Table of contents

Team Members: Pinto, K.C., Arnesen, A.S., Husserl, J., Newman, K., Casanova, L., Rock, C.M., Gurian, P., Gerba, C., Sato, M.I.

Problem Statement

Sewage treatment processes can effectively reduce organic matter and pathogens in effluents, but a byproduct of this process is sewage sludge, which must be effectively treated and disposed of. This is a considerable task; Brazil produces 150-200000 tons/day (Andreoli); the U.S., 7,100,000 (NEBRA). Sewage sludge now frequently enjoys a second life as biosolids, defined by the US Environmental Protection Agency as the “primarily organic solid product produced by wastewater treatment process that can be beneficially recycled.” Because of their high nutrient content and soil conditioning properties, biosolids represent a plentiful, available source of fertilizer and nutrients for farming. Consequently, land application of biosolids to condition soil or fertilize crops or other vegetation has become popular in a number of countries; around half of the biosolids produced in the U.S are being used for this purpose (USEPA).

However, because they originate from sewage sludge, biosolids can contain a number of human enteric pathogens, including bacteria, viruses, and parasites. The levels of these pathogens depend on the pathogen burden in a particular country, the volume of sewage produced, and the type of treatment the biosolids receive, all of which may vary greatly from country to country. The regulatory structure and requirements governing biosolids, including which organisms must be tested and acceptable levels, may also vary (Table 1).

Treatment options for biosolids include thermal treatment, pH treatment, aerobic digestion, anerobic digestion, irradiation, composting, and pasteurization (USEPA). Modern treatment processes are capable of producing biosolids with very low pathogen levels; however, not all biosolids destined for land application receive this level of treatment. The land application of biosolids that have not been treated to remove all human pathogens creates a potential opportunity for direct human contact, including hand contact. Hand contact with biosolid-amended soil followed by hand-to-mouth contact and accidental ingestion creates a route for enteric infection in people who come in close contact with agricultural land. The most obvious risk group for this exposure is workers, who have opportunities to come in direct contact with biosolids either when applying them to land or working on land where application has taken place. Workers are not the only risk group; if agricultural lands are open, accidental incursions or trespassing by others creates an opportunity for direct contact with biosolid-amended soil. However, since biosolids originate from the sewage treatment process, the pathogen levels will depend on the pathogen levels in sewage, which can vary from country to country. Therefore, an assessment of risks from biosolids contact and associated risk management strategies may need to be country specific.

People can come in contact with land applied biosolids on agricultural lands because they are:

  • working applying the biosolids
  • working on agricultural land after application
  • entering agricultural land after application.

What is the risk of illness from pathogens found in human sewage when people come in direct contact with land-applied biosolids on agricultural land, and how does this risk vary across three countries (U.S., Mexico, and Brazil)?

Hazard Identification

Ascaris lumbricoides

Ascaris lumbricoides is the largest nematode (roundworm) parasitizing the human intestine (CDC). The eggs of the worm are found in soil contaminated by human feces or in uncooked food contaminated by soil containing eggs of the worm. A. lumbricoides eggs becomes infected upon exposure to air and oxygen, and people become infected by the accidental ingestion of theses eggs (Cotruvo). This can happen when hands or fingers that have contaminated dirt on them are put in the mouth or by consuming vegetables or fruits that have not been carefully cooked, washed or peeled. Ascaris lives in the intestine and Ascaris eggs are passed in the feces of infected persons. If the infected person defecates outside (near bushes, in a garden, or field) or if the feces of an infected person are used as fertilizer, eggs are deposited on soil (CDC). Ascaris infection is one of the most common intestinal worm infections. It is found in association with poor personal hygiene, poor sanitation, and in places where human feces are used as fertilizer (CDC, Vyas). People infected with Ascaris often show no symptoms. If symptoms do occur they can be light and include abdominal discomfort. Heavy infections can cause intestinal blockage and impair growth in children. Other symptoms such as cough are due to migration of the worms through the body. Ascariasis is treatable with medication prescribed by your health care provider (CDC). Its effects may contribute substantially to child morbidity when associated with malnutrition, pneumonia, enteric diseases and vitamin A deficiency (WHO). Most people recover from symptoms of the infection, even without treatment, but they may continue to carry the worms in their body. Complications can be caused by adult worms that move to certain organs such as the bile duct, pancreas, or appendix. If the worms multiply, they can block the intestine (Vyas). Ascariasis occurs worldwide (Cotruvo), mostly in tropical and subtropical countries. An estimated 807 to 1,221 million people in the world are infected with Ascaris lumbricoides (CDC) approximately 25% of the world´s population annually . An estimated 73% of A. lumbricoides infections are in Asia, while about 12% are located in Africa and 8% in Latin America (O'Lorcain and Holland). Ascariasis occurs in people of all ages, though children are affected more severely than adults (Vyas).

