Are all distributions of risk with the same 95thpercentile equally acceptable?

When evaluating a probabilistic health risk assessment, say at a hazardous waste site, risk managers need a risk management policy that distinguishes an acceptable distribution of risks to individuals in a population from an unacceptable one. If a risk manager decides that the distribution of risk for the status quo is unacceptable, then a risk assessor needs a way to compute cleanup targets, i.e., the risk assessor needs a policy statement against which to estimate distributions of exposure point concentrations which, if engineered at a site, will achieve an acceptable distribution of risk. Some regulatory agencies base acceptability on whether the 95^th percentile of the risk distribution falls at or below a given value, without considering the behavior of the rest of the distribution. As regulatory agencies adopt risk management policies for use with probabilistic risk assessments, we recommend that they base their new policies on two simultaneously binding constraints‐one on an upper percentile and one on the arithmetic mean of the distribution of risk‐in addition to other non‐risk criteria.     

An Evaluation of Risk Drivers from a Sample of Risk Assessments Conducted for the U.S. Army

We evaluate risk drivers at selected U.S. Army installations by developing a database containing contaminant-pathway-receptor combinations that exceed regulatory thresholds for ecological (toxicity quotient greater than one), human health cancer risk (predicted incremental lifetime cancer risk greater than one in ten thousand), and noncancer human health (hazard index greater than one). We compare the risk drivers from the database to reported corrective action objectives from available decision documents. For noncancer hazards, explosives (particularly in ground water) dominate the reported exceedances of regulatory thresholds in the database. PAHs in home-grown produce show the highest number of exceedances of regulatory thresholds for cancer risk. For ecological risks, PAHs in both terrestrial and aquatic environments dominate the exceedances of regulatory thresholds. All available cleanup levels were derived based on human health exposures rather than ecological exposures, except for one site. In general, ecological risks were considered to be “more uncertain,” and that was used as a basis for not relying on backcalculated target levels on the basis of ecological risk. The reverse was true for human health risks: the “conservative” assumptions incorporated into the modeling provided the justification for backcalculating health-protective target levels.     

Bayesian Updating of Model-Based Risk Estimates Using Imperfect Public Health Surveillance Data

Model-based estimation of the human health risks resulting from exposure to environmental contaminants can be an important tool for structuring public health policy. Due to uncertainties in the modeling process, the outcomes of these assessments are usually probabilistic representations of a range of possible risks. In some cases, health surveillance data are available for the assessment population over all or a subset of the risk projection period and this additional information can be used to augment the model-based estimates. We use a Bayesian approach to update model-based estimates of health risks based on available health outcome data. Updated uncertainty distributions for risk estimates are derived using Monte Carlo sampling, which allows flexibility to model realistic situations including measurement error in the observable outcomes. We illustrate the approach by using imperfect public health surveillance data on lung cancer deaths to update model-based lung cancer mortality risk estimates in a population exposed to ionizing radiation from a uranium processing facility.     

Quantification of Changes in Chemical Pesticide Human Health Risk Following the Introduction of Bt Cotton and Herbicide-tolerant Soybean: A Case Study

Human health risks associated with changes in synthetic chemical pesticide use following the introduction of genetically modified crops are quantified in this article. Bacillus thuringiensis ( Bt ) cotton and herbicide-tolerant (HT) soybean, two genetically modified crops, were chosen as the focus for this study based on their global popularity. An innovative multimedia total exposure model, CalTOX, was used to calculate the health risks for two target populations, before and after the introduction of Bt cotton and HT soybean. Major results include the quantification of incremental lifetime cancer risk based on a time-weighted average exposure, and the quantification of hazard ratios for non-cancer effects based on the maximum exposure rate value, both computed over the exposure duration. Results show that human health risks are not intuitively and necessarily reduced with the reduction of pesticide use. For example, more trifluralin was used after the introduction of HT soybeans in Iowa and Minnesota, leading to higher risks. Moreover, the general population may have larger exposures to pesticides when compared with the population living in areas where pesticides are actually applied. This may occur because exposure to pesticides is not only dependent on geographical distance from the contaminated area, but also strongly dependent on other factors, such as the characteristic travel distance and environmental persistence.     

