The International Journal of Life Cycle Assessment https://doi.org/10.1007/s11367-021-01908-y LCI METHODOLOGY AND DATABASES Characterizing antibiotics in LCA—a review of current practices and proposed novel approaches for including resistance O. Nyberg1  · A. Rico2,3 · J. B. Guinée4 · P. J. G. Henriksson5,6,7 Received: 8 December 2020 / Accepted: 30 March 2021 © The Author(s) 2021 Abstract Purpose With antibiotic resistance (ABR) portrayed as an increasing burden to human health, this study reviews how and to what extent toxicological impacts from antibiotic use are included in LCAs and supplement this with two novel approaches to include ABR, a consequence of antibiotic use, into the LCA framework. Methods We review available LCA studies that deal with toxicological aspects of antibiotics to evaluate how these impacts from antibiotics have been characterized. Then, we present two novel approaches for including ABR-related impacts in life cycle impact assessments (LCIAs). The first approach characterizes the potential for ABR enrichment in the environmental compartment as a mid-point indicator, based on minimum selective concentrations for pathogenic bacteria. The second approach attributes human health impacts as an endpoint indictor, using quantitative relationships between the use of antibiotics and human well-being. Results and discussion Our findings show that no LCA study to date have accounted for impacts related to ABR. In response, we show that our novel mid-point indicator approach could address this by allowing ABR impacts to be characterized for envi- ronmental compartments. We also establish cause-effect pathways between antibiotic use, ABR, and human well-being that generate results which are comparable with USEtox and most endpoint impact assessment approaches for human toxicology. Conclusions Our proposed methods show that currently overlooked impacts from ABR enrichment in the environment could be captured within the LCA framework as a robust characterization methodology built around the established impact model USEtox. Substantial amounts of currently unavailable data are, however, needed to calculate emissions of antibiotics into the environment, to develop minimum selective concentrations for non-pathogenic bacteria, and to quantify potential human health impacts from AB use. Keywords Antibiotics · LCA · Resistance · AMR · Antimicrobials · Human health impacts · Resistance · Toxicology · USEtox 1 Introduction Antibiotic (AB) substances are used for treating bacterial infec- tions by killing or inhibiting growth of these organisms (Davies and Davies 2010; Kümmerer 2009). The term AB is broad and Communicated by Matthias Finkbeiner. * O. Nyberg 4 Institute of Environmental Sciences (CML), Leiden oskar.nyberg@su.se University, Einsteinweg 2, Leiden 2333 CC, The Netherlands 5 1 Department of Ecology, Environment and Plant Sciences, Stockholm Resilience Centre, Stockholm University, Stockholm University, Svante Arrhenius väg 20A, Kräftriket, 2B, 10691 Stockholm, Sweden Stockholm, Sweden 6 Beijer Institute of Ecological Economics, Royal Swedish 2 IMDEA Water Institute, Science and Technology Campus Academy of Sciences, Box 50005, 104 05 Stockholm, of  the University of Alcalá, Avenida Punto Com 2, Sweden Alcalá de Henares, 28805 Madrid, Spain 7 WorldFish, Jalan Batu Maung, Batu Maung, 3 Cavanilles Institute of Biodiversity and Evolutionary 11960 Bayan Lepas, Penang, Malaysia Biology, University of Valencia, c/ Catedrático José Beltrán 2, Paterna, 46980 Valencia, Spain Vol.:(012 3456789) The International Journal of Life Cycle Assessment envelops hundreds of different compounds, of which close to is commonly referred to as “the resistome” (D’Costa et al. 2006; 300 are classified as important for human medicine and are cat- Wright 2007; Surette and Wright 2017). Models have been egorized into 30 + groups according to their origin and mode of developed for correlating anthropogenic AB emissions and the action (World Health Organization 2019). ABs have become our resistance development in the environment based on abiotic primary tool for treating and preventing the proliferation of path- parameters (Amos et al. 2015), as well as relationships between ogenic bacterial diseases in a wide range of settings, including AB use in food animal production and human exposure to ABR human medicine, livestock (van den Bogaard and Stobberingh pathogens (van Bunnik and Woolhouse 2017). Nonetheless, it 1999), and aquaculture farms (Sapkota et al. 2008), as well as in remains difficult to establish useful dose-response relationships agriculture to control for bacterial diseases in plants (Stockwell between AB use and the associated impacts, since causalities and Duffy 2012). They are the primary treatment for pneumonia, between AB use, resistance development, and transmission are tuberculosis and gastrointestinal infections, diseases that histori- difficult to pinpoint (Price et al. 2015). cally are thought to have been responsible for 30% of all human Despite the overwhelming human health benefits gained deaths, but they are also essential for post-surgical care (Fair from using pharmaceuticals like ABs, it is imperative to assess and Tor 2014). Inappropriate use of ABs and their environmen- the negative impacts from emissions of these substances to tal release may, however, result in negative consequences (in better improve regulation and manage impacts. Life cycle a concentration dependent context) in the environmental com- assessment (LCA), among other environmental frameworks, partment as they can: (a) infer toxic effects on several living has been used to assess impacts related to the use of ABs. organisms (Carlsson et al. 2009) and (b) modify microbial com- LCA details the environmental impacts related to a product’s munity compositions and affect ecological functions (Costanzo or service’s life cycle. In LCA, human and ecosystem impacts et al. 2005; Grenni et al. 2018). Additionally, negative effects to from the release of chemicals are generally captured by toxico- humans include: (c) indiscriminately killing both pathogenic logical impact categories. Most of these toxicological impact and non-pathogenic bacteria, including bacterial communities assessment methodologies build upon available laboratory that fill useful biological functions (Jernberg et al. 2010; Lange toxicity data and extrapolation methods assigning fate, effect, et al. 2016); (d) induce side effects in humans (Wypych and and exposure pathways to chemical agents. Depending upon Marsland 2018); and (e) promote the development of antibiotic scope, freshwater ecotoxicity, marine ecotoxicity, and/or can- resistant genes (Jernberg et al. 2010; Pérez-Cobas et al. 2013). cer and non-cancer-related human toxicity (European Com- Of the abovementioned impacts, antibiotic resistance (ABR) mission 2010) are either estimated at an intermediary point development in pathogenic bacteria is seen as the most forebod- in the underlying impact pathway by midpoint indicators (e.g. ing for human wellbeing, due to the central role of ABs in mod- number of cancer or non-cancer disease cases in humans or ern medicine (World Health Organization 2014). Bacteria can potentially affected fraction of species in the aquatic environ- develop resistance to ABs either through mutations or by acqui- ment; PAF m3 kg−1), or at the end of the impact pathway as sition of resistance genes from other bacteria through different endpoint indicators (e.g. disability adjusted life-years (DALY) modes of horizontal gene transfer even at low concentrations or potentially disappeared fraction of species to change in con- (i.e., ng/L to few μg/L; Cabello 2006; Grace 2015; Bengtsson- centration (PDF m3 kg−1) (Hauschild and Huijbregts 2015). Palme and Larsson 2016; Jutkina et al. 2018; Klein et al. 2018). In this study, we first review (Sec. 3.1) how impacts from AB ABR genes have existed for millennia (D’costa et al. 2011), use and ABR have been characterized among published LCAs. but the current use of tens of thousands of tonnes of ABs each We then propose two novel impact characterization models to year has resulted in substantial releases of ABs into the environ- capture ABR in LCA (Sec. 3.2) and discuss their respective ment (Van Boeckel et al. 2015; Robinson et al. 2016), thereby strengths and weaknesses. Conclusively (Sec. 4), we summarize generating large areas for interactions between ABs and bacteria the outcomes to suggest best practices for AB use in LCA. which can lead selection for AB resistance genes in environ- mental bacteria (Cabello 2006; Mathew et al. 2007; Rizzo et al. 2013; Larsson 2014; Xiong et al. 2015; Cabello et al. 2016; 2 M ethodology Chuah et al. 2016; Larsson et al. 2018; Osman et al. 2018). There is a growing body of literature connecting ABs released 2.1 R eview of AB use in LCA literature into the environment with the development of ABR in bacterial communities (Heinemann 1999; Wright 2007; Larsson 2014) Relevant literature on previous LCA studies incorporating AB and an increased frequency of ABR genes in environmental or pharmaceuticals was screened using web of science on 10 compartments (Finley et al. 2013), yet the links between envi- March 2021. The search used the phrase “TS = (“LCA” OR ronmental ABR bacteria and the impacts to human health are “LCIA” OR “life cycle assessment” OR “life cycle analysis” not fully understood. The collection of genes coding for ABR OR “life cycle inventory assessment” OR USEtox) AND present in the environment can be viewed as a pool of available TS = (antimicr* OR antibio* OR pharmaceutical* OR micro- genetic material which can be transferred between bacteria and poll*))” while delimiting the search to English-language articles 1 3 The International Journal of Life Cycle Assessment spanning the years 2008–2020. Search denominators and meth- 2.2 Proposing novel approaches to include ABR odology