Enteroviruses

The enteroviruses belong to the Picornaviridae family and are traditionally divided into 9 species based on differences in host range and pathogenic potential (CDC). The species include polioviruses, coxsackievirus (groups A and B), and echoviruses (CDC). Enteroviruses are small, very contagious viruses made of ribonucleic acid (RNA) and protein. The most well known are polioviruses -- the cause of paralytic poliomyelitis, commonly known as polio. While paralytic poliomyelitis is targeted for global eradication through vaccination, the nonpolio enteroviruses continue to be responsible for a wide spectrum of diseases. Infants and young children are hit hardest, however adults are affected as well. Enteroviruses are associated with at least 26 different syndromes and diseases, including coronary heart disease, type 1 diabetes, hand-foot-and-mouth disease, chronic fatigue syndrome/myalgic encephalomyelitis, encephalitis, herpangia, myocarditis, pleurodynia, ADHD, and central nervous system infections such as polio, meningitis, encephalitis, chronic meningoencephalitis, and acute flaccid paralysis. It is possible for an enteroviral infection to result in a multi-organ illness or a series of illnesses in different organs spanning several years. The average incubation period is 3-10 days. The human enteroviruses are ubiquitous viruses that are transmitted from person to person via direct contact with virus shed from the gastrointestinal or upper respiratory tract. Predominantly, Enteroviruses are transmitted via the fecal-oral route (Schwartz). Enteroviruses infect an estimated 50 million people each year in the US and possibly a billion or more worldwide. While ninety percent of enteroviral infections are asymptomatic or result in a mild illness, counting up the fraction of those with serious illness adds up to a large number of people. Nonpolio enteroviruses are responsible for 10-20 million symptomatic infections per year and are more prevalent among children of lower socioeconomic class, probably because of crowding, poor hygiene, and opportunities for fecal contamination. Enteroviral infections are most common in young children (Schwartz). Regarding outbreaks, Coxsackievirus (A16) is the most common cause of hand, foot, and mouth disease (HFMD) in the United States. Coxsackievirus (A24) and Enterovirus (70) have been associated with outbreaks of conjunctivitis. Echoviruses 13, 18, and 30 have caused outbreaks of viral meningitis in the United States. Enterovirus (71) has caused large outbreaks of HFMD worldwide, especially in children in Asia. Some infections from this virus have been associated with severe neurologic disease, such as brainstem encephalitis (CDC). Enteroviruses are acid stable and able to survive exposure to the tough environment of the gastrointestinal tract. They can also survive chlorine, freezing, and can live on surfaces for several days. The virus can be killed with standard disinfectant and heat (CDC).

