Multi-year inter-laboratory exercises for the analysis of illicit drugs and metabolites in wastewater: Development of a quality control system

1 1Multi-year interlaboratory exercises for the analysis of illicit drugs 2and metabolites in wastewater: development of a quality control 3system 4 5 Alexander L.N. van Nuijs1, Foon Yin Lai1, Frederic Been1, Maria Jesus Andres-Costa2, Leon Barron3, 6 Jose Antonio Baz-Lomba4, Jean-Daniel Berset5, Lisa Benaglia6, Lubertus Bijlsma7, Dan Burgard8, Sara 7 Castiglioni9, Christophoros Christophoridis10, Adrian Covaci1, Pim de Voogt11,12, Erik Emke11, Despo 8 Fatta-Kassinos13, Jerker Fick14, Felix Hernandez7, Cobus Gerber15, Iria González-Mariño16, Roman 9 Grabic17, Teemu Gunnar18, Kurunthachalam Kannan19, Sara Karolak20, Barbara Kasprzyk-Hordern21, 10 Zenon Kokot22, Ivona Krizman-Matasic23, Angela Li24, Xiqing Li25, Arndís S.C. Löve26, Miren Lopez de 11 Alda27, Markus R. Meyer28, Herbert Oberacher29, Jake O’Brien30, Jose Benito Quintana16, Malcolm 12 Reid4, Serge Schneider31, Susana Sadler Simoes32, Nikolaos S. Thomaidis33, Kevin Thomas4,30, Viviane 13 Yargeau34, Christoph Ort35 14 1 Toxicological Centre, University of Antwerp, Universiteitsplein 1, 2610 Antwerp, Belgium 15 2 Environmental and Food Safety Research Group (SAMA-UV), Desertification Research Centre CIDE (CSIC-UV16 GV), Av. Vicent Andrés Estellés s/n, Burjassot, Valencia, Spain 17 3 Analytical & Environmental Sciences Division, Faculty of Life Sciences & Medicine, King’s College London, 18 Franklin Wilkins Building, 150 Stamford St., London SE1 9NH, United Kingdom 19 4 Norwegian Institute for Water Research (NIVA), Gaustadalléen 21, 0349 Oslo, Norway 20 5 Institute of Plant Sciences (IPS), University of Bern, Altenbergrain 21, 3013 Bern, Switzerland 21 6 École des Sciences Criminelles, University of Lausanne, Avenue Forel 15, 1015 Lausanne, Switzerland 22 7 Research Institute for Pesticides and Water, University Jaume I, Avda. Sos Baynat s/n, E-12071 Castellón, Spain 23 8 Chemistry Department, University of Puget Sound, Tacoma, WA, 98416, USA 24 9 IRCCS – Istituto di Ricerche Farmacologiche “Mario Negri”, Department of Environmental Health Sciences, Via 25 La Masa 19, 20156 Milan, Italy 26 10 Environmental Pollution Control Laboratory, Aristotle University of Thessaloniki, 54124, Greece 27 11 KWR Watercycle Research Institute, Chemical Water Quality and Health, P.O. Box 1072, 3430 BB Nieuwegein, 28 The Netherlands 29 12 Institute for Biodiversity and Ecosystem Dynamics, University of Amsterdam, P.O. Box 94248, 1090 GE 30 Amsterdam, The Netherlands 31 13 Nireas-International Water Research Center and Civil and Environmental Engineering Department, University 32 of Cyprus, P.O. Box 20537, 1678 Nicosia, Cyprus 33 14 Department of Chemistry, Umeå Unicversity, 901 87 Umeå, Sweden 34 15 School of Pharmacy and Medical Sciences, University of South Australia, Adelaide, Australia, 5001 35 16 Institute for Food Analysis and Research, University of Santiago de Compostela, Constantino Candeira S/N, 36 15782 Santiago de Compostela, Spain 37 17 University of South Bohemia in Ceske Budejovice, Faculty of Fisheries and Protection of Waters, South 38 Bohemian Research Center of Aquaculture and Biodiversity of Hydrocenoses, Zatisi 728/II, CZ-389 25 Vodnany, 39 Czech Republic 40 18 Forensic Toxicology Unit, National Institute for Health and Welfare, P.O.Box 30, 00271 Helsinki, Finland 41 19 Wadsworth Center, New York State Department of Health, and Department of Environmental Health 42 Sciences, School of Public Health, State University of New York at Albany, Empire State Plaza, Albany, NY 43 12201-0509, USA 44 20 Public Health and Environnement Laboratory, UMR 8079 Ecologie Systématique Evolution, Faculty of This is the postprint (accepted manuscript) version of the article published in Trends in Analytical Chemistry. https://doi.org/10.1016/j.trac.2018.03.009 © 2018. This manuscript version is made available under the CC-BY-NC-ND 4.0 license http://creativecommons.org/licenses/by-nc-nd/4.0/ 2 45 Pharmacy, Univ. Paris-Sud, CNRS, AgroParisTech, Université Paris-Saclay, 92296 Châtenay-Malabry, France 46 21 University of Bath, Department of Chemistry, Faculty of Science, Bath BA2 7AY, United Kingdom 47 22 Department of Inorganic and Analytical Chemistry, Poznan University of Medical Sciences, 6 Grunwaldzka 48 Street, 60-780 Poznan, Poland 49 23 Division for Marine and Environmental Research, Rudjer Boskovic Institute, Bijenicka 54, Zagreb, 10000 50 Croatia 51 24 Food Safety Laboratory, Health Sciences Authority, Singapore 52 25 Laboratory for Earth Surface Processes, College of Urban and Environmental Sciences, Peking University, 53 Beijing 100871, China 54 26 Department of Pharmacology and Toxicology, University of Iceland, Hofsvallagata 53, 107 Reykjavik, Iceland 55 27 Water and Soil Quality Research Group, Department of Environmental Chemistry, Institute of Environmental 56 Assessment and Water Research (IDAEA-CSIC), Jordi Girona 18-26, 08034 Barcelona, Spain 57 28 Department of Experimental and Clinical Toxicology, Center for Molecular Signaling (PZMS), Saarland 58 University, 66421 Homburg, Germany 59 29 Institute of Legal Medicine and Core Facility Metabolomics, Medical University of Innsbruck, Muellerstrasse 60 44, 6020 Innsbruck, Austria 61 30 Queensland Alliance for Environmental Health Sciences (QAEHS), University of Queensland, 39 Kessels Road 62 Coopers Plains, Queensland 4108, Australia 63 31 Laboratoire National de Santé, Service de toxicologie analytique et de chimie pharmaceutique, 1 rue Louis 64 Rech, L-3055 Luxembourg 65 32 National Institute of Legal Medicine and Forensic Sciences, South Branch, Rua Manuel Bento de Sousa n◦3, 66 1169-201 Lisbon, Portugal 67 33 Laboratory of Analytical Chemistry, Department of Chemistry, National and kapodistrian of Athens, 68 Panepistimiopolis Zografou, 15771 Athens, Greece 69 34 Department of Chemical Engineering, McGill University, Montreal, Quebec, Canada, H3A0C5 70 35 Eawag, Swiss Federal Institute of Aquatic Science and Technology. Urban Water 71 Management. Überlandstrasse 133, 8600 Dübendorf, Switzerland 