overview is given in the supplementary information in LCA (Fig. S1). The search generated 266 articles that were targeted for screening, of which 80 passed a title screening. Another Potential impacts of ABs reaching the environment are sepa- 37 were rejected following abstract screening. Of remaining rated between toxicological impacts and resistance develop- 43 articles read in full, 27 did not characterize ABs and were ment. Toxicological impacts can be captured in the current excluded. Thus, 16 articles are included in the review, with the toxicological impact model USEtox, applying a three-step addition of the study by Henriksson et al. (2015) that had been approach to derive characterization factors (CFs) for toxic identified prior to screening, and are summarized in Table 1. substances applied in the LCA, considering fate, exposure, Figures were designed using RStudio, Inc, Version 1.1.423 and and effect data (Fantke et al. 2017; Rosenbaum et al. 2011). Microsoft PowerPoint 2016. Table 1 Overview of articles reviewed that considerer AB toxicity (both human toxicity and ecotoxicity) in LCA studies Author Year System LCIA methodology (for Midpoint/endpoint Number of new Ecotoxicity database (for ecotoxicity) characterization AB CFs calcu- antibiotics) lated Munoz et al. 2008 WWTP EDIP97 and USES 2.0 Both 7 USEPA Ecotox d atabasea Munoz et al. 2009 WWTP EDIP97 and USES 2.0 Both 0 USEPA Ecotox d atabasea Hospido et al. 2010 WWTP CML 2 baseline 2000 Midpoint Muñoz et al. 2008 Stone et al. 2010 Swine farming USES-LCA 2.0 Both 0 - Stone et al. 2011 Swine farming EcoIndicator 99 Endpoints 0 - Igos et al. 2012 WWTP EDIP97, EDIP2003, and Endpoint (EDIP2003), 0 Wikipharmab and USEPA ReCiPe + USEtox Midpoint (USEtox) Ecotox database (ECO- SAR) Morais et al. 2013 CFs USEtox Midpoint 6 ECOTOX database 4.0c Igos et al. 2013 WWTP EDIP2003 and USEtox Endpoint (EDIP2003), 0 Wikipharmab and USEPA Midpoint (USEtox) Ecotox database (ECO- SAR) Alfonsin et al. 2014 CFs USES-LCA 2.0 and Midpoint (both USES- 4 USEtox, various USEtox LCA & USEtox) literatured Henriksson et al.* 2015 Aquaculture USEtox Midpoint 24 various literaturee, USEPA ECOTOX databasea Lorenzo-Toja et al. 2016 WWTP USES-LCA 2.0 Both 0 Alfonsín et al. 2014 Ortiz de Garcia et al. 2017 CFs USEtox Midpoint 7 USEPA Ecotox d atabasea, various l iteraturef Tarpani et al. 2018 WWTP USEtox Midpoint 0 Alfonsín et al. 2014 Rahman et al. 2018 WWTP USEtox Midpoint 0 USEtox, Alfonsín et al. 2014 Emara et al. 2018 CFs USEtox, EDIP 97, and Both 0 USEPA Ecotox d atabasea, USES-LCA Wikipharmab, TOX- LINE databasef, USEtox Li et al. 2019 WWTP USEtox Midpoint 11 RIVM e-toxBaseg, USEtox Tarpani et al. 2020 WWTP USEtox Midpoint 0 Alfonsín et al. 2014 * Study acknowledged prior to review, not captured by the literature screening. a http:// cfpub.e pa. gov/e cotox/ b http:// www. wikip harma. org c ECOTOX Database Release 4.0, US Environmental Protection Agency, 2007 d Barceló and Petrovic, 2011; Fent et al., 2006; Isidori et al., 2005; Santos et al., 2010 e Rico et al., 2013; Rico and Van den Brink, 2014 f Dobbins et al., 2009; Iannacone and Alvariño, 2009; Ortiz de García et al., 2014; Santos et al., 2010; Terasaki et al., 2009 g https:// toxnet. nlm.n ih. gov/ newto xnet/ toxlin e. htm h www.e- toxba se. com (not accessible 2020-09-17). WWTP= wastewater treatment plants; CFs = characterization factors 1 3 The International Journal of Life Cycle Assessment For impacts related to resistance development, we choose to freshwater ecotoxicity using USEtox V1.01, while Lorenzo- target: (1) ABR development in the environment as a mid- Toja et al. (2016) characterize emissions using USES-LCA point indicator by sourcing effect data related to concentra- 2.0. The two last studies, Rahman et al. (2018) and Li et al. tions of ABs where resistance development can occur based (2019), perform LCIA with TRACI 2.1, where USEtox is on the risk assessment methodology developed by (Rico the proposed toxicological characterization methodology. et al. 2017), and (2) human health impacts as a result from Four studies deal with more than one toxicological charac- AB use by suggesting a linear dose-response model based on terization method as the focus of these studies is to evaluate statistical correlation connecting AB use, resistance devel- characterization results or LCIA methodologies rather than opment and human health impacts. This was accomplished performing LCAs. by sourcing data from veterinary and medical literature Overall, the scope and methodologies vary substantially while limiting the scope to the EU due to data scarcity. among studies, as do calculated CFs for ABs. Eleven stud- ies apply USEtox methodology for impact characteriza- tions, of which four establish CFs and seven carry out full 3 R esults and discussion LCAs. Among these, Alfonsín et al. (2014) characterize four ABs for both USEtox and USES-LCA. Lorenzo-Toja et al. 3.1 Findings of the review (2016) utilize these USES-LCA CFs, while Tarpani and Azapagic (2018) and Tarpani et al. (2020) use the USEtox Seventeen LCA-related articles deal with ABs, of which two CFs from Alfonsín et al. (2014) for LCAs of wastewater consider indirect toxic impacts related to AB production treatment plants (WWTPs). Muñoz et al. (2008) rank the and transportation (Stone et al. 2010, 2011), three calculate toxicity for 97 pollutants, create seven new CFs for anti- freshwater ecotoxicity CFs for ABs and use them in LCAs biotics, and apply these in an LCA of WWTP. These CFs (Muñoz et al. 2008; Henriksson et al. 2015; Li et al. 2019), are subsequently used by Muñoz et al. (2009) and Hospido two only calculate CFs for ABs (Alfonsín et al. 2014; Ortiz et al. (2010) in other LCAs of WWTPs. Henriksson et al. de García et al. 2017) and eight use available CFs for ABs (2015) and Li et al. (2019) also generate novel CFs for ABs to conduct LCAs (Muñoz et al. 2009; Hospido et al. 2010; using the USEtox methodology (V1.01 and V2.0 respec- Igos et al. 2012, 2013; Lorenzo-Toja et al. 2016; Rahman tively) and use them for their respective LCAs. Meanwhile, et al. 2018; Tarpani and Azapagic 2018; Tarpani et al. 2020). Morais et al. (2013) investigate how pH variation influence Meanwhile, Morais et al. (2013) compare the uncertainty USEtox fate modelling of agents and create novel CFs for six and variability of characterization results at various pH using ABs. Ortiz de García et al. (2017) calculate CFs for seven the USEtox scientific consensus model V1.01 (Rosenbaum additional ABs, and Emara et al. (2018) compare available et al. 2008), and Emara et al. 2018 compare AB-related impact assessment methodologies and CFs for ABs and CFs using different impact assessment methodologies. An other agents. overview of reviewed articles is presented in Table 1. The Of the six studies generating novel freshwater ecotoxic- articles were published from 2008 and onwards, the same ity CFs for ABs, five present fate, exposure, and effect fac- year as the USEtox consensus model was developed. Of the tors in their respective supplementary information. Two of thirteen studies that carried out life cycle inventory assess- these (Henriksson et al. 2015; Ortiz de García et al. 2017) ments (LCIAs), ten evaluated wastewater treatment plant also detail AB’s source toxicity data, but derive their effect interventions and three animal farming. The two studies by factors somewhat differently. Ortiz de García et al. (2017) Munoz et al. (2008, 2009) use both EDIP 97 (Potting and prioritized chronic toxicity data, as recommended by Fantke Hauschild 2006) and USES-LCA (Huijbregts et al. 2000) for et al. (2017), while Henriksson et al. (2015) use both acute LCIA, with EDIP 97 and USES-LCA 2.0 characterization and chronic toxicity data. methodologies respectively thereby characterizing impacts To date, 40 antibiotic agents have been characterized for at both mid-point and endpoint. Stone et al. (2010) use ReC- freshwater ecotoxicity impact with the USEtox methodol- iPe 2008 (Goedkoop et al. 2013) that promotes USES-LCA ogy (Fig. 1; data available in supplementary information, 2.0 characterization methodology for toxicological impacts. Table S1). For some ABs that were characterized more than Meanwhile, Stone et al. 2011 use EcoIndicator99 v2.06 once, large variations exist (e.g. 3.2E + 1 PAF m3 per day (Goedkoop and Spriensma 2001) and Igos et al. (2012) and kg emitted−1 to 1.06E + 7 PAF m3 per day and kg emit- EDIP 97 and EDIP 2003, as well as ReCiPe 2008 combined ted Amoxicillin), while 24 out of 40 antibiotics only were with USEtox. Igos et al. (2013) use EDIP2003 and ReCiPe characterized once. 2008 combined with USEtox. Hospido et al. (2010) declares Despite novel contributions of AB CFs using USEtox, using CLM 2 baseline 2000 for midpoint impact assessment only 40 of the over 300 ABs deemed critically important for (Guinée et al. 2002). Henriksson et al. (2015), Tarpani and human medicine have been characterized for freshwater eco- Azapagic (2018), and Tarpani et al. (2020) all characterize toxicity (Fantke et al. 2017; World Health Organization 2019). 