An environmental risk assessment of contamination of lakes and rivers in Northern Ireland with pharmaceuticals, personal care products and drugs of abuse.

In this work, we present the most spatially comprehensive environmental risk assessment of 114 pharmaceuticals, personal care products and illicit drug residues in the surface waters of Northern Ireland. We sampled at 50 sites spread across five lakes and rivers. We detected 21 different substances in 137 water samples and quantified between 0.7-2967 ng L-1 (n = 834 quantifiable measurements). Within these, the top five most frequently detected compounds were the opioid painkiller tramadol (91%), the cocaine metabolite benzoylecgonine (91%, with cocaine itself at 55% frequency), caffeine (80%), the anti-depressant venlafaxine (76%, which also appears on the EU Water Framework Directive watch list) and the antipsychotic drug carbamazepine (55%). That said, environmental risk quotients (RQs) were low to insignificant overall. Only diclofenac and caffeine presented moderate to high risks (RQ = 3.3 and 24.7, respectively). In comparison to other works, including regulatory body chemical monitoring data, Northern Ireland's inland waters were less heavily contaminated with these substances compared with other river systems in the rest of the UK and the island of Ireland. Preliminary assessment of the potential for generation of antibiotic resistance through use of lowest predicted no-effect concentration for resistance data (PNECRs) showed a statistically elevated risk from trimethoprim in comparison to Environment Agency data monitored over the same timeframe in England. This was particularly noted in the River Lagan in the Greater Belfast area. Statistical analysis of catchments revealed that the sources of these pollutants was likely from both treated and untreated wastewater. This pilot study represents the first baseline study of water contamination with selected pharmaceuticals, personal care products and illicit drugs on which to base future national monitoring programmes.