Chemosphere 84 (2011) 1336–1348

Contents lists available at ScienceDirect

Chemosphere

journal homepage: www.elsevier .com/locate /chemosphere

Detection of pharmaceutically active compounds in the rivers and tap waterof the Madrid Region (Spain) and potential ecotoxicological risk

Y. Valcárcel a,⇑, S. González Alonso a, J.L. Rodríguez-Gil b, A. Gil a, M. Catalá c

a Department of Preventive Medicine, Public Health, Immunology and Medical Microbiology, Faculty of Health Sciences, Rey Juan Carlos University,Avda Atenas s/n, E-28922 Alcorcón, Madrid, Spainb Centre for Toxicology, School of Environmental Sciences, University of Guelph, Bovey Building, Guelph, ON, Canada N1G2W1c Department of Biology and Geology, School of Experimental Science & Technology, Rey Juan Carlos University, Av. Tulipán s/n, E-28933 Móstoles, Madrid, Spain

a r t i c l e i n f o a b s t r a c t

Article history:Received 20 January 2011Received in revised form 5 May 2011Accepted 7 May 2011Available online 8 June 2011

Keywords:Pharmaceutically active compoundsPotential risk characterizationSurface waterTap waterIfosfamideVenlafaxine

0045-6535/$ - see front matter Crown Copyright � 2doi:10.1016/j.chemosphere.2011.05.014

Abbreviations: PhACs, pharmaceutically active comment plant; EU, European Union; DWTP, DrinkingMadrid Region; INE, Instituto Nacional de EstadísticaHQ, Hazard Quotient; DSSTP, DownStream Sewagmethanol; LC gradient, liquid chromatography; US EPtion Agency; MEC, measured environmental conceneffect concentration; UF, uncertainty factor; NOEC, ntion; TU, Toxic Unit; NE, northeast; EC50, half maxQ(SAR), quantitative structure–activity relationship; PConcentration;

PTU, Sum of Toxic Unit; EMEA, Europ

⇑ Corresponding author. Tel.: +34 914 888 891; faxE-mail address: yolanda.valcarcel@urjc.es (Y. Valcá

Concentrations of pharmaceutically active compounds (PhACs) in the order of ng L�1 to lg L�1 have beenreported worldwide in waste, fluvial and even drinking water, raising concern about the efficacy of thecurrently employed waste water treatments in the elimination of this kind of compounds. Despite rank-ing 29th in terms of population, Spain is currently the 8th country on pharmaceutical prescription withan expense of 14 � 109 euros in 2008. In this context, the aim of this study was to determine the presenceof 33 pharmaceutically active compounds in specific points of the main rivers of the Madrid Region (MR)as well as tap water samples from the metropolitan area of Madrid. Additionally, a screening level riskcharacterization by means of the Hazard Quotient (HQ) method was applied. A total of 25 pharmaceuticalcompounds and metabolites were detected in the 10 sampling points downstream the outlet of the majorSTPs of the MR. The highest concentrations were detected for the anticonvulsant carbamazepine and thestimulant caffeine. Concentrations for most of the analyzed compounds exceed levels previously reportedin the literature. Moreover, we report the highest concentration of the cytostatic ifosfamide, detected forthe first time in Spain in surface water. Preliminary risk characterization shows that a total of 16 com-pounds represent at least a low potential hazard based on their scored HQs, with five of them presentin a concentration that exceeds the predicted no effect concentration (PNEC). Toxic Units calculation indi-cates that for all the selected sampling points high hazard is anticipated from the presence of the ana-lyzed compounds in the measured concentrations (TUs > 10). Caffeine and cotinine were detected inall (10) the analyzed tap water samples. Carbamazepine and nicotine were detected in six and venlafax-ine in two samples. No studies venlafaxine in drinking water have been reported. These results clearlypinpoint the need for water quality monitoring and research in urban rivers, as well as the need forimproved water treatment techniques able to eliminate this kind of compounds from the effluent watersas well as from drinking water sources.

Crown Copyright � 2011 Published by Elsevier Ltd. All rights reserved.

1. Introduction

Pharmaceutically active compounds (PhCAs) have been knownto enter the environment for over 20 years (Daughton and Ternes,

011 Published by Elsevier Ltd. All r

pounds; STP, sewage treat-Water Treatment Plant; MR,(National Statistics Institute);e Treatment Plant; MeOH,A, US Environmental Protec-tration; PNEC, predicted noo observed effect concentra-imal effective concentration;EC, Predicted Environmentalean Medicines Agency.

: +34 914 888 955.rcel).

1999). In the last decade, advances in analytical chemistry have al-lowed for the detection of lower concentrations of chemicals inenvironmental samples, dramatically increasing the number ofmicropollutants (i.e. pollutants found in very small traces) de-tected in nature. As the number of detectable PhCAs increased,attention was drawn into this group and they soon becamerecognized as emerging pollutants (Daughton and Ternes, 1999;Heberer, 2002). A decade after this recognition, the increasinglyavailable information, is slowly pushing for the incorporation ofpharmaceuticals into environmental regulation as priority pollu-tants (Bottoni et al., 2010). However, large quantities of additionaldata concerning usage, environmental concentrations, fate, absorp-tion rates and toxicity (especially chronic and mixture toxicity) arestill needed to develop the accurate risk assessments necessary forthe implementation of acceptable management strategies.

ights reserved.

Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348 1337

Concentrations of PhCAs in the ng/L–lg/L range have been re-ported worldwide in effluents from sewage treatment plants (STPs),sediments, surface-, ground-, and even drinking water. PhACs canenter the environment through different routes (Fig. 1) (reviewedby Glassmeyer et al. (2009)). (1) After human consumption, non-metabolised medicines and biologically active metabolites canreach the sewer system. (2) Pharmaceutical industry effluents havebeen demonstrated to contain high levels of PhACs. (3) PhACs maybe excreted by livestock or contained in aquaculture effluents.Incomplete degradation of PhACs from sources (1), (2) and (3) insewage treatment plants (STPs) can lead to their discharge into nat-ural surface waters. (4) Inadequate disposal of unused medicinescan lead to landfill contamination or increased concentrations ofpollutants reaching STPs. (5) Sludge- (compost, biofilm) or man-ure-based soil amendments as well as the use of (6) reclaimed waterare also potential entry routes of PhACs into the environment.Finally, runoff from contaminated fields or poorly insulated landfillscould end up in surface waters and leachates could further contam-inate soils and underground waters.

Although the studies published state that low concentrationsof the pharmaceuticals found do not constitute an imminent riskfor public health (Khetan and Collins, 2007), more informationabout the effects that such substances can have on ecosystemsis becoming available. For instance, apart from affecting the kid-neys of mammals, the anti-inflammatory, diclofenac, has beenassociated with the disappearance of the Oriental White-backedVulture in India and Pakistan (Fent et al., 2006). According to theauthors of the study, this may possibly amount to an ecological

Fig. 1. Map of the study area and sample site localizations (DSSTP) in the Community of Mtreatment plants (STPs).

disaster on a par with the case of DDT. Evidence suggests thatthe beta blocker propranolol has effects on zooplankton and ben-thic organisms. Other effects that have been documented are thefeminisation of male fish caused by exposure to estrogens, thelack of mating among fish (Huggett et al., 2002) and the appear-ance of antimicrobial resistance (Kümmerer, 2004).

Roughly 3000 different PhCAs are commonly used in theEuropean Union (EU) (Gros et al., 2010). Spain is currently theworld’s 8th country in medicines prescription (Glassmeyeret al., 2009), despite being ranked 29th in terms of population(IMS Health 2008; http://www.imshealth.com). Spain spent morethan 14 � 109 € on pharmaceuticals in 2008. Among the differenttherapeutic groups consumed by the Spanish population, twocategories stand out from a public health point of view: antibi-otics and psychoactive medicines. Antibiotics are the third mostused human medicine, and constitute 70% of those used in vet-erinary medicine. In Cars et al. (2001) Spain is shown to be thesecond highest consumer (by volume) of antibiotics in Europe.This could be considered as one of the factors explaining ob-served high rates of microbial resistances (Ministerio de Sanidady Consumo, 1995; Ruiz et al., 2000; Sole et al., 2004).

The observed increase in the number of people with mentalhealth problems during recent years has pushed mental health intothe list of priority issues for the European Commision according tothe Public Health Program (2003–2008). Antidepressives, antipsy-cotics and antiepileptics are some of the top selling pharmaceuticaldrugs through retail pharmacies (October 2008) (IMS Health 2008;http://www.imshealth.com).

adrid and codes of the sampling points and characteristics of the monitored sewage

1338 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

Not included in the most prescribed groups, but associated withincreasingly prevalent pathologies are antihistamines, bronchodil-atators, or drugs used for gastrointestinal pathologies. Of specialconcern are antineoplastics, immunosupressors and thosecompounds with potential endocrine disrupting properties. Finally,nicotine and caffeine, and their respective metabolites cotinine andparaxanthine, have been very frequently detected and representinteresting indicators of anthropogenic contamination (Buergeet al., 2008).

The aim of this study is to determine the presence of 33pharmaceutically active compounds at specific points of themain rivers of the Madrid Region (MR) as well as in tap watersamples of the metropolitan area of Madrid. A screening levelrisk characterization by means of the Hazard Quotient (HQ)method will be applied to the obtained data.

2. Materials and methods

2.1. Description of the sampling site

With an area of 8028 km2 and occupying only 1.6% of the Span-ish territory, the Madrid Region (MR) has an estimated populationof 6 million inhabitants (13.5% of Spain), and is one of the mostdensely populated areas in Europe (676 inhabitants per km2). A to-tal of a 89.3% of the population is concentrated in a conurbationcomposed by the city of Madrid (54.2%) and the municipalities inthe metropolitan area (35.1%) (INEbase, 2007).