1 3 The International Journal of Life Cycle Assessment Fig. 1 Available CFs for AB emissions to freshwater calculated using USEtox, FETP = freshwater ecotoxicity potential. Units are presented in PAF m3 day kg−1 emitted Moreover, the USEtox V2.01 database presents 23 readily ABs. The cause for inconsistencies in freshwater ecotoxic- available human health impacts for ABs, but only four of these ity CFs remains unclear since the underlying toxicity data have CFs for human toxicity attributed, the rest are defined as used for calculating the effect factors remain unavailable. either “n.a” or “0”, which implies toxicity data are labelled However, Morais et al. (2013) show that CFs are sensitive to as “neglected” (USEtox® organic substances database 2.01 differences in abiotic degradation rates as well as ecotoxico- [built 10-July-2017]). An additional nine ABs are character- logical effect (EC50) data, which generally are sourced from ized for human health impacts among the reviewed studies different empirical experiments, as no standardized database (see Table 2 based on data from Emara et al. 2018). The CFs for such data is available for ABs. Nonetheless, the chemi- from Ortiz de García et al. (2017) are reported as “human cal properties and experimental data on ABs that support the toxicity potential – total” (an aggregation of both carcinogenic CFs remain inconsistent across literature. Greater efforts are and non-carcinogenic impact), yet the toxicological input data therefore needed towards generating further modelling and label all of the characterized ABs as non-carcinogenic (NC) experimental data for some compounds, as well as complet- and they are therefore labelled as HTP-NC in Table 2. ing and harmonizing datasets of toxicological properties that To summarize, the use of ABs and the environmental could support effect factors to yield robust CFs. release of AB residues may result in ecotoxicological impacts Based on our review, we conclude that the most severe on animals and humans, changes in microbial communities knowledge gaps include the evaluation of potential human in ecosystems and humans, and resistance development. Eco- health impacts pathways from AB use, resistance development toxicological impacts are addressed in several LCA studies in the environment, and human health impacts associated to using different LCIA methodologies. Most of the reviewed ABR. Four LCA studies included impacts from ABs on human studies look at freshwater ecotoxicity using USEtox, but con- health, but no LCA study to date that is assessing potential clude up to six orders of magnitude difference in CFs for some human health impacts has tried to capture the consequences 1 3 The International Journal of Life Cycle Assessment Table 2 Human toxicity potential characterizations available for ABs created with the USEtox method. Units are reported as “cases per kg emit- ted”. HTP-C: human toxicity potential: carcinogenic, HTP-NC: human toxicity potential: non carcinogenic Antibiotic Source Toxicity type Emission to Emission to Emission to Emission to Emission to Emission to freshwater seawater natural soil agricultural urban air rural air soil Metronidazole USEtox HTP-C 3.30E−07 6.52E−11 1.14E−07 2.98E-07 5.40E-07 4.81E-07 Sulfamethazine USEtox HTP-C 1.08E−07 2.31E−11 2.69E−10 1.45E-09 6.48E-08 4.48E-08 Amoxicillin Ortiz de Garcia HTP-NC 2.11E−08 4.84E−12 2.74E−09 4.50E-09 1.75E-08 1.80E-08 2017 Azithromycin Ortiz de Garcia HTP-NC 6.11E−06 2.10E−08 1.76E−07 9.44E-07 4.60E-06 4.72E-06 2017 CEFACLOR Ortiz de Garcia HTP-NC 1.54E−08 3.17E−12 2.06E−09 3.02E-09 1.24E-08 1.28E-08 2017 Ciprofloxacin Ortiz de Garcia HTP-NC 1.13E−07 2.51E−11 4.46E−08 6.17E-08 1.11E-07 1.15E-07 2017 Clarithromycin Ortiz de Garcia HTP-NC 3.14E−07 1.21E−09 8.60E−08 1.95E-07 2.92E-07 3.01E-07 2017 Levofloxacin Ortiz de Garcia HTP-NC 2.51E−06 6.45E−10 1.20E−06 1.41E-06 2.17E-06 2.26E-06 2017 Norfloxacin Ortiz de Garcia HTP-NC 3.05E−07 6.17E−11 1.09E−07 1.28E-07 1.20E-07 1.25E-07 2017 Sulfamethoxa- Alfonsin 2014 HTP-NC 1.58E−07 1.04E−10 1.03E−08 1.70E-08 3.24E-08 7.03E-09 zole Sulfamethoxa- USEtox HTP-NC 4.70E−07 1.45E−10 1.21E−07 4.35E-07 1.28E-07 7.84E-08 zole Trimethoprim Alfonsin 2014 HTP-NC 5.66E−07 1.54E−10 2.29E−08 3.66E-08 9.16E-08 2.39E-08 Trimethoprim USEtox HTP-NC 2.78E−06 7.67E−10 3.67E−08 3.18E-07 5.64E-07 2.84E-07 of ABR. So far the only impacts considered are direct toxic- to be quantified using causal relationships: (1) the use of mini- ity without addressing ABR, which is briefly mentioned by mum selective concentration (MSC) distributions to character- Igos et al. (2012) and Emara et al. (2018). As described by ize ABR enrichment in the environment as a mid-point impact; Ashbolt et al. (2013), resistance genes which propagate in the and (2) the correlation between ABs used on regional scale environment and become a human health issue adhere to dif- with human health impacts caused by ABR, quantified as ferent pathways than ecotoxicological impacts and will there- DALYs per kg of AB used as an endpoint impact. fore require a different impact assessment approach. Since The first approach, modelling of ABR enrichment in the ABs act as a causative agent for ABR development, spread environment, uses the USEtox® methodology as a starting of resistant bacteria can subsequently occur in the environ- point, as it is the framework recommended by both the Soci- ment and be transmitted to humans. Both latter steps cannot ety of Environmental Toxicology and Chemistry (SETAC) be assessed through physiochemical fate models because bac- and the Joint Research Centre (European Commission teria are the main carriers of these resistance genes and they 2010). In USEtox, Fate factors capture the physiochemical propagate through complex biological interactions. As such, properties of agents and predict their estimated distribution disease transmission and subsequent impacts on human health in environment compartments; exposure factors calculate are not assessed using the exposure and effect pathways cur- bioavailability to aquatic organisms or exposure pathways rently included in the USEtox model (Eq. 1). Without relevant for humans, while effect factors benchmark the actual toxic- pathways to capture the extent of ABR impacts, these are pos- ity of a compound, generally based upon laboratory studies sibly greatly underestimated in LCAs. (Eq. 1 describes human impact characterization and Eq. 3 describes ecotoxicological characterization). 3.2 T wo proposed approaches for addressing ABR ( ) kg in LCA emitted Characterization factor = Fate factor kg in compartmentper day ( ) kg intake Since no cause-effect pathway exists for the impacts of ABR × Exposure factor per kg day in compartment development in LCIA methodology, we below present two ( ) cases kg intake × Effect factor per novel approaches that could potentially allow for ABR impacts day day (1) 1 3 The International Journal of Life Cycle Assessment Again, this approach is aimed at capturing toxicological for a given bacterial taxon. Furthermore, the number of bacteria effects of ABs, and does not prescribe how to incorporate that is included in the EUCAST database is large (> 8) for the ABR impacts. most commonly used antibiotics (Bengtsson-Palme and Lars- son 2016). Questionable is how representative the pathogenic bacteria included in the EUCAST database are for bacterial 3.2.1 Approach 1: characterizing ABR enrichment communities in the environment. In this regard, Bengtsson- in the environment (mid‑point) Palme and Larsson (2016) demonstrate a weak link between taxonomic divergence and sensitivity to antibiotics, and the Approach 1 attempts to quantify the enrichment of resistance study by Tello et al. (2012) finds no significant difference in genes in environmental bacteria based on environmental fate sensitivity between pathogenic bacteria that inhabit human bod- models for ABs and theoretical minimum selective concen- ies exclusively and those that have been reported to also inhabit tration (MSC) distributions as a mid-point impact, similar environmental compartments. Thus, the EUCAST database to USEtox’s ecotoxicity impacts. presents itself as a useful proxy for estimating the susceptibil- The proposed characterization model is based on the ity of environmental communities to antibiotic pressure. methodology developed by Rico et al. (2017), where MSCs Similarly to the formula established by Jolliet et al. (2003) are inferred from minimum inhibitory concentrations (MIC) for the effect factor calculation for ecotoxicity, the effect factor for pathogenic bacteria obtained from the EUCAST database for ABR enrichment in environmental bacteria (EFABR) can (European Committee on Antimicrobial Susceptibility Test- be calculated as: ing 2020). In their study, MSCs were extrapolated from MICs by applying a flat extrapolation factor of 10. These extrapola- 0.5 EFABR = (2) tion factors are derived as the mean MIC/MSC ratio obtained HC50 in experimental studies (Gullberg et al. 2011; Liu et al. 2011), EFABR: ABR effect factor for a given environmental com- but refinements of this extrapolation factor should be imple- partment e.g. freshwater ecosystems (PAF m 3 kg−1). mented as soon as further experimental or modelling data HC50: geometric mean of MSCs for bacteria (kg m−3). become available. Such improvements should account for dif- In ecotoxicity assessments, the effect factor relates to the ferences between broad-spectrum and selective antibiotics, change in PAF as a result of increases in contaminant concentra- considering their mode of action in bacteria. tion. In the proposed ABR enrichment assessment, this will refer MSC data for each antibiotic are used to fit normal distribu- to the change in the fraction of bacterial populations that acquire tions to the log-transformed MSC data, similarly to the spe- a significant increase of resistance genes due to a unit increase of cies sensitivity distribution approach used in ecotoxicological AB exposure concentration. Analogous to the ecotoxicity assess- impact characterization (Posthuma et al. 2001). In an analogous ment, it is based on a linear extrapolation from the H C down to manner, the MSC distributions could be used to extrapolate the 50 HC0 (slope of 0.5), and assumes that the acquisition of resistance hazardous concentration that will promote the development of at the community level increases with AB concentration. ABR in 50% of bacteria ( HC50), which can be calculated as the Finally, the ABR characterization factor (PAF m3 day per geometric mean of the MSC data. A difference to ecotoxicity kg emitted) can be calculated as characterization methodology in USEtox is that HC50 values are generated from the geometric mean of chronic EC ( ) 50 or kg