Assessing exposure and risks of pharmaceuticals in an urban river system

Pharmaceuticals are ubiquitous in the freshwater environment, a result of an increasingly urbanised water cycle. Environmental risk assessments are available for a small proportion of the over 1900 pharmaceuticals in use, raising concern over the potential risks posed by pharmaceuticals with limited data, as effects on non-target organisms have been observed. Experimentally filling these gaps is a large, costly and likely unnecessary task. Risk-based prioritisation is a potential tool for addressing this challenge by identifying which pharmaceuticals may pose risks and are therefore a priority for study. Simple exposure models are commonly used to predict environmental concentrations (PECs), however the suitability of these models for prioritisation is unknown.
A scoping study targeted 95 pharmaceuticals in samples from the Rivers Ouse and Foss in York, UK, 25 were quantified. Measured environmental concentrations (MECs) were compared with simple PECs based on local usage data and dilution factors. MECs and simple PECs were used to prioritise pharmaceuticals and, for the larger River Ouse, different priority lists using the two approaches emerged. This conclusion was based on limited monitoring data, therefore an HPLC-MS/MS quantification method for 33 pharmaceuticals was developed, validated and applied to a year-long monitoring campaign to build a robust monitoring dataset. Significant spatial and temporal trends were observed in both rivers apparently driven by flow, pharmaceutical usage, wastewater treatment removal, and in-stream attenuation. These drivers differently influenced concentrations in either river. The simple PECs and PECs derived from a higher-tier spatial exposure model (LF2000-WQX) were validated against annual average MECs. LF2000-WQX outperformed the simple PEC in both rivers. A re-prioritisation using LF2000-WQX demonstrated that improved predictive power translated into better agreement of prioritisation outcomes with MECs. The use of simple PECs for the prioritisation and risk assessment of pharmaceuticals should be avoided and the use of higher-tier spatial exposure models encouraged.