Characteristics of the sampling points are shown in Fig. 1.Tenout of the 145 sewage treatment plants (STPs) of the MR were se-lected according to the population served. Samples were collectedapproximately 100 m downstream of the point of emission of theSTP. Effluents discharged directly into the five main rivers of MR:Jarama (DSSTP2), Manzanares (DSSTP5 and 6), Guadarrama(DSSTP7), Henares (DSSTP9) and Tagus (DSSTP10); except for Ar-royo del Bodonal (DSSTP1) and Arroyo del Soto (DSSTP8), whichdischarged into creeks flowing into the Jarama and GuadarramaRivers respectively (González Alonso et al., 2010). Our aim is to de-tect maximum PhACs concentrations attained in natural surfacewaters, most likely produced by the localized discharge of STPs.This approach serves to two main goals:

(1) We can evaluate the relationship between pharmaceuticalconsumption and environmental occurrence.

(2) We believe that this approach is more respectful with the‘‘Precautionary Principle’’ and the potential risk estimationis more realistic especially when no comprehensive informa-tion on the occurrence of these substances exists.

River water grab samples were collected in October, 2007. Oneliter of water was collected in pre-rinsed amber glass bottles. Sam-ples were sent to the laboratory for analysis within the next 24 h.

The five main drinking water supply areas of the Madrid metro-politan area were selected for our study. Tap water grab sampleswere taken in pre-rinsed amber glass bottles between 25th and

Table 1Tap water characteristics.

Sample code Water supply area Population served

P-CHAM Colmenar 1724,029P-TIT Getafe 594,744P-USE Islas Filipinas 739,708P-LEG Retamares 433,627P-ALC Valmayor 846,406

30th, October 2007, in public places and private residences(Table 1).

2.2. Pharmaceuticals selected

The PhCAs analyzed in this study have been selected based onseveral criteria, including: interest for environmental and publichealth (tap water), high consumption by resident population andavailable techniques for detection.

Consequently, 33 PhCAs have been analyzed, belonging to thefollowing therapeutic groups: antibacterials, antidepressants/anti-epileptics/anxiolytics, antihistamines, antineoplastic/immunosup-pressants, bronchodilators/anti-asthma, gastrointestinal drugs,local anesthetics, and stimulants (Table 2).

2.3. Sample preparation

Preconcentration of the samples prior to chromatographic anal-ysis was performed by solid phase extraction (SPE), using commer-cial Oasis™ HLB cartridges (200 mg, 6 cc) from Waters (Mildford,MA, USA) and an automated ASPEC XL sample processor fromGilson (Villiers-le-Bel, France) for the purpose. The operationalprocedure has been described in detail elsewhere (Martinez Buenoet al., 2010). Briefly, the cartridges were preconditioned withMeOH (6 mL) and HPLC-grade deionised water (5 mL, pH adjustedto 8 with 20% NH4OH) at a flow rate of 1 mL/min. Aliquots of200 mL of sample (pH adjusted to 8) were then loaded into the car-tridge at a flow rate of 10 mL/min, rinsed with 5 mL of deionisedwater and finally eluted with 2 � 4 mL of MeOH at 1 mL/min.The extracts so obtained were finally evaporated until almost dry-ness, reconstituted with 1 mL of MeOH:water, 10:90 (v/v), filtered,and diluted 1:1 with MeOH:water (10:90) prior to analysis.

2.4. Liquid chromatography-Q-TRAP-mass spectrometry

Analysis of the selected pharmaceuticals was performed by a3200 QTRAP (quadrupole-ion trap) MS/MS system (Applied Biosys-tems, Ontario, Canada) using a turbo ionspray source in positiveand negative modes. Separation was performed in an Agilent Tech-nologies HPLC series 1100, equipped with a reversed-phase C-18analytical column (ZORBAX SB, 250 mm � 3.0 mm I.D.; 5 lm).For analysis in positive mode, the compounds were separatedusing acetonitrile (mobile phase A) and HPLC-grade water with0.1% formic acid (mobile phase B) at a flow rate of 0.2 mL/min. Alinear gradient progressed from 10% A to 100% A in 40 min, afterwhich the mobile phase composition was maintained at 100% Afor 10 min. The re-equilibration time was 15 min. Compounds ana-lyzed in negative mode were separated using acetonitrile (mobilephase A) and HPLC-grade water (mobile phase B) at a flow rateof 0.3 mL/min. LC gradient progressed from 30% A to 100% A in7 min, and was maintained at 100% A for 8 min. The re-equilibra-tion time was 10 min. The injection volume was 20 lL in bothmodes. Specific operational conditions were optimized and havebeen described elsewhere (Martinez Bueno et al., 2010).

Town District Sampling site

Madrid Chamartín Bus station tap waterTitulcia – Private residence tap waterMadrid Usera Private residence tap waterLeganés – Private residence tap waterAlcorcón – University tap water

Table 2Characteristics of analyzed pharmaceuticals.

Therapeutic group Pharmaceutical Cas Quantification limit

Antibacterials Anhydrous Azithromycin 83905-01-5 13Ciprofloxacin 85721-33-1 6Clarithromycin 81103-11-9 6Erythromycin 114-07-8 1Norfloxacin 70458-96-7 10Ofloxacin 82419-36-1 3Sulfamethoxazole 723-46-6 1Tetracycline 60-54-8 23Trimethoprim 738-70-5 2

Antidepressants Amitriptyline 50-48-6 1Citalopram(For Citalopram Hydrobromide) 59729-33-8 1Clomipramine 303-49-1 2Fluoxetine(For Fluoxetine Hydrochloride) 54910-89-3 1Paroxetine 61869-08-7 2Venlafaxine 93413-69-5 1

Antiepileptics Carbamazepine 298-46-4 1Carbamazepine 10,11-Epoxideb) 36507-30-9 1Primidone 125-33-7 5

Antihistamines Loratadine 79794-75-5 1Antineoplastics and immunosuppressants Cyclophosphamide 50-18-0 3

Ifosfamide 3778-73-2 1

AntiprotozoalsAnxiolytic sedatives hypnotics and antipsychotics Diazepam 439-14-5 1Bronchodilators and anti-asthma drugs Salbutamol 18559-94-9 1Gastrointestinal drugs Famotidine 76824-35-6 5

Lansoprazole 103577-45-3 8Omeprazole 73590-58-6 1Ranitidine 66357-35-5 2(For Ranitidine Hydrochloride)

Local anesthetics Mepivacaine 96-88-8 2

Supplementary drugs and other substances Cotinined) 486-56-6 2Nicotine 54-11-5 4Caffeine 58-08-2 4Paraxanthine (1,7-dimethylxanthine)c) 611-59-6 3

Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348 1339

2.5. Screening level risk characterization

The hazard quotient method was used for screening level riskcharacterization. Hazard quotients are defined by the US EPA asthe ratio of the potential exposure to the substance and the levelat which no adverse effects are expected (US EPA, 1997). A hazardquotient (HQ) was calculated for each individual chemical bydividing the measured environmental concentration (MEC) at eachsampling site by the predicted no effect concentration (PNEC). ThisPNEC was derived from the available aquatic toxicity data, apply-ing the pertinent uncertainty factors (UF). This way, HQ = MEC/PNEC, where PNEC = Toxicity benchmark/UF. This method is con-sistent with the European guidelines (EMEA, 2006) where the cal-culation of quotient between Predicted EnvironmentalConcentration (PEC) and the PNEC is needed during Tire A on aPhase II environmental risk assessment of medicinal products forhuman use.

If the HQ < 0.1, no adverse effect is expected. If 0.1 < HQ < 1, thehazard is low, but potential for adverse effects should be consid-ered; and if 1.0 > HQ < 10, some adverse effect or moderate hazardis probable. If HQ > 10, high hazard is anticipated (US EPA, 1998;EMEA, 2006).

As the purpose of this study was to conduct a screening-levelcharacterization, the measured concentrations of each chemicalfor each sampling site were directly applied as the MEC to the cal-culations. When the measured concentration of a compound in aspecific sampling site was below the quantification limit of theanalytical method, a value corresponding to ½ of this limit was

used as the MEC. An exhaustive literature review was performedto collect toxicity data for algae, cladocerans and fish for the stud-ied pharmaceuticals. Based on the accepted exposure scenario forthese kinds of compounds (Chronic exposure from STP’s effluents),European guidelines (EMEA, 2006) recommend the use of chronictoxicity data for the calculation of the PNECs. However, the avail-ability of chronic data, in the way of NOEC for these compoundsis currently limited. In absence of NOEC values, acute toxicity datawas used, and when no experimental toxicity data was available,predicted toxicity values from (Q)SAR models were selected (San-derson et al., 2004). Finally, the lowest value amongst the threegroups was selected as the toxicological benchmark value for thecalculations (Table 3).

Uncertainty factors where applied to this toxicological bench-mark value following the guidelines of the Technical GuidanceDocument on Risk Assessment of the European Union (UE, 1996).A UF of 1000 was applied when at least one short-term L(E)C50from each of the three evaluated trophic levels was available, aUF of 100 was applied when one long-term NOEC was availablefor either cladocerans or fish, and finally a UF of 50 was appliedwhen two long-term NOEC were available for species in two differ-ent trophic levels.

Hazard quotients act as a toxicity-normalized concentrationallowing for the evaluation of chemical mixture toxicity on siteswhere multiple chemicals of potential ecological concern cansimultaneously occur (Jones et al., 1999). In this context, HQs arecommonly referred as Toxic Units (TUs) in order to give a betterrepresentation of the normalized nature of the value.

Table 3Lowest available toxicological data for algae, cladocerans and fish. Predicted no effect concentrations (PNEC) for the lowest toxicological data are presented as well as theuncertainty factor (UF) applied for its calculation in each case. Presented reference corresponds to the data point selected for PNEC calculation.