emitted L C50 values (effect concentration for 50% of tested organisms CFABR = Fate factor kg in compartment per day and lethal concentration of 50% of tested organisms respec- × Exposure factor(kg per kg ) tively) for aquatic organisms from several trophic levels (Fantke bioavailable in compartment 3 −1 et al. 2017). A chronic endpoint is preferred as the chemical fate × Effect factor (PAF m per kg ) and exposure calculations are performed following a steady- (3) state approach, so estimated environmental concentrations where the fate factor describes the distribution of chemicals resemble chronic exposure (Guinée and Heijungs 1993). The in the environment and the exposure factor describes the use of E C50 or L C50 values, as opposed to a no observed effect bioavailable fraction of chemicals that could cause harm to concentration (NOEC) or lowest observed effect concentration freshwater organisms (Fantke et al. 2017), thus utilizing the (LOEC), is mainly supported by the statistical robustness of established modelling framework USEtox for characterizing the 50% response level (Crane and Newman 2000; Larsen and distribution of ABs in the environment while only modifying Hauschild 2007). In the proposed approach, the MSCs approxi- the effect factor. mate the chronic resistance LOEC for bacteria, which are based Following the methodology described above, ABR-HC50 on thousands of data points compiled in the EUCAST database values were generated for 14 ABs to replace ecotoxicological rather than on a single dose-response experiment, so a sufficient H C50 for USEtox input data (Table S5). The USEtox calcula- statistical robustness is assumed for this value as representative tions were set up in the USEtox® 2.1 [built 19-Oct-2017] 1 3 The International Journal of Life Cycle Assessment interface software (available at https:// usetox.o rg/), selecting following a concentration gradient, which could theoretically freshwater emission ecotoxicity and applying default USEtox generate selection for resistance in bacteria at low concen- setting environment. ABR enrichment characterization factors trations, as well as remove resistance at high concentrations could subsequently be generated in harmony with USEtox (e.g. killing bacteria). A strength of this approach is that all 2.1 (Table 3). MIC data can be sourced from the EUCAST database, thus These results show that applying MSC-based H C50 values removing variations in H C50 values due to different data for ABs enables us to derive characterizations that fit the for- sourcing, as highlighted for ecotoxicological data above. merly established LCA framework, thereby complementing Finally, a point on the ecological relevance of this existing freshwater ecotoxicity impacts with ABR specific approach should be raised. The LCA framework accounts impacts. It should be clarified that the CFs generated by for emissions under steady-state conditions (Guinée and this model serve as comparative units of impact related to Heijungs 1993), while AB emissions can be influenced the resistance H C50 for bacteria at the community level, and by pharmacokinetics and environmental processes, and not as a representation of resistance development dynamics ABR development depend on the exposure level as well 1 3 The International Journal of Life Cycle Assessment ◂Fig. 2 Impact pathway overview of AB from an LCIA perspective. (1) Table 3 ABR enrichment CFs using Approach 1 and USEtox 2.1. AB use data is reported as DDDs per 1000 inhabitants per day (DIDs) Ecotoxicological HC50 values are replaced with HC50 values based in respective EU country in the ECDC database (accessed 2020-09- on geometric means of bacterial MSC distributions acquired from the 19). (2) AB use data within the EU region reported as tonnes per year EUCAST database (accessed 11 Nov 2020) in (European Medicines Agency 2017). (3) Assessed with current fate models in USEtox. (4) No correlation between veterinary use of 3GC Antibiotic ABR enrichment charaterization 3 −1 and resistance development in the human healthcare sector. (European factor [PAF m day k g ] Centre for Disease Prevention and Control (ECDC), European Food Safety Authority (EFSA), European Medicines Agency (EMA) 2017). Ampicillin 7.27E + 05 (5) Correlation between veterinary use of 3GC and resistance develop- Amoxicillin 1.89E + 06 ment in the food-animal sector. (European Centre for Disease Prevention Cephalexin 5.75E + 05 and Control (ECDC), European Food Safety Authority (EFSA), Euro- Ciprofloxacin 5.55E + 04 pean Medicines Agency (EMA) 2017). (6) Correlation between human consumption of 3GC and resistance development in human health care Colistin 1.63E + 06 sector. (European Centre for Disease Prevention and Control (ECDC), Doxycycline 3.49E + 05 European Food Safety Authority (EFSA), European Medicines Agency Florfenicol 1.27E + 06 (EMA) 2017). (7) Proposed characterization factor in this paper, based Kanamycin 4.43E + 05 on minimum inhibitory concentrations (MIC) of ABs; predictions of ABR development is used as a comparative endpoint. (8) Connecting Levofloxacin hydrate 2.39E + 07 the ABR present in the environmental compartment to AB treatment Rifampicin 1.16E + 08 failure in the human and veterinary sector; no quantitative data is availa- Sulfamethoxazole 1.00E + 05 ble for this pathway. (9) These steps are aggregated describing the effect Trimethoprim 9.87E + 05 to human health from resistance developed. (10) Cassini et  al. (2019) assessment of treatment failure attributed to ABR can be found for the Erythromycin 2.85E + 05 16 most common pathogen-resistance combinations in Europe. Loss Roxithromycin 3.65E + 05 of human lives and prolonged hospitalizations as an effect are assessed within the same report. (11) No quantitative data available on impacts to veterinary medicine from ABR zoonosis. (12) No quantitative data available. (13) Innes et  al. (2019) report an economic impact models being explored (Ashbolt et al. 2013; Ben et al. 2019), but from of enrofloxacin use and impacts from Campylobacter, Salmonella, is hampered by a lack of relevant data on how AB con- and E. coli bacteria as externalities from AB use, which reaches US$ centrations influence resistance development in human 2200 per kg enrofloxacin used. The model might accommodate calcula- pathogenic bacteria and quantitative pathways describing tions for 3GC in the EU as well. (14) A rational for establishing nutri- tional losses needs to be settled on. We would argue that there are highly environmental exposure to ABR. Hence, below we explore variable regional differences across the world. In high-income regions, an alternative approach relating the use of ABs to human nutrients are easily substituted from another food source in contrast to health impacts from ABR, which goes beyond the expla- low-income regions. However, antibiotic use strategies surely vary nation of mechanistic relationships and the quantitative between small-holder animal husbandry and industrial-scale farming in the latter regions (no use vs. some use respectively). (15) Ecotoxico- determination of each of these pathways. logical modelling according to USEtox, damage to the ecosystem from increased AB concentrations in the environment is characterized as 3.2.2 Approach 2: characterizing human health impacts freshwater ecotoxicity impacts. (16) Weak evidence for damage to eco- for ABR (endpoint) system from ABR (Eckert et al. 2019). (17) End-point measurement of ecotoxicological effects, no clear ecosystem effect from ABR Our alternative approach for establishing linear dose- response relationships between AB use and ABR conse- quences for human well-being is conceptualized in Fig. 2. Ideally, one would quantify each pathway individually, but as on the exposure duration (Ashbolt et al. 2013). This data scarcity currently forces us to exercise a generalized implies that the applicability of this model serves better mass balance approach where only pathways 1, 2, 4, 6, 9, for systems with continuous emissions than systems with and 10 are aggregated into one pathway (Fig. 2). Our pro- erratic AB use (e.g., WWTPs vs. aquaculture farms), but posed linear dose-response model assumes that any use of may be under-representing exposure scenarios that are ABs will contribute to resistance development, which allows prolonged in long periods. Additionally, events related us to circumvent the shortcomings in data connecting envi- to proliferation of resistance genes following ABR devel- ronmental ABR to human health impacts. This implies a opment, across species and exposure to humans, are not loss of ecological relevance, but enables quantification captured within this model. Such quantifications would be of potential impacts from AB use at regional scales. Data highly dependent on the exposure level and duration, and sourced to support the approach are presented in supplemen- the bacteria present in the environment, which would need tary information (Tables S2, S3, and S4 in supplementary to account for more complex pathways. Connecting ABR information). enrichment in the environment to human health impacts The relationship between AB use and ABR develop- (8 in Fig. 2) by incorporating other methods is currently ment is inferred by the Joint Interagency Antimicrobial 1 3 The International Journal of Life Cycle Assessment Consumption and Resistance Analysis (JIACRA) report, ∂x,p,sector,reg = correlation coefficient explaining the rela- which presents logistic regression models correlating use tionship between use in the investigated sector (human or of 21 different ABs from five classes in human and food- veterinary) and resistance development in pathogen p to animal production in the EU and resistance development antibiotic x in the investigated region. in five human pathogens (Salmonella spp., Campylobacter ORx,sector,reg = odds ratio implying the strength of associa- coli, C. jejuni, Escherichia coli, and