Groundwater contamination by microbiological and ...

EI-01885; No of Pages 9

ARTICLE IN PRESS Environment International xxx (2009) xxx–xxx

Contents lists available at ScienceDirect

Environment International j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / e n v i n t

Groundwater contamination by microbiological and chemical substances released from hospital wastewater: Health risk assessment for drinking water consumers Evens Emmanuel a,⁎, Marie Gisèle Pierre b,c, Yves Perrodin d a

Laboratoire de Qualité de l'Eau et de l'Environnement, Université Quisqueya, BP 796 Port-au-Prince, Haïti Laboratoire d'Analyse des Matériaux, Université Quisqueya, BP 796 Port-au-Prince, Haïti Association Haïtienne Femmes Science et Technologie, Université Quisqueya, BP 796 Port-au-Prince, Haïti d Laboratoire des Sciences de l'Environnement, École Nationale des Travaux Publics de l'État, Rue Maurice Audin, 69518 Vaulx-en-Velin, France b c

a r t i c l e

i n f o

Article history: Received 20 March 2008 Accepted 23 January 2009 Available online xxxx Keywords: Hospital effluents Faecal coliforms Organic pollutants Heavy metals Septic tanks Karstic aquifers Groundwater Drinking water Health risks

a b s t r a c t Contamination of natural aquatic ecosystems by hospital wastewater is a major environmental and human health issue. Disinfectants, pharmaceuticals, radionuclides and solvents are widely used in hospitals for medical purposes and research. After application, some of these substances combine with hospital effluents and, in industrialised countries, reach the municipal sewer network. In certain developing countries, hospitals usually discharge their wastewater into septic tanks equipped with diffusion wells. The discharge of chemical compounds from hospital activities into the natural environment can lead to the pollution of water resources and risks for human health. The aim of this article is to present: (i) the steps of a procedure intended to evaluate risks to human health linked to hospital effluents discharged into a septic tank equipped with a diffusion well; and (ii) the results of its application on the effluents of a hospital in Port-au-Prince. The procedure is based on a scenario that describes the discharge of hospital effluents, via septic tanks, into a karstic formation where water resources are used for human consumption. COD, Chloroform, dichlomethane, dibromochloromethane, dichlorobromomethane and bromoform contents were measured. Furthermore, the presence of heavy metals (chrome, nickel and lead) and faecal coliforms were studied. Maximum concentrations were 700 NPP/100 ml for faecal coliforms and 112 mg/L for COD. A risk of infection of 10− 5 infection per year was calculated. Major chemical risks, particularly for children, relating to Pb(II), Cr(III), Cr(VI) and Ni(II) contained in the ground water were also characterised. Certain aspects of the scenario studied require improvement, especially those relating to the characterisation of drugs in groundwater and the detection of other microbiological indicators such as protozoa, enterococcus and viruses. © 2009 Elsevier Ltd. All rights reserved.

1. Introduction In developing countries (DP), the average demand for water by hospitals is 500 L per bed per day (Laber et al., 1999). In addition to this high demand for drinking water is the requirement for specific waters such as physiological solution, sterilised water and serums (Emmanuel et al., 2001). This consumption of water by hospitals, which far exceeds the minimum household consumption of 100 L per inhabitant per day (Gadelle, 1995), gives rise to large volumes of wastewater. Indeed, the chemical substances used in hospitals for healthcare and medical research are frequently found in liquid effluents (Kümmerer, 2001) that are discharged in the same way as classical urban effluents into the communal drainage network without prior treatment (Emmanuel et al., 2005a). Physicochemical and microbiological characterisation studies performed on hospital effluents in several industrialised countries

⁎ Corresponding author. Tel.: +509 3423 4269; fax: +509 2221 4211. E-mail address: [email protected] (E. Emmanuel).

have highlighted the presence of pathogenic microorganisms, some of which are multi-resistant to antibiotics (Leprat, 1998), heavy metals (USEPA, 1989a; Leprat, 1998; Emmanuel et al., 2005a), radioisotopes (Erlandsson and Matsson, 1978), organohalogens, stemming in particular from the use of bleach on organic compounds present in effluents (Emmanuel et al., 2004a), and drug residues (Richardson and Bowron, 1985; Gartisser et al., 1996). Some of these pollutants, especially drug residues and organohalogens, are frequently discharged from sewage plants after having undergone little degradation (Kümmerer, 2001). Platinum originating from the excreted cancerostatic platinum compounds (CPC) cisplatin, carboplatin and oxaliplatin has been found in hospital wastewater (Lenz et al., 2005). Hospital wastewaters are complex mixtures (Boillot and Perrodin, 2008), capable of generating major environmental problems, as they are 5 to 15 times more toxic than classical urban effluents (Panouillères et al., 2007). With the occasional exception of radioelements, nearly all the substances generally identified in the hospitals of industrialised counties are used in the hospitals of developing countries. In some of these countries, the wastewater produced by hospitals is usually

0160-4120/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.envint.2009.01.011

Please cite this article as: Emmanuel E, et al, Groundwater contamination by microbiological and chemical substances released from hospital wastewater: Health risk assessment for drinking water consumers, Environ Int (2009), doi:10.1016/j.envint.2009.01.011

ARTICLE IN PRESS 2

E. Emmanuel et al. / Environment International xxx (2009) xxx–xxx

discharged into septic tanks equipped with diffusion wells. This type of discharge can lead to the pollution of groundwater resources used intensively for drinking water by the population. The impact of septic tanks on groundwater resources has been reported in the literature (McQuillan, 2004). The presence of pollutants produced by hospitals in groundwater therefore makes it necessary to consider possible risks relating to the use of these waters for human consumption. The effects of the compounds concerned on aquatic ecosystems have been the subject of many studies (Kümmerer et al., 1997; Halling-Sørensen et al., 1998; Sprehe et al., 1999; Kümmerer, 2001; Jolibois et al., 2002; Emmanuel et al., 2004a, 2005a,b; Boillot and Perrodin, 2008; Panouillères et al., 2007). However, few have dealt with the global risk for human health linked to the consumption of groundwater treated partially after simultaneous exposure to the different pollutants present in hospital effluents. In the specific geographic context of tropical countries where, in aquatic environments, temperature favours the propagation of pathogenic microbes, there is not only a acute risk of morbidity linked to pathogenic agents, but also the possibility of the appearance of chemical risks (carcinogenic or non carcinogenic) due to the contamination of groundwater resources by chemical substances present in hospital wastewater. The aim of this study is to present: (i) the steps of a procedure formulated to assess human health risks linked to hospital effluents discharged into a septic tank equipped with a diffusion well; and (ii) the results of its application to a hospital in Port-au-Prince. 2. Formulation of a methodology to evaluate health risks linked to hospital effluents 2.1. The general approach of evaluating health risks The National Research Council (1983) defines the assessment of risks as the activity that evaluates the toxic properties of a chemical product and the conditions of human exposure to this product, in view to observing the reality of human exposure and characterising the

nature of the effects that may result. The objective of this approach is to present explicitly to different health authorities, environmental protection organisations and all the parties concerned the elements of analysis on which decision-making should rely. The general approach to evaluating health risks is composed of four main steps: the identification of the danger, the study of the dose-response relation, the estimation of exposure, and the characterisation of the risk (NRC, 1983). Each of these steps corresponds to a research phase that gathers existing data from previous studies and data specifically generated for this study. 2.2. Presentation of study site The site selected in Haiti is an emergency hospital with a capacity of 63 beds. The hospital uses Cidex® to disinfect its medical equipment. This product is composed of 2.4% glutaraldehyde [molecular formula (C5H8O2), structural formula (CHO-(CH2)3-CHO)], and 97.6% inert substances (materials). Surfaces and other equipment are disinfected with chlorine (sodium hypochlorite). The liquid discharges of the different services are discharged into the hospital drainage system which does not collect rainwater discharges. The effluents collected are divided in three septic tanks where they undergo primary treatment that consists in separating large solid materials. The effluents of these tanks are discharged directly into a diffusion well embedded in a matrix consisting of a saturated area and a non saturated area. The groundwater resources are used for drinking water. This aquifer provides more than 50% of the drinking water supply to the population of the Port-au-Prince Urban Community (PPUC) i.e. 3 million people (Emmanuel et al., 2004b). A synthetic description of this scenario is presented in Fig. 1. The scenario highlights the existence of a private drinking water supply network (DWS) and an individual drainage system. It reproduces the DWS mode and the wastewater management of more than 15% of the population of the Urban Community. 54% of its drinking water needs are satisfied while 38% of its needs for solid waste collection are satisfied (OPS/OMS, 2001). The difference for the DWS is ensured by private companies and individual supply systems.