Active ingredient Algae Cladocerans Fish Selected Data UF PNEC Reference

Amitriptyline 0.246* 1.034739 0.989* 0.246* EC50 1000 0.000246 Sanderson et al. (2004)Anhydrous Azithromycin 0.019 120 4.1* 0.019 EC50 1000 0.000019 Harada et al. (2008)Caffeine 339.3 159.62 0.72 0.72 EC50 1000 0.00072 DeYoung et al. (1996)Carbamazepine n/a 25 n/a 25 NOEC 100 0.25 Ferrari et al. (2003)Carbamazepine 10,11-Epoxide n/a n/a n/a – – – – –Ciprofloxacin 6.7 991.485* 8096.293* 6.7 EC50 1000 0.0067 Yang et al. (2008)Citalopram Hydrobromide n/a 0.8 n/a 0.8 NOEC 100 0.008 Henry et al. (2004)Clarithromycin 0.046 3.11* 4.59* 0.046 EC50 1000 0.000046 Yang et al. (2008)Clomipramine 0.163* 0.047* 0.28* 0.047* EC50 1000 0.000047 Sanderson et al. (2004)Cotinine 589.5 n/a n/a 589.5 EC50 1000 0.5895 Blaise et al. (2006)Cyclophosphamide 930 1795* 70* 70* EC50 1000 0.07 Sanderson et al. (2004)Diazepam 16.5 4.3 12.7 4.3 EC50 1000 0.0043 Lilius et al. (1994)Erythromycin 3.898* 7.822* 349 3.898* EC50 1000 0.003898 Sanderson et al. (2004)Famotidine 22* 242* 100 22* EC50 1000 0.022 Sanderson et al. (2004)Fluoxetine Hydrochloride n/a 0.036 0.005 0.005 NOEC 50 0.0001 Foran et al. (2004)Ifosfamide 8.168* 1795* 70* 8.168* EC50 1000 0.008168 Sanderson et al. (2004)Lansoprazole n/a 22 18 18 EC50 1000 0.018 Sanderson and Thomsen (2009)Loratadine 0.049* 0.142* 0.021* 0.049* EC50 1000 0.000049 Sanderson et al. (2004)Mepivacaine 107* 115* 437* 107* EC50 1000 0.107 Sanderson et al. (2004)Metronidazole 12.5 51* 898* 12.5 EC50 1000 0.0125 Lanzky and Halling-Sorensen (1997)Nicotine n/a 0.12 2.9 0.12 NOEC 50 0.0024 Savino and Tanabe (1989)Norfloxacin 18 1232.512* 14040.45* 18 EC50 1000 0.018 Yang et al. (2008)Ofloxacin n/a 10 10 10 NOEC 50 0.2 Robinson et al. (2005)Omeprazole 1.009* 1.068* 12* 1.009* EC50 1000 0.001009 Sanderson et al. (2004)Paraxanthine n/a 100 n/a 100 EC50 1000 0.1 Fernandez et al. (2010)Paroxetine n/a 0.22 n/a 0.22 NOEC 100 0.0022 Henry et al. (2004)Primidone 69* 73* 265* 69* EC50 1000 0.069 Sanderson et al. (2004)Ranitidine Hydrochloride 10* 64* 1076* 10* EC50 1000 0.01 Sanderson et al. (2004)Salbutamol 5.837* 5.73* 1.158* 1.158* EC50 1000 0.001158 Sanderson et al. (2004)Sulfamethoxazole 0.5 0.129* 3.869* 0.129* EC50 1000 0.000129 Sanderson et al. (2004)Tetracycline 1 44.8 0.843* 0.843* EC50 1000 0.000843 Sanderson et al. (2004)Trimethoprim 40 0.09* 795.579* 0.09* EC50 1000 0.00009 Sanderson et al. (2004)Venlafaxine 15.816* 13* 30.581* 13* EC50 1000 0.013 Sanderson et al. (2004)

n/a: Not applicable (no toxicity data was found).* Indicates predicted (Q)SAR values.

1340 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

Sum of Toxic Units (P

TU) for each sampling point were calcu-lated by means of the sum of the pharmaceuticals-specific HQ foreach site in order to estimate the relative toxicity of the watersfrom each site (Finney, 1971).

3. Results and discussion

3.1. PhACs levels in Madrid River waters

Table 4 shows the concentrations obtained for medicines andmetabolites analyzed at the 10 points downstream of STP effluentdischarge (DSSTP) in the main rivers of Madrid region.

Twenty-two medicines and three metabolites were detectedout of 33 analyzed compounds. Concentrations substantially variedamong pharmaceutical groups. The lowest concentration obtainedcorresponded to mepivacaine (local anesthetics) at 9 ng L�1 andthe highest to the antibiotic erythromycin at 3.8 lg L�1.

On totalling the concentrations of all the medications analyzedat the respective sampling points, the highest concentrations wereobtained at DSSTP sampling point 1, corresponding to the Arroyodel Bodonal (Jarama river), with a total concentration of 78.7 lg L�1

(Fig. 2), 67.7 lg L�1 of which was contributed by carbamazepine(antiepileptic drug). DSSTP sampling point 1 receives wastewaterdischarged by the town of Tres Cantos. Despite its relatively smallpopulation (41,064) (INEbase, 2007), this town is home to a leadinguniversity campus (Autonomous University of Madrid), a techno-logical industrial estate with pharmaceutical plants, a hospitaland a large retirement home.

The second and third highest total concentrations were found atDSSTP sampling points 6 and 5, with total concentrations of 65.0

and 57.3 lg L�1 respectively (Fig. 2), and corresponding in bothcases to Manzanares River, the only river traversing Madrid city(3,255,944 inhabitants) (INEbase, 2007). The main contributionto these concentrations being the presence of caffeine and itsmetabolite paraxanthine (stimulants). Ten out of 33 analyzedPhACs were ubiquitous, and nine were found in more than 50%of the samples whereas only seven were under detection limits(Fig. 3, Table 4).

3.1.1. AntibacterialsEight out of 10 substances analyzed were detected (Table 4).

Macrolides erythromycin and clarithromycin showed the highestconcentrations. The frequency of detection of each is shown inFig. 3.

The highest contributions of erythromycin and clarithromycinwere found in DSSTP sampling point 10, corresponding to the Ta-gus river (Fig. 2). This point receives wastewater discharged fromthe town of Aranjuez and the recently built Tagus Hospital, servingaround 70,000 people, including the population of Aranjuez cityand nearby municipalities. These results are consistent with previ-ous studies reporting high concentrations of antibacterials in hos-pital wastewater discharges (reviewed by Heberer (2002)).

The results obtained for this therapeutic group agree well withthe findings of a preliminary study performed by our group inspringtime during the process of selection of the final samplingpoints (Rodriguez-Gil et al., 2010). A clear seasonality is observedfor erythromycin, sulfamethoxazole, trimethoprim and metronida-zole. For erythromycin, extremely high values were obtained inspringtime, close to 4 lg L�1, for DSSTP sampling points 2 and 5.In the rest of the cases, the concentrations found in the present

Table 4Concentration (ng L�1) of pharmaceuticals and metabolites in different sampling points of the rivers of Madrid downstream the main STPs.

Jarama river Manzanares river Guadarrama river Henaresriver

Tagusriver

Therapeutic group Active Ingredient measured DSSTP1 DSSTP2 DSSTP3 DSSTP4 DSSTP5 DSSTP6 DSSTP7 DSSTP8 DSSTP9 DSSTP10 Median SD

Antibacterials Anhydrous Azithromycin 569 <13 <13 478 <13 569 <13 120 120 569 120 264.2Ciprofloxacin 224 <6 <6 215 <6 <6 <6 <6 <6 <6 <3 91.3Clarithromycin 223 132 130 356 200 247 676 551 129 1727 235 490.0Erythromycin 398 160 603 357 <1 <1 721 90 284 3847 320.5 1150.3Metronidazole 1508 1085 1757 727 1140 1251 1572 1834 <3 182 1195.5 630.0Norfloxacin <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 <10 0Ofloxacin 261 60 150 336 208 269 64 552 <3 49 179 167.8Sulfamethoxazole 952 274 378 222 638 450 202 879 32 82 326 319.0Tetracycline <23 <23 <23 <23 <23 <23 <23 <23 <23 <23 <23 0Trimethoprim 690 418 429 202 478 447 400 519 38 61 424 206.2

Antidepressants Amitriptyline <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 0Citalopram Hydrobromide 78 26 59 64 <1 49 141 160 <1 <1 54 56.5Clomipramine <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 0Fluoxetine (For FluoxetineHydrochloride)

<1 <1 <1 <1 <1 <1 <1 <1 <1 <1 <1 0

Paroxetine <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 <2 0Venlafaxine 523 226 614 280 282 509 836 1003 107 100 395.5 305.1

Antiepileptics Carbamazepine 67715 363 557 389 387 2997 522 615 112 121 455.5 21216.9Carbamazepine 10,11-Epoxide 282 45 157 41 100 105 126 131 <1 <1 102.5 84.5Primidone <5 <5 596 <5 647 687 466 492 <5 <5 234.3 309.8

Anxiolytic Sedatives Hypnotics andAntipsychotics

Diazepam 90 12 13 11 20 23 12 13 <1 <1 12.5 25.8

Antihistamines Loratadine <1 3 <1 24 <1 <1 <1 <1 <1 <1 <1 7.4

Antineoplastics and Immunosuppressants Cyclophosphamide <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 <3 0Ifosfamide <1 <1 <1 <1 <1 <1 41 <1 <1 <1 <1 12.8

Bronchiodilators and Anti-asthma DrugsSalbutamol 17 12 28 16 26 27 <1 29 <1 <1 16.5 11.9

Gastrointestinal Drugs Famotidine <5 <5 <5 349 <5 <5 <5 <5 <5 <5 <5 109.6Lansoprazole <8 <8 <8 961 <8 <8 <8 <8 333 <8 <8 308.9Omeprazole 192 <1 <1 30 <1 <1 <1 <1 <1 <1 <1 60.2Ranitidine (For RanitidineHydrochloride)

604 208 1944 628 280 470 323 1248 186 105 396.5 576.7

Local anesthetics Mepivacaine 29 19 21 346 37 33 21 78 12 9 25 102.2

Stimulants Caffeine 1399 1717 1037 2141 12685 13167 969 4722 675 2079 1898 4808.5Cotinine 610 249 218 207 6582 5969 383 921 213 827 496.5 2472.6Nicotine 301 284 552 404 1885 1740 529 421 771 526 527.5 582.4Paraxanthine 2063 722 840 2110 31676 36019 604 3762 316 1482 1772.5 13719.3

Total concentrations (ng L�1) 78 728 6015 10 083 10 894 57 271 65 028 8608 18 140 3328 11 766

aCalculation median value considering half the detection quantification limit.SD: Standard Deviation.

