Risk-management tool for environmental prioritization of pharmaceuticals based on emissions from hospitals

Caterina Zillien*, Cornelis van Loon, Marijn Gulpen, Kevin Tipatet, Birgit Hanssen, Henry Beeltje, Erwin Roex, Rik Oldenkamp, Leo Posthuma, Ad M. J. Ragas

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review


Over the past decade, the health care sector has become increasingly aware of the impact of pharmaceutical emissions to the environment. Yet, it remains unclear which compounds are the most relevant to address and at what point emission control is most effective. This study presents a modelling framework to prioritize pharmaceuticals based on their relative risks for aquatic organisms, using purchase and prescription data from hospitals. The framework consists of an emission prediction module and a risk prioritization module. The emission prediction module accounts for three different routes of intake (oral, intravenous, rectal), for non-patient consumption, and for delayed athome excretion due to relatively long half-lives or prescription durations of selected pharmaceuticals. We showcase the modelling framework with 16 pharmaceuticals administered at two Dutch academic hospitals. Predictions were validated with experimental data from passive sampling in the sewer system. With the exception of metformin, all predictions were within a factor of 10 from measurements. The risk prioritization module ranks each pharmaceutical based on its predicted relative risk for aquatic organisms. The resulting prioritization suggests that emission mitigation strategies should mainly focus on antibiotics and non-steroidal anti-inflammatory drugs (NSAIDs).
Original languageEnglish
Article number133733
JournalScience of the Total Environment
Publication statusPublished - 1 Dec 2019


Dive into the research topics of 'Risk-management tool for environmental prioritization of pharmaceuticals based on emissions from hospitals'. Together they form a unique fingerprint.

Cite this