Fig. 1. Graphic representation of the scenario studied.


ARTICLE IN PRESS E. Emmanuel et al. / Environment International xxx (2009) xxx–xxx

Wastewaters (domestic and industrial) generated by this urban area are most often discharged into a drainage canal or managed by individual drainage systems. An important objective of septic tank design is to achieve hydraulically quiescent conditions that favour settling by gravity of heavy solid particles inside the tank. The settled material forms a layer of sludge on the bottom of the tank that must be removed periodically (Franceys et al., 1992). The removal of solids by settling can achieve high efficiency (Majumder et al., 1960). However, much depends upon the retention time, the inlet and outlet arrangements, and the frequency of sludge extraction. Large surges of flow entering the tank may cause temporarily high concentrations of suspended solids in the effluent owing to disturbance of the solids that have already settled (Franceys et al., 1992). Furthermore, the chemical compounds, especially organohalogenic compounds and drug residues, released by the septic tank and discharged into the natural environment by the diffusion well can be a major source of pollution of the host environment, in particular the soil and groundwater. In the case where the environmental conditions that permit the degradation of these substances do not exist, hospital pollutants risk remaining in the natural environment for a long time, thereby representing a risk in the short, medium and long terms for the species living in these ecosystems (Emmanuel et al., 2002). The scenario, shown in Fig. 1, permits taking into account the possible risk for human health that could result from the ingestion of partially or non treated water taken from the water tables. In this context, the discharge of hospital effluents into the soil and the possible use of the groundwater for DWS needs may contribute, among other things, to the existence of infectious and chronic diseases in the population living in the study area. In the classical version of the methodology used to assess the risk to human health of the action of a substance on a target organ, the risk characterisation phase is performed once exposure has been estimated. A decision-making step has been included at this stage of the study so as to avoid continuing the study if no danger is observed after characterising the exposure. It aims at comparing the maximum concentrations obtained in the hospital well-water to the threshold values prescribed by international regulations for drinking water. For any chemical substance with ratio Cn / Nq b 1 (Cn: pollutant concentration in groundwater; Nq: drinking water quality standard) and for all

3

concentrations of faecal coliforms NPP b 1 for 100 mL, the risk is considered as negligible and the procedure is stopped. On the contrary, for any ratio Cn/Nq N 1 and for any concentration of faecal coliforms NPP N 1 for 100 mL, the method recommends passing on to the following steps of the health risk assessment (Fig. 2). The approach taken is based on the hypothesis that the pollutants contained in the hospital effluents act independently and that no interaction or combined effects occur between them inside the human organism. 2.3. Identification of danger To evaluate the health risk of hospital effluents on the quality of the water intended for human consumption, we have formulated an approach based on detecting in the groundwater the contaminants (“risk tracers”) found most frequently in wastewater generated by hospitals. The risk tracers chosen were: faecal coliforms, metals and organohalogens. The main information reported in the literature on the toxicity of the tracers selected for this study is summarised below. 2.3.1. Faecal coliforms These are indicators or markers of faecal pollution in water. Infectious diseases are mainly transmitted by human and animal excreta, particularly faeces. Contamination can occur via diseased persons and carriers of germs in the community, who contaminate the water supply with pathogenic microorganisms. The consumption of this water or its use to prepare food or for washing and even its inhalation in vapour and aerosol form can lead to infection (OMS, 1994). 2.3.2. Heavy metals Regarding environmental compartments, heavy metals constitute an ecological and human health issue, since heavy metals do not undergo biological degradation, contrary to certain organic pollutants (Emmanuel et al., 2007). Metals exert biological effects that can be beneficial or harmful. Many metals such as Fe, Cu, Co, Mn, Zn, and Cr are essential for humans, and deficiency states with clinical abnormalities have been identified (Caussy et al., 2003). However, high doses of these essential elements can also cause toxic effects. Other metals such as Hg, Pb, Cd, and As are not known to be essential for any animals (Académie des Sciences, 1998). However, whether a metal in the environment causes an adverse effect depends on exposure and

Fig. 2. Flowchart formulated for the method of evaluating health risks relating to hospital effluents and the measures to be taken.


ARTICLE IN PRESS 4


bioavailability and on how much of it enters the body and reaches the critical target organ(s). Therefore, when investigating an outbreak attributable to a metal it is important to gather information on its fate, transport, and distribution in the environment (Caussy et al., 2003). Of the 7 metals (Cr, Cu, Pb, Hg, Ni, Ag, Zn) considered as priority pollutants detected in hospital effluents (USEPA, 1989a,b), 3 of them (Cr, Pb and Ni) were selected for this study. Indeed, copper and zinc are trace elements with known biological functions essential for all living organisms. Given the state of development of healthcare in Haiti, the effluents of clinics can contain micro-droplets of mercury from broken thermometers. Nonetheless, analysing mercury, in spite of its high toxicity for living organisms, cannot be carried out in Haiti given the current performance of the equipment available. As for silver, it was considered as non significant, since the septic tank chosen for this assessment does not receive the water used to rinse X-ray prints loaded with silver residues from the radiology department of the site selected. The toxicity of chromium is directly dependent on the valence state, with hexavalent chromate Cr(VI) and trivalent chromate Cr(III) being of the greatest interest (Académie des Sciences, 1998). Oral bioavailability varies with valence state, with Cr(VI) being more readily absorbed. Cr (VI) can be broken down into Cr(III) within the acidic environment of the stomach (ATSDR, 2000). Acute exposure to chromium is indicated by immediate irritation of the eye, nose, throat, and respiratory tract, which results in burning, congestion, epistaxis, and coughing. Ulceration, bleeding, and erosion of the nasal septum mark chronic exposure. Coughing, chest pains, dyspnea, and chromium-induced asthma indicate exposure to soluble chromium products (Robson, 2003). If chronic exposure is suspected, in conjunction with weight loss, coughing, and hemoptysis, then the development of bronchogenic carcinoma can be expected. Dermatological manifestations include painless, slow-healing ulceration of the fingers, knuckles, and forearms. Ingestion is marked by nausea, vomiting, abdominal pain, prostration, and death associated with uremia (Lewis, 1997). What is more, there is evidence that extreme exposure to chromium causes renal and hepatic damage (Robson, 2003). Nickel is insoluble in water. However, when in very fine particle form, it ionises as Ni(II) in water and in biological fluids such as blood. During exposure by oral ingestion, the major effects observed have been the death of a child after ingestion of 570 mg of nickel/kg (Daldrup et al., 1983) and intestinal problems such as nausea, abdominal cramps and diarrhoeas (Sunderman et al., 1989). Immunological, haematological, hepatic, renal, and genotoxic effects on embryonic development and reproduction have been reported depending on the mode of penetration in the organism (ATSDR, 1993). Both the respiratory tract and immune systems are sensitive targets of Ni(II) toxicity, as it causes chronic bronchitis, emphysema, and impaired lung function (ATSDR, 2003). Drinking water is one of the major sources of human exposure to lead (INERIS, 2002; Fertmann et al., 2004). Lead particularly targets the nervous system, blood and kidney (ATSDR, 1999; INERIS, 2002). Long-term lead exposure may generate irreversible functional and morphological renal changes (Christensen, 1995), distal motor neuropathy and possibly seizures and coma (Robson, 2003). Infants and small children are more sensitive to the effects of lead, which moreover is transported through the placenta to the foetus (Cleymaet et al., 1991; Christensen, 1995). Lead accumulation in foetuses and small children might cause developmental disruption in terms of neurological impairment characterized by a decrease of cognitive faculties, which can be reversible or not, evaluated by psychomotor tests such as the verbal IQ (Intellectual Quotient) test (Académie des Sciences, 1998). The period when IQ is most affected is from birth to about 4 years of age (Watt et al., 2000). 2.3.3. Organohalogens Sodium hypochlorite (NaOCl, CAS no. 7681-52-9 and EC no. 017011-00-1), a solution containing from 12.5 to 25% active chlorine gas