Salmonella spp

Salmonella spp are bacteria of the family Enterobacteriaceae that cause illnesses such as typhoid fever, paratyphoid fever and foodborne illness. Approximately 2000 serotypes cause human disease. Salmonella serotypes Enteritidis, Typhimurium, and Newport account for about half of culture-confirmed Salmonella isolates reported by public health laboratories to the National Salmonella Surveillance System (CDC). The most important sorotypes to human health are Salmonella Typhi, which cause systemic infections Typhoid - an endemic disease in many developing countries - and Salmonella Typhimurium, one of the causative agents of gastroenteritis. Salmonella infections are zoonotic and can be transferred between animals and humans (Cotruvo). Salmonella are usually transmitted to humans by contaminated food, water, or contact with infected animals (CDC, CVE). Salmonella bacteria can survive without a host, under suitable environmental conditions, in waters and soils. They are frequently found in polluted water, contaminated by the excrement of carrier animals and soil (Cotruvo). The most common symptoms of the disease are diarrhea, fever, and abdominal cramps occurring 12 to 72 hours after infection. The illness usually lasts 4 to 7 days. Most people recover without treatment; but some of them need to be hospitalized. In these patients, the Salmonella infection may spread from the intestines to the blood stream, and then to other body sites and cause death unless the person is treated promptly with antibiotics. The disease affects all age groups, but infants, the elderly, and persons with compromised immune systems are more likely to develop a severe illness (CDC, CVE). Salmonella outbreaks may directly or indirectly have an association with water (Cotruvo). Approximately 42,000 cases of salmonellosis are reported in the United States, every year and it is estimated that approximately 400 persons die each year with acute salmonellosis (CDC). In Brazil, during the period 1999-2007, most of the outbreaks of bacterial diarrhea in the State of São Paulo were due to Salmonella spp, and S. enteritidis represents 43.2% of these outbreaks (CVE). Many cases are not diagnosed or reported, so the actual number of infections may be greater (CDC). The incidence of salmonellosis varies considerably between countries and within countries (case rates 10–>250/100 000 human population). In most of the world, the prevalence of salmonellosis depends on the water supply, waste disposal, food production preparation practices, and climate (Cotruvo).

Exposure Assessment

A schematic of a possible exposure scenario is shown in Figure 1.

Figure 1. Pathways for human exposure to pathogens in land-applied biosolids

The exposure scenario being assessed is direct hand contact followed by hand-to-mouth contact, resulting in accidental ingestion. Two groups are considered in this scenario: Workers applying biosolids to land and people coming into incidental contact with cropland where biosolids have been applied. Workers are likely to be highly exposed individuals (HEI), coming into direct contact with biosolids during application, without any pathogen attenuation that might occur over time after application. The key differences between these two groups are 1) the possible elapsed time between application and contact and 2) the quantity of soil ingested. This exposure scenario can be described by the following equations:

\begin{align*} bs_{frac} \times conc_{trt} = conc_{soil_{t_0}} \end{align*}

Equation 1. Concentration of pathogens in biosolids-amended soil at time 0 (time of application)

\begin{align*} conc_{soil_{t_0}} \times e^{-rate \times t_i} = conc_{soil_{t_i}} \end{align*}

Equation 2. Concentration of pathogens in biosolids-amended soil at time t after application

 

Equation 3. Dose of pathogens resulting from ingestion of biosolids-amended soil

Pathogens in biosolids

A literature search was conducted of both published and unpublished literature on occurrence of pathogens in biosolids in the U.S., Mexico, and Brazil. Experts in these countries were queried as to the existence of unpublished research or monitoring data, which was located for Mexico and Brazil. The occurrence of Salmonella, enterovirus, and Ascaris in biosolids varies between countries (Table 1). Pathogen content for biosolids in the U.S. is generally low; Ascaris eggs are often undetectable (Pepper et al.). Levels in Brazil are higher; Ascaris eggs are detectable in many sewage sludge samples taken from wastewater treatment plants in the country. Individual Brazil sites were modeled with a log normal distribution; Brazil concentrations overall were modeled using a fixed effects model that selected a waste water treatment plant (WWTP) at random. Literature searches found very little reliable data about pathogen content of treated biosolids in Mexico.

Table 1. Occurrence of Salmonella, Enterovirus, Adenovirus, and Ascaris ova in biosolids in the United States, Mexico, and Brazil
Country ' Pathogen Distribution Mean/g SD
Brazil (Bastos et al., 2012, Razzolini, Salvador) WWTP 1 Ascaris log-normal 0.670 0.550
WWTP 2     0.230 0.120
WWTP 3     2.300 1.510
WWTP 4     0.590 0.650
WWTP 5     3.790 1.960
WWTP 1 Enterovirus log-normal 0.867 0.882
WWTP 2     0.750 1.122
WWTP 3     4.280 5.433
WWTP 4     3.217 1.869
WWTP 5     0.817 0.915
WWTP 6     1.733 2.214
WWTP 1 Salmonella log-normal 0.328 1.115
WWTP 2     17.183 12.468
WWTP 3     0.006 0.000
WWTP 4     2.405 2.354
WWTP 5     0.148 0.492
United States (Pepper, et al.)   Ascaris point estimate 0.250 0,.50
  Enterovirus log-normal 0.419 0.200
  Salmonella log-normal 3.240 2.600
Mexico (Magos-Navarro)   Salmonella Log-normal 63.5 9.26

