Portal de Periódicos da CAPES

Radiology and the Climate Crisis: Opportunities and Challenges— Radiology In Training

2021; Radiological Society of North America; Volume: 300; Issue: 3 Linguagem: Inglês

10.1148/radiol.2021210851

1527-1315

Bryan Buckley, Peter J. MacMahon,

Radiation Dose and Imaging

HomeRadiologyVol. 300, No. 3 PreviousNext Reviews and CommentaryFree AccessPerspectivesRadiology and the Climate Crisis: Opportunities and Challenges—Radiology In TrainingBryan W. Buckley , Peter J. MacMahonBryan W. Buckley , Peter J. MacMahonAuthor AffiliationsFrom the Department of Radiology, Mater Misericordiae University Hospital, Eccles Street, Dublin 7, Ireland.Address correspondence to B.W.B. (e-mail: [email protected]).Bryan W. Buckley Peter J. MacMahonPublished Online:Jul 13 2021https://doi.org/10.1148/radiol.2021210851MoreSectionsPDF ToolsImage ViewerAdd to favoritesCiteTrack CitationsPermissionsReprints ShareShare onFacebookTwitterLinked In Dr Bryan Buckley is a third-year resident at the Mater Misericordiae University Hospital in Dublin, Ireland. He is a recipient of the 2019 Royal College of Surgeons Ireland Faculty of Radiologists Publication Award. His research interests include environmental sustainability, artificial intelligence, and medical education.Download as PowerPointOpen in Image Viewer SummaryRadiology trainees have an important role to play in advancing more sustainable and environmentally conscious radiology practices through research, education, and local advocacy.IntroductionThe COVID-19 pandemic continues to challenge our perspective of “normal” within every health care system around the world. The willingness to adapt and rapidly change normal practices has been a crucial component of our response to the pandemic. Within radiology specifically this has been no different, with drastic changes to some areas of practice, including remote reporting, new services aimed specifically at the needs of patients with COVID-19, and even the redeployment of radiology staff. With vaccination programs progressing worldwide, there is now hope that an end is in sight. The ability of health care and radiology to adapt, as it has over the past year, must be harnessed for the next challenge to world health—the climate crisis. As radiology trainees, we must lead in pressing the issue.Energy-intensive Health CareThe contribution of health care activity to global greenhouse gas emission is substantial. If the health care sectors of the United States, Canada, Australia, and England were combined and ranked as an independent nation, they would be estimated to be the seventh largest emitter of greenhouse gases in the world (1). Specifically within the United States, estimation of health care contribution to the overall greenhouse gas emissions was estimated at 10% in 2013 (2).With the increasing awareness of the impact of human activity on the environment, there has been a noticeable societal shift toward a demand for environmentally sustainable business practices. Research carried out in 2015 in the United Kingdom revealed that 92% of the public believed it is important for the health system to work in a more sustainable way, with one in four believing it should be a top priority (up from 19% 2 years prior) (3). Anesthesiology and surgery have led in the field of sustainability within medicine. Within radiology specifically, there has been a surprising lack of research on sustainability, as technology is energy-intensive and plays a substantial role in our specialty.The largest contributors to the energy usage within the radiology department are large advanced diagnostic imaging equipment (ie, those used with MRI, CT, US, and nuclear medicine). MRI has the largest energy consumption when production and in-use phases are considered; the energy used for each individual study performed with an MRI scanner is comparable to that used when cooling a three-bedroom house with central air conditioning for 1 day or the desalination of 7000 gallons of fresh water (4). US was demonstrated to use one-twentieth of the overall energy used with either CT or MRI. Even when imaging equipment is not in active use or is in a nonproductive idle state, it continues to require energy. For example, approximately two-thirds of energy usage for CT occurs during this nonproductive idle state, indicating a substantial energy inefficiency (5). Elsewhere within the radiology department, an analysis of 32 radiology reporting stations demonstrated overall power consumption of 53 170 kWh per annum, or the equivalent of the annual energy consumption of 12 family households in Switzerland (6). Other devices that are regularly used within a radiology department, from desktop computers to air conditioning systems, also contribute to the overall environmental impact of a service.Although every imaging study performed must be clinically justified, it must also be performed as an efficient allocation of a limited resource. Models such as “utilization management” have been proposed as a systemic approach to evaluating the appropriateness of clinical care and achieving best possible patient outcomes by using the most appropriate health resources. However, these only examine economic impacts. In choosing any one modality over another, consideration of the ecologic impact should also factor into decision making around the allocation of these resources.The Ecologic Cost of Big DataOne overlooked aspect in assessing the environmental impact of the radiology