Enterococci) (Euro- tion between use of AB x and resistance development for pean Centre for Disease Prevention and Control (ECDC), each pathogen p in the investigated sector and region. European Food Safety Authority (EFSA), and European Since odds ratio values range between 0 to infinity, we Medicines Agency (EMA) 2017). This report calculates express this coefficient as the square root of ln OR to not odds ratios based on logistic regression analysis of AB use suggest an overrepresentation of the odds for resistance data and resistance development data to suggest statisti- development in a particular sector. This correlation coef- cally correlated associations between human and animal ficient is subsequently used to imply an effect from AB use consumption of antibiotics, and resistance development on resistance development, expressed as in selected human pathogens. Odds ratios are explained by Szumilas (2010, p227) as the representation of “the ABFx,p,reg = ABUx,hum,reg × x,p,hum,reg + ABUx,vet,reg × x,p,vet,reg chance that an outcome will occur given a particular expo- (5) sure compared to the odds of the outcome occurring in the where absence of that exposure”. ABFx,p,reg = total resistance developed in pathogen p from In LCA, the endpoint of toxicological impact to humans use of antibiotic x in the investigated region (resistance per is expressed as disability adjusted life years (DALY), a met- kg year−1). ric that accounts for the years lost due to premature mor- ABUx,hum,reg = total use of antibiotic x in human health tality and productive life due to disability, and is globally sector in the investigated region (kg year−1). scalable (Murray and Lopez 1994). We therefore argue that ∂x,p,hum,reg = correlation coefficient explaining relationship the common unit for ABR impacts to human health would between human use and resistance development in pathogen be the same, given that it is a well-established concept in p to antibiotic x in the investigated region. LCA’s cause-effect endpoint pathways. Data available on ABUx,vet,reg = total use of antibiotic x in veterinary sector the impacts to human health have been published by Cassini in the investigated region (kg year−1). et al. (2019), who produced an extensive report on the human ∂x,vet,region = correlation coefficient explaining relationship health impacts from AB-resistant pathogens within the EU between veterinary sector use and resistance development in health care system, attributing DALYs to 16 pathogen-ABR pathogen p to antibiotic x in the investigated region. combinations. Next, the pathways between ABF to ABR related impacts We subsequently use the impact pathway between AB use need to be defined (9 in Fig. 2). This step needs to be aggre- and DALYs to account for AB use in human and veterinary gated with the subsequent effect; loss of lives, and prolonged medicine respectively (1 and 2 in Fig. 2) in the EU, by estab- hospitalization (10 in Fig. 2), here expressed as lishing a causal relationship between AB use and ABR devel- DALYx,p,reg = disability adjusted life years attributed to opment (4 and 6 in Fig. 2) together with the impacts caused resistance to AB x in pathogen p per region and year. by the subsequent failure to treat infections due to ABR (9 This gives us a mass balance approach attributing DALYs and 10 in Fig. 2). We argue that the use and effect pathways per kg AB used in the investigated region expressed as should be parameterized according to the region of interest DALY DALY (country or continental scale), since availability and enforce- x,p1,reg x,p2,reg DALY per kg ABx = + ment of local and regional policy will arguably shape the use ABFx,p ,reg ABF 1 x,p2,reg (Eq.6) and misuse patterns of ABs at each level (Laxminarayan and DALYx,pn,reg Malani 2007; Søgaard Jørgensen et al. 2020). +⋯ + ABFx,pn,reg To build a model for expressing DALYs as a product of AB use, we rely on published odds ratio data from the JIA- To demonstrate our proposed cause-effect pathway, we CRA report (given that there is a significant correlation (e.g. allocate DALYs to the use of 3rd-generation cephalosporin CI does not cross 1 and p < 0.05)) to express a correlation (3GC) in EU, a group of beta-lactam antibiotics to which coefficient between the use and resistance development for 18 ABs belong. This class of ABs is classified as critically the investigated AB as important for human medicine (World Health Organiza- tion 2019), yet is still used within food-animal production √ lnOR = x,p,sector,reg x,p,sector,reg (4) to some extent (European Medicines Agency 2019). 3GC resistance is also among the top contributors to pathogen- where ABR related mortalities in the EU (Cassini et al. 2019). 1 3 The International Journal of Life Cycle Assessment Following the pathways of Fig. 2, we initially establish from 8.62*10−12 DALY per kg −8 emitted to 1.98*10 DALY quantities of 3GC used in the EU for 2015. According to the per kgemitted depending on emission compartment selected). ECDC Database (accessed 2020-09-17), close to 270 tonnes This example shows that causal relationships are pos- of 3GC were used for human treatment (both community sible to infer between AB use and DALYs, but overlooks and hospital use) in 2015 (Table S2), while the European several important aspects and suffers from data scarcity. Medicines Agency (2017) reports 13.9 tonnes 3GC used in Nonetheless, we manage to establish a C FABR for 3GC that animal husbandry that same year, totalling about 284 tonnes associates seven orders of magnitude higher DALYs per annually. kg of AB compared to a CF for a related AB which only Data are available on odds ratios from the JIACRA report considers direct toxicity impacts on humans. It should be (European Centre for Disease Prevention and Control (ECDC), noted, that 3GC is the only AB with enough data available European Food Safety Authority (EFSA), European Medicines to describe this pathway in a European setting currently. Agency (EMA) 2017) for human consumption of 3rd- and For instance, Cassini et al. (2019) do attribute DALYs to 4th-generation cephalosporin antibiotics (3GC) and resistance another pathogen resistant to 3GC, but there is no avail- development in human pathogen E. coli. Note that use of 3rd- able odds ratio for this combination in the JIACRA report generation and 4th-generation cephalosporins is reported as a (European Centre for Disease Prevention and Control sum, but will be addressed as 3GC in this example for simplic- (ECDC), European Food Safety Authority (EFSA), Euro- ity. The reported odds ratio for this specific AB-pathogen com- pean Medicines Agency (EMA) 2017) and characterizing bination is 1.94 (CI 1.47–2.54, p-value < 0.001), which implies additional ABs using this linear dose-response concept that odds are 94% for resistance to develop in E. coli for each will require substantial data collection at regional lev- increased defined daily dose of 3GC in the human sector. For els. However, as data are continuously being generated the veterinary sector, however, the report shows no statistically and reported, we would expect that the coverage of this significant correlation between animal consumption of 3GC approach could be expanded in the future (Limmathurot- and resistance development in the human pathogen E. coli sakul et al. 2019). (odds ratio 4.13 CI 0.78–21.08, p-value < 0.094), and ∂x,vet,region Our endpoint approach is designed to characterize will hence be accounted for as “0”. AB use in both human health and veterinary sectors, to Cassini et al. (2019) report that the most critical pathogen account for total use of antibiotics within a region follow- resistance related–infections to human health is 3GC resistant ing the One Health concept, acknowledging that ABs can Escherichia coli. This pathogen caused a median number of promote ABR regardless of sector. Ideally, refining Eq. 5 37.2 DALYs per 100,000 population reported in 2015; 191 to include emissions of ABs at a production stage would 883 DALYs across Europe this year given a population of improve the model even further, but data on the amounts 515.8 million. We input data into our model Eq. 7 according of ABs emitted during production are largely nonexistent, to Eqs. 4, 5, and 6: while they are reported to be substantial in some areas DALYx,p,reg DALY per kg ABxemitted = √ √ (7) ABU OR OR x,hum,reg × ln x,p,hum,reg + ABU × ln x,p,vet,reg x,vet,reg which yields the following for our pathogen-resistance com- (Larsson 2014; el Balkiny 2014). Despite that this dose- bination E. coli infections resistant to 3GC example (Eq. 8): response concept is an oversimplification of the complex 191883 DALY 0.87 DALYs per kg 3GCemitted = (8) 2.7 105kg 0.81406 1.39 104∙ × + ∙ kg × 0 To put this in perspective, we compare our 0.873 DALYs cause-effect pathways connecting AB use, resistance kg−1 AB with the single available characterization of a development, dissemination, and human health impacts, cephalosporin-class AB, Cefaclor, a 2nd-generation ceph- we still had to aggregate stages in the pathway. Also, alosporin, characterized by Ortiz de García et al. (2017) considering the limited data availability on ABR-related using USEtox methodology (i.e. non-homologous method- human impacts, only 16 pathogen-resistance combinations ology as to the case above). Ortiz de García et al. (2017) are available from Cassini et al. (2019), which limits the report that the median attributed DALYs are 5.48*10−8 applicability of this approach. ABR frequencies and dam- DALY per kgemitted as non-carcinogenic impact (ranging age to human health caused by economic and nutritional 1 3 The International Journal of Life Cycle Assessment losses from livestock mortalities are neither considered in existing impact assessment models to generate characteri- our example (11–14 in Fig. 2), though Innes et al. (2019) zation factors for ABR enrichment in the environment at have proposed