Dietary exposure and risk assessment to cadmium of the adult population of Jiangsu province,China: Comparing between semi-probabilistic and fully probabilistic approaches

With rapid industrialization and modernization in China, long-term dietary exposure to cadmium (Cd) has the potential to cause health risks. The traditional probabilistic approach for long-term dietary assessment was a semi-probabilistic method that had limitations in qualifying the variation in concentration data. Thus, this article estimated the long-term dietary exposure to Cd of the adults of Jiangsu province using semi-probabilistic and fully probabilistic approaches, and compared the two models. Cd concentration data were obtained from the national food contamination monitoring program 2001–2009. Food consumption data were gathered from the National Diet and Nutrition Survey conducted in 2002. The provisional tolerable daily intakes (PTDI) of 0.8 μg kg^–1 bw for Cd was applied in the risk assessment. The estimates of semi-probabilistic approach were within the provisional tolerable daily intakes (PTDI) at mean and all percentiles. However, the P95 of exposure with fully probabilistic approach for all population groups were above PTDI. Rice and rice products followed by wheat flour and wheat flour products, crustaceans, pak-choi, beans, and bean products contributed most to the Cd intake for adults. The fully probabilistic estimates could indicate a possible public health concern on Cd for adults in Jiangsu province.     

Equine Risk Assessment for Insecticides Used in Adult Mosquito Management

Since West Nile virus (WNV) was introduced to New York City in 1999, it has subsequently spread through the Americas, creating human and animal health risks. Our equine risk assessment focused on three pyrethroid insecticides (phenothrin, resmethrin, and permethrin), pyrethrins, and two organophosphate insecticides (malathion and naled). Piperonyl butoxide, a synergist commonly used in pyrethroids, was also assessed. The objective was to use deterministic and probabilistic risk assessment methodologies to evaluate health risks to horses from vector management tactics used for control of adult mosquitoes. Our exposure estimates were derived from the Kenaga nomogram for food deposition, AgDRIFT® for deposition onto soil and hair, AERMOD for ambient air concentrations, and PRZM-EXAMS for water concentrations. We used the risk quotient (RQ) method for our assessment with the RQ level of concern (LOC) set at 1.0. RQs were determined by comparing the exposure to no-observable-effect-levels. Acute deterministic RQs ranged from 0.0004 for phenothrin to 0.2 for naled. Subchronic deterministic RQs ranged from 0.001 for phenothrin to 0.6 for naled. The probabilistic assessment revealed estimates of deterministic acute and subchronic RQs were highly conservative. Our assessment revealed that risks to horses from adult mosquito insecticides are low and not likely to exceed the LOC.     

Application of a Probabilistic Risk Assessment Methodology to a Lead Smelter Site

Exposure of children to lead in the environment was assessed at the Murray Smelter Superfund site using both a deterministic risk assessment approach, the Integrated Exposure Uptake Biokinetic (IEUBK) model, and a probabilistic approach, the Integrated Stochastic Exposure (ISE) model. When site-specific data on lead in environmental media were input as point estimates into the IEUBK model, unacceptable risks were predicted for children living within five of eight study zones. The predicted soil cleanup goal was 550?ppm. Concentration and exposure data were then input into the ISE model as probability distribution functions and a one-dimensional Monte Carlo analysis (ID MCA) was run to predict the expected distribution of exposures and blood lead values. Uncertainty surrounding these predictions was examined in a two-dimensional Monte Carlo analysis (2-D MCA). The ISE model predicted risks that were in the same rank order as those predicted by the IEUBK model, although the probability estimates of exceeding a blood lead level of 10?µg/dl (referred to as the P10) from the ISE model were uniformly lower than those predicted by the IEUBK model. The 2-D MCA allowed evaluation of the confidence around each P10 level, and identified the main sources of both uncertainty and variability in exposure estimates. The ISE model suggested cleanup goals ranging from 1300 to 1500 ppm might be protective at this site.     