services has been the rapid growth in the need for storage in radiologic data centers. As of 2020, there were 28.5 million individual imaging studies in the Irish national radiologic data storage system (the National Integrated Medical Imaging System, or NIMIS). This amounts to over 1 petabyte of radiologic data, with an approximate growth in data storage of 23% every year over the past 5 years. Fifty-three percent of that data is dominated by CT studies. This is reflected in the increase of the average size of a CT study stored in the National Integrated Medical Imaging System as measured in megabytes. In 2011, the average size was 66 megabytes but has grown year-on-year by approximately 20%, with the current average size of a CT study now standing at 160 megabytes. This reflects the increasing complexity of the average CT study now performed, with more thin-section reconstructions, multiple reformats, and other postprocessing to create additional imaging series. The storage and transmission of large volumes of medical imaging requires energy.Data storage servers and cooling systems account for the greatest shares of direct electricity use within a data center, approximately 86% of the energy usage (7). Total water consumption in United States data centers was estimated in 2014 at 626 billion liters of water per year (over 75% directly related to electricity generation), with this estimated at the time to grow to 660 billion liters of water per year in 2020 (7). Data centers across the world may emit as much carbon dioxide as the global aviation industry (8). Furthermore, with the increasing interest and expectations around artificial intelligence, particularly within radiology as a specialty, the need for higher computing power is only going to grow. The computational effort required to solve problems increases exponentially with the size of data sets, so with an ultimate goal of continuous-learning artificial intelligence models, this will not be solved with the addition of more hardware (9). Further research and promotion of more computationally efficient algorithms, as well as hardware requiring less energy, is needed (10).Aside from the initial environmental costs associated with putting such infrastructure in place, every chest radiograph or brain CT scan that is stored in a data center (including the inherent need for duplicate copies and redundancy) places an energy demand on that system. While any one individual study likely has miniscule energy requirements, even for very long-term storage, any study or part of a study that may serve no future purpose to the health care system creates a needless carbon footprint. For most health care systems, such as ours in Ireland, these data are often retained indefinitely. If the volumes and size of data that is required to be maintained continues to grow at current rates, this may need to be addressed with clear guidelines for the minimization of excess or redundant data that is of no perceived future benefit. One such example would be the multiple scans performed during CT perfusion imaging and CT-guided procedures or the multiplanar reconstructions that are created for ease of reviewing by the radiologist. These data or some part of them could be dispensed with at some defined point in time and the carbon footprint of a service reduced with no impact on future patient care while providing cost savings to the health care system.Greening RadiologyWhile medicine as a whole may be slow to adopt sustainable business practices and respond to societal shifts and patients’ expectations in this regard, radiology is uniquely placed to contribute due to its inherent technology base. These changes must be driven at a local level and, as trainees, we can play our part in this regard through advocacy and education. All of the factors contributing to the specialty’s environmental impact have not yet been considered, and more research is needed, specifically on the environmental impacts of data storage and processing. The COVID-19 pandemic has demonstrated that health care systems and radiology departments are adaptable and can reconfigure to deal with crises. This same ingenuity will be needed to overcome the climate crisis.Within radiology, we are well accustomed to the “as low as reasonably achievable” principle, or ALARA, which dictates that the radiation exposure to a patient undergoing a particular study or procedure should be as low as reasonably possible. A similar approach should be taken to environmental impact where appropriate. For example, when it is agreed that an ovarian cyst can be followed with US rather than MRI (the former of which uses one-twentieth of the energy required compared with that needed for a single MRI examination), this optimizes the ecologic as well as the economic impact without affecting patient outcomes.Similarly, a debate around data creation and storage is warranted. Not all data are created equal. As a specialty, radiology is arguably in its infancy when it comes to data creation. Limiting the excess data that is routinely created and often stored indefinitely makes sense from both an economic and ecologic perspective. The recent interest in abbreviated MRI protocols that limit “excess” sequences in answering a clinical question are very welcomed in this regard. On the storage end, the multiple reconstructions and iterations of the raw data that are routinely created in CT imaging are unlikely to serve a clinical use in the future and do not need