Land application of biosolids can involve spraying or spreading onto soil surface, and tilling into soil after surface application. Tilling results in a top layer that is a mix of topsoil and biosolids. Adjusted incidental ingestion of biosolids for occupational workers and the general population ranges uniformly from 0.05 to 0.62 grams/person/day. All discrete calculations assumed consumption of 0.48 grams/person/day (Kumar, USEPA).

Pathogens in land-applied biosolids may undergo inactivation over time, affecting risks from ingestion. A literature search was conducted for inactivation rates of the pathogens of interest in soil, biosolids, or biosolids-amended soil, under field or laboratory conditions. The pathogens remaining in the soil for exposure assessment are described by Equation 2, shown earlier. Decay rates in biosolids for the three pathogens are included in the model (Table 2). For enterovirus and Ascaris, the rate of inactivation is described by a log-linear decay model (Equation 2):

\begin{align*} conc_{soil}_{t_0} \times e^{-rate \times t_i} = conc_{soil}_{t_i} \end{align*}

Table 2. Inactivation rates for pathogens in Biosolids
Pathogen Model ln(rate) Rate(g/day) ln(max) Max rate(g/day)
Ascaris log-normal -9.48 7.67E-05 -9.34 8.76E-05
(Griffiths, Jackson)          
Enterovirus log-normal -5.30 0.005 -2.81 0.06
(Enriquez et al., Henis, John, Lyon, Medema et al.)          
Salmonella log-normal -4.65 0.00959 -2.86 0.0575
(Enriquez et al., Henis, John, Lyon, Medema et al.)          
           

Inactivation of Salmonella is best described by a two-stage die-off model that accounts for inactivation under varying environmental conditions (Equations 4 and 5) (Rogers et al.).

\begin{align*} N_t = N_0 e^{-k_1t}  for t < t' \end{align*}

\begin{align*} N_t = N_0e^{-k_1t'}e^{-k_2(t-t')} for t >= t' \end{align*}

One city has been chosen from each country to provide data for representative environmental conditions that will be used in the inactivation model for that country. The environmental conditions will be represented by Topeka, Kansas (U.S.), Monterrey (Mexico), and Porto Feliz (Brazil). Representative temperature and soil moisture conditions for each country and the corresponding parameters for this model are shown in Table 3.

Table 3. Country-specific parameters for a two-stage model for Salmonella inactivation and regrowth (Rogers, et al., 2011)
Location Conditions k1 k2 t’
Porto Feliz, Brasil: moist & warm 80% FC, 25 C 0.33 0.065 14
Topeka, Kansas: moist & cool 80% FC, 10 C 0.39 0.041 21
Monterry, Mexico: dry & warm 60% FC, 25 C 0.84 0.09 5.35
 

Variability in exposure was explored by Monte Carlo analysis to model the distributions of Salmonella in biosolid-amended soils under varying environmental conditions at different time points post-application (Figure 2).

Figure 2. Monte Carlo analysis of Salmonella levels in biosolid amended soils after application

Dose Response Assessment

Adjusted incidental ingestion of biosolids for occupational workers and the general population ranges uniformly from 0.05 to 0.62 grams/person/day. All discrete calculations assumed consumption of 0.48 grams/person/day (Kumar, USEPA). The dose of a pathogen at time t is equal to the concentration of pathogen per gram of amended soil multiplied by the grams of soil ingested per contact.

Assumptions

As with any risk assessment, a number of assumptions underlie the model selected. These assumptions are below. Additional assumptions for the model are shown in Table 4.