(Cl2), has a wide range of biomedical applications related to its biocide properties (Brondeau et al., 2000; Emmanuel et al., 2004a). When NaOCl is added to water and wastewater, the solution reacts readily with biological materials (including proteins and nucleotide bases) to produce a variety of organic halogenated compounds (USEPA, 1989a) named AOX (adsorbable organic halogens), which are mostly lipophilic, persistent, and toxic in aquatic environments (SalinojaSalonen and Jokela, 1991). Hospital effluents reveal the presence of organochlorine compounds in high concentrations (Leprat, 1998). Up to 10 mg/L of AOX was found in the effluents of the hospitalization services of a university hospital centre (Gartisser et al., 1996). Bromodichloromethane, Chloroform and 1,1-Dichloromethane were considered as priority pollutants detected in hospital wastewater (USEPA, 1989a,b). Epidemiological studies examining the health effects of populations exposed to chlorination by-products have elevated rates of bladder (Cantor et al., 1998; McGeehin et al., 1993), colon-rectum (Hildesheim et al., 1998), and brain (Cantor et al., 1999; Wilkins et al., 1979; Lawrence et al., 1984; Flaten, 1992) cancers. As summarized in a recent meta-analysis by Morris et al. (1992), higher exposure to chlorination by-products in drinking water may be related to an approximately 10 to 40% excess risk of cancers of the bladder and colon-rectum. 2.4. Exposure assessment Study of groundwater contamination essentially requires geological and hydrogeological investigations of the site in which the hospital examined is situated. The geology and hydrogeology of the region in which the hospital in this study is located are dominated by a karstic aquifer. The main characteristics of the latter are that they have irregular pores, cracks, fractures and conduits of various shapes and dimensions. This type of physically and geometrically heterogeneous structure gives rise to complex hydraulic conditions, with hydraulic parameters subject to considerable variations in time and space. After a precipitation, the rapid and turbulent replenishment of the groundwater occurs via the drainage of high volumes of non-filtered water through large channels (Denić-Jukić and Jukić, 2003). Data relating to boring the well to supply the hospital with water: the different geological formations of the non saturated area and the well shaft plan are shown in Fig. 3. Filter screens are located as several points along the shaft. These accessories involve the catchment of several aquifers during pumping hours. These filter screens can also be used to drain water when the pump is idle, thereby possibly leading to a transfer of hospital pollutants to deeper levels of the groundwater. This information and the general data collected during the geological and hydrogeological studies of the region in which the site is located were used to formulate a theoretical synthesis of the circulation of water flows and thus put forward the hypothesis of a probable connection between the hospital effluents discharged into the diffusion well of the hospital and the hospital well-water. However, given the complexity of the transfer of pollutants in the karstic area, it was not possible to model and quantify this exchange exactly. Under these conditions, and taking into account the possibility of taking measurements on the well water, the assessment of exposures was performed by analysing the groundwater rather than analysing the water of the diffusion well followed by modelling the transfers in the soil to the groundwater. This choice, which may appear obvious in terms of the realism of the final estimated exposure concentrations, cannot always be performed in risk assessments since it may, in certain cases, change the procedure used for a given source of pollution (in this case the discharge of hospital effluents via the diffusion well) in an assessment of the risks linked to the groundwater in general (whatever the origin of its pollution). In this case, since the connection between the discharge of hospital effluents and the groundwater was highly


ARTICLE IN PRESS E. Emmanuel et al. / Environment International xxx (2009) xxx–xxx

5

Fig. 3. Well casing plan for the hospital water supply borehole.

probable, and no other major pollution had been identified in the area, the assessment of the risks relating to hospital effluents could therefore be performed by analysing the groundwater chosen. 3. Materials and methods Several liquid effluent sampling campaigns were performed from 2002 to 2005 on one of the hospital's 3 septic tanks (tank serving the hospitalisation department with a capacity of 22 beds) and on the DWS well water of the hospital. The samples of the different campaigns were carried out at the following points: 1. the discharge line of the hospital DWS well, 2. the inlets of the diffusion well (effluents of the septic tank), The effluents were sampled between 7:30 and 11:30 a.m. All the samples were placed in plastic containers with a volume of 1 l. These recipients were rinsed 3 times with the water to be examined. To fill the recipients, we used an improved manual sampling method consisting in preparing an average sample over 100 min (1 h 40 min) with a sampling time step of 100 ml every 10 min. pH measurements were performed on all the samples taken using this method. The recipients containing the samples taken at the points chosen were carefully labelled and conserved at 4 °C. Once taken, the samples were transported to the laboratory in less than an hour. The French and European protocols described in “L'analyse de l'eau” (Rodier, 1996) and the American protocols described in “Standard Methods for water and wastewater” (APHA, 1995) were used for analysing the parameters measured. These parameters correspond to easily dosable selected risk tracers and to the different additional measurements (pH, total suspended matters (TSM), Conductivity, COD, Nitrates, etc.) intended to improve knowledge of the pollution levels of the liquid compartments in question. The analysing of AOX in the hospital well water was done by coulombometry as per standard NF EN 1485. The organohalogenic solvents, especially bromochloromethane, chloroform, dichlorobromomethane, dibromochloromethane, bromoform, and dichloromethane were dosed by HS/GC/MS as per standard NF EN ISO 10301. Faecal coliforms were determined by using the microplate method (NF T 90-433). The metals were measured using protocol ISO 11 885 on

samples filtered at 0.45 µm, treated with pure nitric acid (pH b 2) and studied under ICP-AES (Inductively Coupled Plasma-Atom Emission Spectroscopy). Chrome, nickel and lead were determined respectively at the following wavelengths: 267.716; 231.604; 220.353 nm. Cr(VI) analysing was done as per the HACH 8023 method by using a HACH 2010 spectrophotometer on filtered samples. The analysing principle is based on the reduction of Cr(VI) to Cr(III) linked to the oxidation of diphenylcarbazide into diphenylcarbazone. The latter forms with the resulting Cr(III) a pink-violet compound absorbent at 540 nm. Total chrome was measured using the filtered samples, as per the HACH 8024 method, using a HACH 2010 spectrophotometer. This is an alkaline oxidation method with hypobromite in the visible region at 540 nm. 3.1. Risk characterisation 3.1.1. Definition of populations exposed by studying the type of exposure identified The populations concerned by this risk assessment were the staff and patients residing at the hospital (about 200 people including about 50 infants) as well as people living nearby and who consume the same groundwater. A surface area of 20 ha was used for the assessment (i.e. a total of about 4000 people (Lhérisson, 1999) including 1600 infants under 10 years of age). The main exposure path identified and studied was the consumption of drinking water for all the parameters selected. 3.1.2. Chemical risks The equations proposed by INVS (2000) were applied to estimate the risks linked to heavy metals and organohalogens. Indeed, for a given chemical substance and exposure path, the general equation for calculating the maximum daily intake (MDI), administered by exposure vector “i”, is as follows (INVS, 2000): MDIi =

Ci⁎Q i⁎ET BW⁎WT

ð1Þ

Where MDI (Maximum Daily Intake) is the proportion of substance absorbed per day of exposure, Ci is the concentration of the toxic


ARTICLE IN PRESS 6


Table 1 Results of the physicochemical and bacteriological characterisation of the effluents of the septic tank Parameters

Unit

Average

Minima

Maxima

Standard deviation

N

Limit of detection

pH Conductivity Chlorides TSM NO3 COD Pb Ni Crtotal Faecal c.