to be sent for data storage.Being comfortable with and aware of sustainable business practices will likely become an increasingly important factor in patient decision making and expectations. Considering the environmental impact of our decisions, making recommendations that limit carbon emissions, and advocating for business practices that can reduce them will improve our specialty and better serve our patients of the future.Disclosures of Conflicts of Interest: B.W.B. disclosed no relevant relationships. P.J.M. disclosed no relevant relationships.AcknowledgmentsThe authors express their gratitude to the staff at the National Integrated Medical Imaging System, or NIMIS, for their help in researching this article.References1. Sherman JD, MacNeill A, Thiel C. Reducing pollution from the health care industry. JAMA 2019;322(11):1043–1044. Crossref, Medline, Google Scholar2. Eckelman MJ, Sherman J. Environmental impacts of the U.S. health care system and effects on public health. PLoS One 2016;11(6):e0157014. Crossref, Medline, Google Scholar3. Public opinion survey 2015: Sustainability and the NHS, Public Health and Social Care system – Ipsos Mori survey. Public Health England. https://www.england.nhs.uk/greenernhs/wp-content/uploads/sites/51/2021/02/Sustainability-and-the-NHS-Public-opinion-survey-2015.pdf. Published March 2016.Accessed November 16, 2020. Google Scholar4. Martin M, Mohnke A, Lewis GM, Dunnick NR, Keoleian G, Maturen KE. Environmental Impacts of Abdominal Imaging: A Pilot Investigation. J Am Coll Radiol 2018;15(10):1385–1393. Crossref, Medline, Google Scholar5. Heye T, Knoerl R, Wehrle T, et al. The Energy Consumption of Radiology: Energy- and Cost-saving Opportunities for CT and MRI Operation. Radiology 2020;295(3):593–605. Link, Google Scholar6. Hainc N, Brantner P, Zaehringer C, Hohmann J. “Green Fingerprint” Project: evaluation of the power consumption of reporting stations in a radiology department. Acad Radiol 2020;27(11):1594–1600. Crossref, Medline, Google Scholar7. Shehabi A, Smith S, Sartor D, et al. United States Data Center Energy Usage Report. LBNL-1005775. Lawrence Berkeley National Laboratory,Berkeley, California. https://www.osti.gov/biblio/1372902/. Published 2016. Accessed November 16, 2020. Google Scholar8. Pearce F. Energy Hogs: Can World’s Huge Data Centers Be Made More Efficient? Yale Environment 360. https://e360.yale.edu/Features/Energy-Hogs-Can-Huge-Data-Centers-Be-Made-More-Efficient. Published April 3, 2018. Accessed January 10, 2021. Google Scholar9. Pianykh OS, Langs G, Dewey M, et al. Continuous learning AI in radiology: implementation principles and early applications. Radiology 2020;297(1):6–14. Link, Google Scholar10. Strubell E, Ganesh A, McCallum A. Energy and Policy Considerations for Deep Learning in NLP. Association for Computational Linguistics. https://www.aclweb.org/anthology/P19-1355/. Published 2019. Accessed November 16, 2020. Google ScholarArticle HistoryReceived: Apr 9 2021Revision requested: Apr 27 2021Revision received: May 6 2021Accepted: May 12 2021Published online: July 13 2021Published in print: Sept 2021 FiguresReferencesRelatedDetailsCited ByClimate Change and Radiology: Impetus for Change and a Toolkit for ActionMaura Brown, Julia Hyde Schoen, Jonathan Gross, Reed A. Omary, Kate Hanneman, 18 April 2023 | Radiology, Vol. 307, No. 4The evolving call to action for including climate change and environmental sustainability themes in health professional education: A scoping reviewMeagan E.Brennan, Diana L.Madden2023 | The Journal of Climate Change and Health, Vol. 9All Specialties in Radiology Must Address the Climate CrisisJonathan S. Gross, Cassandra L. Thiel, 8 February 2022 | Radiology, Vol. 303, No. 2Sustainability in interventional radiology: are we doing enough to save the environment?Pey LingShum, Hong KuanKok, JulianMaingard, KevinZhou, VivienneVan Damme, Christen D.Barras, Lee-AnneSlater, WinstonChong, Ronil V.Chandra, AshuJhamb, MarkBrooks, HamedAsadi2022 | CVIR Endovascular, Vol. 5, No. 1Accompanying This ArticleAuthor Interview: Radiology and the Climate Crisis: Opportunities and Challenges—Radiology In TrainingApr 11 2023Default Digital Object SeriesRecommended Articles Increasing Opportunities for Trainees to Engage in Global Health Radiology: Radiology In TrainingRadiology2021Volume: 300Issue: 2pp. E320-E322Cardiac Imaging Trends from 2010 to 2019 in the Medicare PopulationRadiology: Cardiothoracic Imaging2021Volume: 3Issue: 5Small Steps toward a More Sustainable and Energy-efficient Operation of MRIRadiology2023Volume: 307Issue: 4The Impact of the COVID-19 Pandemic on the Radiology Research Enterprise: Radiology Scientific Expert PanelRadiology2020Volume: 296Issue: 3pp. E134-E140External Factors That Influence the Practice of Radiology: Proceedings of the International Society for Strategic Studies in Radiology MeetingRadiology2017Volume: 283Issue: 3pp. 845-853See More RSNA Education Exhibits Climate Change and Radiology: What Radiologists Need to KnowDigital Posters2022Pandemic Preparedness: Streamlining the Breast Imaging Patient Care and Teaching Experience During COVID-19Digital Posters2020Misplaced Gas — Sinister Finding in Septic Diabetics: A Pictorial Review of Emphysematous InfectionsDigital Posters2019 RSNA Case Collection Hydrogen peroxide ingestionRSNA Case Collection2020High Pressure Injection InjuryRSNA Case Collection2021Amoebic abscess RSNA Case Collection2020 Vol. 300, No. 3 Video SummaryMetrics Altmetric Score PDF download