Passive sampling of wastewater as a tool for the long-term monitoring of community exposure: Illicit and prescription drug trends as a proof of concept
Graphical abstract
Introduction
Monitoring drug use has traditionally been performed by questionnaire-based surveys and police statistics. Estimating population drug use through the analysis of wastewater samples has been established as an approach for monitoring patterns of community drug use (Castiglioni et al., 2014). Wastewater-based epidemiology (WBE) studies for drugs have provided valuable information, showing spatial and temporal differences across different countries (Ort et al., 2014b, Thomas et al., 2012). More recently, WBE results have been also compared with other sources of information confirming its potential as a complementary approach for obtaining a more accurate picture of the drug use situation (Baz-Lomba et al., 2016b, Been et al.,, Zuccato et al., ). Furthermore, WBE has recently been applied to assess the community level exposure of humans to a range of environmental stressors (Gracia-Lor et al., 2016, Rousis et al., 2016, Rousis et al., 2017) as well as their combined response to such stressors (Ryu et al., 2016). The generation of community level exposure data that can be compared with other complementary sources of data has a clear potential within environmental epidemiology.
Despite good agreement with the other sources of data, it is possible that wastewater data may be typified by low temporal representativeness and high spatial variability due to the use of different substances and spatial and temporal trends in availability. WBE results therefore need to be carefully interpreted (Baz-Lomba et al., 2016b). Thus far, the temporal coverage of most of the WBE studies performed has been typically limited to a one-week sampling program (Ort et al., 2014b). In order to more accurately estimate the representative mean annual substance use by WBE a recent study recommended the use of stratified random sampling schemes (typically 56 samples per year) (Ort et al., 2014a). Furthermore, different sampling strategies based on the sampling frequency or composite sampling mode have been evaluated in order to decrease the sampling uncertainty (Ort et al., 2010). However, increasing the sampling frequency implies an additional costs together with the power and space requirements of an automated sampling device and such a frequency may still prove inadequate in certain circumstances such as the short-term changes in use patterns or variations in concentrations associated with other external factors such as precipitation (Ort et al., 2014a).
Passive sampling devices (PSD) are an alternative sampling tool to overcome some of the above-mentioned issues. PSD have been demonstrated as a good alternative for the monitoring of drugs and other micropollutants in wastewater providing time-integrated estimates that compensate for fluctuating concentrations (Harman et al., 2011b, Kaserzon et al., 2014). PSD may also decrease the limits of quantification compared with traditional sampling and be used as a screening tool for the detection of emerging compounds present at very low concentrations (Alvarez et al., 2014). Furthermore, the use of PSD over a specified period can be performed without human intervention, without any power requirements and at low cost. For example, the annual monitoring of drugs in wastewater can be performed with as few as 26 PSD (Harman et al., 2011b).
The polar organic chemical integrative sampler (POCIS) has thus far been applied for the analysis of over 300 polar organic substances in water (Harman et al., 2012, Morin et al., 2012). This includes a number of pharmaceuticals and illicit drugs, as well as other polar contaminants such as pesticides (Gonzalez-Rey et al., 2015, Jones-Lepp et al., 2004, Metcalfe et al., 2011). The physicochemical properties of the compounds will determine whether they accumulate in the sampler based on the different solute-solvent-sorbent interactions (i.e. the version of POCIS presented in this study has a good selectivity for compounds with log Kow approximately between 2 and 4) (Harman et al., 2011b). The theory and modelling of chemical uptake by POCIS have been explained in detail elsewhere (Alvarez et al., 2004, Huckins et al., 1993, Vermeirssen et al., 2012). There are however few studies that consider modelling uptake rates for real in situ environmental exposures. Depending on sampler design, PSD can be used in either equilibrium or time-weighted average (TWA) modes to give concentrations of the desired analyte. In contrast to PSDs for hydrophobic compounds, where sampling rates (RS) can be modelled by physicochemical properties such as molecular weight, RS for POCIS must first be calibrated experimentally. Laboratory generated RS can vary significantly between different studies depending on the different calibration methods and conditions used and standardization of the different calibration methods has been recommended in order to reduce these discrepancies (Harman et al., 2012, Morin et al., 2012). Furthermore, RS generated in the laboratory under controlled exposure conditions may not be representative of the actual values under different and variable environmental conditions that can lead to biased data when calculating TWA concentrations (Miller et al., 2016).
