Objective:
This pilot study aims to assess the environmental impact of Positron Emission Tomography/Computed Tomography (PET/CT) imaging, in the context of efforts to mitigate climate change. This is studied by investigating the energy usage, waste generation, and the radiopharmaceutical production.

Methods:
Power measurements of the PET/CT scanner were conducted over four weeks and during different power settings (system on, computers off, software shutdown and energy saving mode) with a power analyzer. Various outcome measures, such as energy and cost savings per hour, energy consumption per PET/CT radiopharmaceutical procedure, and total annual energy consumption in two distinct scenarios, were calculated. Additional power measurements of reporting stations and desktop computers were performed in off-mode, sleeping mode and active mode. The data for waste disposal was collected by counting and weighing consumables used for a Fluorine-18 Fluorodeoxyglucose ([18F]FDG) procedure. Waste streams were identified and the data categorized per waste stream. The workflow for producing the [18F]FDG radiopharmaceutical was mapped out and data on the energy consumption of the cyclotron and cleanrooms and waste generation was collected. All data was finally expressed in CO2-equivalents (CO2-eq) by using emission factors.

Results:
Of the four different PET/CT power settings, the energy saving mode showed the most significant energy reduction compared to system on, resulting in ∼30% energy savings and a 0.94 kg CO2-eq emission reduction per hour. Furthermore, the energy consumption of various radiopharmaceutical procedures showed results ranging from 3.03 (2.90-3.09) to 5.15 (5.00-5.41) kWh/procedure (1.90 to 1.12 kg CO2-eq). The reporting stations and desktop computers showed both a ∼71% energy reduction in shutdown mode compared to system on. Correct waste separation of plastics and paper showed a ∼64%
CO2-eq reduction per procedure. For radiopharmaceutical production, the impact of the cyclotron’s energy consumption was found the largest (0.44 kg CO2-eq/procedure) compared to HVAC energy consumption (0.13 kg CO2-eq/procedure) and waste disposal (0.003 kg CO2-eq/procedure).The total environmental impact of [18F]FDG PET/CT imaging was measured at 2.01 kg CO2-eq per procedure, with PET/CT scanner energy consumption (62%) and cyclotron energy consumption (22%) being the primary contributors.

Conclusion:
This study showed that the environmental impact of PET/CT imaging is substantial and that it can be reduced by implementing energy saving strategies during non-operational hours, improving waste segregation for recycling and optimizing the energy efficiency of cyclotrons and cleanrooms. Furthermore, hospitals and manufacturers can even reduce the impact further by addressing opportunities such as using eco-friendly or recycled materials and incorporating green energy sources for power supply. Future research should focus on increasing energy-efficiency of scanners, cyclotron and HVAC systems, reducing the scan length and reducing low-value scanning to further decrease the environmental impact of PET/CT imaging while maintaining high diagnostic standards.

The operating room and central sterile supply department (CSSD) are two
resource-intensive healthcare departments. Reducing the environmental and financial impact of these departments is of great importance in the global debate on sustainability. Removing unnecessary items from instrument trays may provide a partial solution. However, improvements to tray compositions are often time-consuming, not permanent and not using the full reduction potential. This research compares different methods of medical instrument tray optimization and combines different aspects to create a new methodology. Three models compared and evaluated objective and subjective instrument use percentages and different suggestions for tray optimizations. New tray compositions, based on objective use rates, were discussed with the medical specialists and attuned to their requirements. The result
was then reviewed by the OR-assistants, before making any final adjustments. This methodology is tested for abdominal instrument trays at the gynaecology department of the Leiden University
Medical Center (LUMC). The new trays were evaluated over a test period and missing items were registered during follow-up. Risks to patient safety as a result of missing individual instruments
and chances of missing instruments for different reduction methods are discussed and quantified. Mean use rates of the designated instrument sets are 28.4% (SD=6.43%) for open surgery (n=16) and 47.6% (SD=8.16%) for minimally invasive procedures (n=12). A 37% reduction of instruments is reached across three abdominal trays by removing unnecessary items. Weight of the tray contents is reduced by 31%. During the evaluation (n=7 procedures), mean instrument use for abdominal procedures increased from 28.4% (SD=6.43%) to 46.47% (SD=10.96%) after tray optimization. A reduction based on use rates with a 10% cut-off or based on the recommendations of the medical specialists, induces an 8.7% or 3.9% chance of missing any instrument during the procedure, respectively. Tray content reductions based on OR-assistant suggestions and objective use rates with a 0% cut-off are safe, but do not utilize the complete reduction potential. Different tray optimization methods were compared. Use rates were measured in the OR and expert recommendations were discussed in group sessions. A combination of objective use rates and subjective expert considerations can lead to a significant reduction of the amount of unnecessary items in the medical instrument cycle without harming patient safety. Group consensus amongst medical specialists suggests the most radical reduction of instruments on the tray, but increases chances of missing instruments during surgery. Recommendations based on use rates have to be supplemented by reduction efforts based on the clustering of equivalent instruments. Future work should focus on ways to scale reduction efforts to an autonomous hospital-wide system and teach optimization models to mimic human expert input. It should also focus on exact measurements in the CSSD to estimate reductions in costs and CO2-footprint reliably.