By Kristina D. Mena
Summary: Risk assessments for microbiological contamination in various water sources are vital to protect public health. Using USEPA’s recommended level of acceptable risk, different studies use different pathogens as central components. According to the author, this practice needs to continue to fully ensure public safety.
Despite current water treatment advancements, microbial waterborne disease continues to occur in the United States. This may be due to a temporary breakdown in water treatment or water contamination after treatment. Exposure to contaminated drinking water by a community may lead to a recognizable outbreak of disease; however, these cases represent only a fraction of the total associated with waterborne microorganisms. Often, such cases are undetected; the ill person may not seek medical treatment, for example, or perhaps the source pathogen isn’t identified. Epidemiological data—on the causes, distribution and control of disease in populations—obtained from waterborne outbreaks provide information on human health impacts of microbial-contaminated water. It is, however, difficult to understand the public health significance associated with frequent exposure to low levels of contamination. A risk assessment methodology (based on the framework developed by the National Academy of Sciences1) has been developed and applied to predict human health consequences from exposure to pathogens in the environment, including water.2-5 Policy makers can use information obtained from a risk assessment to establish guidelines that address microbial water quality issues, and the U.S. Environmental Protection Agency (USEPA) first used risk assessment to develop the Surface Water Treatment Rule for Giardia.6
Monitoring for coliform bacteria, as indicators of microbial water quality, has proven to be inadequate at assuring treatment efficacy for pathogens and preventing waterborne outbreaks.7 Laboratory studies show no correlation between absence of coliform bacteria in water and absence of other (potentially disease-causing) microorganisms, such as enteric viruses and protozoan parasites. Using available data on specific pathogens, risk assessment is used to predict their public health impact. There are four steps to the risk assessment process—hazard identification, dose-response assessment, exposure assessment and risk characterization. In the hazard identification step, a microorganism is investigated to characterize it as a pathogen. Laboratory data and any available epidemiological data are evaluated to determine the microorganism’s ability to cause disease as well as describe all possible health outcomes resulting from exposure. Other issues addressed here include microorganism transmission routes, and the role of host factors such as immunity and response to multiple exposures. Table 1 lists some waterborne microorganisms whose roles as “hazards” could be evaluated in a risk assessment.
Using different studies
Dose-response assessment involves determining the relationship between dose of the microorganism (hazard) and the incidence or extent of the adverse health effect. This step may use data from animal or human studies (depending on what’s available) that may have used infection or illness as a health endpoint. If animal studies were used, extrapolation from animals to humans is required. It’s also necessary to extrapolate from high to low doses since relatively high concentrations of microorganisms are used in dose-response studies involving healthy human volunteers so high frequencies of infections can be observed with a minimum number of study participants. Mathematical models that may represent a microorganism-host interaction have been described elsewhere2 and two models in particular were shown to adequately reflect the infection process (see Table 2).
Several assumptions are made when using these models. It’s assumed a person is exposed to a random distribution of pathogenic microorganisms and one microorganism is capable of initiating infection. It’s also assumed each exposure is statistically independent of another. The appropriate model is selected using the method of maximum likelihood to determine which best fits the dose-response data for a particular microorganism. Parameters are subsequently defined.2 If illness and death are health endpoints in question, morbidity and mortality ratios (obtained through the hazard identification step) can be incorporated in the model to determine these probabilities. Annual risks can be determined (assuming a person is exposed to a constant amount of microorganisms daily) using the following equation:
Annual Risk of Infection (Pannual) = 1 – (1 – Pi)365
Setting the goal
The goal of exposure assessment is to determine the amount of water (such as drinking water) a person consumes as well as the number of microorganisms in the water. The USEPA uses a 2 liters per person per day (2 L/person/day) exposure for drinking water.8 To estimate exposure to microorganisms, a review of published literature describing occurrence studies of specific microorganisms in water can provide quantitative data. Unfortunately, this type of information is lacking for microorganisms in water. A microorganism’s ability to survive environmental stressors and its susceptibility to inactivation by water treatment are also considered here. Exposure assessment also distinguishes the population exposed, such as size and age distribution of population, for example.
Risk characterization uses information obtained from the previous steps to estimate health risks associated with exposure to a particular microbial hazard. All assumptions and uncertainties (such as dose, exposure frequency, population affected, etc.) that contributed to the risk assessment procedure are described here. The probability of becoming infected, ill or even dying after consuming various amounts of a particular pathogen can be estimated by using the appropriate model in Table 2.
Risk assessment has been conducted for specific waterborne microorganisms, such as rotavirus and Giardia.9,10 One study found rotavirus to be the most common cause of viral gastroenteritis worldwide and expressed the highest infectivity of any waterborne virus. Several waterborne outbreaks were associated with fecal contamination or inadequate water treatment. The beta-Poisson model with defined dose-response parameters for rotavirus was used in this assessment.4 Concentrations of rotavirus detected in drinking water and surface waters [assuming water treatment achieved 99.99 percent (4-log) reduction11] were obtained and used in calculating human health risks associated with drinking water exposure. The USEPA recommends annual risks of infection for waterborne microorganisms not exceed 1 in 10,000.12 When calculating risks for rotavirus in drinking water (assuming 2 L/person/day exposure) where polluted surface waters were the source and a 99.99 percent reduction of viruses was assumed, annual risk of infection didn’t meet this recommendation of infection greater than 1 in 1,000. The authors concluded that to meet USEPA’s 1 in 10,000 goal, 5 to 6 logs of virus removal would be necessary.