Y.Valcárcel

etal./Chem

osphere84

(2011)1336–

13481341

Fig. 2. Concentrations totalled by therapeutic group at each sampling point.

Fig. 3. Frequency of detection across all samples analyzed for each of the compounds.

1342 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

study, performed in autumn, were clearly higher. Metronidazolealso showed extreme variations, with springtime values in the ngrange, to autumn values in the lg range. Tetracycline was underdetection limits for both sampling campaigns.

Concentrations obtained for the studied anti-infective agentsare higher than those reported in the literature reviewed by

Mompelat et al. (2009), except for sulfamethoxazole, since Kolpinet al. (2002) detected this compound at 1.9 lg L�1 in US streams.

Among the studies carried out in Spain, the antibacterialsappearing at the highest frequency in the Ebro River (Gros et al.,2007) were azithromycin, trimethoprim, erythromycin and sulfa-methoxazole. In a study recently published by Fernandez et al.

Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348 1343

(2010) in the MR only sulfamethoxazole and trimethoprim wereanalyzed, resulting in total concentrations much lower than thoseobtained by our group. Likewise, the study by Martinez Bueno et al.(2010) also carried out in MR, analyzed the same substances andobtained lower concentrations. They detected neither tetracyclinenor norfloxacin.

According to Carballa et al. (2008), in 2003 12.7 tons of sulfa-methoxazole, 8.1 tons of trimethoprim and 0.4 tons of erytrhomy-cin were consumed in Spain. Trimethoprim is mainly excreted inurine unchanged (Sweetman and Martindale, 2008). A study per-formed in rivers Taff and Ely (South Wales, UK) (Kasprzyk-Hordernet al., 2008), also found trimethoprim, and demonstrated that it ishighly persistent, since its concentration 25 km downstream dis-charge point showed reductions from only 56–12% of its originalconcentration. Low removals (around 60%), have been reportedfor sulfamethoxazole and trimethoprim, confirming its persistencein the aquatic environment (Bendz et al., 2005; Carballa et al.,2004). Macrolides, such as erythromycin, clarithromycin and zai-thromycin were, as well, quite persistent, as they were not elimi-nated in the majority of points studied by Gros et al. (2007). Inthe study carried out by Hirsch et al. (1999) in Germany, tetracyc-lins were not detected. This result is not surprising since they read-ily precipitate with cations such as calcium and accumulate insewage sludge or sediments (Daughton and Ternes, 1999).

3.1.2. Antidepressives, antiepileptics and anxyoliticsSix of the 10 analyzed substances were detected (Table 4). In

this large group, the antiepileptic also used as an antidepressive,carbamazepine, was the PhAC found in highest concentrations,67.7 lg L�1 (median 456 ng L�1), with presence in 100% of sam-pling points (Fig. 3). The highest contribution to the elevated con-centration of carbamazepine corresponds to DSSTP sampling point1 (67.7 lg L�1) (Arroyo del Bodonal, Tres Cantos), a waste-domi-nated stream (Fig. 2). As remarked above, a technological industrialestate with pharmaceutical plants and a geriatric hospital arefound near this point.

Carbamazepine is a highly consumed medication, although lessin Spain than elsewhere in Europe. The annual amount of carbam-azepine prescribed in Spain during 2003 was estimated at20.0 tons/year (Carballa et al., 2008). High environmental concen-trations of this pharmaceutical have been reported very frequentlyworldwide.

We have recently reported a similar concentration for carbam-azepine in DSSTP sampling point 1 in springtime (González Alonsoet al., 2010). The highest concentrations reported in Madrid rivershave been 202 ng/L (Martinez Bueno et al., 2010), 104 ng L�1

(Fernandez et al., 2010) and 532 ng L�1 (Rodriguez-Gil et al.,2010) corresponding all of them to different points along JaramaRiver. The values we report surpass the highest concentrations re-ported worldwide (around 1.3 lg L�1), (Hua et al., 2006; Zhanget al., 2008; Zuehlke et al., 2004). DSSTP sampling point 1 is locatedin a waste-dominated stream, Arroyo del Bodonal, flowing intoJarama River. Such high repeated measurements could indicatean abnormality in the discharge of carbamazepine by Tres CantosSTP and could be responsible, in part, of the high concentrationsfound along Jarama River by the authors referred above. Althougha relatively small population (41,064 inhabitants, INEbase, 2007),Tres Cantos houses a large University Campus (AutonomousUniversity of Madrid), a technological park with pharmaceuticalindustry, a hospital and a large geriatric home that could be relatedwith this level. In the study of Gros et al. (2007) performed in theEbro River (northeast Spain) carbamazepine levels showed a med-ian value of 48 ng L�1. Population density differences could ac-count for the lower levels found in the Ebro River.

Carbamezepine has a extremely low removal rate (lower than7%) in sewage treatment plants, and is an excellent tracer

substance for pharmaceutical agents in the environment(Moldovan, 2006). Carbamazepine as an environmental pollutanthas been shown to be ubiquitously present in the aquaticenvironment and has been found in STP effluents, surface waters,groundwater and occasionally in drinking water all over the world.

Primidone, like carbamazepine, is not attenuated during bankinfiltration and has been detected in samples from municipal sew-age influents and effluents and in surface waters (up to 635 ng L�1)in Germany (Heberer, 2002).

Within the group of antidepressives, only venlafaxine andcitalopram were detected. Venlafaxine is a structurally novel anti-depressant for oral administration and its occurrence in surfacewaters has only been reported in recent works. In Spain, it has onlybe recently reported by our group (González Alonso et al., 2010)and confirmed by the team of Martinez Bueno et al. (2010). Theconcentrations obtained in both studies were considerably lowerthan those presented here, albeit, unlike Martinez-Bueno’s, ourprevious work also reported the presence of venlafaxine at all sam-pling points. Likewise, it has been found in studies carried out in US(Kolpin et al., 2002; Schultz and Furlong, 2008), Canada (Vander-ford and Snyder, 2006; Lajeunesse et al., 2008) and by the recentwork by Schultz et al. (2010) in two US effluent-impacted streams,with concentrations ranging from some ng L�1 to 1 lg L�1.

Effexor (commercial name of venlafaxine) was the 10th greatestselling pharmaceutical worldwide in 2006 (IMS Health 2006). Pre-scription data of this medication differ strongly among countries.In 2007 (http://www.mscm.es), venlafaxine, was the 8th activeingredient with highest consumption according to total cost forthe Spanish National Health System (134.78 million euros).

No ecotoxicological data of venlafaxine is available, but its highoccurrence in the aquatic environment (and in drinking water, aswill be presented below) makes it necessary and urgent to learnits actual environmental risk.

The quantity of citalopram prescribed in Spain during 2003 was1.6 tons and the Predicted Environmental Concentrations in Spainrange from 20 to 60 ng L�1 (Carballa et al., 2008). The concentra-tions obtained by Martinez Bueno et al. (2010) are lower thanthe estimations of Carballa et al. and far lower than the levels re-ported here. Our data are closer to Carballa’s estimations, and aresimilar to Schultz and Furlong (2008) in the Pecan Creek (Texas,Isa), which range from 50 to 90 ng L�1. Lajeunesse et al. (2008) ob-tained values ranging from 3.4 to 15.5 ng L�1 for the St. LawrenceRiver in Montreal (Canada), which were also lower than concentra-tions reported for Madrid rivers.

The anxyolitics diazepam was found in very low concentrations,from 11 ng L�1 to 90 ng L�1 (median 12 ng L�1), with a frequency ofdetection of 80% (Fig. 3). In the MR concentrations obtained byMartinez Bueno et al. (2010) range from 1 ng L�1 to 13 ng L�1. Inrivers the highest reported concentration is 30 ng L�1 (Terneset al., 2001), whereas 40 ng L�1 have been reported for effluentwater.

3.1.3. AntihistaminesLoratadine, which was only found in two points (3 and

24 ng L�1) (Table 4). In Spain this compound has only been ana-lyzed previously by Martinez Bueno et al. (2010), who report itin only one point at concentrations similar to those presented here,2–10 ng L�1.

3.1.4. Antineoplastic and immunosuppressantCommonly called ‘‘cytostatics’’, are frequently used in chemo-

therapy. Two of them were analyzed, ifosfamide and cyclophos-phamide, but only ifosfamide was detected in DSSTP samplingpoint 7 (Guadarrama River), at a concentration of 41 ng L�1

(Table 4). None of the studies carried out previously in Spain havedetected any of these PhACs in surface water; to our knowledge

1344 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

our study is the first. These medications have been normally de-tected in hospital effluent in the ng L�1 range. The presence ofifosfamide in DSSTP sampling point 7 could originate by the dis-charge of Guadarrama hospital.

Reported concentrations in surface water ranged from below50 pg L�1 to 10.1 ng L�1 (Buerge et al., 2006). The highest levelsof cyclophosphamide (10 ng L�1) were obtained in an Italian riverby Zuccato et al. (2000). Kümmerer et al. (2000) calculated a Pre-dicted Environmental Concentration (PEC) of 0.8 ng L�1 for ifosfa-mide in German surface waters.

Due to their large pharmacological potency, cytostatics presentteratogenic and mutagenic properties and seem to undergo a neg-ligible degradation in sewage treatment processes. Accordingly, re-search on their occurrence and fate may be interesting regardingtheir potential risk for humans (especially in highly sensitivegroups such as infants, children and pregnant women) and theenvironment.

