Despite the major impact of healthcare on the environment, the sustainability of (new) treatments is currently understudied. We took the first steps towards analyzing the environmental impact of an MR-guided interventional radiology treatment, i.e., the MR-HIFU treatment of uterine fibroids, by evaluating the CO2 emission of energy use by the MRI scanner and the MR-HIFU device, the CO2 emission associated with medication use, and by evaluating the amount of solid waste produced during a single treatment. We are aware that our results do not represent a full LCA of uterine fibroid MR-HIFU and that they underestimate the total use. In an ideal situation, different environmental impact categories would be part of an LCA and all elements of the MR-HIFU treatment life cycle would be analyzed. Nevertheless, we believe publishing this data is important as it may contribute to future comparative LCAs. More importantly, with this study, we aimed to underline the need to support and perform LCAs for treatments, and to provide further insight into the challenges of performing an LCA in a healthcare setting, and specifically in interventional radiology [27, 28].

The operating mechanism of an MR-HIFU treatment is induced focused energy. Therefore, the energy consumed by the MR-HIFU device is an important contributor to total CO2 emission during treatment. In our analyses, we used the conversion factor “grey energy,” a representative Dutch combination of coal, gas, and nuclear energy, without considering the energy used to build the production facilities [24]. The emission of kWh energy generated by water, wind, or solar energy is 0.004, 0.014, and 0.061 kg CO2/kWh, respectively [24]. In our hospital, all energy is CO2 neutrally generated. Therefore, the exact amount of energy needed to perform the MR-HIFU treatment may seem less relevant. However, even “green” energy must be generated and may have an impact on other environmental impact categories.

The CO2 emission of pharmaceuticals is understudied, and industry LCA publications cannot be verified, as they require access to confidential manufacturing processes [1, 29]. Medication accounts for around 25% of the UK’s total healthcare CO2 emissions [30].

To estimate the CO2 emission of the medication administered during an MR-HIFU treatment, we used CO2 emission data from previous studies. These studies often did not include the emission of packaging [6, 25, 26]. McAlister et al. calculated that 90% of morphine CO2 emissions were caused by sterilization and packaging. Therefore, sterilization and packaging should not be neglected [29]. An important advantage of MR-HIFU treatment over uterine fibroid surgery is that there is no need for anesthetic gases, which are a major contributor to the CO2 emissions from medication in general [6, 30].

Hospitals in the USA generate 3.4 billion pounds of solid waste annually [7]. The procurement supply chain causes most of the CO2 emission. Therefore, decisions made at the product manufacturing stage could reduce environmental impact. Clements et al. analyzed the packaging waste from single-use products used for interventional radiology procedures [31]. Of the 72 products analyzed, 55% of their total weight consisted of waste and 76% could potentially be safely replaced by reusable products. We did not include the amount of recyclability in the scope of our research because no recycling policy was available in our hospital, or in the healthcare system on interventions, but this aspect should be included in future analyses.

Ideally, a comparative study would have been conducted, comparing LCAs of different uterine fibroid treatments. However, at this time, such a study is not feasible due to several hurdles, including the lack of an LCA database with relevant health inventories, the lack of standardization of how a healthcare treatment LCA should be performed, and what boundaries are considered acceptable. As a result, the only studies available at this stage are those such as ours that include only part of an environmental analysis. Thiel et al. performed a waste analysis after different types of hysterectomy with an average mass of 13.7 kg—much higher than the 1.2 kg we collected after MR-HIFU [7]. Chua et al. performed an LCA of all energy, materials, and waste used in a radiological intervention room [32], reporting on 98 interventions including embolizations, but the indications for these procedures were unclear and they may not have been performed as treatment of uterine fibroids. For future comparisons, it is necessary to standardize which elements of the treatments should be included and to provide as disaggregated data as possible.

The main limitation of our study is that we did not perform a full cradle-to-grave LCA of uterine fibroid MR-HIFU and therefore it should be considered as preliminary. Ideally, all life cycle phases of a product, process, or system are included since interventions in one phase may have consequences in another [5]. At the same time, all LCA studies are incomplete to some extent because boundaries must be set to limit the amount of data and analysis required. We only included the three components of the MR-HIFU treatment (Fig. 3) for which we had the resources to examine: those that were in our own center of influence and were expected to differ most from other uterine fibroid treatments. Current LCA databases lack data on healthcare processes, products, and systems data and, together with LCIA software, are not publicly available. The reason for this seems to be a lack of awareness and transparency by vendors, hospital mechanics, and energy suppliers. This hurdle could be overcome if sustainability became as important as clinical effectiveness and cost-effectiveness in healthcare.

Unfortunately, it was not possible to measure the energy consumed by both the MR scanner and the MR-HIFU device during the MR-HIFU treatment. Therefore, we used the second-best option, which is often done in LCAs and concerns literature sources. It is important to note that these literature sources are limited and often contain unvalidated data. We included energy consumption data from a comparable MRI scanner that we used during our treatments from a master’s thesis [23]. In other publications, different MRI scanners were used, but the amount of energy per mode was comparable and therefore considered legitimate [13, 14].

Moreover, the amount of energy calculated in this study is an underestimation of the total energy used. For example, the use of heating, ventilation, and air conditioning used in the MRI room can also be relevant contributors [32]. The production of the MRI scanner was excluded from this analysis. Since the MRI scanner has an average lifespan of 15 years, it is questionable whether the emission from the construction of the MRI scanner contributes significantly to the CO2 emission of a single treatment [13].

For medication use, the main limitation is the fact that, for most drugs, there is no LCA data available or is it not possible to retrieve such data due to drug patents. Furthermore, we did not analyze the amount of unused medication. Unused and disposed medication is an important contributor to the negative environmental impact of healthcare in general and, if not correctly disposed of, can contribute to water contamination [4].

Our amount of waste measured is most likely an underestimation, as we did not include sharp materials, although the additional weight is expected to be small. In addition, since the waste analyses were prospective, bias due to the treatment team’s knowledge of the amount of waste produced could not be ruled out. However, the team involved in the MR-HIFU treatments was not specifically engaged with sustainability.

Timely action is urgently needed to reduce the environmental impact of healthcare. In light of this study and the challenges that remain, we would make some suggestions for change so that the interventional radiology community can take responsibility and make a positive contribution.

First, we should determine the appropriate indication for treatment and treat only when necessary and beneficial. We should minimize the use of materials, substitute them with more environmentally friendly products, move away from certain heat-trapping anesthetic gases, maximize instrument reuse or encourage (research into) re-usable instruments, and reduce off-hour energy consumption [7, 21, 22, 31, 33]. Secondly, the hospital purchasing departments need to focus on the environmental impact of the products they purchase, and hospitals should collaborate with suppliers who are willing to provide information on the environmental impact of their products; authorities and hospitals should feel the societal responsibility to be part of this change.

Thirdly, we should feel the need to perform LCAs and should be willing to contribute to them. Standardizing the way in which LCAs of treatments in healthcare are carried out, including clarity on how boundaries are set, should be developed by the community and high-quality healthcare LCA databases and LCIA software should be made available on an open-access basis. To create this healthcare database, the community as a whole should be willing to work open access and it should become a mandatory element of future funding and grants. Finally and importantly, research grants focused on sustainability could contribute to the transformation process. They could provide the much-needed financial contribution to support the necessary actions outlined above. In the Netherlands, a specific healthcare LCI database is currently being developed with government funding.