a model to account for economic losses in a mid-point level, and a correlation between AB use and a US setting which could possibly be included at further DALYs for endpoint impacts. The mid-point characteri- development of this model. Additional economic costs zation approach for ABR enrichment in the environment from prolonged hospitalizations due to resistant infec- provides a robust comparative model for assessing AB use tions is neither included in our model, which is focusing or removal from wastewater, agriculture, or industrial pro- on health impacts, but a similar approach could possibly cesses. Further development of this methodology would be included using life cycle costing (Estevan and Schaefer benefit from refining the current MIC to MSC extrapola- 2017). Moreover, we limit ourselves to a European setting tions since a fair amount of AB MIC data are available. For where AB use could be expected to be fairly well regu- the endpoint approach, we use 3GC as an example to prove lated, including proper reporting on and administration of the concept of our theory, but few causal relationships and drugs, destruction of excess drugs, full treatment cycles of data limitations challenge the practical usefulness of this patients and animals, and possible preliminary screenings approach at present. We would therefore recommend cau- for ABR genes. The current practical applicability to LCA tion when interpreting human health impacts from ABs in of this suggested approach could be questioned since there LCA studies until more holistic methodologies and bet- is no ability to compartmentalize emissions and suggest ter data become available. Future LCAs including ABs impacts related to various environments. The assumptions should ideally adopt a One Health approach and could made for the linear dose-response concept completely dis- benefit from complementary environmental risk assess- regard the beneficial aspects of AB use since many lives ments, allowing for the dynamics of AB use and emission are saved each year by these pharmaceuticals and modern in all relevant sectors to be accounted for. medicine relies upon functional antibiotics, but the reason- ing of human health benefits holds true for the many other Supplementary information The online version contains supplemen- chemicals as well, and could be discussed with a broader tary material available at https://d oi.o rg/1 0.1 007/s 11367-0 21-0 1908-y. audience. Conclusively, applying this type of simplified Acknowledgements This work was undertaken as part of the CGIAR linear dose-response concept for AB use while circum- Research Programs on Fish Agri-Food Systems (FISH) led by World- venting stochastic dynamics of ABR development and the Fish and on Climate Change, Agriculture and Food Security (CCAFS). contribution of resistance from the environmental com- These programs are supported by contributors to the CGIAR Trust partment will imply statistical inference without causality. Fund. Since there is no available methodology to extrapolate how Funding P. J. G. Henriksson is partially funded by FORMAS Sea- environmental ABR impacts human health in an LCA con- Win project (2016-00227) and partially funded by FORMAS Inequal- text, we had to create two separate pathways, one prospective ity and the Biosphere project (2020-00454). A. Rico is supported by method which looks at the onset of ABR in environmental bac- a Ramón y Cajal grant provided by the Spanish Ministry of Science teria using the established characterization model USEtox, and and Innovation (RYC2019-028132-I). Open access funding provided by Stockholm University. a second retrospective method based on statistical correlations between AB use and human health impacts. Since these mod- Open Access This article is licensed under a Creative Commons Attri- els are based on different assumptions with little commonality, bution 4.0 International License, which permits use, sharing, adapta- they express different strengths and shortcomings that need tion, distribution and reproduction in any medium or format, as long to be considered before implementation. We view these two as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes novel AB characterization methodologies as steppingstones to were made. The images or other third party material in this article are further refinements and discussions on the holistic assessment included in the article’s Creative Commons licence, unless indicated of ABs and resistance development in LCA. otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will 4 C onclusions need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://c reati vecom mons.o rg/l icens es/b y/4.0 /. ABs have been assumed to contribute to ecotoxic- ity and human toxicity in seventeen LCAs, but the lat- ter impact category is sparsely characterized. Moreover, ABR impacts as a consequence of AB use have not yet References been accounted for in LCA, which suggests that the full impacts of AB use in LCAs are severely underestimated. Alfonsín C, Hospido A, Omil F, Moreira MT, Feijoo G (2014) PPCPs in wastewater - update and calculation of characterization factors In response, we present two approaches that acknowl- for their inclusion in LCA studies. J Clean Prod 83:245–255. edge these impacts and that can be readily included in https://d oi.o rg/1 0.1 016/j. jclep ro.2 014.0 7. 024 1 3 The International Journal of Life Cycle Assessment Amos GC, Gozzard E, Carter CE, Mead A, Bowes MJ, Hawkey PM, el Balkiny H (2014) Determination of veterinary pharmaceuticals in Zhang L, Singer AC, Gaze WH, Wellington EMH (2015) Vali- production wastewater by TLC-densitometry. Anal Chem Lett dated predictive modelling of the environmental resistome. ISME 4:319–328. https:// doi.o rg/1 0. 1080/2 22979 28. 2014. 925823 J 9:1467–1476. https:// doi. org/ 10.1 038/i smej. 2014. 237 Emara Y, Lehmann A, Siegert MW, Finkbeiner M (2018) Modeling Ashbolt NJ, Amézquita A, Backhaus T, Borriello P, Brandt KK, Collignon pharmaceutical emissions and their toxicity-related effects in P, Coors A, Finley R, Gaze WH, Heberer T, Lawrence JR, Larsson life cycle assessment (LCA): a review. Integr Environ Assess DGJ, McEwen SA, Ryan JJ, Schönfeld J, Silley P, Snape JR, Van Manag 15:6–18. https:// doi.o rg/1 0.1 002/i eam. 4100 den Eede C, Topp E (2013) Human health risk assessment (HHRA) Estevan H, Schaefer B (2017) Life Cycle Costing - state of the art for environmental development and transfer of antibiotic resistance. report. ICLEI – Local Gov Sustain Eur Secr 50. Available Environ Health Perspect 121:993–1001. https://d oi.o rg/1 0.1 289/e hp. at https:// sppre gions. eu/ filea dmin/ user_ upload/ Life_ Cycle_ 120631 6 Costin g_S oA_R eport.p df Barceló D, Petrovic M, eds. (2011) The Ebro river basin. Springer European Centre for Disease Prevention and Control (ECDC), Euro- Science & Business Media. pean Food Safety Authority (EFSA), European Medicines Ben Y, Fu C, Hu M, Liu L, Wong MH, Zheng C (2019) Human health Agency (EMA) (2017) ECDC/EFSA/EMA second joint report risk assessment of antibiotic resistance associated with antibiotic on the integrated analysis of the consumption of antimicrobial residues in the environment: A review. Environ Res 169:483– agents and occurrence of antimicrobial resistance in bacteria 493. https:// doi.o rg/ 10.1 016/j. envres. 2018. 11.0 40 from humans and food-producing animals – Joint Interagency Bengtsson-Palme J, Larsson DGJ (2016) Concentrations of antibiot- Antimicrobial Consumption and Resistance Analysis (JIA- ics predicted to select for resistant bacteria: proposed limits for CRA) Report. EFSA Journal 2017;15(7):4872, 135 pp. https:// environmental regulation. Environ Int 86:140–149. https:// doi. doi. org/ 10.2 903/j.e fsa.2 017.4 872 org/1 0.1 016/j. envint. 2015. 10. 015 European Commission (2010) Joint Research Centre - Institute for Cabello FC (2006) Heavy use of prophylactic antibiotics in aquacul- Environment and Sustainability: International Reference Life ture: a growing problem for human and animal health and for the Cycle Data System (ILCD) Handbook - review schemes for environment. Environ Microbiol 8:1137–1144. https://d oi.o rg/1 0. Life Cycle Assessment. EUR 24710 EN. Luxembourg. Pub- 1111/j.1 462-2 920.2 006.0 1054.x lications Office of the European Union; 2010. https:// doi.o rg/ Cabello FC, Godfrey HP, Buschmann AH, Dölz HJ (2016) Aquacul- 10. 2788/ 39791 ture as yet another environmental gateway to the development European Medicines Agency (2017) Sales of veterinary antimicro- and globalisation of antimicrobial resistance. Lancet Infect Dis bial agents in 30 European countries in 2015: trends from 2010 16:e127–e133. https://d oi.o rg/1 0.1 016/S 1473-3 099(16)0 0100-6 to 2015. Seventh ESVAC report. EMA/184855/2017 Carlsson G, Örn S, Larsson DGJ (2009) Effluent from bulk drug pro- European Medicines Agency (2019) Sales of veterinary antimicro- duction is toxic to aquatic vertebrates. Environ Toxicol Chem bial agents in 31 European countries in 2017: trends from 2010 28:2656. https://d oi.o rg/1 0. 1897/0 8- 524.1 to 2017. Ninth ESVAC Report. EMA/294674/2019 Cassini A, Högberg LD, Plachouras D et al (2019) Attributable deaths European Committee on Antimicrobial Susceptibility Testing (2020) and disability-adjusted life-years caused by infections with anti- Data from the EUCAST MIC distribution website. http://w ww. biotic-resistant bacteria in the EU and the European Economic eucast.o rg. Accessed 11 Nov 2020 Area in 2015: a population-level modelling analysis. Lancet Infect Fantke P (ed.), Bijster M, Guignard C, Hauschild M, Huijbregts M, Dis 19:56–66. https://d oi.o rg/ 10. 1016/ S1473- 3099(18) 30605-4 Jolliet O, Kounina A, Magaud V, Margni M, McKone T, Post- Chuah LO, Effarizah ME, Goni AM, Rusul G (2016) Antibiotic Appli- huma L, Rosenbaum RK, van de Meent D, van Zelm R (2017) cation and Emergence of Multiple Antibiotic Resistance (MAR) USEtox® 2.0 Documentation (Version 1). http:// usetox. org in Global Catfish Aquaculture. Curr Environ Heal reports 3:118– Fair RJ, Tor Y (2014) Antibiotics and bacterial resistance in the 127. https:// doi. org/ 10. 