One of the primary uncertainties associated with the calculation of POCIS RS, which in turn fundamentally affects the reliability of POCIS derived TWA concentrations, is the influence of environmental factors, such as the rate of water flow (Kaserzon et al., 2013, Li et al., 2010b), temperature (Li et al., 2010a), pH (Li et al., 2011) and biofouling (Harman et al., 2009). Different approaches have been proposed, such as the use of external RS corrections (Alvarez et al., 2007), the performance reference compounds (PRC) approach used for hydrophobic PSD (Huckins et al., 2002) and more recently the development of the diffusive gradient in thin-film sampler for organics (Challis et al., 2016). All of these approaches have challenges, but a comprehensive method for relating the uptake in POCIS to environmental factors remains elusive (Harman et al., 2012). In-situ calibration of POCIS has been proposed as an alternative strategy to generate more reliable and constant RS for a specific site, however only a few papers have published in-situ RS values (Harman et al., 2011b, Jacquet et al., 2012, Mazzella et al., 2010, Zhang et al., 2008). However, in-situ calibration is also not without its challenges; largely due to the extra costs and the need for more extensive water sampling, compared with laboratory methods. An overall lack of understanding of the sorption phenomena for different compounds means that it is prudent to consider POCIS derived water concentration data as semi-quantitative (Harman et al., 2011a, Miège et al., 2015). Dalton and colleagues (Dalton et al., 2014) have described the variability of in-situ Rs associated with environmental factors (Morin et al., 2012) and more recently, Poulier and colleagues (Poulier et al., 2014) estimated that the uncertainty related to their POCIS data for several pesticides might be as high as 138%.
The aim of the current study was to evaluate whether POCIS are a suitable and cost-effective alternative to grab sampling for the long-term monitoring of substance use and potentially exposure at community level through WBE. The in-situ RS was determined for a number of pharmaceuticals (atenolol, citalopram, carbamazepine, oxazepam, metoprolol, 2-ethylidene-1,5-dimethyl-3,3-diphenylpyrrolidine (EDDP) and morphine) and illicit drugs (cocaine, benzoylecgonine and methamphetamine) in POCIS to estimate substance use trends over 2012 and 2013. The reproducibility of RS was estimated by overlapping three in-situ calibrations to mitigate the potential confounding effects and impact of different environmental conditions. The accuracy of the POCIS RS was evaluated by assessing two different uncertainty levels, taking into account the coefficient of variation of the three RS calculated during the three different calibration periods and the repeatability for each of the triplicates deployed during subsequent long-term (2-yr) monitoring.
Section snippets
Chemicals, materials and POCIS samplers
Information on chemicals, materials and POCIS samplers is provided in the Supporting Information.
Wastewater and POCIS extraction and analysis
Information on wastewater and POCIS extraction and analysis is provided in the Supporting Information.
Quality assurance
Information on quality assurance is provided in the Supporting Information.
In situ calibration study design
All samples were collected at the VEAS wastewater treatment plant (WWTP) in Oslo (Norway). VEAS treats sewage for approximately 600,000 people of which the city contributes about 70.5% and the adjoining areas representing the
Sample rates: three in situ calibrations in one
All of the target compounds were detected and quantified in both POCIS and wastewater samples with the exception of methamphetamine in POCIS-Set 1 (after 18 days of exposure), which suffered from high ion suppression due to the complexity of the matrix. Water concentrations for all of the target compounds were mainly influenced by the wastewater flow rates due to the heavy periods of rain during calibration. When comparing the water flow-corrected loads, cocaine, and its main urinary metabolite
Conclusions
POCIS TWA concentrations have been shown as a good complementary tool for the monitoring of certain pharmaceuticals and drugs present in wastewater when performing in-situ calibration. The poor knowledge of modelling uptake and the use of proper exposure corrections are the main issues related with the estimation of POCIS TWA concentrations and were solved by performing three overlapped in-situ calibrations under different environmental conditions. Furthermore, in-situ data allowed the
Acknowledgements
The authors would like to thank Pia Ryrfors, VEAS WWTP, for providing access to the sampling area and the best information regarding the wastewater samples. Jose Antonio Baz Lomba acknowledges the EU International Training Network SEWPROF (Marie Curie-FP7-PEOPLE, grant number 317205) for his Early Stage Researcher grant.