U µS/cm mg/L mg/L mg/L mg/L µg/L µg/L µg/L NPP/100 mL

7.67 313 179 24.57 3.28 510 12.5 100 300 7.5 × 104

7.43 297 172 12.80 0.89 425 10 30 180 2 × 105

7.86 324 191 33.20 4.65 618 15 180 440 119 × 104

0.17 11.63 7.70 8.47 1.51 70 3.54 62.85 105 36 × 104

5 5 5 5 5 5 5 5 5 5

– – – – – – 0.012 0.006 0.006

substance in the polluted medium “i”, Q the quantity of this vector brought into daily contact with the organism by the route considered (expressed in L/day for water media), ER is the exposure rate (without unit), ET is the exposure time (in years), BW is the body weight (in kg) and WT the time of weighing. TP is the duration (in years) over which the amount is weighed. In this formula, by convention, the time of weighing is identical to the exposure time (TW = EW) to reach the threshold: MDI is close to an annual average without consideration of the total period of exposure. The calculation of risk for man consists in correlating the exposure level data with the dose-response relations. The risks are estimated differently according to whether the substances act or not with an effect threshold. To calculate the MDI, a total consumption of 2 L/day was used for adults, while 0.75 L/day was used for infants. Body weights of 70 kg and 10 kg respectively were allocated to adults and infants under 10 years old. 3.1.2.1. Substances acting with a threshold effect. For compounds acting with a threshold effect, a danger quotient (DQ) was characterised as a function of Eq. (2), which expressed the ratio between the exposure assessment (MDI) and the Acceptable Daily Intake (ADI). MDI and ADI are expressed in mg/kg/day. R=

MDI ADI

R MDI ADI

ð2Þ risk Maximum daily intake Acceptable daily intake

This numerical value is not, strictly speaking, a risk and the assessment here is qualitative: a ratio lower than 1 means that the population exposed is theoretically out of danger, whereas a quotient higher than 1 means that the toxic effect can occur without it being possible to predict its probability (INVS, 2000). Three levels were considered in the framework of this assessment to interpret the results: Rb1 R=1 RN1

L: Low M: Moderate H: High

SCRmex = SIRmex ⁎n

ð4Þ

3.1.3. Biological risks The “Beta-Poisson” distribution model (Haas et al., 1999) was used to characterise the biological risk generated by E. coli: − α d 1=α 2 −1 P ðdÞ = 1 − 1 + N50 d N50 α

ð5Þ

Exposure dose Average infecting dose equal to 8.60 × 107 for E. coli Parameter of probability function equal to 0.1778 for E. coli.

4. Results and discussion 4.1. Physicochemical and bacteriological characterisation of the septic tank effluents The pH of the samples studied during the different campaigns fluctuated between 7.43 and 7.86, underscoring the existence of a slightly alkaline environment. The variation noted in the pH measurements of hospital effluents is less than 1 unit pH. The results of the physicochemical characterisation of the septic tank effluents studied are summarised in Table 1. The results obtained for electric conductivity [297–324 µm/cm] indicate the presence of mineral substances in moderate concentrations. Values for chlorides vary from 172 to 191 mg/L. These concentrations are higher than the values [30–100 mg/L] usually found in domestic wastewater (Metcalf and Eddy, 1991). On the other hand they are lower than the values [63–359 mg/L] obtained by Emmanuel et al. (2004a) for the effluents of a tropical disease unit of a hospital in a large city in southeast France. The concentrations obtained for metals are mostly lower than the threshold values set by international regulations on discharges of wastewater containing these pollutants (European Commission, 1998a). The lead, nickel and chrome contents of the wastewaters studied are far higher than the concentrations measured for the hospital effluents by Emmanuel et al. (2005a). In the samples of effluents from the septic tank, COD concentrations vary from 425 to 618, much higher than the threshold value of 125 mg/L recommended by Directive 98/15/EC for industrial wastewaters. The information available in the literature highlights the very low concentration of bacterial flora present in the effluents (Bernet and Fines, 2000). In this study, the maximum value obtained for bacterial flora was 119 × 104 NPP/100 ml. This number is lower than the 1 × 108 NPP/100 ml generally found in urban wastewater (Metcalf and Eddy, 1991), and higher than the flora 2.4 × 103 NPP/100 ml counted in the hospital effluents by Emmanuel et al. (2005a). The low concentration of bacterial flora and the ecotoxicity of hospital effluents have been attributed by some authors to the presence of drugs and disinfectants in them (Deloffre-Bonnamour, 1995). During the sampling campaign of 2002, 5 samples were collected at the inlet and outlet of the septic tank in the framework of this study. The results obtained from this campaign permit appreciating the capacity of the septic tank to retain certain pollutants contained in the effluents studied. The efficiency of the septic tank for the primary treatment of hospital effluents was assessed by using Eq. (6), Efficience =

ðCxe − Cxs Þ ⁎100 Cxe

ð6Þ

Where Cxe in the concentration of parameter “x” at the inlet of the septic tank (input), and Cxs is the concentration of the same parameter at its outlet (effluent). Table 2 supplies the first results on the efficiency of the septic tank in retaining certain substances and organic material in particular. Table 2 Efficiency of the septic tank

3.1.2.2. Substances acting without a threshold effect. For carcinogenic and mutagenic substances, acting without a threshold effect, the risk assessment is genuinely quantitative. The probability of occurrence of cancer for the lifetime of the subjects exposed, which is added to the basic risk not linked to this exposure, is called surplus individual risk (SIR): it is calculated for each path, by multiplying the USR (unit surplus risk) by the “lifetime” total average daily dose (INVS, 2000): SIRmex = MDImex ⁎USRmex

The product of this risk by the number of the people exposed provides the surplus collective risk (SCR). It represents an estimation of the number of surplus cancers linked to the exposure studied which should occur during the lifetime of this group of individuals.

ð3Þ

Parameters

Unity

Average concentration at the inlet of the septic tank

Average concentration at the outlet of the septic tank

Efficiency %

Conductivity TSM Chlorides NO3 COD Pb Ni Crtotal

µS/cm mg/L mg/L mg/L mg/L µg/L µg/L µg/L

330 40.24 188 4.03 611 31.25 128 234

313 24.57 179 3.28 510 12.5 100 300

5 39 5 17 17 39 45 –


ARTICLE IN PRESS E. Emmanuel et al. / Environment International xxx (2009) xxx–xxx Table 3 Results of the physicochemical and bacteriological characterisation of the groundwater Parameters

Unit

Average

Minima

Maxima

Standard deviation

N

pH Conductivity Chlorides NO3 COD Pb Ni Crtotal Cr6+ Faecal c.