1007/ s40572-0 16- 0091-2 21st century. Perspect Medicin Chem 25–64. https:// doi. org/ Costanzo SD, Murby J, Bates J (2005) Ecosystem response to antibiot- 10.4 137%2 FPMC. S14459 ics entering the aquatic environment. Mar Pollut Bull 51:218– Fent, K, Weston AA, Caminada D (2006) Ecotoxicology of human 223. https:// doi.o rg/1 0. 1016/j. marpo lbul.2 004. 10. 038 pharmaceuticals. Aquat. Toxicol. 76: 122-159. https://d oi.o rg/ Crane M, Newman MC (2000) What level of effect is a no observed 10. 1016/j. aquato x. 2005. 09. 009 effect? Environ Toxicol Chem 19:516–519. https:// doi. org/ 10. Finley RL, Collignon P, Larsson DGJ, Mcewen SA, Li XZ, Gaze 1002/ etc. 56201 90234 WH, Reid-Smith R, Timinouni M, Graham DW, Topp E (2013) D’Costa VM, King CE, Kalan L, Morar M, Sung WWL, Schwarz C, The scourge of antibiotic resistance: the important role of the Froese D, Zazula G, Calmels F, Debruyne R, Golding GB, Poinar environment. Clin Infect Dis 57:704–710. https:// doi. org/ 10. HN, Wright GD (2011) Antibiotic resistance is ancient. Nature 1093/ cid/ cit355 477:457–461. https:// doi. org/ 10.1 038/ natur e10388 Goedkoop M, Heijungs R, Huijbregts M, De Schryver A, Struijs JV, D’Costa VM, McGrann KM, Hughes DW, Wright GD (2006) Sampling Van Zelm R (2013) A life cycle impact assessment method the antibiotic resistome. Science 311:374–377. https://d oi.o rg/1 0. which comprises harmonised category indicators at the mid- 1126/ scienc e. 11208 00 point and the endpoint level. First edition (version 1.08). The Davies J, Davies D (2010) Origins and evolution of antibiotic resist- Hague, Ministry of VROM. ance. Microbiol Mol Biol Rev 74:417–433. https:// doi. org/ 10. Goedkoop M, Spriensma R (2001) The Eco-indicator 99 - a damage 1128/ MMBR. 00016- 10 oriented method for life cycle impact assessment. PRé Consult Dobbins LL, Usenko S, Brain RA, Brooks BW (2009) Probabilistic b.v, Amersfoort, Netherlands ecological hazard assessment of parabens using Daphnia magna Grace D (2015) Review of evidence on antimicrobial resistance and and Pimephales promelas. Environ Toxicol Chem 28:2744–2753. animal agriculture in developing countries. Evidence on Dde- https:// doi. org/1 0. 1897/ 08- 523.1 mand, UK. https:// doi. org/ 10. 12774/ eod_c r. june20 15. graced Eckert EM, Quero GM, Di Cesare A, Manfredini G, Mapelli F, Borin Grenni P, Ancona V, Caracciolo AB (2018) Ecological effects of anti- S, Fontaneto D, Luna GM, Corno G (2019) Antibiotic distur- biotics on natural ecosystems: a review. Microchem J 136:25–39. bance affects aquatic microbial community composition and https:// doi. org/ 10. 1016/j.m icroc. 2017. 02. 006 food web interactions but not community resilience. Mol Ecol Guinée J, Heijungs R (1993) A proposal for the classification of toxic 28:1170–1182. https:// doi. org/ 10. 1111/m ec.1 5033 substances within the framework of life cycle assessment of 1 3 The International Journal of Life Cycle Assessment products. Chemosphere 26:1925–1944. https://d oi.o rg/1 0.1 016/ Jutkina J, Marathe NP, Flach CF, Larsson DGJ (2018) Antibiotics 0045- 6535(93) 90086-K and common antibacterial biocides stimulate horizontal trans- Guinée JB, Gorrée M, Heijungs R, Huppes G, Kleijn R, de Koning A, fer of resistance at low concentrations. Sci Total Environ 616– van Oers L, Sleeswijk AW, Sangwon S, Udo de Haes HA, de 617:172–178. https:// doi.o rg/ 10. 1016/j.s citot env. 2017. 10. 312 Bruijn JA, van Duin R, Huijbregts MA (2002) Handbook on life Klein EY, Van Boeckel TP, Martinez EM, Pant S, Gandra S, Levin cycle assessment: operational guide to the ISO standards. Kluwer SA, Goossens H, Laxminarayan R (2018) Global increase and Academic Publishers, Dordrecht geographic convergence in antibiotic consumption between Gullberg E, Cao S, Berg OG, Ilbäck C, Sandegren L, Hughes D, 2000 and 2015. Proc Natl Acad Sci 201717295. https:// doi. Andersson DI (2011) Selection of resistant bacteria at very low org/ 10. 1073/ pnas.1 7172 95115 antibiotic concentrations. PLoS Pathog 7:1–9. https://d oi.o rg/1 0. Kümmerer K (2009) Antibiotics in the aquatic environment - a 1371/j ourn al. ppat. 100215 8 review - part I. Chemosphere 75:417–434. https://d oi. org/ 10. Hauschild MZ, Huijbregts MAJ (2015) Life cycle impact assessment 1016/j. chemo sphere.2 008. 11.0 86 - the complete world of life cycle assessment. Springer, Nether- Lange K, Buerger M, Stallmach A, Bruns T (2016) Effects of antibiot- lands, Dordrecht ics on gut microbiota. Dig Dis 34:260–268. https:// doi. org/ 10. Heinemann JA (1999) How antibiotics cause antibiotic resistance. Drug 1159/0 0044 3360 Discov Today 4:72–79. https://d oi.o rg/1 0.1 016/S 1359-6 446(98) Larsen HF, Hauschild M (2007) GM-troph: a low data demand eco- 01294-X toxicity effect indicator for use in LCIA. Int J Life Cycle Assess Henriksson PJG, Rico A, Zhang W, Ahmad-Al-Nahid S, Newton R, 12:79–91. https:// doi. org/ 10. 1065/l ca20 06. 12. 288 Phan LT, Zhang Z, Jaithiang J, Dao HM, Phu TM, Little DC, Larsson DGJ (2014) Antibiotics in the environment. Ups J Med Sci Murray FJ, Satapornvanit K, Liu L, Liu Q, Haque MM, Kruijs- 119:108–112. https:// doi. org/ 10.3 109/ 03009 734. 2014. 896438 sen F, De Snoo GR, Heijungs R, Van Bodegom PM, Guinée JB Larsson DGJ, Andremont A, Bengtsson-Palme J, Brandt KK, de Roda (2015) Comparison of Asian aquaculture products by use of Husman AM, Fagerstedt P, Fick J, Flach CF, Gaze WH, Kuroda statistically supported life cycle assessment. Environ Sci Tech- M, Kvint K, Laxminarayan R, Manaia CM, Nielsen KM, Plant nol 49:14176–14183. https:// doi. org/ 10. 1021/ acs.e st. 5b046 34 L, Ploy MC, Segovia C, Simonet P, Smalla K, Snape J, Topp Hospido A, Carballa M, Moreira M, Omil F, Lema JM, Feijoo G (2010) E, van Hengel AJ, Verner-Jeffreys DW, Virta MPJ, Welling- Environmental assessment of anaerobically digested sludge reuse in ton EM, Wernersson AS (2018) Critical knowledge gaps and agriculture : potential impacts of emerging micropollutants. Water research needs related to the environmental dimensions of anti- Res 44:3225–3233. https:// doi.o rg/1 0. 1016/j.w atres. 2010. 03. 004 biotic resistance. Environ Int 117:132–138. https:// doi. org/ 10. Huijbregts MA, Thissen U, Guinée JB, Jager T, Kalf D, Van de Meent 1016/j. envint. 2018. 04. 041 D, Ragas AM, Sleeswijk AW, Reijnders L (2000) Priority assess- Laxminarayan R, Malani A (2007) Extending the cure: policy responses ment of toxic substances in life cycle assessment. Part I: calcu- to the growing threat of antibiotic resistance. Earthscan. lation of toxicity potentials for 181 substances with the nested Li Y, Zhang S, Zhang W, Xiong W, Ye Q, Hou X, Wang C, Wang P multi-media fate, exposure and effects model USES–LCA. (2019) Life cycle assessment of advanced wastewater treatment Chemosphere 1;41(4):541–73. https:// doi. org/ 10. 1016/ S0045- processes: Involving 126 pharmaceuticals and personal care 6535(00)0 0030-8 products in life cycle inventory. J Environ Manage 238:442–450. Iannacone J, Alvariño L (2009) Aquatic risk assessment of seven phar- https:// doi.o rg/ 10. 1016/j. jenvm an. 2019.0 1. 118 maceutical products on Daphia magna. Ecol Apl 8:71. https:// Limmathurotsakul D, Sandoe JAT, Barrett DC, Corley M, Hsu LY, doi. org/1 0.2 1704/ rea. v8i1-2. 384 Mendelson M, Collignon P, Laxminarayan R, Peacock SJ, How- Igos E, Benetto E, Venditti S, Köhler C, Cornelissen A (2013) Com- ard P (2019) ‘Antibiotic footprint’ as a communication tool to aid parative and integrative environmental assessment of advanced reduction of antibiotic consumption. J Antimicrob Chemother wastewater treatment processes based on an average removal of 1–6. https:// doi. org/ 10. 1093/j ac/ dkz185 pharmaceuticals Elorri Igos, Enrico Benetto, Silvia Venditti, Liu A, Fong A, Becket E, Yuan J, Tamae C, Medrano L, Maiz M, Christian Köhler. Water Sci Technol 67:387–394. https:// doi. Wahba C, Lee C, Lee K, Tran KP, Yang H, Hoffman RM, Salih org/1 0. 2166/ wst. 2012. 581 A, Miller JH (2011) Selective advantage of resistant strains at Igos E, Benetto E, Venditti S, Kohler C, Cornelissen A, Moeller R, trace levels of antibiotics: a simple and ultrasensitive color test Biwer A (2012) Is it better to remove pharmaceuticals in decen- for detection of antibiotics and genotoxic agents. Antimicrob tralized or conventional wastewater treatment plants? A life cycle Agents Chemother 55:1204–1210. https://d oi.o rg/1 0.1 128/A AC. assessment comparison. Sci Total Environ 438:533–540. https:// 01182- 10 doi.o rg/ 10.1 016/j.s cito tenv.2 012.0 8.0 96 Lorenzo-Toja Y, Alfonsín C, Amores MJ, Aldea X, Marin D, Moreira Innes GK, Randad PR, Korinek A, Davis MF, Price LB, So AD, MT, Feijoo G (2016) Beyond the conventional life cycle inven- Heaney CD (2019) External societal costs of antimicrobial resist- tory in wastewater treatment plants. Sci Total Environ 553:71–82. ance in humans attributable to antimicrobial use in livestock. https:// doi. org/ 10. 1016/j. scito tenv. 2016. 02.0 73 Annu Rev Public Health 41:141–157. https:// doi. org/ 10. 1146/ Mathew AG, Cissell R, Liamthong S (2007) Antibiotic Resistance in annur ev- publh ealth-0 40218- 043954 Bacteria Associated with Food Animals: A United States Per- Isidori M, Lavorgna M, Nardelli A, Parrella A, Previtera L, Rubino M spective of Livestock Production. Foodborne Pathog Dis 4:115– (2005) Ecotoxicity of naproxen and its phototransformation prod- 133. https:// doi. org/ 10. 1089/ fpd.2 006.0 066 ucts. Sci Total Environ 348:93–101. https:// doi. org/ 10. 1016/j. Morais SA, Delerue-Matos C, Gabarrell X (2013) Accounting for the scitot env. 2004.1 2. 