Soil Cleanup Goal for Dioxin Using Probabilistic Risk Assessment Techniques

State regulators in Florida recently approved a first-of-its-kind probabilistic risk assessment (PRA) for determining an alternative residential Soil Cleanup Target Level (SCTL) for dioxin (32 ng/kg TEQ). The default residential SCTL (7 ng/kg TEQ) is based on a single, deterministic calculation with numerous conservative assumptions, resulting in an overly conservative value far beyond the regulatory mandate (i.e., 10^?6 increase in cancer risk). Conversely, this PRA used a Monte Carlo simulation to estimate risk for all members of a large population using a combination of scientific data and professional judgment, with final details developed during negotiations with regulators. The simulation parameters were defined probabilistically and reflect the ranges of values for the following exposure variables: body weight, exposure duration, exposure frequency, fraction from contaminated source, soil ingestion rate, and relative bioavailability. Other variable and uncertain parameters were treated deterministically per direction from the regulators. The state also required that a pre-supposed high-risk subpopulation be analyzed separate from the full receptor population. Despite the conservativeness of the alternative SCTL, this PRA represents a significant step toward more realistic estimates of human health risks caused by environmental contaminant exposure.     

A comparison of the predicted risks of developing osteosarcoma for dogs exposed to 238PuO2 based on average bone dose or endosteal cell dose.

Dose-response relationships observed in laboratory animals can be used to identify possible human risk factors and may also be used in a quantitative manner when human data are not available. This paper presents an analysis of the dose dependency of osteosarcoma incidence in beagle dogs given a single inhalation exposure to a monodisperse aerosol of 238PuO2. We were particularly interested in comparing the predicted risks that were based on average bone dose with those based on endosteal cell dose and in evaluating the advantages of using a more biologically relevant cell-specific dose in risk estimation. The endosteal cell dose was calculated using the method of Marshall et al. (Health Phys. 35, 91-101, 1978), as extended to account for exposure by inhalation. The relationship between dose and time to tumor was analyzed by the proportional hazards regression model. The probability of developing osteosarcoma was strongly dependent on dose for dogs receiving low doses, but this was not true for dogs receiving high doses. The predicted risk based on endosteal cell dose was not consistently higher or lower than the risk based on average bone dose at various times after exposure, because the relationship between these two doses was not linear with respect to time. Also, as a result of the nonlinear relationship between these two doses, the risk estimated based on endosteal cell dose would not be a fixed factor of that based on the average dose. Random errors in the measured initial lung burden had a relatively large impact on the predicted risk based on endosteal cell dose, and the difference between the estimated risk of developing osteosarcoma based on endosteal cell dose and that based on average bone dose is likely to be within the error margins of the estimated risks.     

The need for new methods to backcalculate soil cleanup targets in interval and probabilistic cancer risk assessments

Ordinary algebra may be used to backcalculate health‐based cleanup targets in deterministic risk assessments, but it does not work in interval or probabilistic risk assessments. Equations with interval or random variables do not follow the rules of ordinary algebra. This paper explains the need for more sophisticated methods to backcalculate soil cleanup targets when using interval or random variables.     

A quantitative evaluation of ten approaches to setting site‐specific cleanup objectives

Over the past several years, a number of reviews of various approaches employed or proposed by different jurisdictions for setting contaminated site cleanup objectives have been conducted. These methods are normally divided into two groups: absolute and relative. The absolute approach establishes a numerical concentration for a contaminant in soil that is applied to all sites, regardless of site‐specific factors. The relative approach generally employs a risk assessment modeling technique to derive a soil cleanup number based on site‐specific and chemical‐specific factors that will result in an acceptable exposure or risk. In the latter case, the numerical concentrations may vary from site to site, but the risk is the same. In the former, the numerical concentration is the same from site to site, but the risk, although unstated, varies. Numerous reviews of these methods have been undertaken, but all have been qualitative in nature, examining theoretical bases and not comparing the actual performance of each method in the assessment of one site. We have undertaken to derive cleanup guidelines for a single, existing contaminated site in Canada using ten different methods. Five of these were absolute methods (British Columbia Assessment Criteria, Alberta Soil Guidelines, Ontario Decommissioning Guidelines, Quebec ABC's, and New Jersey Acceptable Soil Contaminant Levels) and five were relative methods (U.S. Environmental Protection Agency Public Health Evaluation Manual, U.S. Army Primary Pollutant Limit Values, California Site Mitigation Decision Tree, California Technical Standard, and AERIS). The resulting analysis compares relative methods to absolute, examines the strengths and weaknesses of each approach, and discusses the similarities and differences of the outputs of the various approaches when employed to set cleanup objectives.     