72 Corresponding author: 73 Prof. Dr. Alexander L.N. van Nuijs 74 Toxicological Centre, University of Antwerp 75 Universiteitsplein 1 76 2610 Antwerp, Belgium 77 e-mail: [email protected] 78 tel: +32 (0)3 265 24 98 Highlights First worldwide inter-laboratory exercise for analysis of illicit drugs in wastewater Results revealed (pre-)analytical issues for certain analytes Six years of exercises have resulted in optimized procedures and protocols Quality control system will make wastewater-based epidemiology results more reliable 3 79 Abstract 80 This study presents the development of a worldwide inter-laboratory testing scheme for the analysis 81 of seven illicit drug residues in different matrices (standard solutions, tapand wastewater). By 82 repeating this exercise for six years with participation of 37 laboratories from 25 countries, the 83 testing scheme was substantially improved based on experiences gained across the years (e.g. matrix 84 type, sample conditions, spiking levels). From the exercises, (pre-)analytical issues (e.g. pH 85 adjustment, filtration) were revealed for some analytes which resulted in formulation of best86 practice protocols, both for inter-laboratory setup and analytical procedures. The results illustrate 87 the effectiveness of the inter-laboratory testing scheme in assessing laboratory performance in the 88 framework of illicit drug analysis in wastewater. The exercise proved that measurements of 89 laboratories were of high quality (> 80% satisfactory results for 6 out of 7 analytes) and that 90 analytical follow-up is important to assist laboratories in improving robustness of wastewater-based 91 epidemiology results. 92 93 Keywords 94 Illicit drugs; wastewater; inter-laboratory testing; wastewater-based epidemiology; quality assurance 4 95 1. Introduction 96 The measurement of the human excretion products of illicit drugs in influent wastewater has been 97 recognized as an alternative and complementary approach for estimating the consumption of illicit 98 drugs within communities, i.e. the catchment of wastewater treatment plants (WWTPs) [1-3]. The 99 principle behind wastewater-based epidemiology (WBE) derives from the fact that parent 100 compounds and/or their human metabolites (i.e., drug residues) are excreted in urine and faeces 101 following illicit drug use and end up in urban sewer systems [3]. The ability of WBE to provide useful 102 and timely information on temporal (daily, weekly, monthly, and annually) and spatial (withinand 103 between-countries) variations in illicit drug consumption has been demonstrated [4-15]. The 104 European Monitoring Centre for Drug and Drug Addiction (EMCDDA) has recently acknowledged the 105 added value of WBE to socio-epidemiological methods, such as population surveys, seizure data and 106 crime statistics, in generating useful and relevant data on population drug use [3]. 107 108 With the aim to improve and optimize WBE, a Europe-wide collaboration was initiated in 2010. Seven 109 European institutions – University of Antwerp (BE), Eawag (CH), University Jaume I (ES), Mario Negri 110 Institute (IT), KWR Watercycle Research Institute (NL), Norwegian Institute for Water Research NIVA 111 (NO), and University of Bath (UK) - established the research group SCORE (Sewage analysis CORe 112 group Europe) [16]. The ultimate goals of SCORE are (a) to collaborate in the field of WBE to provide 113 reproducible data; (b) to improve and harmonize the analytical procedures used in different 114 laboratories to analyze drug residues in wastewater samples; and (c) to perform international studies 115 comparing illicit drug consumption in communities across the world. To this end, SCORE has 116 coordinated monitoring studies and exercises to assure the quality of reported data based on agreed 117 best-practices tackling sampling, storage and analysis. Important results from this collaboration are 118 multi-city studies demonstrating the usefulness of WBE on an international level to obtain the most 119 recent data on illicit drug consumption [17-18]. 120 121 In order to further optimize and fine-tune WBE, it is imperative to gain knowledge on the sources of 122 uncertainties that are associated with the approach. In 2013, SCORE performed a thorough 123 evaluation on the uncertainties of WBE using the best-practice protocols and data that were 124 available from the comparative Europe-wide WBE research [19]. One of the cornerstones of WBE is 125 to accurately quantify concentrations of drug residues in wastewater samples by means of reliable 126 analytical procedures [20]. This requires fully validated analytical procedures before routine analysis 127 can be initiated and participation in external quality control schemes is, where possible, highly 128 recommended. External quality control through inter-laboratory exercises are based on the 5 129 distribution of the same test samples (in our case prepared by NIVA) to all participants. The latter 130 analyse all test samples without any knowledge of the concentrations of target analytes and return 131 their results to the coordinator of the exercise (in our case Eawag, who does not analyse test samples 132 and does not know the nominal spike value until final compilation of results). The coordinator 133 converts the submitted results into objective scores that reflect the performance of individual 134 laboratories and the group. These scores can alert participants of unexpected problems and can 135 result in actions to be taken [21]. 136 137 SCORE initiated inter-laboratory exercises in 2011 in order to develop a quality control scheme for 138 laboratories that analyze illicit drug residues in wastewater for WBE purposes. Since its debut, the 139 testing scheme has been carried out annually with increasing participation of different laboratories, 140 also extending the network outside Europe. The objectives of the presented interlaboratory exercise 141 are (a) to illustrate the results of the six-year inter-laboratory testing scheme; (b) to evaluate 142 advancements achieved over these years and to identify issues still to be resolved; (c) to formulate 143 recommendations for future inter-laboratory exercises and (d) to propose a robust quality control 144 system to improve the analytical performance of laboratories analyzing illicit drugs in wastewater. 