3.1.5. Bronchodilatators and anti-asthma drugsSalbutamol (albuterol in the US) was found ranging from 12 to

29 ng L�1 in seven out of 10 sampling points (Table 4). It was alsodetected by Martinez Bueno et al. (2010), albeit in lower concen-trations, 2–4 ng L�1. Ternes (1998) detected lower concentrations,20 ng L�1, in municipal sewage effluents, whereas only sporadicdetections have been reported for surface waters (Hirsch et al.,1999).

3.1.6. Gastrointestinal drugsThe lower concentrations corresponded to omeprazol, 30 ng L�1

and the highest to ranitidine 1.9 lg L�1.Ranitidine is an antagonist of H2, one of the histamine receptors

that inhibit stomach acid production and is frequently used to treatgastroesofageal reflux. Martinez Bueno et al. (2010) report a max-imum concentration of 63 ng L�1 in the MR, and Gros et al. (2007)found 142 ng L�1 in the Ebro River. Other international studies,such as the work on Rivers Lambro and Po, in Italy (Zuccatoet al., 2006) detected a maximum concentration of 38 ng L�1,although the highest value in the bibliography was reported byKolpin et al. (2002), at 580 ng L�1. Ranitidine is documented asone of the substances with a lower percentage of elimination inSTPs (<30%), belonging to the group of the substances with highpersistence and/or stability in waters (Castiglioni et al., 2006).The high concentrations obtained at our sampling points evidencethis characteristic.

3.1.7. Local anestheticMepivacaine, was found in concentrations ranging from 9 to

346 ng L�1 (Table 4) in 100% of the samples (Fig. 3).

3.1.8. StimulantsAll the substances of the group of stimulants nicotine, caffeine

and their metabolites cotinine and paraxanthine, respectively,were found in all samples (Table 4, Fig. 3). As mentioned by Buergeet al. (2008) nicotine, cotinine, caffeine and paraxanthine can rep-resent interesting indicators of anthropogenic contamination as aconsequence of their widespread detection in the environmentand their potential relationship with water contamination levels,giving an indication of the presence of other chemicals related tohuman activities.

The highest concentrations of caffeine were detected in thepoints corresponding to Madrid city effluents (Manzanares river)(Fig. 2). The geographical layout of concentration ranges coincideswith the distribution of the population in the MR. The largest con-centrations have been found downstream the STPs with a biggervolume of influent sewage and the lower concentrations corre-spond with less populated areas. These results are in line with

Martinez Bueno et al. (2010) who also report an increase in caf-feine concentration in the most densely populated zones (centraland southern). Huerta-Fontela et al. (2008a) reported slightlylower concentrations (mean 1.1 lg L�1) in their study in North-eastern Spain (Llobregat, Segre, Foix and Besòs basins) carriedout also in autumn. Far lower concentrations have been reportedin the MR along the Henares–Jarama–Tagus river system (13.2–314.3 ng L�1) (Fernandez et al., 2010) or in other parts of the world(reviewed by Buerge et al. (2003)). Only Ferreira and da Cunha(2005) have reported similar levels in rivers near urban areas inBrasil (160–357 lg L�1). Paraxanthine, main metabolite of caffeine,was found at all of the points at concentrations in the lg L�1 range.Its ubiquity is linked to caffeine since nearly 84% of caffeine isdemethylated to yield paraxanthine (Guerreiro et al., 2008).

Paraxanthine has been detected in other study in MR surfacewater in ranging from 0.3 to 27.8 ng L�1 (Fernandez et al., 2010).Although similar concentration were found in NE-Spain byHuerta-Fontela et al. (2008b) or Benotti and Brownawell (2007)in Jamaica Bay, up to lg L�1concentrations.

The river water concentrations found in our study for caffeineand paraxanthine can be compared to STP effluent levels reportedin previous studies (Buerge et al., 2003; Huerta-Fontela et al.,2008a), which indicates that Madrid STP effluents discharge verylarge quantities of these compounds.

There are fewer studies evaluating the presence of nicotine andits metabolite cotinine in STP influent/effluent or surface waters.Most studies prefer cotinine as a surrogate parameter of contami-nation by nicotine due to a higher stability and half-life. Nicotineconcentrations found in our study (range from 301 ng L�1 to1.9 lg L�1) is higher than those previously reported in Spain (Huer-ta-Fontela et al., 2008a; Martinez Bueno et al., 2010) and elsewhere(Benotti and Brownawell, 2007). Cotinine concentrations in ourstudy ranged from 207 ng L�1 to 6.6 lg L�1. The highest concentra-tion of cotinine was found in DSSTP sampling point 5 in Manzan-ares River (Fig. 2), downstream the discharge of Butarque STP,the plant serving the highest population equivalent. However, pre-vious work in Spain (Martinez Bueno et al., 2010; Huerta-Fontelaet al., 2008a) as well as the US (Focazio et al., 2008) reported coti-nine concentrations one order of magnitude lower. Our river waterconcentrations can be compared to those reported for STP effluentsin Spain (Huerta-Fontela et al., 2008b), Switzerland (Buerge et al.,2008), and the US (Benotti and Brownawell, 2007). This fact con-firms that our sampling strategy leads to the detection of elevatedlevels.

3.2. Screening level risk characterization

Chronic data were only available for six compounds, namelyofloxacin, citalopram hydrobromide, fluoxetine hydrochloride, par-oxetine, carbamazepine and nicotine. No toxicological data wereavailable for the carbamazepine’s metabolite carbamazepine 10,11-epoxide. For a total of seventeen out of the 33 studied com-pounds the benchmark value applied for the HQ calculation wasthat of the (Q)SAR prediction due either to the lack of experimentaltoxicity data for the selected groups or because predicted valuerepresented the lowest E(L)C50. For nine of the compounds theonly available data was that predicted by (Q)SAR models (Table 5).

A total of five of the analyzed pharmaceuticals are present in atleast one of the sampling sites in a concentration that exceeds thepredicted no effect concentration (PNEC) (Table 5). Anhydrousazithromycin, clarithromycin and caffeine showed HQ > 10 in atleast one of the sampling sites, implying that a high hazard is antic-ipated from their presence in the measured concentrations. Forsulfamethoxazole and trimethoprim 1.0 > HQ < 10 and thus, someadverse effect or moderate hazard is probable. Finally, erythromy-cin, carbamazepine, loratadine, metronidazole, paraxanthine,

Table 5Hazard Quotients (HQ) calculated for each compound and sampling site. HQs for the median measured environmental concentration measured for each compound as well as cumulative HQ (Toxic Units) for each sampling point arepresented. Bold font represents HQ > 1.

Therapeutic group/Active Ingredient HQ

DSSTP1 DSSTP2 DSSTP3 DSSTP4 DSST‘P5 DSSTP6 DSST‘P7 DSSTP8 DSSTP9 DSSTP10 MEDIAN

AntibacterialsAnhydrous Azithromycin 29.9 3.42E�01 3.42E�01 25.2 3.42E�01 29.9 3.42E�01 6.32 6.32 29.9 27.6Ciprofloxacin 3.34E�02 4.48E�04 4.48E�04 3.21E�02 4.48E�04 4.48E�04 4.48E�04 4.48E�04 4.48E�04 4.48E�04 3.28E�02Clarithromycin 4.85 2.87 2.83 7.74 4.35 5.37 14.7 12.0 2.80 37.5 5.11Erythromycin 1.02E�01 4.10E�02 1.55E�01 9.16E�02 1.28E�04 1.28E�04 1.85E�01 2.31E�02 7.29E�02 9.87E�01 9.68E�02Metronidazole 1.21E�01 8.68E�02 1.41E�01 5.82E�02 9.12E�02 1.00E�01 1.26E�01 1.47E�01 0.00E + 00 1.46E�02 1.00E�01Norfloxacin 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04 2.78E�04Ofloxacin 1.31E�03 3.00E�04 7.50E�04 1.68E�03 1.04E�03 1.35E�03 3.20E�04 2.76E�03 7.50E�06 2.45E�04 1.04E�03Sulfamethoxazole 7.38 2.12 2.93 1.72 4.95 3.49 1.57 6.81 2.48E�01 6.36E�01 2.53Tetracycline 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02 1.36E�02Trimethoprim 7.67 4.64 4.77 2.24 5.31 4.97 4.44 5.77 4.22E�01 6.78E�01 4.71

AntidepressantsAmitriptyline 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03 2.03E�03Citalopram Hydrobromide 9.75E�03 3.25E�03 7.38E�03 8.00E�03 6.25E�05 6.13E�03 1.76E�02 2.00E�02 6.25E�05 6.25E�05 8.00E�03Clomipramine 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02 2.13E�02Fluoxetine (For Fluoxetine Hydrochloride) 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03 5.00E�03Paroxetine 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04 4.55E�04Venlafaxine 4.02E�02 1.74E�02 4.72E�02 2.15E�02 2.17E�02 3.92E�02 6.43E�02 7.72E�02 8.23E�03 7.69E�03 3.04E�02

AntiepilepticsCarbamazepine 2.71E�01 1.45E�03 2.23E�03 1.56E�03 1.55E�03 1.20E�02 2.09E�03 2.46E�03 4.48E�04 4.84E�04 1.82E�03Carbamazepine 10,11-Epoxide n/a n/a n/a n/a n/a n/a n/a n/a n/a n/a n/aPrimidone 3.62E�05 3.62E�05 8.64E�03 3.62E�05 9.38E�03 9.96E�03 6.75E�03 7.13E�03 3.62E�05 3.62E�05 8.64E�03

Anxiolytic sedatives hypnotics and antipsychoticsDiazepam 2.09E�02 2.79E�03 3.02E�03 2.56E�03 4.65E�03 5.35E�03 2.79E�03 3.02E�03 1.16E�04 1.16E�04 3.02E�03

AntihistaminesLoratadine 1.02E�02 6.12E�02 1.02E�02 4.90E�01 1.02E�02 1.02E�02 1.02E�02 1.02E�02 1.02E�02 1.02E�02 2.76E�01

Antineoplastics and immunosuppressantsCyclophosphamide 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05 2.14E�05Ifosfamide 6.12E�05 6.12E�05 6.12E�05 6.12E�05 6.12E�05 6.12E�05 5.02E�03 6.12E�05 6.12E�05 6.12E�05 5.02E�03