A Probabilistic Assessment of Household Exposures to MTBE from Drinking Water

The oxygenate methyl tertiary butyl ether (MTBE) has been added to reformulated gasoline in the U.S. to meet national ambient air quality standards. Although MTBE has provided significant health benefits in terms of reduced criteria and toxic air pollutants, detections of MTBE in some groundwater and drinking water sources have raised concerns about potential environmental contamination and human exposures. In this paper, we examine the frequency and concentration of MTBE detections in drinking water sources in California from 1995 to 1999, and provide a preliminary analysis of the distribution of household exposures to MTBE from water-related activities. Using published data on the toxicity and possible cancer hazard posed by MTBE, we estimate the likely cancer and non-cancer risks for the general population in California from past and potential future MTBE exposures. More highly exposed subgroups were also addressed. Our findings indicate that less than 2% of all sampled drinking water sources in California had detectable levels of MTBE in 1999, with average MTBE drinking water concentrations ranging from 0.09 to 4.9?ppb for this year. Both the detection rate for MTBE and average MTBE concentrations have remained relatively stable since 1995, despite increased sampling of drinking water sources in California. The probabilistic exposure analysis suggests that drinking water exposures to MTBE are unlikely to pose a significant health risk for the general population or more highly exposed individuals in California.     

Assessment of the Chemical Quality of Recreational Bathing Water in Argentina by Health Risk Analysis

Del Azul creek (Argentina) is a natural water body used for recreational bathing in which heavy metals and pesticides have been detected. The aim of the study is to estimate the probabilistic non-cancer and cancer risks by recreational bathing, applying U.S. Environmental Protection Agency models for aggregated (exposure through accidental oral intake of water and dermal contact) and cumulative risks (combined exposure to groups of substances) for bathers of 5, 10, 15, and 20 years old. Bathing in Del Azul creek does not generate risks at the concentrations and the exposure scenarios considered. Cypermethrin and arsenic and heptachlor were the riskiest non-cancer and cancer substances, respectively. Our study highlights the importance of considering both routes of exposure because of the great significance of the dermal route and because of the variability of population characteristics, as it has been stated in other studies. These considerations are highly significant for Argentina, where the quality of recreational water for other than microbiological causes is frequently evaluated based only on the harmful oral intake and applied to a hypothetical individual representative of the population.     

Impact of Random Variables Probability Distribution on Public Health Risk Assessment from Contaminated Soil

Probabilistic methods are now being applied increasingly to public health risk assessment instead of the deterministic, conservative, point estimates. An essential part of the probabilistic methods is the selection of probability distribution functions to represent the uncertainty of the random variables considered. We study the effect of selection of different probability distribution functions on the probabilistic outcome using the first-order reliability method (FORM). An example of cancer risk resulting from dermal contact with benzo(a)pyrene (BaP)-contaminated soil is given. Cancer potency factor, soil concentration, and fraction of skin area exposed were assigned normal, lognormal, and uniform probability distribution functions, and the effect of probability of exceeding a target risk level (termed the probability of failure) and sensitivity measures were studied. We investigated the question: what happens when one assumes different distribution shapes with the same mean and standard deviation? The results indicate that the selection of a probability distribution function for the random variables had a moderate impact on the probability of failure when the target risk is at the 50th percentile level, while the impact was much larger for a 95th target risk percentile. We conclude that the probability distribution will have a large impact because in most cases the regulatory threshold risk is at the tail end of the risk distribution. The impact of the distributions on probabilistic sensitivity, however, showed a reversed trend, where the impact was slightly more appreciable for the 50th percentile than for the 95th percentile. The selection of distribution shape did not, however, alter the order of probabilistic sensitivity of the basic random variables.     

Critique on the Use of the Standardized Avian Acute Oral Toxicity Test for First Generation Anticoagulant Rodenticides

Avian risk assessments for rodenticides are often driven by the results of standardized acute oral toxicity tests without regards to a toxicant's mode of action and time course of adverse effects. First generation anticoagulant rodenticides (FGARs) generally require multiple feedings over several days to achieve a threshold concentration in tissue and cause adverse effects. This exposure regimen is much different than that used in the standardized acute oral toxicity test methodology. Median lethal dose values derived from standardized acute oral toxicity tests underestimate the environmental hazard and risk of FGARs. Caution is warranted when FGAR toxicity, physiological effects, and pharmacokinetics derived from standardized acute oral toxicity testing are used for forensic confirmation of the cause of death in avian mortality incidents and when characterizing FGARs’ risks to free-ranging birds.     