068 dissociating properties of organic chemicals in LCIA: an uncer- Jernberg C, Löfmark S, Edlund C, Jansson JK (2010) Long-term tainty analysis applied to micropollutants in the assessment of impacts of antibiotic exposure on the human intestinal micro- freshwater ecotoxicity. J Hazard Mater 248–249:461–468. https:// biota. Microbiology 156:3216–3223. https:// doi. org/ 10. 1099/ doi. org/1 0.1 016/j. jhazm at.2 013. 01.0 02 mic.0.0 40618-0 Muñoz I, José Gómez M, Molina-Díaz A, Huijbregts MAJ, Fernández-Alba Jolliet O, Margni M, Charles R, Humbert S, Payet J, Rebitzer G, AR, García-Calvo E (2008) Ranking potential impacts of priority and Rosenbaum R (2003) IMPACT 2002+: a new life cycle impact emerging pollutants in urban wastewater through life cycle impact assessment methodology. Int J Life Cycle Assess 8:324–330. assessment. Chemosphere 74:37–44. https:// doi. org/ 10. 1016/j. https://d oi.o rg/1 0. 1007/ BF0297 8505 c hemo sphere.2 008. 09. 029 1 3 The International Journal of Life Cycle Assessment Muñoz I, Rodríguez A, Rosal R, Fernández-Alba AR (2009) Life cycle issue. Trans R Soc Trop Med Hyg 110:377–380. https:// doi. org/ assessment of urban wastewater reuse with ozonation as tertiary 10.1 093/t rstmh/ trw048 treatment. A focus on toxicity-related impacts. Sci Total Environ Rosenbaum RK, Bachmann TM, Gold LS, Huijbregts MAJ, Jolliet O, 407:1245–1256. https://d oi.o rg/1 0.1 016/j.s citot env.2 008.0 9.0 29 Juraske R, Koehler A, Larsen HF, MacLeod M, Margni M, McK- Murray CJL, Lopez AD (1994) Quantifying disability: data, methods one TE, Payet J, Schuhmacher M, Van De Meent D, Hauschild and results. Bull World Health Organ 72:481–494 MZ (2008) USEtox - the UNEP-SETAC toxicity model: recom- Ortiz de García S, García-Encina PA, Irusta-Mata R (2017) The poten- mended characterisation factors for human toxicity and freshwa- tial ecotoxicological impact of pharmaceutical and personal ter ecotoxicity in life cycle impact assessment. Int J Life Cycle care products on humans and freshwater, based on USEtoxTM Assess 13:532–546. https://d oi.o rg/1 0.1 007/s 11367-0 08-0 038-4 characterization factors. A Spanish case study of toxicity impact Rosenbaum RK, Huijbregts MAJ, Henderson AD, Margni M, McKone scores. Sci Total Environ 609:429–445. https://d oi.o rg/1 0.1 016/j. TE, Van De Meent D, Hauschild MZ, Shaked S, Li DS, Gold LS, scitot env. 2017.0 7.1 48 Jolliet O (2011) USEtox human exposure and toxicity factors for Ortiz de García SA, Pinto Pinto G, García-Encina PA, Irusta-Mata R comparative assessment of toxic emissions in life cycle analysis: (2014) Ecotoxicity and environmental risk assessment of phar- sensitivity to key chemical properties. Int J Life Cycle Assess maceuticals and personal care products in aquatic environments 16:710–727. https://d oi. org/ 10. 1007/s 11367- 011-0 316-4 and wastewater treatment plants. Ecotoxicology 23:1517–1533. Santos LHMLM, Araújo AN, Fachini A, Pena A, Delerue-Matos C, https:// doi. org/ 10. 1007/s 10646-0 14- 1293-8 M.C.B.S.M. Montenegro (2010) Ecotoxicological aspects related Osman KM, Kappell AD, Elhadidy M, Elmougy F, El-Ghany WAA, to the presence of pharmaceuticals in the aquatic environment. Orabi A, Mubarak AS, Dawoud TM, Hemeg HA, Moussa IMI, J Hazard Mater 175:45–95. https:// doi. org/ 10. 1016/j. jhazm at. Hessain AM, Yousef HMY (2018) Poultry hatcheries as potential 2009. 10.1 00 reservoirs for antimicrobial-resistant Escherichia coli: A risk to Sapkota A, Sapkota AR, Kucharski M, Burke J, McKenzie S, Walker public health and food safety. Sci Rep 8:1–14. https://d oi.o rg/ 10. P, Lawrence R (2008) Aquaculture practices and potential human 1038/ s41598- 018-2 3962-7 health risks: current knowledge and future priorities. Environ Pérez-Cobas AE, Artacho A, Knecht H, Ferrús ML, Friedrichs A, Ott Int 34:1215–1226. https://d oi. org/ 10. 1016/j. envint. 2008.0 4.0 09 SJ, Moya A, Latorre A, Gosalbes MJ (2013) Differential effects Søgaard Jørgensen P, Folke C, Henriksson PJG, Malmros K, Troell M, of antibiotic therapy on the structure and function of human gut Zorzet A (2020) Coevolutionary governance of antibiotic and microbiota. PLoS One 8. https:// doi. org/ 10. 1371/ journ al. pone. pesticide resistance. Trends Ecol Evol 35:484–494. https:// doi. 00802 01 org/1 0.1 016/j. tree.2 020. 01. 011 Posthuma L, Suter II GW, Traas TP (2001) Species sensitivity dis- Stockwell VO, Duffy B (2012) Use of antibiotics in plant agriculture. tributions in ecotoxicology. CRC Press. https://d oi.o rg/1 0.1 201/ Rev sci tech Off int Epiz 31:199–210. https://d oi. org/ 10. 20506/ 978142 00323 14 rst.3 1.1. 2104 Potting J, Hauschild MZ (2006) Spatial differentiation in life cycle Stone JJ, Aurand KR, Dollarhide CR, Jinka R, Thaler RC, Clay DE, impact assessment - the EDIP2003 methodology. Int J LCA Clay SA (2011) Determination of environmental impacts of anti- 1:11–13. https://d oi.o rg/1 0. 1065/l ca20 06. 04.0 14 microbial usage for US Northern Great Plains swine-production Price LB, Koch BJ, Hungate BA (2015) Ominous projections for global facilities: a life-cycle assessment approach. Int J Life Cycle antibiotic use in food-animal production. Proc Natl Acad Sci U Assess 16:27–39. https://d oi. org/ 10.1 007/s 11367- 010-0 241-y S A 112:5554–5555. https:// doi.o rg/ 10. 1073/ pnas. 150531 2112 Stone JJ, Dollarhide CR, Jinka R, Thaler RC, Hostetler CE, Clay DE Rahman SM, Eckelman MJ, Onnis-Hayden A, Gu AZ (2018) Com- (2010) Life cycle assessment of a modern northern great plains parative life cycle assessment of advanced wastewater treatment U.S. swine production facility. Environ Eng Sci 27:1009–1018. processes for removal of chemicals of emerging concern. Envi- https:// doi. org/ 10.1 089/ ees. 2010.0 051 ron Sci Technol 52:11346–11358. https:// doi. org/ 10. 1021/ acs. Surette MD, Wright GD (2017) Lessons from the environmental anti- est. 8b0003 6 biotic resistome. Annu Rev Microbiol 71:309–329. https:// doi. Rico A, Jacobs R, Van den Brink PJ, Tello A (2017) A probabilistic org/1 0.1 146/ annur ev-m icro-0 90816- 093420 approach to assess antibiotic resistance development risks in Szumilas M (2010) Explaining odds ratios. J Can Acad Child Adolesc environmental compartments and its application to an intensive Psychiatry 19:227–229 (PMID: 20842279) aquaculture production scenario. Environ Pollut 231:918–928. Tarpani RRZ, Alfonsín C, Hospido A, Azapagic A (2020) Life cycle https:// doi. org/ 10. 1016/j. envpol. 2017. 08. 079 environmental impacts of sewage sludge treatment methods for Rico A, Phu TM, Satapornvanit K, Min J, Shahabuddin AM, Henriks- resource recovery considering ecotoxicity of heavy metals and son PJG, Murray FJ, Little DC, Dalsgaard A, Van den Brink PJ pharmaceutical and personal care products. J Environ Manage (2013) Use of veterinary medicines, feed additives and probiotics 260. https://d oi. org/ 10. 1016/j. jenvm an.2 019.1 09643 in four major internationally traded aquaculture species farmed in Tarpani RRZ, Azapagic A (2018) Life cycle environmental impacts of Asia. Aquaculture 412–413:231–243. https:// doi. org/1 0.1 016/j. advanced wastewater treatment techniques for removal of phar- aquac ulture. 2013. 07.0 28 maceuticals and personal care products (PPCPs) 215:258–272. Rizzo L, Manaia C, Merlin C, Schwartz T, Dagot C, Ploy MC, Michael https:// doi.o rg/1 0. 1016/j. jenvm an.2 018.0 3.0 47 I, Fatta-Kassinos D (2013) Urban wastewater treatment plants as Tello A, Austin B, Telfer TC (2012) Selective pressure of antibiotic hotspots for antibiotic resistant bacteria and genes spread into the pollution on bacteria of importance to public health. Environ environment: A review. Sci Total Environ 447:345–360. https:// Health Perspect 120:1100–1106. https:// doi. org/ 10. 1289/ ehp. doi. org/ 10. 1016/j. scito tenv. 2013. 01.0 32 11046 50 Rico A, Van den Brink PJ (2014) Probabilistic risk assessment of Terasaki M, Makino M, Tatarazako N, (2009). Acute toxicity of para- veterinary medicines applied to four major aquaculture species bens and their chlorinated by-products with Daphnia magna and produced in Asia. Sci Total Environ 468–469:630–641. https:// Vibrio fischeri bioassays. J. Appl. Toxicol. 29:242–247. http:// doi.o rg/ 10. 1016/j. scito tenv. 2013.0 8.0 63 dx.d oi. org/ 10. 1002/j at. 1402 Robinson TP, Bu DP, Carrique-Mas J, Fèvre EM, Gilbert M, Grace D, van Boeckel TP, Brower C, Gilbert M, Grenfell BT, Levin SA, Rob- Hay SI, Jiwakanon J, Kakkar M, Kariuki S, Laxminarayan R, inson TP, Teillant A, Laxminarayan R (2015) Global trends in Lubroth J, Magnusson U, Thi Ngoc P, Van Boeckel TP, Woolhouse antimicrobial use in food animals. Proc Natl Acad Sci 112:5649– MEJ (2016) Antibiotic resistance is the quintessential One Health 5654. https://d oi. org/1 0. 1073/ pnas.1 50314 1112 1 3 The International Journal of Life Cycle Assessment van Bunnik BAD, Woolhouse MEJ (2017) Modelling the impact of Wypych TP, Marsland BJ (2018) Antibiotics as instigators of microbial curtailing antibiotic usage in food animals on antibiotic resist- dysbiosis: implications for asthma and allergy. Trends Immunol ance in humans. R Soc Open Sci 4. https://d oi.o rg/1 0.1 098/r sos. 39:697–711. https:// doi. org/1 0. 1016/j. it. 2018.0 2. 008 161067 Xiong W, Sun Y, Zhang T, Ding X, Li Y, Wang M, Zeng Z (2015) van den Bogaard AE, Stobberingh EE (1999) Antibiotic usage in ani- Antibiotics, Antibiotic Resistance Genes, and Bacterial Com- mals. Drugs 58:589–607. https:// doi. org/ 10. 2165/ 00003 495- munity Composition in Fresh Water Aquaculture Environment 199958 040- 00002 in China. Microb Ecol 70:425–432. https:// doi. org/ 10. 1007/ WHO (2014) Antimicrobial resistance: global report on surveillance. s00248- 015-0 583-x World Health Organization World Health Organization (2019) Critically important antimicrobials Publisher’s Note Springer Nature remains neutral with regard to for human medicine, 6th revision jurisdictional claims in published maps and institutional affiliations. Wright GD (2007) The antibiotic resistome: the nexus of chemical and genetic diversity. Nat Rev Microbiol 5:175–186. https:// doi. org/ 10. 1038/ nrmic ro1614 1 3