145 146 2. Setup of the inter-laboratory exercises 147 2.1. Target analytes 148 A total of seven illicit drug residues were targeted in the inter-laboratory testing scheme. These 149 included cocaine (COC), benzoylecgonine (BE, cocaine metabolite), 3,4-methylenedioxy150 methamphetamine (MDMA), amphetamine (AMP), methamphetamine (METH), 11-nor-9-carboxy151 tetrahydrocannabinol (THC-COOH, THC metabolite), and 6-monoacetylmorphine (6-MAM, heroin 152 metabolite). These analytes are widely regarded as the main urinary biomarkers of the worldwide 153 most consumed illicit drugs (COC, MDMA, AMP, METH, cannabis and heroin) and are the focus of 154 most bioanalytical and WBE initiatives around the world [22]. Certified spiking solutions of each of 155 the target analytes were supplied by Cerilliant Corporation (Round Rock, Texas, USA). All spiking 156 solutions were supplied in sealed glass ampoules at 1 mg/mL in methanol. 157 158 2.2. Design of the exercises 159 The basis of the inter-laboratory testing scheme was to compare the performance of the analytical 160 procedures employed by participating laboratories. Two separate modules were included to evaluate 161 in each laboratory (a) the use of correct analytical reference standards and the performance of the 6 162 instrumental analysis (Module 1), and (b) the performance of entire analytical procedures applied to 163 the analysis of wastewater, including sample preparation (Module 2). 164 165 For Module 1, a methanol solution containing the seven target analytes was used. For Module 2, 166 samples of tap water and wastewater spiked with the seven analytes were employed. Participants 167 were asked to use their own in-house developed and validated analytical procedures for the analysis 168 of the samples. Replicate analysis of each sample was requested (n = 5 for Module 1 and n = 3 for 169 Module 2). Commonly, sample pre-treatment consisted of filtration followed by solid-phase 170 extraction for Module 2 samples. All laboratories employed liquid chromatography coupled to mass 171 spectrometry using mass-labelled internal standards to perform detection and quantification of the 172 analytes. More information on different techniques, including sample preparation procedures, used 173 for this type of analyses can be found in Castiglioni et al. (2013) and Hernandez et al. (in press) [19174 20]. 175 Analyte stability in various matrices and conditions is a crucial aspect of any inter-laboratory exercise 176 as it can substantially affect the outcomes of the analyses, particularly in the absence of certified 177 reference material in target matrices. Stability of illicit drugs in wastewater has been the subject of 178 numerous investigations, which were recently reviewed by McCall et al. (2016) [23]. Detailing the 179 results from all these studies goes beyond the scope of the present paper, however, a brief overview 180 regarding the analytes targeted in this inter-laboratory exercise is reported here. Both COC and BE 181 have been shown to be stable in wastewater over multiple weeks when stored refrigerated (4 °C and, 182 ideally, -20 °C), at low pH and in the dark. Similarly, MDMA, AMP and METH have been shown to be 183 stable under similar conditions. THC-COOH and 6-MAM, on the other hand, have been shown to be 184 very sensitive to temperature and, for THC-COOH, low pH. 185 186 2.3. Preparation of test samples 187 All test samples were prepared by the Norwegian Institute for Water Research (NIVA). Figure 1 and 188 Table 1 give an overview of the type of test samples included in each year (2011-2016) and the 189 nominal spiking levels used. The two modules together comprised three matrices (i.e., methanol, tap 190 water and wastewater) spiked at different concentrations for each of the target analytes. Spiking 191 concentrations for all matrices changed from year to year to avoid bias and ensure legitimate results. 192 Certified spiking solutions (1 mg/mL in methanol) were diluted to prepare working solutions at 100 193 µg/mL or 10 µg/mL in methanol. The working solutions were then used to prepare different test 194 samples. 195 The methanol solution (Module 1) containing the analytes was prepared from each of the 100 µg/mL 196 working solutions. Aliquots (1 mL) of this methanol sample were then transferred to separate glass 7 197 vials and capped. Each vial was accurately weighed and stored at -20 °C ahead of shipment to the 198 participants. Participants were asked to weigh the samples at arrival and to report deviations from 199 the weight at preparation. 200 Spiked wastewater and tap water samples (Module 2) were prepared in a 20 L high-density 201 polyethylene (HDPE) plastic container pre-washed with tap water and methanol. Twenty litres of cold 202 tap water or fresh wastewater from VEAS WWTP in Oslo (Norway) were poured into the container, 203 spiked with different volumes of the 10 µg/mL working standard solutions to obtain relevant 204 concentrations (at ng/L range) and stirred for 2 h to homogenize the mixture. In 2012, one of the 205 wastewater samples was used as it is; no spiking with target analytes occurred. 206 Samples from Module 2 were acidified to adjust the pH to 3.5 in 2012 and 2013. This pH adjustment 207 was agreed upon by the organizers of the exercise as at that time it was assumed that acidification of 208 samples was the best way to prevent degradation of the analytes [19]. In 2014-2016, no pH 209 adjustment of the tap water was performed because of the new insight into the negative effect of 210 low pH on the stability of THC-COOH in wastewater [23-24]. The changes in used matrices and pH 211 conditions across the years of the inter-laboratory exercise were the result of experiences of 212 previous years and of advancements made in the field of WBE. 213 Aliquots of at least 250 mL were placed in HDPE containers and stored at -20 °C before shipping to 214 the participants. As real wastewater was used, and which likely contained unknown concentrations 215 of the target analytes, it was not possible to use a genuine “blank” wastewater sample and nominal 216 values could thus not be reported. Instead, a total value, comprising background concentrations (x) 217 and the spiked level, was computed (Table 1). 