U µS/cm mg/L mg/L mg/L µg/L µg/L µg/L µg/L NPP/100 mL

7.354 298 244 120 82 25 100 326 7 629

6.7 267 200 109 59 3.26 b LD 18 b LD 300

8.01 330 231 148 112 40 250 470 10 989

0.45 22 13 13 17 17 96 88 5 204

10 10 10 10 10 5 5 5 5 10

LD: Limit of detection. The efficiency of the septic tank in treating hospital effluents is poor. Indeed, regarding the retention of biodegradable substances their efficiency is in the region of 35% (Gschlössl et al., 1999). Only 17% of COD concentration is retained. Likewise for NO3, it is 39% for TSM (Total suspended matters) and Pb, 45% for Ni. For chromium, it has been observed that the concentration in the influent is lower than the concentration in the effluent of the septic tank. The results obtained for Pb and Ni are probably due to the distance separating the inlet pipe from the bottom of the tank. No efficiency was detected for total chrome. Considering the volume used by the hospitals, this distance can also influence the settling of solids and biochemical reaction cycles.

4.2. Results of physicochemical and bacteriological analyses of groundwater The results of the physicochemical and bacteriological analyses of groundwater are summarised in Table 3. The pH of the samples studied during the campaigns varied from 6.74 to 8.01. Although higher than 1 unit of pH, this variation lies within the limits proposed by (OMS, 1994) for drinking water. COD, a non conventional pollutant, is sometimes used to characterise the global concentration of organic pollutants (Rodier, 1996). COD can be used to provide data on the existence of organic substances that can only be oxidised by aerobic biological processes (USEPA, 1993). High concentrations of COD were measured [59–112 mg/L] in the groundwater. Although this parameter was not considered directly as a risk tracer, it should nonetheless be emphasised that its minimal concentration was far higher than the threshold value of 5 mg/L prescribed by the Belgian standard (DGRNE, 1998) for water intended for human consumption, probably expressing the presence of high concentrations of organic substances in the groundwater of the site studied. A maximal concentration of 0.09 mgAOX/L was measured in the hospital well water. Concentrations of bromochloromethane and dichloromethane were both b 1 µg/L. As for the other organohalogenic solvents studied, the following maximal concentrations were obtained: chloroform [1.2 µg/L], dichlorobromomethane [2.6 µg/L], dibromochloromethane [5.2 µg/L] and bromoform [4.6 µg/L]. Metals Pb [10–40 µg/L], Ni [15–250 µg/L] and Cr [18–470 µg/L] were detected in the hospital well water. Maximal concentrations of metals measured in this water were higher than those obtained for the same parameters in the septic tank (Table 1). Most of the heavy metals present in water occur in ionic forms. Heavy metal ions are acknowledged to be highly toxic and can accumulate in water and soils (Siegel, 2002; Bradl, 2004; Qin et al., 2006; Bhattacharyya and Gupta, 2007). Since the values obtained for lead, nickel and chromium in this study are very high; consumers of this water are exposed to major health problems. Despite the relative abundance of clay in the soils of Port-au-Prince, several other factors may explain the high concentrations of heavy metals in the groundwater studied. Indeed, clay minerals in soils act as natural scavengers by removing and accumulating the contaminants in the water that passes through the soil, through ion exchange and adsorption (Bhattacharyya and Gupta, 2007). However, inorganic colloids, metal speciation, metal concentration, pH, solid, solution mass ratio and contact time are also very important in controlling the adsorption of heavy metals and their distribution between soil and water (Bradl, 2004). The information reported in the literature on the individual behaviour of selected heavy metals states that pH plays an important role in the adsorption and precipitation behaviour

7

of chromium, lead and nickel in soils. Cr(III) adsorption increases with increasing pH and soil organic matter content whereas there is a decrease of competing cations and dissolved organic ligands in the solution. The increased adsorption of Cr(III) with increasing pH is caused by the cation exchange reactions of hydrolyzed substances (Bradl, 2004). The adsorption of Cr (VI) onto various adsorbents is a function of pH. Adsorption increases with decreasing pH due to the protonation of hydroxyl groups (Rai et al., 1989). Bhattacharyya and Gupta (2007) show that is not possible to carry out adsorption experiments with Pb(II) at pHN 6.0, or for Ni(II) at pHN 8.0, due to precipitation of the metals as hydroxides, thereby introducing uncertainty into the interpretation of the results. Since the pH of the septic tank effluents ranged from 7.43 to 7.86, it appears necessary in the future to carry out experimental studies on soil pH and pursue research into the geological and chemical characteristics of the selected site in order to understand the different mechanisms governing the transfer of heavy metals to the groundwater. Nitrate contamination is generally observed in low yielding wells and in close proximity to potential point waste sources but may also arise from diffuse sources. Most nitrates found in natural waters is of anthropogenic origin, originating from organic and inorganic sources, the former including wastewater, waste discharges and the latter comprising mainly artificial fertilizers. Nitrate is present naturally in groundwater in low concentrations, typically in the range 5–9 mg/L NO3 (OMS, 1996). The nitrate concentrations measured [100 to 148 mg/L] were much higher than those usually detected in natural waters, as well as the threshold value of 50 mg/L for drinking water (OMS, 1996; European Commission, 1998b). Degradation of groundwater quality by nitrates and its consequences on human health have been the subject of many studies (Fraser and Chilvers, 1981; Laftouhi et al., 2003; Saadi and Maslouhi, 2003; Ibnoussina et al., 2006). Experimental studies have shown that the presence of high concentrations of nitrates in groundwater is linked to the nature of the soil on the one hand and, on the other, to the effect of continuous discharges of wastewater and successive irrigations (Ibnoussina et al., 2006). Poss and Saragoni (1992) described a leaching phenomenon, using experimental tests in a column, that highlights the effect of the quantity of water on nitrate transport, resulting in an appreciable loss in the upper soil layers and accumulation in the deeper ones, where the nitrates finally enter drainage water. Ibnoussina et al. (2006) showed that the aquifers of sandy soil with low clay and organic contents are particularly sensitive to nitrate contamination. Since the clay content of the soil of the site studied is high, there should be no sign of nitrates being released into the groundwater. Therefore it appears that preferential paths exist in the non saturated area of the site studied (Fig. 1) leading, during heavy rainfalls, to rapid leaching of the nitrates stored in the soil into the groundwater. In addition, excess consumption of nitrates in the human diet can lead to health risks that include methamoglobinaemia in infants (blue baby syndrome) and possible carcinogenic hazards. The toxicity of nitrates to humans is thought to result solely from its reduction to nitrites. Nitrites are involved in the oxidation of normal haemoglobin into methaemoglobin which is unable to transport oxygen to the body's tissues (OMS, 1996). According to international regulations E. coli must be non detectable in a sample of 100 ml (OMS, 1996; European Commission, 1998b). A number of faecal coliforms (E. coli) varying from 300 to 989 NPP/100 were counted in the hospital well water. In addition to this high bacterial contamination, the literature reports considerable circulation of Cryptosporidium sp. oocystes identified in the surface water and in the water supply intended for human consumption in several districts of Port-au-Prince (Brasseur et al., 2002). The strong presence of E. coli in the well water highlights the existence of a major source of faecal contamination. The effluents from septic tanks, agricultural organic waste and dumps are the sources most usually considered as the potential causes of groundwater contamination. Soils can be effective in removing microorganisms by predation, filtration and absorption (Dussart-Baptista et al., 2003). Also, soils contribute to the adsorption of nitrates contained in wastewater and the effluents of irrigation (Ibnoussina et al., 2006). They are also one of the factors controlling the adsorption of heavy metal ions (Kerndorf and Schnitzer, 1980). However, natural, unprotected areas, such as with karstic, sandy and gravely terrains and extremely vulnerable fractured aquifers, allow the rapid movement of contaminants into the groundwater with minimal attenuation, leading to the possibility of high risk situations. Although the presence of clayey soils, tillage and peat will, in many instances, retard the vertical migration of microbes, preferential secondary flow paths such as cracks in clay materials can bypass the filtering effect of such soils. The values measured in the hospital well water emphasise substantial contamination of the groundwater resources of the study area and lead to the assumption of a high