Determining acceptable risk
Another group developed a risk assessment model to estimate risk of infection after exposure to treated waters contaminated with different levels of Giardia cysts. Giardia is considered the most identifiable waterborne agent in the United States13,14 and is associated with a long duration of diarrhea in infected individuals. Using the exponential model (with defined dose-response parameters for Giardia) and survey data on occurrence of Giardia cysts in polluted and pristine waters, levels of water treatment necessary to achieve the acceptable level of risk of infection (1 in 10,000 annually) were determined. Assuming a water utility used those particular source waters, a 3- to 5-log removal/inactivation of cysts would have to occur to meet the USEPA’s recommendation.
The USEPA uses risk of infection rather than illness in its recommendation due to a variety of factors that contribute to probability of illness, such as host susceptibility and variation in virulence of strains of microorganisms. Using infection as the health endpoint of interest is therefore more protective of our immunocompromised populations—infants, the elderly, pregnant women, transplant recipients, among others—which are more likely to develop severe illness from an infection. Health risks may be overestimated for those who ingest less than the assumed amount of water (2 L/person/day) and those with pre-existing immunity to the particular pathogen of interest. Health risks, however, may be underestimated since infections can lead to secondary transmission, which can magnify the impact of exposure to waterborne microorganisms. Risk estimates have been compared to epidemiological data obtained from waterborne outbreaks, and results indicate the probability models can successfully predict health outcomes from exposure to microorganisms.5
The methodology described here can be applied to assess health risks for all types of populations as well as to evaluate microbial water quality for different sources of water (drinking water, recreational water, etc.). More data are needed on the occurrence of microorganisms in water. Plus, the application of sensitive laboratory detection techniques during surveillance studies of viruses and protozoa would lead to better estimates of exposure. Due to the great variability in occurrence, infectivity and pathogenicity of waterborne microorganisms, it’s pertinent that risk assessments are performed for specific microorganisms so the public health impact from microbial-polluted waters can be fully identified.
- National Academy of Sciences, “Risk assessment in the federal government: managing the process,” National Academy Press, Washington, D.C., 1983.
- Haas, C.N., “Estimation of risk due to low doses of microorganisms: a comparison of alternative methodologies,” American Journal of Epidemiology, 118:573-582, 1983.
- Regli, S., et al., “Modeling the risk from Giardia and viruses in drinking water,” Journal of American Water Works Association, 83:76-84, 1991.
- Haas, C.N., et al., “Risk assessment of virus in drinking water,” Risk Analysis, 13:545-552, 1993.
- Haas, C.N., J.B. Rose and C.P. Gerba, Quantitative Microbial Risk Assessment, John Wiley & Sons Inc., New York, 1999.
- U.S. Environmental Protection Agency, “National primary drinking water regulations: filtration and disinfection; turbidity; Giardia lamblia, viruses, Legionella, and heterotrophic bacteria,” Federal Register, 54(124):27486-27541, 1989.
- Sobsey, M.D., et al., “Using a conceptual framework for assessing risks to health from microbes in drinking water,” Journal of American Water Works Association, 85:44-48, 1993.
- Macler, B.A., and S. Regli, “Use of microbial risk assessment in setting United States drinking water standards,” International Journal of Food Microbiology, 18:245-256, 1993.
- Gerba, C.P., et al., “Waterborne rotavirus: a risk assessment,” Water Research, 30:2929-2940, 1996.
- Rose, J.B., C.N. Haas and S. Regli, “Risk assessment and control of waterborne giardiasis,” American Journal of Public Health, 81:709-713, 1991.
- U.S. Environmental Protection Agency, Guidance Manual for Compliance with Filtration and Disinfection Requirements for Public Water Systems Using Surface Water Sources, EPA Report No. 570/9-88-018, USEPA, Washington, D.C., 1991.
- Macler, B.A., “Acceptable risk and U.S. microbial drinking water standards,” Safety of Water Disinfection, ed. by G.F. Craun, ILIS Press, Washington, D.C., pp. 619-626, 1993.
- Craun, G.F., “Cause of waterborne outbreaks in the United States,” Water Science and Technology, 24:17-20, 1991.
- Kramer, M.H., et al., “Surveillance for waterborne disease outbreaks: United States, 1993-1994,” Morbidity Mortality Weekly Report, 45(SS-1):1-15, 1996.
About the author
Dr. Kristina D. Mena is assistant professor of environmental sciences at the University of Texas Health Science Center at Houston School of Public Health. She earned a master of science in public health (MSPH) degree from the University of South Florida and a doctorate in environmental microbiology from the University of Arizona. Her research interests include laboratory detection of waterborne and foodborne pathogens, and application of microbial risk assessment to address related public health issues. She can be contacted at (915) 747-8514, (915) 747-8512 (fax) or email: email@example.com