References (50)
- et al.
Spatial and temporal trends in occurrence of emerging and legacy contaminants in the Lower Columbia River 2008-2010
Sci. Total Environ.
(2014) - et al.
Passive Sampling Techniques in Environmental Monitoring
(2007) - et al.
Target and suspect screening of psychoactive substances in sewage-based samples by UHPLC-QTOF
Anal. Chim. Acta
(2016) - et al.
Recent trends in the availability and use of amphetamine and methamphetamine in Norway
Forensic Sci. Int.
(2015) - et al.
Testing wastewater to detect illicit drugs: state of the art, potential and research needs
Sci. Total Environ.
(2014) - et al.
Atrazine contamination at the watershed scale and environmental factors affecting sampling rates of the polar organic chemical integrative sampler (POCIS)
Environ. Pollut.
(2014) - et al.
Enantiomer profiling of high loads of amphetamine and MDMA in communal sewage: a Dutch perspective
Sci. Total Environ.
(2014) - et al.
A passive sampling method for detecting analgesics, psycholeptics, antidepressants and illicit drugs in aquatic environments in the Czech Republic
Sci. Total Environ.
(2014) - et al.
Occurrence of pharmaceutical compounds and pesticides in aquatic systems
Mar. Pollut. Bull.
(2015) - et al.
Passive sampling of perfluorinated chemicals in water: flow rate effects on chemical uptake
Environ. Pollut.
(2013)
The effects of dissolved organic matter and pH on sampling rates for polar organic chemical integrative samplers (POCIS)
Chemosphere
Contaminants in the coastal karst aquifer system along the Caribbean coast of the Yucatan Peninsula, Mexico
Environ. Pollut.
Position paper on passive sampling techniques for the monitoring of contaminants in the aquatic environment – achievements to date and perspectives
Trends Environ. Anal. Chem.
An in situ intercomparison exercise on passive samplers for monitoring metals, polycyclic aromatic hydrocarbons and pesticides in surface waters
TrAC Trends Anal. Chem.
Chemical calibration, performance, validation and applications of the polar organic chemical integrative sampler (POCIS) in aquatic environments
Trac Trends Anal. Chem.
Can POCIS be used in Water Framework Directive (2000/60/EC) monitoring networks? A study focusing on pesticides in a French agricultural watershed
Sci. Total Environ.
Monitoring population exposure to pesticides based on liquid chromatography-tandem mass spectrometry measurement of their urinary metabolites in urban wastewater: a novel biomonitoring approach
Sci. Total Environ.
Wastewater-based epidemiology to assess human exposure to pyrethroid pesticides
Environ. Int.
Comparing illicit drug use in 19 European cities through sewage analysis
Sci. Total Environ.
Analysis of emerging contaminants in sewage effluent and river water: comparison between spot and passive sampling
Anal. Chim. Acta
Development of a passive, in situ, integrative sampler for hydrophilic organic contaminants in aquatic environments
Environ. Toxicol. Chem.
Comparison of pharmaceutical, illicit drug, alcohol, nicotine and caffeine levels in wastewater with sale, seizure and consumption data for 8 European cities
BMC Public Health
Development and calibration of an organic-diffusive gradients in thin films aquatic passive sampler for a diverse suite of polar organic contaminants
Anal. Chem.
Wastewater Analysis and Drugs: a European Multi-city Study, European Monitoring Centre for Drugs and Drug Addiction
Cited by (0)
- 1
Current address: Queensland Alliance for Environmental Health Science (QAEHS), University of Queensland, 39 Kessels Road, Coopers Plains QLD 4108, Australia.