Table 4 Comparison of measured or estimated maximal concentrations with the threshold values Parameters

Units

Measured maximal concentrations (Mmc)

Threshold values (Tv)

References (Tv)

Ratios Mmc/Tv

Bromoform Chloroform Dibromochloromethane Dichlorobromomethane Crtotal Cr(VI) Ni Pb Faecal coliforms

µg/L µg/L µg/L µg/L µg/L µg/L µg/L µg/L NPP/100 mL

4.6 1.2 5.2 2.6 470 10 250 40 700

100 200 100 60 50 50 20 10 NPP b 1 in 100 mL

OMS (1996) OMS (1996) OMS (1996) OMS (1996) OMS (1996) EC (1998b) OMS (1996) OMS (1996) OMS (1996)

b1 b1 b1 b1 N1 b1 N1 N1 N1


ARTICLE IN PRESS 8


Table 5 Calculated risk for substances with threshold effect (metals) Substances CAS #

ADI Risk quotient Level MDI (mg/kg day)− 1 (mg/kg day)− 1 Adult Infant

Cr(III) Ni Pb

16065-83-1 2.03 7440-02-0 2.57 7439-92-1 0.18

5.32 2.89 0.46

Adult Infant Adult Infant 1.5 0.02 0.0035

1 129 50

4 145 132

M E E

E E E

health risk for the consumers of water from this water table. Given the importance of this aquifer to the water supply of PPUC (Emmanuel et al., 2004b), it will be necessary in future to identify the different sources of pollution, extend the physicochemical characterisation of these waters and identify the mechanisms governing the transfer of pollutants from the surface to the groundwater. 4.3. Assessment of dangers for human health This step, which was introduced in the general methodology of health risk assessment (NRC, 1983), consists in comparing (Table 4) the concentrations measured in the groundwater for the risk tracers selected to the guide values for drinking water prescribed by international regulations. With the exception of organohalogenic solvents and Cr(VI), all the other physicochemical parameters were found at values higher than the threshold values for water intended for human consumption. The ratio between the maximal number of faecal coliforms present in the groundwater and the bacteriological quality standard of water intended for human consumption was far greater then 1. These results confirm the existence of a danger to the health of the population consuming this groundwater and thus the need to pursue these studies. 4.4. Characterisation of risks for human health 4.4.1. Microbiological risks The infectious risk calculated for the faecal coliforms measured in the groundwater gave a result of 10− 5 infection per year (1 person per 100 000). This risk is lower than that of 10− 4 infection per year per person, considered in the United States as the tolerable risk level linked to the consumption of drinking water (Hass, 1996). Furthermore, in a tropical country where temperature favours the development and growth of pathogenic germs, 10− 5 infection per year per person appears to be non negligible risk. The approach presented here leads to a quantitative assessment of infectious risks. It should be improved regarding the risk linked to Cryptosporidium and to the enteroccocus which are currently very efficient indicators for estimating faecal pollution. Indeed, in certain districts of Port-au-Prince risk of infection by oocystes of Cryptosporidium ranging from 1% to 5% and from 1% to 97% have been calculated respectively for the immunocompetent population; and for the immunodepressed population according to the load of oocystes contained in the water consumed (Bras et al., 2007). It is therefore advisable, in the framework of managing risk to human health related to the contamination of the fresh water resources of Port-au-Prince by hospital effluents, to henceforth verify these initial results by measuring other indicators of faecal pollution of water such as faecal enteroccocus, Cryptosporidium sp., other parasites and the entero-viruses. 4.4.2. Chemical risks Regarding lead and its inorganic derivatives, the USEPA (1989a,b) and ATSDR (1999) do not propose any value for carcinogenic effects. Age, state of health, ponderal load in lead, and length of exposure are all factors that influence lead metabolism and complicate the establishment of these values (INERIS, 2002). The USEPA (1998a,b) notes that there is insufficient information to determine the carcinogenic effects of Cr(III) in water and in foods. As for nickel, little is known as yet about human exposure by the oral path to water contaminated by this metal. Thus risk calculations for metals were performed by using the method usually employed for noncarcinogenic substances, i.e. substances acting with a threshold effect. Table 5 shows the risk levels calculated for non-carcinogenic substances. With the exception of chrome, which has a moderate risk for adults, all the other metals have a high risk for both adults and children. In spite of the prevailing uncertainty about whether or not substances such as Cr(III) and Ni(II) have carcinogenic properties, the results obtained show that the population is exposed to a considerable chemical risk.

5. Conclusion The quality of a risk assessment depends on the validity of the different data used to perform it: physicochemical, toxiclogical, epidemiological data, etc. as well as the construction of realistic scenarios (Zmirou and Perrodin, 1999). However, numerous uncertainties remain regarding the approach pursued. In the case of the scenario studied, mention can be made in particular of the choice of pollutants and the toxicological data on

the non or carcinogenic character of the pollutant minerals in the drinking water. These uncertainties are almost always present in health risk management. Risk assessment remains a scientific activity that permits predicting the probable effects of pollutants in human beings. Nonetheless, it is obvious that the results of these assessments permit the adoption of policies designed to avert a worst-case situation. The scenario presented here leads to a quantitative assessment of human health risks. In the case of the scenario studied, it should be noted that the degradation of groundwater is due to human activities. The content measured for the mineral pollutants is far higher than the values naturally present in these water resources. It is necessary in the future to validate these initial results by further and more in-depth assessments of risks to human health including, among other things, the determination of other biological indicators of faecal pollution of water, especially by Cryptosporidium spp., faecal enterococcus and the enteroviruses, coupled with epidemiological studies. References Académie des Sciences. Contamination des sols par les éléments traces: les risques et leur gestion. Rapport N° 42. Paris: Lavoisier Tec&Doc; 1998. 440 pp. APHA. Standard Methods for the Examination of Water and Wastewater. Washington D.C.: APHA, AWWA, WEF, 1995, 19th ed. 1220 pp. ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological Profile for Nickel. Altanta, GA: U.S. Department of Health and Human Services; 1993. (http:// www.atsdr.cdc.gov). ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological profile for lead. Atlanta, GA: U.S. Department of Health and human Services; 1999 (http://www. atsdr.cdc.gov). ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological profile for chromium. Atlanta, GA: U.S. Department of Health and human Services; 2000 (http://www.atsdr.cdc.gov). ATSDR (Agency for Toxic Substances and Disease Registry). Toxicological profile for nickel. Atlanta, GA: U.S. Department of Health and human Services; 2003 (http:// www.atsdr.cdc.gov). Bernet S, Fines M. Effluents du CHU de CAEN: Etude qualitative et quantitative de la flore microbienne et recherche de bactéries multirésistantes. Poster. Quatrième journée du Réseau Régional d'Hygiène de Basse-Normandie, Caen; 2000. Bhattacharyya KG, Gupta SS. Adsorptive accumulation of Cu(II), Co(II), Pb(II), and Ni(II) from water on montmorillonite: influence of acid activation. J Colloid Interface Sci 2007;310:411–24. Boillot, C., Perrodin, Y., Joint-action ecotoxicity of binary mixtures of glutaraldehyde and surfactants used in hospitals: use of the toxicity index model and isoblogram representation. Ecotoxicology and Environmental Safety, 2008;71(1):252–259. Bradl HB. Adsorption of heavy metal ions on soils and soils constituents. J Colloid Interface Sci 2004;277:1–18. Bras A, Emmanuel E, Lilite O, Brasseur P, Pape J-W, Raccurt C. Evaluation du risque dû à Cryptosporidium sp présent dans l'eau de boisson à Port-au-Prince, Haïti. Environ Risques Santé 2007;6(5):355–64. Brasseur P, Eyma E, Verdier RI, Agnamey P, Liautaud B, Dei-Cas E, et al. Circulation des oocystes de Cryptosporidium dans les eaux de surface et de distribution par adduction publique à Port-au-Prince. In: Emmanuel E, Vermande P, editors. Actes du Colloque International « La gestion intégrée de l'eau en Haïti »; 2002. p. 173–6. Brondeau MT, Falcy M, Jargot S, Miraval S, Protois JC, Reynier M, et al. Eaux et extraits de Javel. Fiche toxicologique N° 157 de l'INRS. Cahiers de notes documentairesHygiène et sécurité du travail, vol. 178; 2000. p. 11–5. Cantor KP, Lynch CF, Hildesheim ME, Dosemeci M, Lubin J, Alavanja M, et al. Drinking water source and chlorination byproducts. I. Risk of bladder cancer. Epidemiology 1998;9:21–8. Cantor KP, Lynch CF, Hildesheim ME, Dosemeci M, Lubin J, Alavanja M, et al. Drinking water source and chlorination byproducts in Iowa. III. Risk of brain cancer. Am J Epidemiol 1999;150:552–60. Caussy D, Gochfeld M, Gurzau E, Neagu C, Ruedel H. Lessons from case studies of metals: investigating exposure, bioavailability, and risk. Ecotoxicol Environ Saf 2003;56:45–51. Christensen JM. Human exposure to toxic metals: factors influencing interpretation of biomonitoring results. Sci Total Environ 1995;166:89–135. Cleymaet R, Collys K, Retief DH, Michotte Y, Slop D, Taghon E, et al. Relation between lead in surface tooth enamel, blood, and saliva from children residing in the vicinity of a nonferrous metal plant in Belgium. Br J Ind Med 1991;48:702–9. Daldrup T, Haarhoff K, Szathmary SC. Fatal nickel sulfate poisoning. Beitr Gerichtl Med 1983;41:141–4. Deloffre-Bonnamour, N., Les rejets des établissements de santé: des effluents liquides aux déchets solides. Mémoire de Maîtrise, Université Claude Bernard-Lyon1, Institut Universitaire Professionnalisé, Génie de l'Environnement-Ecodéveloppement, Lyon; 1995. Denić-Jukić V, Jukić D. Composite transfer functions for karst aquifers. J Hydrol 2003;274:80–94. DGRNE (Direction Générale des Ressources Naturelles et de l'Environnement). “Arrêté de l'Exécutif régional wallon du 20 juillet 1989 relatif à la qualité de l'eau distribuée