218 219 2.4. Participants and sample shipping 220 The inter-laboratory exercises were organized by SCORE and were open to interested participants 221 from any institution. In order to participate to the exercise, laboratories were required to register 222 (without any payment) following an invitation sent out by SCORE or through the SCORE website [16]. 223 Over the period between 2011 and 2016, a total of 37 laboratories from 25 countries participated in 224 the exercises (for more details on participation in each year, see Table 1). Most of the participating 225 laboratories (81%) were located in Europe, while the rest (19%) was spread over different continents 226 (North-America, Asia and Oceania) (Figure 2). The participants located within the European Union 227 received the test samples, shipped on ice, during the following 24-48 hours while for the remaining 228 participants from the other continents the average transport time was 2-4 days. Temperature during 229 shipment was not recorded, but participants were asked to not analyse samples if defrosted upon 230 reception (responsibility if the participant). 231 8 232 2.5. Evaluation of results 233 Participating laboratories were required to report measured concentrations of the target analytes in 234 each sample type provided. Results of individual replicates were submitted. Furthermore, 235 participants had to clearly highlight when concentrations were not quantifiable (i.e., below limits of 236 quantification) or when the analysis for a certain compound was not performed. Limits of 237 quantification for each participant were estimated with a fixed protocol and compared to self238 assessed limit of quantifications. It was established at a signal-to-noise ratio of 10 using the 239 quantifier transition from chromatograms of samples spiked at the lowest validation level tested. The 240 estimated limits of quantification were for all participating laboratories within the same order of 241 magnitude and comparable to what was reported by each lab based on validation data. Since 2015, 242 one spiking level was used to evaluate whether the analytical procedures of participants had limit of 243 quantifications that are relevant in the context of WBE studies. If participants could not report values 244 for this sample, they were notified that their analytical procedures did not reach relevant sensitivity. 245 First, the mean concentration (m) of replicates for each participant and for each sample type was 246 calculated. Secondly, after testing for normality, a Grubbs’ test was performed to identify outliers 247 which were excluded from further analysis. From the remaining means, the group’s mean [i.e., mean 248 of means (M)] and the group’s standard deviation (SD) were computed. To evaluate the performance 249 of each participant ( ), z-scores ( ) for every analyte and sample type were calculated as follows: 𝑖 𝑧 𝑖 250 𝑧 𝑖 = 𝑚 𝑖 ‒ 𝑀 𝑆𝐷 251 Following the ISO standard, a laboratory passed the inter-laboratory exercise when its |z| ≤ 2 [21, 252 25]. Participants with results that were identified as outliers (Grubb’s test) or had |z|-values > 2 were 253 individually notified about the deviation and were allowed to recheck their submitted values for 254 inconsistencies or errors. Note that no detail ( , M) was supplied with the notification of the 𝑧 𝑖 255 deviation in order to maintain impartiality. If these laboratories were able to supply a viable 256 explanation (such as transcription errors), they were allowed to resubmit corrected results. If 257 accepted, newly submitted values were used to compute updated values for , M, SD and . 𝑚 𝑖 𝑧 𝑖 258 The purpose of this iterative process lies in the goal of SCORE to advance and improve WBE. The 259 inter-laboratory exercise was therefore used to assist laboratories in optimizing their analytical 260 procedures and improve the overall performance. 261 262 3. Results and Discussion 263 3.1. Assigned value: group’s mean vs. nominal concentration 264 15 473 Sciences and Engineering Research Council of Canada (NSERC), Ministry of Education, Youth and 474 Sports of the Czech Republic (projects CENAKVA and CENAKVA II), EU Marie Skłodowska-Curie 475 Fellowship (APOLLO 749845) and the Swiss National Science Foundation (SNSF, P2LAP2_164892). 476 The following persons are acknowledged for help in sample analysis: Marijan Ahel, Evroula Hapeshi, 477 Popi Karaolia, Esther López-García, Nicola Mastroianni, Cristina Postigo, Inés Racamonde, Rosario 478 Rodil, Isaac Rodríguez, Tania Rodríguez-Álvarez, Ivan Senta, , and Senka Terzic, . 16 References 1. van Nuijs ALN, Castiglioni S, Tarcomnicu I, Postigo C, Lopez de Alda M, Neels H, Zuccato E, Barcelo D, Covaci A. Illicit drug consumption estimations derived from wastewater analysis: a critical review. Sci Total Environ. 2011a;409:3564-77 2. Castiglioni S, Thomas KV, Kasprzyk-Hordern B, Vandam L, Griffiths P. Testing wastewater to detect illicit drugs: state of the art, potential and research needs. Sci Total Environ. 2014;487:613-20 3. European Monitoring Centre for Drugs and Drug Addiction. Assessing illicit drugs in wastewater: advances in wastewater-based drug epidemiology, Insights 22, Publications Office of the European Union, 2016, Luxembourg 4. Harman C, Reid M, Thomas KV. In situ calibration of a passive sampling device for selected illicit drugs and their metabolites in wastewater, and subsequent year-long assessment of community drug usage. Environ Sci Technol. 2011;45:5676-82 5. van Nuijs ALN, Mougel JF, Tarcomnicu I, Bervoets L, Blust R, Jorens PG, Neels H, Covaci A. Sewage epidemiology--a real-time approach to estimate the consumption of illicit drugs in Brussels, Belgium. Environ Int. 2011b;37:612-21 6. Nefau T, Karolak S, Castillo L, Boireau V, Levi Y. Presence of illicit drugs and metabolites in influents and effluents of 25 sewage water treatment plants and map of drug consumption in France. Sci Total Environ. 