  • Fraction of pathogens that are transferred to soil when land applied with biosolids is 100% of those pathogens in the biosolid (Brooks et al., Haas et al.)
  • Adjusted incidental ingestion of biosolids for occupational workers and the general population ranges uniformly from 0.05 to 0.62 grams/person/day. All discrete calculations assumed consumption of 0.48 grams/person/day (Kumar, USEPA).
  • The decay of pathogens in land applied biosolids can be described by log-linear (Ascaris and enterovirus) and two-stage dieoff (Salmonella) models.
  • The mixing of topsoil and biosolids during application results in a final concentration of biosolids in amended soils ranging from 1-50%.
  • Exposure to biosolids occurs 30 or more days after land application for the general population
  • Enivronmental conditions in each of the three countries are represented by one city in an agricultural area, for which temperature and soil moisture data are available.
  • The decay of pathogens in land applied biosolids follows an exponential decay model, with the value of k varying according to an exponential distribution based on literature decay values for each pathogen.
  • Exposure to biosolids occurs 0, 15 and 30 days after land application.
  • The mixing of topsoil and biosolids during application results in a final concentration of biosolids in amended soils ranging from 1-50%.
  • US Ascaris concentration is point value at lower threshold of detection .
Table 4. Assumptions in model formation
Parameter Value Unit
Application Method Slinger and disk incorporation None
Biosolids application rate 2.57 dry tons biosolids/acre
Temperature 83 Fahrenheit
Time of soil ingestion after application (general public) 31 days
Time of soil ingestion after application (occupational) 0 days
 

Pathogen-specific dose-response models were selected for each pathogen under consideration. The parameters of the models chosen for each organism are shown in Table 5. The dose-response relationship for Salmonella is best described by the β-Poisson model. The parameters of the Salmonella model were determine by pooling several dose-response experiments using different strains of Salmonella (Haas, et al., Hornick, Hornick, McCullough, McCullough, McCullough, Meynell and Meynell). Since there is a lack of data on specific types of Salmonella present in biosolids, and the diverse inputs into biosolids may mean that diverse strains are present, a model representing several different strains is preferred for representing the dose-response relationship. The relationship for enterovirus can potentially be described by three different models: an exponential model derived from a porcine challenge study of two types of enterovirus (Cliver), a β-Poisson model from a human challenge study of echovirus 12 (Schiff), and a β-Poisson model from a human challenge study of rotavirus (Ward). Risk from all three dose-response models was compared, and it was found that porcine enterovirus models dose response more conservatively than human echovirus model but less dramatically than the human rotavirus model (Figure 3).

Figure 3. Comparative change in risk over time for 3 candidate dose response models for enterovirus infection

Ascaris is best described by the β-Poisson model. There are no human or animal challenge dose-response studies for Ascaris or helminth eggs; the dose response model was obtained from an epidemiologic study of A. lumbricoides infection in children under 15 conducted in the Mezquital Valley, Mexico (Mara, Navarro et al.).

Table 5. Dose Response Model Parameters for Selected Pathogens (Cliver, Haas, et al., Navarro, et al.)
Agent Best fit model Optimized parameters LD50/ID50 Host Agent strain Route Dose units endpoint
Salmonella (pooled) β-Poisson α= 0.3126 4.92×101 human multiple oral CFU positive stool culture
    N50 = 2.36×104            
Enterovirus exp k = 3.74×10-3 1.85×102 pig porcine, PE7-05i oral PFU infection
Ascaris β-Poisson α= 0.104 - human Ascaris lumbricoides oral HO stool parasite tests
 

Model Parameters

The starting point was a deterministic model, followed by incorporation of Monte Carlo simulation to estimate risk of infection. First, a general model was created that did not incorporate environmental factors in the analysis of pathogen decay rates. An extension of the initial model was then created in which pathogen decay rates were a function of environmental variables, including temperature and soil moisture (based on rainfall data). Model parameters are shown in Table 6.

Table 6. Parameters of final risk model
Description Symbol Value Units
Concentration of microbes in treated biosolids conc_trt -- microbe specific unit / g
Biosolid fraction in topsoil (Gerba et al., 2002) bs_frac 0.01-0.5 ratio of biosolid to soil
Time of soil ingestion after biosolids application (assumption) tsoilconc. -- days
Rate of decay (Table 2) rate -- fraction of microbes/day
Infective fraction (assumption) finf 1 fraction of microbes
Amount of soil ingested (USEPA) soil_ingest 0.05-0.62 g/day/person
Concentration in soil after application= (calculated) conc_soil_t0 -- microbes/g
Concentration in soil on day of contact=30 (calculated) conc_soil_ti -- microbes/g
Dose (calculated) dose -- microbes
Risk of infection (calculated)   -- risk
 

Literature values were used for the distribution of pathogen concentrations in biosolids in the US, and a fixed effects model was created based on wastewater treatment plant sampling data for pathogen concentrations in Brazil.

