Assessing the Carbon Footprint of Hemodialysis: A First Step Toward Environmentally Sustainable Kidney Care

Climate change has been declared the single biggest health threat facing humanity.¹ Its negative effects on health are broad, extending to all human organ systems, including the kidneys. Increased ambient temperatures and heat extremes translate into more AKI, kidney stones, and urinary tract infections, which may evolve into CKD and kidney failure.² Higher temperatures, along with more flooding events, increase diarrheal, rodent- and mosquito-borne disease spread, all major causes of AKI in low-income and tropical regions.² Strenuous work and prolonged exposure to high temperatures have been identified as probable contributors to epidemics of CKD in diverse world regions, a health crisis that has disproportionately affected underprivileged populations.²

The science is clear—no degree of further temperature rise is safe to health. We must urgently curb emissions to avert catastrophic health effects and prevent millions of avoidable deaths. Although every sector needs to play its part in limiting climate change, the responsibility is particularly great for the health care sector, which exists to protect and improve health. Health care is also a major part of the problem, contributing an estimated 4%–6% of global greenhouse gas emissions.³ Indeed, if global health care were a country, it would be the world’s fifth largest emitter.⁴

Within health care, the carbon footprint of hemodialysis (HD) appears to be particularly high, due to both a high use of energy, water, and consumables, and the repeated nature of dialysis treatments. However, studies quantifying this carbon effect are sparse.

In this issue of JASN, Sehgal et al. examine greenhouse gas emissions arising from 209,481 in-center HD treatments across 15 facilities in Ohio, United States. They do so by utilizing lifecycle analysis, a methodology that estimates the total amount of carbon dioxide (CO₂) directly and indirectly generated by an activity over all its life stages, from the extraction of raw materials such as gas and iron ore, to the manufacture and distribution of goods utilized, through the treatment itself and the end-of-life of all products.

They show average per-treatment emissions of 58.9 kg carbon dioxide equivalents (CO₂-eq), with a three-fold variation between facilities. The largest contributors were electricity and gas usage and the transport of both patients and staff, with transportation, gas, and water usage being the three contributors that varied the most between facilities.

In so doing, the authors confirm HD treatments at these facilities have a sizeable carbon footprint, comparable to that of 1.5 hours of being imaged in an magnetic resonance imaging scanner.

This study is important because it is the first of its kind originating from the United States. The United States has the second highest incidence of treated kidney failure of all countries and utilizes thrice-weekly HD as the predominant form of KRT.⁵ This makes it responsible for a major share of HD emissions globally, while simultaneously providing it with substantial purchasing power. In turn, the latter confers to it an ability to exert pressure for change on dialysis industry groups. As such, engagement of, and leadership from, the United States kidney care sector in carbon reduction efforts is critical.

Three previous studies from the United Kingdom,⁶ Australia,⁷ and Morocco⁸ have estimated greenhouse gas emissions related to HD. As noted by Sehgal et al., comparing these studies is difficult because they differ in the emissions contributors they include and exclude from consideration. For instance, the Australian and Moroccan studies consider the carbon effect of pharmaceuticals, whereas the UK study and Seghal et al. do not.

However, some emissions contributors were examined by all studies, including energy use. Strikingly, Sehgal et al. showed emissions from energy use of 23.9 kg CO₂-eq per treatment, compared with 5.1–12.1 kg CO₂-eq in the other studies. Although they suggest this might be explained by regional variations in the amount of energy used and the carbon intensity of local energy sources, they also showed that electricity use varied three-fold and natural gas use eight-fold between the facilities they studied, all of which were from the same region.

This highlights a limitation of Sehgal et al.’s study, which is that electricity and natural gas use data were obtained from utility bills rather than from measuring the power usage of the individual pieces of equipment in these facilities. Although this enabled the authors to identify high energy usage, it provided no information on the underlying contributors. They do suggest that, on the basis of manufacturer specifications, the dialysis equipment in their facilities should have used less than a quarter of the total energy consumed, with the rest therefore used by the heating and cooling of each facility, along with other nondialysis-related equipment. The breakdown of this remaining 75% of energy use remains unexplained.

Water was a lesser contributor to carbon emissions than energy in Sehgal et al.’s study, but it is a precious resource and thus requires conserving for its own sake. It is important, then, that Sehgal et al. demonstrated high per-treatment usage (0.6 m³ or 600 L), with five-fold variation between participating facilities. As with energy, however, water data were on the basis of utility bills rather than being measured. This means further work must still be done to understand whether it is the reverse osmosis systems, dialysis machines, or indeed the bathrooms or gardens in these facilities driving water usage.

As we progress from estimating environmental effect in HD to conclusively addressing it, granular resource usage data will be required. Without this, designing and implementing improvement initiatives will be difficult. We suggest that energy and water submetering equipment be installed in all new dialysis facility builds and retrofitted into older facilities where possible. This will not only facilitate future research, but also allow individual dialysis facilities, organizations, and networks to monitor and compare ongoing usage. In turn, this will enable the setting of environmental improvement targets and key performance indicators across the sector and the tracking of improvements (or otherwise).

Seghal et al. also showed that patient and staff transport was responsible for a major portion of carbon emissions (28.3%) across the facilities they studied, and was the third-largest cause of between-facility emissions variation. Similarly, transportation was responsible for 24% and 22% of total emissions in the UK and Moroccan studies, respectively, compared with only 9% in the Australian study. As suggested by Seghal et al. the effect of transportation on emissions is likely to be largely determined by the geographic areas serviced by facilities and the availability of public transport options. Although these are factors are not easily modified by clinicians or indeed dialysis organizations and networks, these groups do have the ability to influence the uptake of home dialysis therapies, which in turn could mitigate emissions from transport.

Importantly, however, there has been very limited study of carbon emissions from either home HD or peritoneal dialysis,⁹ and no study comparing emissions across the three dialysis modalities. Given peritoneal dialysis has a very different resource consumption and waste generation profile to HD, and both home therapies are typically performed more frequently than in-center dialysis, such studies are very much needed.

For all future carbon-emissions research in dialysis, we believe the involvement of environmental scientists is critical to ensure robust data collection and analysis at a fine level of detail, thereby allowing the important question of “could we reduce the effect if we changed?”… Ideally, questions such as this will be asked and answered by a collaboration of clinicians, engineers, equipment manufacturers, industry groups, and more.

Health care needs to be on the pathway to net-zero carbon emissions, not as an aspirational goal, but rather because every kilogram of cumulative carbon emissions results in irreversible warming over a human time scale. In treating patients today, we must ensure we are maintaining intergenerational equity and not harming people in the future. Furthermore, rather than decarbonization coming at a high economic cost, it is instead likely to reduce the cost of treatment through the efficient use of energy and consumables, which in turn will allow redirection of savings to clinical care and better outcomes for patients.

Disclosures

None.

Funding

None.

Author Contributions

K. Barraclough wrote the original draft; and S. McAlister reviewed and edited the manuscript.