reviewed and 21 have been approved for inclusion in the five subregistries of the NER. These subregistries include (in the order of establishment) the Trichloroethylene (TCE), Dioxin and Benzene Subregistries (all with established sites); and the Radioactive Substances and Chromium Subregistries (currently in the site selection phase).

As the site selection process continues, further refinements to the methods used will be needed to integrate improved environmental data into the process of exposure assessment. The process described within this paper has been structured to be flexible so that updating and validating the methods used will be relatively simple. By continually assessing and improving the site selection process, and thus improving exposure assessment, the health information collected for the NER should bring us closer to understanding the relationships between environmental exposures to contaminants and potential health effects related to those exposures.

References

* Agency for Toxic Substances and Disease Registry (ATSDR) (1993), HazDat Database, Version 1.5. U.S. Dept. of Health and Human Services, Public Health Service, Atlanta, GA.

* ATSDR (1992), Public Health Assessment Guidance Manual, U.S. Dept. of Health and Human Services, Public Health Service, Atlanta, GA.

* ATSDR (1988), Policies and Procedures for Establishing a National Registry of Persons Exposed to Hazardous Substances: The National Exposure Registry, U.S. Dept. of Health and Human Services, Public Health Service, Atlanta, GA.

* Akland, G.G. (1991), Organizational components and structural features of EPA's new human exposure research program, J. Exposure Anal. and Environ. Epid. 1(2):129-141.

* Brown, S.L. (1987), Exposure assessment, In: Tardiff, R.G. and J.V. Rodricks (eds.), Toxic Substances and Human Risk: Principles of Data Interpretation, Plenum Press, New York, NY. Pp. 377-390.

* Burg, J.R. (1989), Policies and procedures for establishing the National Exposure Registry, J. Am. Col. Tox. 8(5):949-954.

* Burke, T., H. Anderson, N. Beach, S. Colome, R.T. Drew, M. Firestone, F.S. Hauchman, T.O. Miller, D.K. Wagener, L. Zeise and N. Tran (1992), Role of exposure databases in risk management, Arch. Environ. Health 47(6):421-429.

* Fields, T. (1991), Magnitude of toxic substances pollution problem: Superfund sites in regions, states, and localities close to HBCU/MI institutions. Presented at the Conference on Health Effects of Toxic Substances and Contingency Planning, September 26, 1991, Atlanta, GA.

* Gift, J.S., R.E. Grissom and J.M. Straight (1991), Biological markers: Monitoring populations exposed to hazardous substances, J. Environ. Health 54(2):22-26.

* Gochfeld, M. (1991), Setting the research agenda for chromium risk assessment, Environ. Health Perspect. 92:3-5.

* Lioy, P.J. (1990), Assessing total human exposure to contaminants: A multidisciplinary approach, Environ. Sci. Technol. 24:938-945.

* Michaud, J.M., S.R. Ripple and D.J. Paustenbach (1991), Human health risks associated with contaminated sites: Critical factors in the exposure assessment. In: Kostecki, P.T. and E.J. Calabrese (eds.), Hydrocarbon Contaminated Soils and Groundwater, Lewis Publishers, Chelsea, MI. Pp. 283-292.

* National Research Council (1987), Biological markers in environmental health research, Environ. Health Perspect. 74:3-9.

* Ott, W. (1982), Concepts of human exposure to air pollution, Environ. Inter. 7:179-196.

* Sexton, K., S.G. Selevan, D.K. Wagener and J.A. Lybarger (1992), Estimating human exposures to environmental pollutants: Availability and utility of existing databases, Arch. Environ. Health 47(6):398-407.

* Shalat, S.L. (1990), Can we use distance to study disease? In: Andrews, J.S., L.O. Askew, J.A. Bucsela, D.A. Hoffman, B.L. Johnson and C. Xinteras (eds.), Proceedings of the Fourth National Environmental Health Conference -- Environmental Issues: Today's Challenge for the Future, Assoc. of State and Territorial Health Officials, ATSDR, Center for Environmental Health and Injury Control, Centers for Disease Control, pp. 33-38.

* Stevens, J.B. and D.L. Swackhamer (1989), Environmental pollution: A multimedia approach to modeling human exposure, Environ. Sci. Technol. 23:1180-1186.

* U.S. EPA (1991a), FRDS: Federal Reporting Data System, Office of Groundwater and Drinking Water, Washington, D.C.

* U.S. EPA (1991b), STORET: Storage and Retrieval of U.S. Waterways Parametric Data, Office of Water, Washington, D.C.

Figure 3.

Outline of the site selection process for the National Exposure Registry

Phase I -- Identification of Potential Sites

Step 1. Potential sites identified (primary contaminant present)

1. VIEW database 2. HazDat database 3. ATSDR regional representatives 4. Other federal and state and local officials 5. Other databases (STORET, FRDS, etc.)

Step 2. ATSDR public health assessments for potential sites reviewed (Primary criteria for review applied)

1. Documented contamination of media (one or more) 2. Level of contamination 3. Size of population potentially exposed 4. Current potential exposure 5. Past potential exposure (duration and time elapsed since last exposure)

Step 3. Candidate sites identified

Phase II -- Data Collection

Step 4. ATSDR site files and additional information for candidate sites reviewed

(Secondary criteria for review applied)

1. Presence of sensitive subpopulations 2. Availability of biomonitoring data 3. Number of secondary or potentially confounding contaminants 4. Reported health problems 5. Assessment of participation

Step 5. Candidate sites ranked using the NER site-scoring protocol

Phase III -- Site Visit/Data Validation

Step 6. Site visit

1. Meetings with state, county and local officials 2. Review of state, county and local files 3. Tour of site and affected neighborhoods

Step 7. All data collected are reviewed for appropriate quality assurance/quality control procedures

Phase IV -- Final Selection

Step 8. Preparation and review of rationale for selection of candidate site(s)

Step 9. Site(s) selected

Step 10. Environmental dose reconstruction implemented

Ginger L. Gist, Ph.D., ATSDR, 1600 Clifton Rd., MS E-31, Atlanta, GA 30333.