ARTICLE IN PRESS E. Emmanuel et al. / Environment International xxx (2009) xxx–xxx par réseau, modifié par l'arrêté de l'Exécutif régional wallon du 21 février 1991”. DGRNE, Division de l'Eau. Namur, 1998. Dussart-Baptista L, Massei N, Dupont J-P, Jouenne T. Transfer of bacteria-contaminated particles in a karst aquifer:evolution of contaminated materials from a sinkhole to a spring. J Hydrol 2003;284:285–95. Emmanuel, E., Blanchard, J.-M., Keck, G., Perrodin, Y., Caractérisation chimique, biologique et écotoxicologique des effluents hospitaliers. Déchets Sciences et Techniques, revue francophone d'écologie industrielle, 2001, 22:31–33. Emmanuel E, Perrodin Y, Keck G, Blanchard J-M, Vermande P. Effects of hospital wastewater on aquatic ecosystem. Proceedings of the XXVIII Congreso Interamericano de Ingenieria Sanitaria y Ambiental. Cancun, México, 27–31 de octubre; 2002. CDROM. Emmanuel E, Blanchard J-M, Keck G, Vermande P, Perrodin Y. Toxicological effects of sodium hypochlorite disinfections on aquatic organisms and its contribution to AOX formation in hospital wastewater. Environ Int 2004a;30:891–900. Emmanuel E, Joseph O, Fanfan PN, Vermande P, Winiarski T. Trend analysis of the groundwater salinity of the Cul-de-sac aquifer in Haiti. Proceedings of the XXIX Congreso Interamericano de Ingenieria Sanitaria y Ambiental. San Juan, Puerto Rico, 22–28 de agosto; 2004b. CDROM. Emmanuel E, Perrodin Y, Blanchard J-M, Keck G, Vermande P. Ecotoxicological risk assessment of hospital wastewater: a proposed framework for raw effluents discharging into urban sewer network. J Hazard Mater 2005a;A117:1–11. Emmanuel E, Hanna K, Bazin C, Keck G, Clement B, Perrodin Y. Fate of glutaraldehyde in hospital wastewater and combined effects of glutaraldehyde and surfactants on aquatic organisms. J Environ Int 2005b;31:399–406. Emmanuel E, Angerville R, Joseph O, Perrodin Y. Human health risk assessment of lead in drinking water: a case study from Port-au-Prince Haiti. Int J Environ Pollut 2007;31(3/4):280–91. Erlandsson B, Matsson S. Medically used radionucides in sewage sludge. Water Air Soil Pollut 1978;9:199–206. European Commission. Directive 98/15/EEC amending Council Directive 91/271/EEC: Urban wastewater treatment. Brussels; 1998a. Off. J. of European Communities N° L 67/29–30 (7 March 1998). European Commission. DIRECTIVE 98/83/CE du 3 novembre 1998 relative à la qualité des eaux destinées à la consommation humaine, Bruxelles,1998b. JO L 330 du 5.12.1998, p. 32. Fertmann R, Hentschel S, Dengler D, Jan U, Lommel A. Lead exposure by drinking water: an epidemiological study in Hamburg. Int J Hyg Environ Health 2004;207:235–44. Flaten TP. Chlorination of drinking water and cancer incidence in Norway. Int J Epidemiol 1992;21:6–15. Franceys R, Pickford J, Reed R. A Guide to the Development of on-Site Sanitation. Geneva: WHO92 4 154443 0; 1992. 229 pp. Fraser P, Chilvers C. Health aspects of nitrate in drinking water. Sci Total Environ 1981;18:103–16. Gadelle F. Le monde manquera-t-il bientôt d’eau? Sécheresse, 1995, vol. 6, pp. 11–15. Gartisser St, Brinkler L, Erbe T, Kümmerer K, Willmund R. Contamination of hospital wastewater with hazardous compounds as defined by § 7a WHG. Acta Hydrochim Hydrobiol 1996;24(2):90–7. Gschlössl T, Scheleypen P, Steinmann C, Melzer A. Constructed wetlands and lagoons for wastewater treatment — a field of cooperation between biologist and engineer. Limnologica 1999;29(1):103–4. Hass CN. Acceptable microbial risk. J Am Water Works Assoc 1996;88(12). Haas CN, Rose JB, Gerba CP. Quantitative Microbial Risk Assessment. New York: Wiley; 1999. 449 pp. Halling-Sørensen B, Nielsen N, Lanzky PF, Ingerslev F, Holten-Lützhøft, Jørgensen SE. Occurrence, fate and effects of pharmaceutical substances in the environment — a review. Chemosphere 1998;36:357–93. Hildesheim ME, Cantor KP, Lynch CF, Dosemeci M, Lubin J, Alavanja M, et al. Drinking water source and chlorination byproducts. II. Risk of colon and rectal cancers. Epidemiology 1998;9:29–35. Ibnoussina M, Mohamed El Haroui, Abdellatif Maslouhi. Expérimentation et modélisation de la lixiviation de l'azote nitrique dans un sol sableux. C. R. Geosciences 2006;338:787–94. INERIS (Institut National de l’Environnement Industriel et des Risques). Plomb et ses dérivés, in Fiche de données toxicologiques et environnementales des substances chimiques. Paris: INERIS; ERIS-DRC-01-25590-ETSC-Api/SD-N 00df257, p.90. INVS (Institut National de Veille Sanitaire). Guide pour l'analyse du volet sanitaire des études d'impact. Saint-Maurice Cédex: Institut de Veille Sanitaire; 2000. 49 pp. Jolibois B, Guerber M, Vassal S. Glutaraldehyde in hospital wastewater. Arch Environ Contam Toxicol 2002;42:137–44. Kerndorf H, Schnitzer M. Sorption of metals on humic acid. Geochim Cosmochim Acta 1980;44:1701–8. Kümmerer K. Drugs in the environment: emission of drugs, diagnostic aids and disinfectants into wastewater by hospitals in relation to other sources — a review. Chemosphere 2001;45:957–69. Kümmerer K, Meyer M, Steger-Hartmann T. Biodegradability of the anti-tumour agent Ifosfamide and its occurrence in hospital effluents and communal sewage. Water Res 1997;11:2705–10. Laber J, Raimund H, Shrestha R. Two-stage constructed wetland for treating hospital wastewater in Nepal. Water Sci Technol 1999;40:317–24. Laftouhi N, Vanclooster M, Jalal M, Witam O, Aboufirassi M, Bahir M, et al. Groundwater nitrate pollution in the Essaouira Bassin (Morocco). C R Geosci 2003;335:307–17. Lawrence C, Taylor P, Trock B, Reilly A. Trihalomethanes in drinking water and human colorectal cancer. J Natl Cancer Inst 1984;72:563–8. Lenz K, Hann S, Koellensperger G, Stefanka Z, Stingeder G, Stingeder N, et al. Presence of cancerostatic platinum compounds in hospital wastewater and possible elimination by adsorption to activated sludge. Sci Total Environ 2005;345:141–52.