Risk Characterization

The final models compare the risks of each of the three pathogens between two countries. Using 1 in 10,000 (1×10-4) as the acceptable risk level, Enterovirus risk is minimal in both the US and Brazil. The risk of Ascaris infection in Brazil is above the benchmark value even 30 days after application (Figure 4). The risks of Salmonella (Figure 6), and Enterovirus (Figure 5) infection are under the acceptable risk threshold in the U.S. and Brazil, but NOT in Mexico. This is true from time 0 to 35 days post-application. The time point 30 days post-application is of interest because U.S. biosolids regulations mandate no contact with an application site for 30 days after application of Class B biosolids (USEPA). These results suggest that 30 days is insufficient to reduce the risk of Class B biosolid application in some countries, mainly driven by the risk of Ascaris infection.

Figure 4. Monte Carlo analysis of Ascaris infection risks at time 0, 15 and 30 days post-application

Figure 5. Monte Carlo analysis of Enterovirus infection risks at time 0, 15 and 30 days post-application

Figure 6. Salmonella infection risks at time 0, 15 and 30 days post-application (error bars too wide to display on graph)

Sensitivity Analysis

Sensitivity analysis shows that risk distributions are highly sensitive to:

  • Biosolid fraction in soil
  • Amount of soil ingested
  • Concentration of pathogen in treated biosolid before application.
  • Salmonella and enterovirus are sensitive to day of consumption

The risks of Salmonella infection under varying environmental conditions were modeled (Figure 7). In the U.S. and Brazil, the risk of infection declines between 0, 15, and 30 days, with the risk in Brazil remaining slightly higher that the risk in the U.S. The risk of Salmonella infection in Mexico was consistently higher than Brazil or the U.S. at all three time points, starting at 10-3 at day 0. The results suggest that the environmental conditions in Mexico may be such that the inactivation of Salmonella over 30 days is barely sufficient to bring infection risk under the 10-4 threshold.

Figure 7. Risks of Salmonella in biosolids over time under varying environmental conditions influencing the rate of bacterial inactivation

Risk Management & Communication

Risk management options for the land application of biosolids include:

  • Requirements for treatment resulting in pathogen reduction
  • Access restrictions on land where biosolids are land applied, allowing time for pathogen inactivation before people are permitted to have contact with land
  • Personal protective equipment for workers to reduce pathogen contact during application of biosolids

These results suggest that the main driver for infectious disease risks from biosolids exposure is the presence of Ascaris and the resulting risk of Ascaris infection, which is endemic in many areas of the world. However, effective control of Ascaris risk in biosolids may be challenging to achieve solely by focusing on biosolids treatment and contact restrictions that allow for inactivation over time after land application. One important higher-level control measure would be mitigation of disease in the population; in countries with high Ascaris burden and thus high inputs into sewage, it may be difficult to control Ascaris solely through the sewage treatment process. However, improved treatment techologies, such as thermophilic digestion, should also be prat of a plan to reduce the risks of biosolids application. Continued campaigns to reduce contributions from animals and more stringent site access restrictions are also important. In a country like Brazil, highly treated Class A biosolids may be perceived as having requirements that are too restrictive and difficult to achieve, and thus most biosolids produced are of Class B quality. However, it appears that Class B biosolids regulations are not protective enough for Ascaris in countries like Brazil and Mexico. It is necessary to and review/revise regulations to allow beneficial reuse while continuing to protect public health. Overall, this risk assessment suggests that:

  • 30 days is insufficient to reduce the risk of Class B biosolid application.
  • After 30 days:
    • Ascaris risk remains greater than 10-4.
    • Salmonella risk is minimal in the US and Brazil but NOT in Mexico.
  • Enterovirus risk is minimal in both the US and Brazil.
  • Further research on contamination of soil from biosolids throughout a multi-year period is warranted to reduce uncertainty in the risk estimate.

 

 

References