2013;461-462:712-22 7. Mackuľak T, Skubák J, Grabic R, Ryba J, Birošová L, Fedorova G, Spalková V, Bodík I. National study of illicit drug use in Slovakia based on wastewater analysis. Sci Total Environ. 2014;494-495:158-65 8. Ort C, Eppler JM, Scheidegger A, Rieckermann J, Kinzig M, Sörgel F. Challenges of surveying wastewater drug loads of small populations and generalizable aspects on optimizing monitoring design. Addiction. 2014;109:472-81 9. Ostman M, Fick J, Näsström E, Lindberg RH. A snapshot of illicit drug use in Sweden acquired through sewage water analysis. Sci Total Environ. 2014;472:862-71. 17 10. Been F, Bijlsma L, Benaglia L, Berset JD, Botero-Coy AM, Castiglioni S, Kraus L, Zobel F, Schaub MP, Bücheli A, Hernández F, Delémont O, Esseiva P, Ort C. Assessing geographical differences in illicit drug consumption--A comparison of results from epidemiological and wastewater data in Germany and Switzerland. Drug Alcohol Depend. 2016;161:189-99 11. Kankaanpää A, Ariniemi K, Heinonen M, Kuoppasalmi K, Gunnar T. Current trends in Finnish drug abuse: Wastewater based epidemiology combined with other national indicators. Sci Total Environ. 2016;568:864-74 12. Krizman I, Senta I, Ahel M, Terzic S. Wastewater-based assessment of regional and temporal consumption patterns of illicit drugs and therapeutic opioids in Croatia. Sci Total Environ. 2016;566567:454-62. 13. Lai FY, O'Brien JW, Thai PK, Hall W, Chan G, Bruno R, Ort C, Prichard J, Carter S, Anuj S, Kirkbride KP, Gartner C, Humphries M, Mueller JF. Cocaine, MDMA and methamphetamine residues in wastewater: Consumption trends (2009-2015) in South East Queensland, Australia. Sci Total Environ. 2016;568:803-9 14. Zuccato E, Castiglioni S, Senta I, Borsotti A, Genetti B, Andreotti A, Pieretti G, Serpelloni G. Population surveys compared with wastewater analysis for monitoring illicit drug consumption in Italy in 2010-2014. Drug Alcohol Depend. 2016;161:178-88. 15. Mastroianni N, López-García E, Postigo C, Barceló D, López de Alda M. Five-year monitoring of 19 illicit and legal substances of abuse at the inlet of a wastewater treatment plant in Barcelona (NE Spain) and estimation of drug consumption patterns and trends. Sci Total Environ. 2017;609:916-926. 16. SCORE (2010) Sewage analysis CORE group Europe. URL: http://score-cost.eu. Accessed: 2017-0907. (Archived by WebCite® at http://www.webcitation.org/6tIO1NrbC) 17. Thomas KV, Bijlsma L, Castiglioni S, Covaci A, Emke E, Grabic R, Hernández F, Karolak S, KasprzykHordern B, Lindberg RH, Lopez de Alda M, Meierjohann A, Ort C, Pico Y, Quintana JB, Reid M, Rieckermann J, Terzic S, van Nuijs ALN, de Voogt P. Comparing illicit drug use in 19 European cities through sewage analysis. Sci Total Environ. 2012;432:432-9 18 18. Ort C, van Nuijs ALN, Berset JD, Bijlsma L, Castiglioni S, Covaci A, de Voogt P, Emke E, FattaKassinos D, Griffiths P, Hernández F, González-Mariño I, Grabic R, Kasprzyk-Hordern B, Mastroianni N, Meierjohann A, Nefau T, Ostman M, Pico Y, Racamonde I, Reid M, Slobodnik J, Terzic S, Thomaidis N, Thomas KV. Spatial differences and temporal changes in illicit drug use in Europe quantified by wastewater analysis. Addiction. 2014;109:1338-52 19. Castiglioni S, Bijlsma L, Covaci A, Emke E, Hernández F, Reid M, Ort C, Thomas KV, van Nuijs ALN, de Voogt P, Zuccato E. Evaluation of uncertainties associated with the determination of community drug use through the measurement of sewage drug biomarkers. Environ Sci Technol. 2013;47:145260 20. Hernández F, Castiglioni S, Covaci A, de Voogt P, Emke E, Kasprzyk-Hordern B, Ort C, Reid M, Sancho JV, Thomas KV, van Nuijs ALN, Zuccato E, Bijlsma L. Mass spectrometric strategies for the investigation of biomarkers of illicit drug use in wastewater. Mass Spectrom Rev. in press (doi: 10.1002/mas.21525) 21. Thompson M, Ellison SL, Wood R. The international harmonized protocol for the proficiency testing of analytical chemistry laboratories. Pure Appl Chem. 2006; 78, 145-196 22. Baselt R. Disposition of toxic drugs and chemicals in man. 11th edition, Biomedical Publications, Seal Beach, CA, 2017, ISBN 978-0-692-77499-1 23. McCall AK, Bade R, Kinyua J, Lai FY, Thai PK, Covaci A, Bijlsma L, van Nuijs ALN, Ort C. Critical review on the stability of illicit drugs in sewers and wastewater samples. Water Res. 2016;88:933-47 24. Senta I, Krizman I, Ahel M, Terzic S. Assessment of stability of drug biomarkers in municipal wastewater as a factor influencing the estimation of drug consumption using sewage epidemiology. Sci Total Environ. 2014;487:659-65 25. ISO13528:2015(E). Statistical methods for use in proficiency testing by interlaboratory comparisons, ISO, 2015, Geneva, Switzerland 26. Causanilles A, Baz-Lomba JA, Burgard DA, Emke E, Gonzalez-Marino I, Krizman-Matasic I, Li A, Love ASC, McCall AK, Montes R, van Nuijs ALN, Ort C, Quintana JB, Senta I, Terzic S, Hernandez F, de 19 Voogt P, Bijlsma L. Improving wastewater-based epidemiology to estimate cannabis use: focus on the initial aspects of the analytical procedure. Anal. Chim. Acta 2017;988:27-33. 27. González-Mariño I, Quintana JB, Rodríguez I, Cela R. Determination of drugs of abuse in water by solid-phase extraction, derivatisation and gas chromatography-ion trap-tandem mass spectrometry. J Chromatogr A. 2010;1217:1748-60 28. Chen C, Kostakis C, Irvine RJ, Felgate PD, White JM. Evaluation of pre-analysis loss of dependent drugs in wastewater: stability and binding assessments. Drug Test Anal. 2013;5:716-21. 20 Figure 1. Inter-laboratory overview and scheme of the sample preparation and shipment for Module 2. 21 Figure 2. Map with location of the participants of the inter-laboratory exercises 22 Figure 3. Deviation of the assigned value (= group’s mean) from the nominal value (= spiking level) for the standard solution (top) and the tap water samples (bottom) in relation to the assigned value for the seven analytes. The dotted line represents 25% deviation. Entries with deviations > 25% are marked with the year of the inter-laboratory exercise. 