9

Lhérisson G. Les caractéristiques générales de l'agglomération de Port-au-Prince métropolitain. In: Holly G, editor. Les problèmes environnementaux de la région métropolitaine de Port-au-Prince. Port-au-Prince: Commission pour la commémoration du 250e anniversaire de la fondation de la ville de Port-au-Prince; 1999. p. 17–26. Leprat P. Les rejets liquides hospitaliers, quels agents et quelles solutions techniques? Les Assises Nationales QUALIBIO 1998 « Santé et Environnement Hospitalier », Caen; 1998. p. 10–3. Lewis R. Metals. In: Ladou J, editor. Occupational and environmental medicine. New York: McGrawHill; 1997. p. 405–39. (Chapter 27). Majumder N, Prakasam T, Suryaprakasam M. A critical study of septic tank performance in rural areas. India: The institution of engineers; 1960. McGeehin MA, Reif JS, Becher JC, Mangione EJ. Case-control study of bladder cancer and water disinfection methods in Colorado. Am J Epidemiol 1993;138:492–501. McQuillan D. Ground-water quality impacts from on site septic system. Proceedings, National Onsite Wastewater Recycling Association, 13th Annual Conference, Albuquerque, NM, November 7–10; 2004. p. 6–18. Metcalf, Eddy. Inc. Wastewater Engineering, Treatment, disposal, and reause. 3rd ed. New York: McGraw-Hill; 1991. Revised by Tchobanoglous G., Burton F.L. Irwin. Morris RD, Audet AM, Angelillo IF, Chalmers TC, Mosteller F. Chlorination by-products and cancer: a meta-analysis. Am J Public Health 1992;82:955–62. NRC (National Research Council). Risk Assessment in the Federal Government: Managing the Process. Washington DC: National Academy Press; 1983. 191 pp. OMS (Organisation Mondiale de la Santé). Directives de qualité de l'eau de boisson. Recommandations, 2è ed.Genève: OMS; 1994. p. 8–30. OMS (Organisation Mondiale de la Santé). Directives de qualité de l'eau de boisson. Critères d'hygiène et documents à l'appui, 2ème édGenève: OMS; 1996. 1024 pp. OPS/OMS (Organisation Panaméricaine de la Santé/Organisation Mondiale de la Santé) Synthèse des taux de couverture des besoins en approvisionnement en eau potable et assainissement au 31 décembre 2000. Rapport WASAM 2000, OPS/OMS, Port-auPrince, 2001, 80 pp. Panouillères M, Boillot C, Perrodin Y. Study of the combined effects of a paracetic acidbased disinfectant and surfactants contained in hospital effluents on Daphnia magna. Ecotoxicology 2007;16:327–40. Poss R, Saragoni H. Leaching of nitrate, calcium and magnesium under maize cultivation on oxisol in Togo. Fertil Res 1992;33:123–33. Qin F, Wen B, Shan XQ, Xie YN, Liu T, Zhang SZ, et al. Mechanisms of competitive adsorption of Pb, Cu, and Cd on peat. Environ Pollut 2006;144:669–80. Rai D, Eary LE, Zachara JM. Environmental chemistry of chromium. Sci Total Environ 1989;86(1–2):15–23. Richardson ML, Bowron JM. The fate of pharmaceutical chemicals in the aquatic environment. J Pharm Pharmacol 1985;37:1–12. Robson M. Methodologies for assessing exposures to metals: human host factors. Ecotoxicol Environ Saf 2003;56:104–9. Rodier J. L'analyse de l'eau. 8e éd. Paris: DUNOD; 1996. 1384 pp. Saadi Z, Maslouhi A. Modeling nitrogen dynamics in the unsaturated soils for evaluating the nitrate contamination of the M'nasra groundwater. Adv Environ Res 2003;7 (4):803–23. Salinoja-Salonen MS, Jokela JK. Measurement of organic halogen compounds in urine as an indicator of exposure. Scand J Work Environ Health 1991;17:75–8. Siegel FR. Environmental geochemistry od potentially toxic heavy metals. Hiedelberg: Springer-Verlag; 2002. Sprehe M, Geiβen SU, Vogelpohl A. Treatment of AOX-containing wastewater from hospitals — degradation of iodized X-ray contrast medium. Korresp Abwasser 1999;46(4):548–58. Sunderman Jr FW, Hopfer SM, Sweeney KR. Nickel absorption and kinetics in human volunteers. Proc Soc Exp Biol Med 1989;191:5–11. USEPA (United States Environmental Protection Agency). Preliminary data summary for the hospitals point source category. Office of water regulations and standards. Washington, D.C.: Office of water, United States Environmental Protection Agency; 1989a. 76 pp. (EPA 440/1-89/060-n). USEPA (United States Environmental Protection Agency). Risk assessment guidance for superfund. Volume I: Human health evaluation manual (Part A). Washington, D.C.: United States Environmental Protection Agency; 1989b. 248 pp. USEPA (United States Environmental Protection Agency). Training manual for NPDES permit writers; 1993. Washington, D.C.: United States Environmental Protection Agency; EPA/833/B-93/003. USEPA (United States Environmental Protection Agency). Toxicological review for trivalent chromium (CAS No. 16065-83-1). Support of Summary Information on the Integrated Risk Information System (IRIS). Washington, D.C.: United States Environmental Protection Agency; 1998a. 51 pp. USEPA (United States Environmental Protection Agency). Toxicological review for hexavalent chromium (CAS No. 18540-29-9). Support of Summary Information on the Integrated Risk Information System (IRIS). Washington, D.C.: United States Environmental Protection Agency; 1998b. 77 pp. Watt GCM, Britton A, Gilmour HG, Moore MR, Murray GD, Robertson SJ. Public health implications of new guidelines of lead in drinking water: A case study in an area with historically high water lead levels. Food Chem Toxicol 2000;38:573–9. Wilkins J, Reiches N, Kruse C. Organic chemical contaminants in drinking water and cancer. Am J Epidemiol 1979;110:420–48. Zmirou D, Perrodin Y. Approche méthodologique de l'évaluation des risques pour la santé, des déchets stabilisés. Application à un scenario, Déchets — Sciences et Techniques, 16; 1999. p. 10–3.