Bronchiodilators and anti-asthma drugsSalbutamol 1.47E�02 1.04E�02 2.42E�02 1.38E�02 2.25E�02 2.33E�02 4.32E�04 2.50E�02 4.32E�04 4.32E�04 2.25E�02

Gastrointestinal drugsFamotidine 1.14E�04 1.14E�04 1.14E�04 1.59E�02 1.14E�04 1.14E�04 1.14E�04 1.14E�04 1.14E�04 1.14E�04 1.59E�02Lansoprazole 2.22E�04 2.22E�04 2.22E�04 5.34E�02 2.22E�04 2.22E�04 2.22E�04 2.22E�04 1.85E�02 2.22E�04 3.59E�02Omeprazole 1.90E�01 4.96E�04 4.96E�04 2.97E�02 4.96E�04 4.96E�04 4.96E�04 4.96E�04 4.96E�04 4.96E�04 1.10E�01Ranitidine (For Ranitidine Hydrochloride) 6.04E�02 2.08E�02 1.94E�01 6.28E�02 2.80E�02 4.70E�02 3.23E�02 1.25E�01 1.86E�02 1.05E�02 3.97E�02

Local anestheticsMepivacaine 2.71E�04 1.78E�04 1.96E�04 3.23E�03 3.46E�04 3.08E�04 1.96E�04 7.29E�04 1.12E�04 8.41E�05 2.34E�04

StimulantsCaffeine 1.94 2.38 1.44 2.97 17.6 18.3 1.35 6.56 9.38E�01 2.89 2.64Cotinine 1.03E�03 4.22E�04 3.70E�04 3.51E�04 1.12E�02 1.01E�02 6.50E�04 1.56E�03 3.61E�04 1.40E�03 8.42E�04Nicotine 1.25E�01 1.18E�01 2.30E�01 1.68E�01 7.85E�01 7.25E�01 2.20E�01 1.75E�01 3.21E�01 2.19E�01 2.20E�01Paraxanthine 2.06E�02 7.22E�03 8.40E�03 2.11E�02 3.17E�01 3.60E�01 6.04E�03 3.76E�02 3.16E�03 1.48E�02 1.77E�02Toxic Units 52.9 12.8 13.2 41.0 33.9 63.5 23.1 38.1 11.2 73.0

n/a: Not applicable (No toxicity data was available for the calculation).

Y.Valcárcel

etal./Chem

osphere84

(2011)1336–

13481345

Table 6PhACs concentrations (ng/L) in tap water.

Therapeutic group Active Ingredient measured P-ALC P-CHAM P-LEG P-TIT P-USE Median Frequency (%)

Antidepressants Venlafaxine <1 <1 <1 9 <1 <1 20

Antiepiletics Carbamazepine 5 <1 4 4 <1 4 60

Stimulants and Other Substances Caffeine 75 15 26 30 31 30 100Cotinine 34 14 25 27 19 25 100Nicotine 100 36 <4 <4 72 36 60

1346 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

omeprazole, ranitidine hydrochloride and nicotine represent a lowpotential hazard based on their scored HQs (0.1 < HQ < 1). Theseresults are however preliminary and should be taken with cautiondue to several factors. The exposure data used for the calculation isbased on a single sampling event, and this data does not accuratelyrepresent the long term exposure of the organisms present in thestudied ecosystems. This is especially important when chronicexposures are assumed for the calculation. The applied exposuredata only represents concentrations in water, and factors such asreal uptake rates, bioconcentration or bioaccumulation should betaken into account in further studies. Available dose–response datafor pharmaceutical compounds is currently limited, especially forthe case of chronic data. Reliable (Q)SAR estimates for fish, Daphniaand algal toxicity are available for chemicals with a non-specificmode of action, however their applicability to pharmaceuticalsshould be taken with caution. Lastly, further data is needed as wellfor those compounds for which uncertainties about potential endo-crine disrupting effects exist.

The cumulative HQ orP

TU shows that for all the selected sam-pling points high hazard is anticipated from the presence of theanalyzed compounds in the measured concentrations, becausethe

PTU exceed the value of 10 in all cases, ranging from 12.2 to

73.0. As for the case of individual HQs, theP

TU data obtained inthis study should be taken as preliminary. A quotient addition ap-proach assumes that toxicities are additive or approximately addi-tive (US EPA, 1998). In the case of the studied pharmaceuticalssimilar modes of action of compounds in the same therapeuticgroup could account for certain level of overestimation of the risk.At the same time however, possible synergic interactions betweencompounds with different modes of action could lead to underes-timations of risk.

3.3. PhACs in tap water of the main supply areas of Madrid

Five tap water samples, from the MR’s main areas of supplywere analyzed for the same suite of pharmaceuticals describedabove and five out of the 33 selected compounds were found. Ta-ble 6 shows concentrations for each PhAC. Caffeine and cotininewere detected in all the tap water samples with concentrationsranging from 15 to 75 ng L�1 and 14 to 34 ng L�1, respectively.The presence of these two compounds in drinking water is a stepbeyond the consideration of caffeine and cotinine as indicators ofanthropogenic pollution in surface- and groundwater systems.The frequent presence of both may be mainly due to its anthropicnature and elevated consumption as well as to chemical stability ofthese compounds (Ferreira, 2005).

The other PhACs above the detection limit were nicotine, car-bamazepine and velafaxine. The antidepressive venlafaxine wasfound in one sample. No studies of this compound in drinkingwater have been reported. Carbamazepine concentrations detectedin Madrid tap water are close to the quantification limit and similarto those reported by other authors in finished water (post secondchlorination before distribution) (Heberer et al., 2004; Togola andBudzinski, 2008) and tap water (Kim et al., 2007). Caffeine concen-trations found in our study (median 35 ng L�1) are slightly lower

than values reported by Huerta-Fontela et al. (2008b) in finishedwater (above 125 ng L�1) although similar to those found by Togolaand Budzinski (2008) in France and Stackelberg et al. (2007) in theUS, both in finished water.

The levels detected for nicotine are especially high compared tothe study by Huerta-Fontela et al. (2008b) who observed a degra-dation of 100% of this compound at DWTP. However, similar valueswere found in finished water in the US (Stackelberg et al., 2004).The median concentration of detectable cotinine in tap waterwas 23.8 ng L�1 which is in agreement with previously reportedvalues in the US (Stackelberg et al., 2004), but lower than thatfound in finished water in NE-Spain (above 320 ng L�1) (Huerta-Fontela et al., 2008b).

Despite low concentrations of PhACs, their presence in Madridtap waters highlights an inadequate elimination in the DWTPs.More and more studies aim to determine the percentage of elimi-nation of PhACs in DWTPs with different treatments (Hua et al.,2006; Huerta-Fontela et al., 2008a). The results show that, for somepharmaceutical compounds, primary (ozone) and secondary (chlo-rine) oxidation treatments applied in DWTPs are not suitable(Benotti et al., 2009). Neither are coagulation and sand filtrationprocesses (Hua et al., 2006; Kim et al., 2007). New advanced oxida-tion processes, such as Photo-Fenton photooxidation, are yieldingpromising results (Rodriguez-Gil et al., 2010).

In relation to the possible effects on human health derived fromthe presence of these substances in human drinking water, (Joneset al., 2005) several studies show that the likelihood of acute ef-fects (short term) of such low PhACs concentrations in drinkingwater is very low. Nevertheless, the reports by the Hazardous Sub-stances Data Bank (http://toxnet.nlm.nih.gov) for the detectedcompounds state that nothing is currently known on potentiallong-term effects at low exposure levels.

4. Conclusions

Twenty-five pharmaceuticals and metabolites have been de-tected in the main rivers of the MR downstream of effluent dis-charge of the 10 most important STPs. Likewise, five substanceshave been detected in the low nanogram/litre level in tap waterof the region.

The highest concentrations in river water were measured forthe antiepileptics carbamazepine and the stimulant caffeine,followed by the antibacterial metronidazole, the gastrointesti-nal drug ranitidine, and the antidepressive venlafaxine.Clarithromycine, trimethoprim, venlafaxine, carbamazepine,paraxanthine, ranitidine, mepivacaine, caffeine, and nicotinewere detected in 100% of river water samples.Preliminary risk characterization shows that a total of five ofthe analyzed pharmaceuticals are present in at least one ofthe sampling sites in a concentration that exceeds the predictedno effect concentration (PNEC), with three of them showing anHQ > 10. Another 8 compounds represent a low potential haz-ard based on their HQs (0.1 < HQ < 1). Toxic Units calculationindicates that for all the selected sampling points high hazard

Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348 1347

is anticipated from the combined presence of the analyzed com-pounds in the measured concentrations (

PTU > 10).

Regarding drinking water, caffeine and cotinine were present inall (10) of samples, carbamazepine and nicotine in 6 and venla-faxine in 2. No studies of this compound (venlafaxine) in drink-ing water have been reported. The highest concentrations werefor nicotine and caffeine.

These preliminary results, based on single grab sample analysis,clearly demonstrate the need for water quality monitoring and re-search in urban rivers using composite and periodical samplings,since the discharge of pharmaceuticals and their metabolites canfluctuate daily, monthly or seasonally. Despite the limitations ofthis type of study, the results also show the need for improvedsewage treatment techniques able to eliminate this kind of com-pounds from STP effluent in order to maintain ecosystem qualityand preserve reservoirs for human drinking water.

Also, the presence of a number of these compounds in drinkingwater (mainly nicotine and caffeine) samples generates concernabout the efficacy of the applied water treatment techniques. Thepresence of these compounds in a watershed like Madrid’s, wherethe drinking water is mainly collected in headwater streams, raisesconcern about the potential risks in other areas where waters frommedium or low sections of major rivers are used as drinking watersources.

Based on the detected concentrations and the results from thepreliminary risk characterization, we consider some of the ana-lyzed compounds as potential candidates to be included in PriorityPollutants Lists. Reference values for these compounds should beassessed and included in water quality guidelines.

