Department Environmental Social Sciences

Micropollutants (e.g. pharmaceuticals) are increasingly detected in water bodies: their elimination in conventional wastewater treatment plants (WWTP) is incomplete. Mostly, we are facing considerable uncertainty about aquatic ecosystem effects, with some exceptions. For instance, negative effects on fish reproduction have been linked to the exposure of estrogens (female hormones). In such a complex decision situation with high uncertainty and many stakeholders involved, decisions cannot be made by scientists alone, but participatory and multidisciplinary approaches are needed. For instance, technologies to eliminate micropollutants at WWTP are expensive, and society will have to decide, whether potential risks justify the high costs.

Micropollutants can be reduced by new technologies at WWTP or source control. Hospitals that are important point sources could be one target of measures. Options include treating all hospital wastewater or partial treatment (e.g. oncology wards to remove cytostatics (anti-cancer drugs)). Moreover, only the urine could be collected and treated separately, since the urine contains ca. 60–70% of the medicals that are excreted by humans. Urine source separation has been intensively researched in the project Novaquatis. At European level, the reduction of pharmaceutical residues from wastewater has been researched in the projects PILLS and noPILLS.

This project focused on a relevant, well-defined decision situation with two exemplary Swiss case studies: a cantonal hospital and a cantonal psychiatric clinic. MCDA was used to combine expert predictions (e.g., costs, pharmaceutical mass flows, ecotoxicological risk, pathogen removal) with the subjective preferences of 26 stakeholders. The example also served to study more general research questions related to the practical application of MCDA in environmental decision problems.

In the cantonal hospital, 68 technical and organizational options were analyzed, and 50 in the psychiatric clinic. The organizational alternatives included the separate collection of urine, the technical measures e.g. treatment of the wastewater with activated carbon, ozonation, or reverse osmosis. Expert predictions were used to estimate the performance of each alternative, e.g. concerning pathogen removal, reduction of the pharmaceutical load and ecotoxicological risk as well as the associated costs. The predictions also included estimates of the effort for nurses and patients, and public perception.

Interviews with 26 stakeholders from the hospital and with representatives of the authorities were carried out. They were asked to state their preferences and give valuations concerning important trade-offs in this decision. The standard method to elicit weights was modified by introducing a ‘Reversed-Swing Method’ that seemed more realistic for this case. The MCDA procedure was a compromise between time demand and elaborateness; which was tested with careful sensitivity analyses.

While the general hospital contributed substantially to the total load of pharmaceuticals that arrives at the communal wastewater treatment plant, the contribution from the psychiatric clinic was relatively small. This indicates that careful analysis of each specific case is required. The Top-100 most heavily used pharmaceuticals also strongly differed between the two hospitals, with only 37 overlapping pharmaceuticals. Moreover, only a part of the Top-100 pharmaceuticals actually contributed to the environmental risk of the emitted mixture of pharmaceuticals; specifically four pharmaceuticals (amiodarone, ritonavir, clotrimazole, and diclofenac) contributed most strongly to this risk. The load, i.e. the amount of consumed pharmaceuticals is thus insufficient to predict the environmental risk.

When comparing the different options for pharmaceutical removal with MCDA in the general hospital, those technical options that remove the total load of pharmaceuticals performed systematically better than cheaper options, which can only partially removal the load of pharmaceuticals. The outcome was less clear for the psychiatric clinic. The results concerning the best-performing options were very stable across stakeholders, even though individual people did have different preferences. Stakeholder involvement with MCDA proved promising, and the feedback of our interview partners was generally very positive. The results indicate that the protection of water resources is of high importance. Point source options at hospitals can be very well accepted if there is a reasonable trade-off between pharmaceutical removal and costs.