Probabilistic Analysis of Human Health Risks Associated with Background Concentrations of Inorganic Arsenic: Use of a Margin of Exposure Approach

Substantial evidence exists from epidemiological and mechanistic studies supporting a sublinear or threshold dose–response relationship for the carcinogenicity of ingested arsenic; nonetheless, current regulatory agency evaluations have quantified arsenic risks using default, generic risk assessment procedures that assume a linear, no-threshold dose–response relationship. The resulting slope factors predict risks from U.S. background arsenic exposures that exceed certain regulatory levels of concern, an outcome that presents challenges for risk communication and risk management decisions. To better reflect the available scientific evidence, this article presents the results of a Margin of Exposure (MOE) analysis to characterize risks associated with typical and high-end background exposures of the U.S. population to arsenic from food, water, and soil. MOE values were calculated by comparing a no-observable-adverse-effect-level (NOAEL) derived from the epidemiological literature with exposure estimates generated using a probabilistic (Monte Carlo) model. The plausibility and conservative nature of the exposure and risk estimates evaluated in this analysis are supported by sensitivity and uncertainty analyses and by comparing predicted urinary arsenic concentrations with empirical data. Using the more scientifically supported MOE approach, the analysis presented in this article indicates that typical and high-end background exposures to inorganic arsenic in U.S. populations do not present elevated risks of carcinogenicity.     

A Simulation-Assessment Modeling Approach for Analyzing Environmental Risks of Groundwater Contamination at Waste Landfill Sites

An integrated simulation-assessment modeling approach for analyzing environmental risks of groundwater contamination is proposed in this paper. It incorporates an analytical groundwater solute transport model, an exposure dose model, and a fuzzy risk assessment model within a general framework. The transport model is used for predicting contaminant concentrations in subsurface, and the exposure dose model is used for calculating contaminant ingestion during the exposure period under given exposure pathways. Both models are solved through the Monte Carlo simulation technique to reflect the associated uncertainties. Based on consideration of fuzzy relationships between exposure doses and cancer risks, risk levels of different exposure doses for each contaminant can be calculated to form a fuzzy relation matrix. The overall risks can then be quantified through further fuzzy synthesizing operations. Thus, probabilistic quantification of different risk levels (possibilities) can be realized. Results of the case study indicate that environmental risks at the waste landfill site can be effectively analyzed through the developed methodology. They are useful for supporting the related risk-management and remediation decisions.     

Assessing Risks of Plant-Based Pharmaceuticals: II. Non-Target Organism Exposure

Plant-based pharmaceuticals potentially offer a cleaner method of producing a protein for drug manufacturing than traditional methods because plants are free of mammalian infectious agents. However, in the open environment they have the potential for intra-and inter-species gene flow, protein exposure to the public and non-target organisms, and they also have the potential to contaminate livestock feed. This study used probabilistic approaches to quantify the non-target organism risks associated with three pharmaceutical proteins produced in field-grown maize. The risk assessment for plant-based pharmaceuticals was conducted for four receptor species used as surrogates for a wider range of species. Body weights and maize consumption rates for each species were modeled from currently available information and used to calculate the exposure based on expression levels of three proteins. The acute dietary exposure for the receptor species was a single-day event in which the total maize consumption came from the recombinant maize. The non-target organism risk assessment demonstrated that risks will vary between species and between proteins, based primarily on differences in toxic endpoint and consumption rates. It also shows the utility of probabilistic, quantitative risk assessment methodologies and the importance of assessing risks from plant-based pharmaceuticals on a case-by-case basis.     

Human health‐based soil cleanup guidelines for diesel fuel no. 2

Soil cleanup guidelines were developed for diesel fuel No. 2 that are protective of human health. Guidelines were conservatively based on a residential land use scenario. This scenario estimates human health risks associated with long‐term exposure to site soil via the inhalation, dermal, and ingestion routes of exposure. Lifetime dermal cancer studies were selected as the basis for deriving a safe level of diesel fuel in soil. Soil cleanup guidelines for diesel fuel No. 2 ranged from 1166 to 11,287 mg/kg for adult or child residents and represent contaminant levels that pose acceptable health risks for both present and proposed future uses of a site.