Finally, engagement via participation in the current pharmaceu-tical waste recycling programs should be a priority in both regionaland national levels in order to minimize the most easily managedsource of this kind of pollution.

Acknowledgements

This work has been possible thanks to a ‘‘great team’’. The re-search was designed by Dra. Valcárcel assisted by Dr. Catala. Fieldwork has been co-directed by Dra. Valcárcel and Dra. GonzálezAlonso and performed Dra. González Alonso and Raúl Romo. Fig-ures have been elaborated by J.L. Rodríguez Gil and Dr. GonzálezAlonso and Dr. Jon San Sebastian help in the design of Fig. 1. Allthe authors have participated in bibliographic research and manu-script writing and have approved the final document.

The authors thank Dr. Agüera, of the Pesticide Residues Re-search Group of Almería University leaded by Dr. Fernandéz-Alba,who performed the analysis of the pharmaceuticals. Finally, wethank Leilan Baxter for critical review of the final manuscript.

References

Bendz, D., Paxeus, N.A., Ginn, T.R., Loge, F.J., 2005. Occurrence and fate ofpharmaceutically active compounds in the environment, a case study: HojeRiver in Sweden. Journal of Hazardous Materials 122 (3), 195–204.

Benotti, M.J., Brownawell, B.J., 2007. Distributions of pharmaceuticals in an urbanestuary during both dry- and wet-weather conditions. Environmental Scienceand Technology 41 (16), 5795–5802.

Benotti, M.J., Trenholm, R.A., Vanderford, B.J., Holady, J.C., Stanford, B.D., Snyder,S.A., 2009. Pharmaceuticals and endocrine disrupting compounds in USdrinking water. Environmental Science and Technology 43 (3), 597–603.

Blaise, C., Gagné, F., Eullaffroy, P., Férard, J.-F., 2006. Ecotoxicity of selectedpharmaceuticals of urban origin discharged to the Saint-Lawrence river(Québec, Canada): a review. Brazilian Journal of Aquatic Sciences andTechnology 10 (2), 29–51.

Bottoni, P., Caroli, S., Barra, A., 2010. Pharmaceuticals as priority watercontaminants. Toxicological and Environmental Chemistry 92 (3), 549–565.

Buerge, I.J., Buser, H.R., Poiger, T., Muller, M.D., 2006. Occurrence and fate of thecytostatic drugs cyclophosphamide and ifosfamide in wastewater and surfacewaters. Environmental Science and Technology 40 (23), 7242–7250.

Buerge, I.J., Kahle, M., Buser, H.R., Muller, M.D., Poiger, T., 2008. Nicotine derivativesin wastewater and surface waters: application as chemical markers fordomestic wastewater. Environmental Science and Technology 42 (17), 6354–6360.

Buerge, I.J., Poiger, T., Muller, M.D., Buser, H.R., 2003. Caffeine, an anthropogenicmarker for wastewater contamination of surface waters. Environmental Scienceand Technology 37 (4), 691–700.

Carballa, M., Omil, F., Lema, J.M., 2008. Comparison of predicted and measuredconcentrations of selected pharmaceuticals, fragrances and hormones inSpanish sewage. Chemosphere 72 (8), 1118–1123.

Carballa, M., Omil, F., Lema, J.M., Llompart, M., Garcia-Jares, C., Rodriguez, I., Gomez,M., Ternes, T., 2004. Behavior of pharmaceuticals, cosmetics and hormones in asewage treatment plant. Water Research 38 (12), 2918–2926.

Cars, O., Mölstad, S., Melander, A., 2001. Variation in antibiotic use in the EuropeanUnion. Lancet 357 (9271), 1851–1853.

Castiglioni, S., Bagnati, R., Fanelli, R., Pomati, F., Calamari, D., Zuccato, E., 2006.Removal of pharmaceuticals in sewage treatment plants in Italy. EnvironmentalScience and Technology 40 (1), 357–363.

Daughton, C.G., Ternes, T.A., 1999. Pharmaceuticals and personal care products inthe environment: agents of subtle change? Environmental Health Perspectives107 (6), 907–938.

DeYoung, D.J., Bantle, J.A., Hull, M.A., Burks, S.L., 1996. Differences in sensitivity todevelopmental toxicants as seen in Xenopus and Pimephales embryos. Bulletinof Environmental Contamination and Toxicology 56 (1), 143–150.

EMEA, 2006. Guideline on the Environmental Risk Assessment of MedicinalProducts for Human Use (EMEA/CHMP/SWP/4447/00 corr 1⁄). EuropeanMedicines Agencies, London.

Fent, K., Weston, A.A., Caminada, D., 2006. Ecotoxicology of humanpharmaceuticals. Aquatic Toxicology 76, 122–159.

Fernandez, C., Gonzalez-Doncel, M., Pro, J., Carbonell, G., Tarazona, J.V., 2010.Occurrence of pharmaceutically active compounds in surface waters of theHenares–Jarama–Tajo River system (Madrid, Spain) and a potential riskcharacterization. The Science of the Total Environment 408 (3), 543–551.

Ferrari, B., Paxeus, N., Lo, G.R., Pollio, A., Garric, J., 2003. Ecotoxicological impact ofpharmaceuticals found in treated wastewaters: study of carbamazepine,clofibric acid, and diclofenac. Ecotoxicology and Environmental Safety 55 (3),359–370.

Ferreira, A.P., 2005. Caffeine as an environmental indicator for assessing urbanaquatic ecosystems. Cadernos de Saúde Pública 21 (6), 1884–1892.

Ferreira, A.P., da Cunha, C.L., 2005. Anthropic pollution in aquatic environment:development of a caffeine indicator. International Journal of EnvironmentalHealth Research 15 (4), 303–311.

Finney, D.M., 1971. Probit Analysis. Cambridge University Press, Cambridge.Focazio, M.J., Kolpin, D.W., Barnes, K.K., Furlong, E.T., Meyer, M.T., Zaugg, S.D.,

Barber, L.B., Thurman, M.E., 2008. A national reconnaissance forpharmaceuticals and other organic wastewater contaminants in the UnitedStates – II) untreated drinking water sources. The Science of the TotalEnvironment 402 (2–3), 201–216.

Foran, C.M., Weston, J., Slattery, M., Brooks, B.W., Huggett, D.B., 2004. Reproductiveassessment of Japanese medaka (Oryzias latipes) following a four-weekfluoxetine (SSRI) exposure. Archives of Environmental Contamination andToxicology 46 (4), 511–517.

Glassmeyer, S.T., Hinchey, E.K., Boehme, S.E., Daughton, C.G., Ruhoy, I.S., Conerly, O.,Daniels, R.L., Lauer, L., McCarthy, M., Nettesheim, T.G., Sykes, K., Thompson, V.G.,2009. Disposal practices for unwanted residential medications in the UnitedStates. Environment International 35 (3), 566–572.

González Alonso, S., Catalá, M., Romo Maroto, R., Rodríguez-Gil, J.L., Valcárcel, Y.,2010. Pollution by psychoactive pharmaceuticals in the Rivers of Madridmetropolitan Area (Spain). Environmental International 36 (2), 195–210.

Gros, M., Petrovic, M., Barcelo, D., 2007. Wastewater treatment plants as a pathwayfor aquatic contamination by pharmaceuticals in the Ebro river basin (northeastSpain). Environmental Toxicology and Chemistry 26 (8), 1553–1562.

Gros, M., Petrovic, M., Ginebreda, A., Barcelo, D., 2010. Removal of pharmaceuticalsduring wastewater treatment and environmental risk assessment using hazardindexes. Environment International 36 (1), 15–26.

Guerreiro, S., Toulorge, D., Hirsch, E., Marien, M., Sokoloff, P., Michel, P.P., 2008.Paraxanthine, the primary metabolite of caffeine, provides protection againstdopaminergic cell death via stimulation of ryanodine receptor channels.Molecular Pharmacology 74 (4), 980–989.

Harada, A., Komori, K., Nakada, N., Kitamura, K., Suzuki, Y., 2008. Biological effects ofPPCPs on aquatic lives and evaluation of river waters affected by differentwastewater treatment levels. Water Science and Technology 58 (8), 1541–1546.

Heberer, T., 2002. Occurrence, fate, and removal of pharmaceutical residues in theaquatic environment: a review of recent research data. Toxicology Letter 131(1–2), 5–17.

Heberer, T., Mechlinski, A., Fanck, B., Knappe, A., Massmann, G., Pekdeger, A., Fritz,B., 2004. Field studies on the fate and transport of pharmaceutical residues inbank filtration. Ground Water Monitoring and Remediation 24 (2), 70–77.

Henry, T.B., Kwon, J.W., Armbrust, K.L., Black, M.C., 2004. Acute and chronic toxicityof five selective serotonin reuptake inhibitors in Ceriodaphnia dubia.Environmental Toxicology Chemistry 23 (9), 2229–2233.

1348 Y. Valcárcel et al. / Chemosphere 84 (2011) 1336–1348

Hirsch, R., Ternes, T., Haberer, K., Kratz, K.L., 1999. Occurrence of antibiotics inthe aquatic environment. The Science of the Total Environment 225 (1–2), 109–118.

Hua, W., Bennett, E.R., Letcher, R.J., 2006. Ozone treatment and the depletion ofdetectable pharmaceuticals and atrazine herbicide in drinking water sourcedfrom the upper Detroit River, Ontario, Canada. Water Research 40 (12), 2259–2266.

Huerta-Fontela, M., Galceran, M.T., Martin-Alonso, J., Ventura, F., 2008a. Occurrenceof psychoactive stimulatory drugs in wastewaters in north-eastern Spain. TheScience of the Total Environment 397 (1–3), 31–40.

Huerta-Fontela, M., Galceran, M.T., Ventura, F., 2008b. Stimulatory drugs of abuse insurface waters and their removal in a conventional drinking water treatmentplant. Environmental Science and Technology 42 (18), 6809–6816.