23 Figure 4. Relative standard deviation of the group in relation to the assigned value M (logarithmic scale) for the three matrices [standard solution (blue), tap water (green) and wastewater (red)] and seven analytes. All years (2011-2016) included. 24 Figure 5. Boxplot showing the difference in the group’s RSD for the three different matrices (MEOH = standard solution; TW = tap water; WW = wastewater) in 2013 and 2014 for all analytes. 2011 z counts −4 −2 0 2 4 0 1 2 311 0 1 0 469ng/mL 18% 11 [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 3 412 0 1 0 30ng/mL 13% 13 [36] MeOH 2013 −4 −2 0 2 4 0 1 2 3 4 513 1 1 0 330ng/mL 9% 15 23 [400] MeOH 2014 −4 −2 0 2 4 0 1 2 3 418 1 1 0 559ng/mL 12% 20 9 [600] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 5 625 0 0 0 34ng/mL 16% MeOH [40] 2016 −4 −2 0 2 4 0 1 2 3 4 5 6 724 0 1 0 28ng/mL 17% 15 [25] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs Cocaine (COC) counts −4 −2 0 2 4 0 1 2 3 4 514 0 1 0 98ng/L 27% 22 [100] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 718 0 1 0 114ng/L 10% 20 [150] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 621 0 1 0 101ng/L 22% 22 [150] Water(3) −4 −2 0 2 4 0 1 2 3 4 5 6 7 823 0 2 0 80ng/L 30% 35, 37 [100] Water(1) counts −4 −2 0 2 4 0 1 2 3 415 0 0 0 53ng/L 32% [50] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 617 0 1 1 52ng/L 12% 9 5 [60] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 6 7 8 920 0 2 0 59ng/L 25% 1, 10 [100] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 722 1 2 0 45ng/L 25% 37 14, 35 [50] Water(2) counts −4 −2 0 2 4 0 1 2 3 412 0 0 1 70ng/L 24% 13 [x+8] Wastewater(1) −4 −2 0 2 4 0 1 2 3 414 0 1 0 208ng/L 34% 15 [x+100] Wastewater(2) −4 −2 0 2 4 0 1 2 3 418 0 1 0 113ng/L 11% 5 [x+150] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 5 6 719 0 2 1 34ng/L 39% 1, 10 8 [60] Water(1) z −4 −2 0 2 4 0 1 2 3 4 5 6 7 8 9 10 11 12 22 1 2 0 5.7ng/L 43% 37 5, 31 [5] Water(3) z counts −4 −2 0 2 4 0 1 2 3 411 0 1 1 64ng/L 20% 7 13 [x] Wastewater(2) z −4 −2 0 2 4 0 1 2 314 0 1 0 160ng/L 28% 15 [x+50] Wastewater(1) z −4 −2 0 2 4 0 1 2 3 4 5 617 0 1 1 53ng/L 15% 9 5 [x+60] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:07 / CO p4 8 2011 z counts −4 −2 0 2 4 0 1 2 3 4 511 0 1 0 507ng/mL 20% 11 [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 3 411 0 1 0 95ng/mL 7% 13 [120] MeOH 2013 −4 −2 0 2 4 0 1 2 3 4 514 0 1 0 612ng/mL 19% 10 [800] MeOH 2014 −4 −2 0 2 4 0 1 2 3 4 5 6 7 818 1 2 0 703ng/mL 11% 20 3, 25 [900] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 5 6 724 0 2 0 55ng/mL 14% MeOH 8, 25 [60] 2016 −4 −2 0 2 4 0 1 2 3 4 5 6 7 825 0 1 0 19ng/mL 20% 32 [20] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs MDMA counts −4 −2 0 2 4 0 1 2 3 4 5 614 0 1 0 271ng/L 22% 13 [300] Water(2) −4 −2 0 2 4 0 1 2 3 4 520 0 0 0 323ng/L 19% [400] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 7 821 0 2 0 219ng/L 15% 8, 25 [260] Water(3) −4 −2 0 2 4 0 1 2 3 4 5 6 725 0 1 0 136ng/L 23% 35 [150] Water(1) counts −4 −2 0 2 4 0 1 2 3 4 5 614 0 1 0 84ng/L 26% 13 [90] Water(1) −4 −2 0 2 4 0 1 2 3 4 520 0 0 0 64ng/L 25% [80] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 622 0 1 0 94ng/L 19% 25 [120] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 7 825 0 1 0 70ng/L 23% 35 [75] Water(2) counts −4 −2 0 2 4 0 1 2 3 412 0 0 1 95ng/L 29% 13 [x+42] Wastewater(1) −4 −2 0 2 4 0 1 2 3 4 513 0 1 1 311ng/L 23% 13 10 [x+300] Wastewater(2) −4 −2 0 2 4 0 1 2 3 4 5 6 719 0 1 0 318ng/L 19% 25 [x+400] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 5 621 0 2 0 73ng/L 16% 8, 25 [90] Water(1) z −4 −2 0 2 4 0 1 2 3 4 5 6 722 1 2 1 8.5ng/L 26% 37 26, 32 21 [8] Water(3) z counts −4 −2 0 2 4 0 1 2 39 0 0 4 9.5ng/L 28% 4, 10, 13, 24 [x] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 513 0 1 1 114ng/L 24% 13 10 [x+90] Wastewater(1) z −4 −2 0 2 4 0 1 2 3 4 5 618 0 2 0 81ng/L 18% 18, 25 [x+80] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:08 / CO p5 8 2011 z counts −4 −2 0 2 4 0 1 2 3 4 5 611 0 1 0 513ng/mL 20% 11 [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 3 4 5 613 0 0 0 45ng/mL 16% [56] MeOH 2013 −4 −2 0 2 4 0 1 2 3 414 0 1 0 557ng/mL 15% 6 [700] MeOH 2014 −4 −2 0 2 4 0 1 2 3 4 520 1 0 0 627ng/mL 10% 20 [750] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 5 6 7 8 924 0 2 0 98ng/mL 13% MeOH 25, 29 [120] 2016 −4 −2 0 2 4 0 1 2 3 4 5 6 726 0 0 0 41ng/mL 17% [40] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs Amphetamine (AMP) counts −4 −2 0 2 4 0 1 2 3 4 5 615 0 0 0 252ng/L 21% [250] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 7 819 0 1 0 148ng/L 19% 25 [200] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 623 0 0 0 170ng/L 18% [200] Water(3) −4 −2 0 2 4 0 1 2 3 4 5 625 0 1 0 142ng/L 20% 32 [140] Water(1) counts −4 −2 0 2 4 0 1 2 3 415 0 0 0 86ng/L 26% [80] Water(1) −4 −2 0 2 4 0 1 2 3 4 518 1 1 0 56ng/L 20% 7 25 [70] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 623 0 0 0 127ng/L 19% [160] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 725 0 1 0 71ng/L 18% 32 [70] Water(2) counts −4 −2 0 2 4 0 1 2 313 0 0 0 389ng/L 23% [x+118] Wastewater(1) −4 −2 0 2 4 0 1 2 3 4 514 0 0 1 663ng/L 24% 10 [x+250] Wastewater(2) −4 −2 0 2 4 0 1 2 3 4 519 1 0 0 308ng/L 13% 25 [x+200] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 523 0 0 0 67ng/L 24% [80] Water(1) z −4 −2 0 2 4 0 1 2 3 4 5 6 7 8 924 0 1 1 14ng/L 31% 37 21 [12] Water(3) z counts −4 −2 0 2 4 0 1 2 313 0 0 0 168ng/L 29% [x] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 414 0 0 1 493ng/L 26% 10 [x+80] Wastewater(1) z −4 −2 0 2 4 0 1 2 3 419 1 0 0 207ng/L 12% 25 [x+70] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:08 / CO p6 8 2011 z counts −4 −2 0 2 4 0 1 2 3 411 0 1 0 457ng/mL 27% 11 [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 3 4 512 0 1 0 97ng/mL 10% 23 [128] MeOH 2013 −4 −2 0 2 4 0 1 2 3 414 0 1 0 129ng/mL 13% 16 [200] MeOH 2014 −4 −2 0 2 4 0 1 2 3 4 5 618 1 2 0 121ng/mL 11% 20 3, 25 [150] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 5 6 7 825 0 1 0 63ng/mL 18% MeOH 