Huggett, D.B., Brooks, B.W., Peterson, B., Foran, C.M., Schlenk, D., 2002. Toxicity ofselect beta adrenergic receptor-blocking pharmaceuticals (B-blockers) onaquatic organisms. Archives of Environmental Contamination and Toxicology43, 229–235.

INEbase, 2007. National Statistics Institute on-line database. <http://www.ine.es/>(accessed February 2009).

Jones, D.S., Barnthouse, L.W., Suter, G.W., Efroymson, R.A., Field, J.M., Beauchamp,J.J., 1999. Ecological risk assessment in a large river-reservoir: 3. EnvironmentalToxicology and Chemistry 18 (4), 599–609.

Jones, O.A., Lester, J.N., Voulvoulis, N., 2005. Pharmaceuticals: a threat to drinkingwater? Trends in Biotechnology 23 (4), 163–167.

Kasprzyk-Hordern, B., Dinsdale, R.M., Guwy, A.J., 2008. The occurrence ofpharmaceuticals, personal care products, endocrine disruptors and illicitdrugs in surface water in South Wales, UK. Water Research 42 (13), 3498–3518.

Khetan, S.K., Collins, T.J., 2007. Human pharmaceuticals in the aquatic environment:a challenge to Green Chemistry. Chemical Reviews 107, 2319–2364.

Kim, S.D., Cho, J., Kim, I.S., Vanderford, B.J., Snyder, S.A., 2007. Occurrence andremoval of pharmaceuticals and endocrine disruptors in South Korean surface,drinking, and waste waters. Water Research 41 (5), 1013–1021.

Kolpin, D.W., Furlong, E.T., Meyer, M.T., Thurman, E.M., Zaugg, S.D., Barber, L.B.,Buxton, H.T., 2002. Pharmaceuticals, hormones, and other organic wastewatercontaminants in US streams, 1999–2000: a national reconnaissance.Environmental Science and Technology 36 (6), 1202–1211.

Kümmerer, K., 2004. Resistance in the environment. The Journal of AntimicrobialChemotherapy 54 (2), 311–320.

Kümmerer, K., Al-Ahmad, A., Bertram, B., Wiessler, M., 2000. Biodegradability ofantineoplastic compounds in screening tests: influence of glucosidation and ofstereochemistry. Chemosphere 40 (7), 767–773.

Lajeunesse, A., Gagnon, C., Sauve, S., 2008. Determination of basic antidepressantsand their N-desmethyl metabolites in raw sewage and wastewater using solid-phase extraction and liquid chromatography–tandem mass spectrometry.Analytical Chemistry 80 (14), 5325–5333.

Lanzky, P.F., Halling-Sorensen, B., 1997. The toxic effect of the antibioticmetronidazole on aquatic organisms. Chemosphere 35 (11), 2553–2561.

Lilius, H., Isomaa, B., Holmstrom, T., 1994. A comparison of the toxicity of 50reference chemicals to freshly isolated rainbow-trout hepatocytes and Daphniamagna. Aquatic Toxicology 30 (1), 47–60.

Martinez Bueno, M., Hernando, M., Herrera, S., Gómez, M., Fernández-Alba, A.R.,Bustamante, I., García-Calvo, E., 2010. Pilot survey of chemical contaminantsfrom industrial and human activities in river waters in Spain. InternationalJournal of Environmental Analytical Chemistry 90 (3), 321–343.

Ministerio de Sanidad y Consumo, 1995. Dirección General de Aseguiramiento yPlanificación Sanitaria. Agencia de Evaluación de Tecnologías Sanitarias.Medicina Clínica 1995 (106), 267–279.

Moldovan, Z., 2006. Occurrences of pharmaceutical and personal care products asmicropollutants in rivers from Romania. Chemosphere 64 (11), 1808–1817.

Mompelat, S., Le, B.B., Thomas, O., 2009. Occurrence and fate of pharmaceuticalproducts and by-products, from resource to drinking water. EnvironmentInternational 35 (5), 803–814.

Robinson, A.A., Belden, J.B., Lydy, M.J., 2005. Toxicity of fluoroquinolone antibioticsto aquatic organisms. Environmental Toxicology Chemistry 24 (2), 423–430.

Rodriguez-Gil, J.L., Catalá, M., González Alonso, S., Romo Maroto, R., Valcarcel, Y.,Segura, Y., Molina, R., Melero, J.A., Martinez, F., 2010. Heterogeneous photo-Fenton treatment for the reduction of pharmaceutical contamination in Madridrivers and ecotoxicological evaluation by a miniaturized fern spores bioassay.Chemosphere 80 (4), 381–388.

Ruiz, A., Ruiz-Tovar, M., Pérez, B., Díaz, P., López, R., 2000. Non-hospitalconsumption of antibiotics in Spain: 1987–1997. The Journal of AntimicrobialChemotherapy 45 (3), 395–400.

Sanderson, H., Johnson, D.J., Reitsma, T., Brain, R.A., Wilson, C.J., Solomon, K.R., 2004.Ranking and prioritization of environmental risks of pharmaceuticals in surfacewaters. Regulatory Toxicology and Pharmacology 39 (2), 158–183.

Sanderson, H., Thomsen, M., 2009. Comparative analysis of pharmaceuticals versusindustrial chemicals acute aquatic toxicity classification according to the UnitedNations classification system for chemicals. Assessment of the (Q)SARpredictability of pharmaceuticals acute aquatic toxicity and theirpredominant acute toxic mode-of-action. Toxicology Letters 187 (2), 84–93.

Savino, J.F., Tanabe, L.L., 1989. Sublethal effects of phenanthrene, nicotine, andpinane on Daphnia pulex. Bulletin of Environmental Contamination andToxicology 42 (5), 778–784.

Schultz, M.M., Furlong, E.T., 2008. Trace analysis of antidepressant pharmaceuticalsand their select degradates in aquatic matrixes by LC/ESI/MS/MS. AnalyticalChemistry 80 (5), 1756–1762.

Schultz, M.M., Furlong, E.T., Kolpin, D.W., Werner, S.L., Schoenfuss, H.L., Barber, L.B.,Blazer, V.S., Norris, D.O., Vajda, A.M., 2010. Antidepressant pharmaceuticals intwo US effluent-impacted streams: occurrence and fate in water and sediment,and selective uptake in fish neural tissue. Environmental Science andTechnology 44 (6), 1918–1925.

Sole, J., Rodriguez, G., Grahit, V., Juncadella, E., 2004. Consumo de antibióticos y suposible relación con la resistencia bacteriana en la región sanitaria Costa dePonent: Análisis evolutivo durante los períodos inicial y final de la últimadécada. Atención Primaria 34 (3), 128–133.

Stackelberg, P.E., Furlong, E.T., Meyer, M.T., Zaugg, S.D., Henderson, A.K., Reissman,D.B., 2004. Persistence of pharmaceutical compounds and other organicwastewater contaminants in a conventional drinking-water-treatment plant.The Science of the Total Environment 329 (1–3), 99–113.

Stackelberg, P.E., Gibs, J., Furlong, E.T., Meyer, M.T., Zaugg, S.D., Lippincott, R.L.,2007. Efficiency of conventional drinking-water-treatment processes inremoval of pharmaceuticals and other organic compounds. The Science of theTotal Environment 377 (2–3), 255–272.

Sweetman, S.C., Martindale, W., 2008. Martindale: Guía Completa de ConsultaFarmacoterapéutica. Pharma Editores, Barcelona.

Ternes, T., Bonerz, M., Schmidt, T., 2001. Determination of neutral pharmaceuticalsin wastewater and rivers by liquid chromatography-electrospray tandem massspectrometry. Journal of Chromatography (A) 938 (1–2), 175–185.

Ternes, T., 1998. Occurrence of drugs in German sewage treatment plants andrivers. Water Research 32 (11), 3245–3260.

Togola, A., Budzinski, H., 2008. Multi-residue analysis of pharmaceuticalcompounds in aqueous samples. Journal of Chromatography (A) 1177 (1),150–158.

UE, 1996. Technical Guidance Document in support of Council Directive 93/67/EECon Risk Assessment for New Notified Substances and Commission Regulation(EC) 1488/94 on Risk Assessment for Existing Substances (Part II:Environmental Risk Assessment). Office for Official Publications of theEuropean Communities, Luxembourg.

US EPA, 1997. Terms of Environment: Glossary, Abbreviations and Acronyms. EPAPublication No.175-B-97-001. Washington, DC, US Environmental ProtectionAgency.

US EPA, 1998. Guidelines for Ecological Risk Assessment. EPA Publication No. 630/R-95/002F. Washington, DC, US Environmental Protection Agency.

Vanderford, B.J., Snyder, S.A., 2006. Analysis of pharmaceuticals in water by isotopedilution liquid chromatography/tandem mass spectrometry. EnvironmentalScience and Technology 40 (23), 7312–7320.

Yang, L.H., Ying, G.G., Su, H.C., Stauber, J.L., Adams, M.S., Binet, M.T., 2008. Growth-inhibiting effects of 12 antibacterial agents and their mixtures on thefreshwater microalga Pseudokirchneriella subcapitata. EnvironmentalToxicology Chemistry 27 (5), 1201–1208.

Zhang, Y., Geissen, S.U., Gal, C., 2008. Carbamazepine and diclofenac: removal inwastewater treatment plants and occurrence in water bodies. Chemosphere 73(8), 1151–1161.

Zuccato, E., Calamari, D., Natangelo, M., Fanelli, R., 2000. Presence of therapeuticdrugs in the environment. Lancet 355 (9217), 1789–1790.

Zuccato, E., Castiglioni, S., Fanelli, R., Reitano, G., Bagnati, R., Chiabrando, C., Pomati,F., Rossetti, C., Calamari, D., 2006. Pharmaceuticals in the environment in Italy:causes, occurrence, effects and control. Environmental Science and PollutionResearch International 13 (1), 15–21.

Zuehlke, S., Duennbier, U., Heberer, T., 2004. Determination of polar drug residues insewage and surface water applying liquid chromatography-tandem massspectrometry. Analytical Chemistry 76 (22), 6548–6554.