28 [80] 2016 −4 −2 0 2 4 0 1 2 3 4 525 0 1 0 49ng/mL 16% 32 [50] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs Methamphetamine (METH) counts −4 −2 0 2 4 0 1 2 3 4 514 1 0 0 53ng/L 22% 23 [50] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 619 0 1 0 83ng/L 19% 25 [100] Water(2) −4 −2 0 2 4 0 1 2 3 4 522 0 1 0 145ng/L 21% 25 [180] Water(3) −4 −2 0 2 4 0 1 2 3 4 5 6 7 825 0 1 0 119ng/L 18% 32 [120] Water(1) counts −4 −2 0 2 4 0 1 210 2 0 3 12ng/L 30% 11, 23 4, 10, 18 [10] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 6 718 0 2 0 21ng/L 25% 7, 16 [25] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 6 722 0 1 0 67ng/L 23% 10 [90] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 725 0 1 0 62ng/L 19% 32 [60] Water(2) counts −4 −2 0 2 4 0 1 2 3 4 512 0 1 0 422ng/L 22% 10 [x+49] Wastewater(1) −4 −2 0 2 4 0 1 2 3 413 1 0 1 563ng/L 22% 23 10 [x+50] Wastewater(2) −4 −2 0 2 4 0 1 2 3 4 517 0 2 1 174ng/L 16% 16, 25 21 [x+100] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 5 6 722 0 1 0 40ng/L 26% 10 [50] Water(1) z −4 −2 0 2 4 0 1 2 3 4 5 622 1 1 2 6.1ng/L 30% 37 15 17, 21 [6] Water(3) z counts −4 −2 0 2 4 0 1 2 312 1 0 0 318ng/L 18% 10 [x] Wastewater(2) z −4 −2 0 2 4 0 1 2 312 1 1 1 522ng/L 18% 23 18 10 [x+10] Wastewater(1) z −4 −2 0 2 4 0 1 2 3 4 5 6 718 0 1 1 114ng/L 20% 25 21 [x+25] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:08 / CO p7 8 2011 z counts −4 −2 0 2 4 0 1 2 39 0 1 0 525ng/mL 11% 11 [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 310 0 1 1 199ng/mL 11% 10 18 [226] MeOH 2013 −4 −2 0 2 4 0 1 2 311 1 1 0 837ng/mL 11% 23 20 [1000] MeOH 2014 −4 −2 0 2 4 0 1 2 3 4 5 6 7 8 9 10 11 16 1 2 0 1090ng/mL 8% 20 11, 21 [1000] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 5 6 719 0 2 2 204ng/mL 11% MeOH 5, 8 28, 29 [200] 2016 −4 −2 0 2 4 0 1 2 3 4 5 619 0 1 0 134ng/mL 14% 5 [125] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs THC−COOH counts −4 −2 0 2 4 0 1 28 0 0 2 40ng/L 47% 15, 23 [400] Water(2) −4 −2 0 2 4 0 1 2 312 1 0 3 53ng/L 43% 1 9, 15, 20 [500] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 6 718 0 1 1 353ng/L 21% 11 15 [450] Water(3) −4 −2 0 2 4 0 1 2 3 417 0 2 1 247ng/L 24% 27, 32 21 [300] Water(1) counts −4 −2 0 2 4 0 1 27 0 0 4 24ng/L 78% 9, 15, 20, 23 [100] Water(1) −4 −2 0 2 4 0 1 2 311 1 0 4 30ng/L 42% 1 2, 9, 15, 20 [200] Water(1) −4 −2 0 2 4 0 1 2 3 4 5 618 0 1 1 262ng/L 22% 11 15 [350] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 616 0 2 2 121ng/L 28% 27, 37 5, 21 [150] Water(2) counts −4 −2 0 2 4 0 1 2 3 4 59 0 1 2 213ng/L 10% 16 4, 23 [x+75] Wastewater(1) −4 −2 0 2 4 0 1 28 0 0 2 136ng/L 71% 15, 23 [x+400] Wastewater(2) −4 −2 0 2 4 0 1 2 312 0 0 4 80ng/L 68% 7, 9, 18, 20 [x+500] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 5 619 0 0 1 211ng/L 22% 15 [250] Water(1) z −4 −2 0 2 4 0 1 2 3 4 517 1 1 1 44ng/L 40% 37 27 21 [50] Water(3) z counts −4 −2 0 2 4 0 1 2 3 4 59 0 1 2 88ng/L 18% 18 4, 23 [x] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 48 0 0 2 71ng/L 70% 15, 23 [x+100] Wastewater(1) z −4 −2 0 2 4 0 1 210 0 0 7 49ng/L 52% 1, 2, 7, 9, 17, 18, 20 [x+200] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:08 / CO p8 8 2011 z counts −4 −2 0 2 4 0 1 2 311 0 0 0 472ng/mL 19% [500] MeOH 2012 z −4 −2 0 2 4 0 1 2 37 0 1 0 32ng/mL 8% 17 [56] MeOH 2013 −4 −2 0 2 4 0 1 2 310 0 0 0 195ng/mL 22% [300] MeOH 2014 −4 −2 0 2 4 0 1 2 3 414 0 1 0 195ng/mL 14% 11 [250] MeOH 2015 −4 −2 0 2 4 0 1 2 3 4 517 0 2 0 127ng/mL 14% MeOH 16, 25 [180] 2016 −4 −2 0 2 4 0 1 2 3 4 517 0 1 0 61ng/mL 19% 11 [60] MeOH nr. of labs passed outliers |z|>2 <LOQ mean of all labs |z|<2 [nominal spike value] rsd of all labs |z|<2 lab IDs lab IDs lab IDs 6-MAM counts −4 −2 0 2 4 0 1 2 38 0 0 2 39ng/L 39% 11, 13 [90] Water(2) −4 −2 0 2 4 0 1 2 3 414 0 0 0 123ng/L 26% [180] Water(2) −4 −2 0 2 4 0 1 2 3 4 5 616 0 1 0 223ng/L 22% 25 [300] Water(3) −4 −2 0 2 4 0 1 2 3 417 0 1 0 145ng/L 26% 32 [160] Water(1) counts −4 −2 0 2 4 0 1 2 36 0 0 3 11ng/L 68% 10, 11, 13 [30] Water(1) −4 −2 0 2 4 0 1 2 313 0 1 0 71ng/L 20% 5 [90] Water(1) −4 −2 0 2 4 0 1 2 3 4 516 0 1 0 147ng/L 26% 25 [210] Water(2) −4 −2 0 2 4 0 1 2 3 417 0 1 0 75ng/L 28% 32 [80] Water(2) counts −4 −2 0 2 4 0 1 27 0 0 1 97ng/L 26% 13 [x+88] Wastewater(1) −4 −2 0 2 4 0 1 27 0 0 3 60ng/L 28% 10, 11, 13 [x+90] Wastewater(2) −4 −2 0 2 4 0 1 2 314 0 0 0 71ng/L 34% [x+180] Wastewater(2) z −4 −2 0 2 4 0 1 2 3 4 517 0 0 0 110ng/L 31% [150] Water(1) z −4 −2 0 2 4 0 1 2 3 415 1 0 2 6.4ng/L 44% 37 5, 20 [5] Water(3) z counts −4 −2 0 2 4 0 1 2 37 0 0 1 10ng/L 34% 13 [x] Wastewater(2) z −4 −2 0 2 4 0 1 26 0 0 3 24ng/L 25% 10, 11, 13 [x+30] Wastewater(1) z −4 −2 0 2 4 0 1 2 3 413 0 0 1 24ng/L 46% 15 [x+90] Wastewater(1) Version 'v2b_TrAC' generated 2018−03−09 16:17:08 / CO p9 8 BE MeOH Lab ID normalized concentrations (with mean of means per year after removing outliers) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 Version 'v2b_TrAC' generated 2018−03−09 16:16:53 / CO ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ●● ● ●● ● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2011 MeOH (n.labs = 12); 1 = 511ng/L 2012 MeOH (n.labs = 13); 1 = 61ng/L 2013 MeOH (n.labs = 15); 1 = 402ng/L 2014 MeOH (n.labs = 21); 1 = 440ng/L 2015 MeOH (n.labs = 26); 1 = 20ng/L 2016 MeOH (n.labs = 26); 1 = 30ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p10 8 BE Water Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:16:55 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) 2013 W1 (n.labs = 15); 1 = 46ng/L 2013 W2 (n.labs = 15); 1 = 148ng/L 2014 W1 (n.labs = 20); 1 = 28ng/L 2014 W2 (n.labs = 20); 1 = 95ng/L 2015 W1 (n.labs = 23); 1 = 39ng/L 2015 W2 (n.labs = 23); 1 = 80ng/L 2015 W3 (n.labs = 23); 1 = 121ng/L 2016 W1 (n.labs = 26); 1 = 143ng/L 2016 W2 (n.labs = 26); 1 = 74ng/L 2016 W3 (n.labs = 26); 1 = 12ng/L p11 8 BE Wastewater Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:16:56 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 WW1 (n.labs = 13); 1 = 181ng/L 2013 WW2 (n.labs = 13); 1 = 154ng/L 2014 WW1 (n.labs = 15); 1 = 397ng/L 2014 WW2 (n.labs = 15); 1 = 496ng/L 2015 WW1 (n.labs = 20); 1 = 201ng/L 2015 WW2 (n.labs = 20); 1 = 271ng/L analyses) mean (boxes=quartiles of triplicate outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p12 8 COC MeOH Lab ID normalized concentrations (with mean of means per year after removing outliers) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 Version 'v2b_TrAC' generated 2018−03−09 16:16:56 / CO ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2011 MeOH (n.labs = 12); 1 = 469ng/L 2012 MeOH (n.labs = 13); 1 = 30ng/L 2013 MeOH (n.labs = 15); 1 = 330ng/L 2014 MeOH (n.labs = 20); 1 = 559ng/L 2015 MeOH (n.labs = 25); 1 = 34ng/L 2016 MeOH (n.labs = 25); 1 = 28ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2 2*sd (after removing outliers => |z|=2) p13 8 AMP Water Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:00 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 W1 (n.labs = 15); 1 = 86ng/L 2013 W2 (n.labs = 15); 1 = 252ng/L 2014 W1 (n.labs = 20); 1 = 56ng/L 2014 W2 (n.labs = 20); 1 = 148ng/L 2015 W1 (n.labs = 23); 1 = 67ng/L 2015 W2 (n.labs = 23); 1 = 127ng/L 2015 W3 (n.labs = 23); 1 = 170ng/L 2016 W1 (n.labs = 26); 1 = 142ng/L 2016 W2 (n.labs = 26); 1 = 71ng/L 2016 W3 (n.labs = 25); 1 = 14ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p20 8 AMP Wastewater Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:00 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● 2013 WW1 (n.labs = 13); 1 = 389ng/L 2013 WW2 (n.labs = 13); 1 = 168ng/L 2014 WW1 (n.labs = 14); 1 = 493ng/L 2014 WW2 (n.labs = 14); 1 = 663ng/L 2015 WW1 (n.labs = 20); 1 = 207ng/L 2015 WW2 (n.labs = 20); 1 = 308ng/L analyses) mean (boxes=quartiles of triplicate outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p21 8 METH MeOH Lab ID normalized concentrations (with mean of means per year after removing outliers) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 Version 'v2b_TrAC' generated 2018−03−09 16:17:01 / CO ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● 2011 MeOH (n.labs = 12); 1 = 457ng/L 2012 MeOH (n.labs = 13); 1 = 97ng/L 2013 MeOH (n.labs = 15); 1 = 129ng/L 2014 MeOH (n.labs = 21); 1 = 121ng/L 2015 MeOH (n.labs = 26); 1 = 63ng/L 2016 MeOH (n.labs = 26); 1 = 49ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p22 8 METH Water Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:01 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ●● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 W1 (n.labs = 12); 1 = 12ng/L 2013 W2 (n.labs = 15); 1 = 53ng/L 2014 W1 (n.labs = 20); 1 = 21ng/L 2014 W2 (n.labs = 20); 1 = 83ng/L 2015 W1 (n.labs = 23); 1 = 40ng/L 2015 W2 (n.labs = 23); 1 = 67ng/L 2015 W3 (n.labs = 23); 1 = 145ng/L 2016 W1 (n.labs = 26); 1 = 119ng/L 2016 W2 (n.labs = 26); 1 = 62ng/L 2016 W3 (n.labs = 25); 1 = 6.1ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p23 8 METH Wastewater Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:02 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● 2013 WW1 (n.labs = 13); 1 = 422ng/L 2013 WW2 (n.labs = 13); 1 = 318ng/L 2014 WW1 (n.labs = 14); 1 = 522ng/L 2014 WW2 (n.labs = 14); 1 = 563ng/L 2015 WW1 (n.labs = 19); 1 = 114ng/L 2015 WW2 (n.labs = 19); 1 = 174ng/L analyses) mean (boxes=quartiles of triplicate outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p24 8 THC-COOH MeOH Lab ID normalized concentrations (with mean of means per year after removing outliers) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 Version 'v2b_TrAC' generated 2018−03−09 16:17:02 / CO ● ● ● ● ● ●● ●● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● 2011 MeOH (n.labs = 10); 1 = 525ng/L 2012 MeOH (n.labs = 12); 1 = 199ng/L 2013 MeOH (n.labs = 13); 1 = 837ng/L 2014 MeOH (n.labs = 19); 1 = 1090ng/L 2015 MeOH (n.labs = 21); 1 = 204ng/L 2016 MeOH (n.labs = 20); 1 = 134ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p25 8 THC-COOH Water Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:03 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 W1 (n.labs = 8); 1 = 24ng/L 2013 W2 (n.labs = 8); 1 = 40ng/L 2014 W1 (n.labs = 12); 1 = 30ng/L 2014 W2 (n.labs = 13); 1 = 53ng/L 2015 W1 (n.labs = 19); 1 = 211ng/L 2015 W2 (n.labs = 19); 1 = 262ng/L 2015 W3 (n.labs = 19); 1 = 353ng/L 2016 W1 (n.labs = 19); 1 = 247ng/L 2016 W2 (n.labs = 18); 1 = 121ng/L 2016 W3 (n.labs = 19); 1 = 44ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p26 8 THC-COOH Wastewater Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:04 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 WW1 (n.labs = 10); 1 = 213ng/L 2013 WW2 (n.labs = 10); 1 = 88ng/L 2014 WW1 (n.labs = 8); 1 = 71ng/L 2014 WW2 (n.labs = 8); 1 = 136ng/L 2015 WW1 (n.labs = 10); 1 = 49ng/L 2015 WW2 (n.labs = 12); 1 = 80ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2xsd (after removing outliers => |z|=2) p27 8 6-MAM MeOH Lab ID normalized concentrations (with mean of means per year after removing outliers) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 Version 'v2b_TrAC' generated 2018−03−09 16:17:04 / CO ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ●● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2011 MeOH (n.labs = 11); 1 = 472ng/L 2012 MeOH (n.labs = 8); 1 = 32ng/L 2013 MeOH (n.labs = 10); 1 = 195ng/L 2014 MeOH (n.labs = 15); 1 = 195ng/L 2015 MeOH (n.labs = 19); 1 = 127ng/L 2016 MeOH (n.labs = 18); 1 = 61ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p28 8 6-MAM Water Lab ID normalized concentrations (with mean of means per year after removing outliers) Version 'v2b_TrAC' generated 2018−03−09 16:17:04 / CO 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 0.0 0.5 1.0 1.5 2.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ●● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 2013 W1 (n.labs = 6); 1 = 11ng/L 2013 W2 (n.labs = 8); 1 = 39ng/L 2014 W1 (n.labs = 14); 1 = 71ng/L 2014 W2 (n.labs = 14); 1 = 123ng/L 2015 W1 (n.labs = 17); 1 = 110ng/L 2015 W2 (n.labs = 17); 1 = 147ng/L 2015 W3 (n.labs = 17); 1 = 223ng/L 2016 W1 (n.labs = 18); 1 = 145ng/L 2016 W2 (n.labs = 18); 1 = 75ng/L 2016 W3 (n.labs = 16); 1 = 6.4ng/L mean (boxes=quartiles of triplicate analyses) outlier (Grubbs, p<0.001, repeated) excluded (after removing outliers, |z|>2) 2*sd (after removing outliers => |z|=2) p29 8