Carbon Footprint of General, Regional, and Combined Anesthesia for Total Knee Replacements

Forbes McGain, F.A.N.Z.C.A, F.C.I.C.M., Ph.D.; Nicole Sheridan, F.A.N.Z.C.A.; Kasun Wickramarachchi, B.Sc., M.P.H., M.D.; Simon Yates, M.D., Brandon Chan, M.B.B.S.; Scott McAlister, B.Sc., P.grad., Dip.Sci., M.Eng.Sci.


Anesthesiology. 2021;135(6):976-991. 

In This Article


The carbon footprints of anesthesia for a knee replacement were similar for general, spinal, and combination approaches, with significant overlap between the CIs. There was considerable within-group variation for general and combination anesthesia (a twofold difference in minimal-maximal carbon dioxide equivalent emissions), but only 50% difference for spinal anesthesia. The three major components of carbon dioxide equivalent emissions across all groups were (with approximations) single-use equipment (20 to 25%, mainly plastics), electricity for the patient air warmer (15%), and pharmaceuticals (8%). Carbon dioxide equivalent emissions from sevoflurane use for general anesthesia (32% total) and combination anesthesia (17% total) were considerable. Carbon dioxide equivalent emissions for cleaning reusable equipment were more than 25% total for spinal, and 20% for combined anesthesia. Oxygen use was about 15% of carbon dioxide equivalent emissions for spinal anesthesia. Importantly, the duration of anesthesia was 20% longer for spinal versus general anesthesia. Procedure duration contributes to carbon dioxide equivalent emissions, particularly electricity for the air warmer.

Inhalational anesthesia is known to have higher carbon dioxide equivalent emissions than total intravenous anesthesia.[35,36] For general anesthesia, the use of low flow (minimum 6 ml liquid sevoflurane/h) rather than total intravenous anesthesia increased the carbon dioxide equivalent emissions by 1.2 kg carbon dioxide equivalent emissions/h. There is, however, sparse evidence comparing the carbon footprint of general and spinal anesthesia.[13,37] Spinal anesthesia had a high carbon footprint, partially attributable to cleaning reusable equipment and compression of liquid oxygen, the carbon dioxide equivalent emissions for which were elevated due to the electricity mix of 75% brown coal for Melbourne, Australia. It is unclear internationally what standard oxygen administration is during spinal anesthesia, but flow rates of greater than 6 l/min may be atypical. For cleaning reusable equipment, we assumed worst case steam sterilizer efficiency,[25,27] recognizing that potential efficiency improvements[25,38] could reduce carbon dioxide equivalent emissions by 0.5 kg carbon dioxide equivalent emissions/h just for anesthesia alone. The modeled carbon dioxide equivalent emissions for cleaned reusables in Australia are similar to China, but double the United States, and quadruple Europe/United Kingdom, because of different energy mixes.[15]

Our small, single-center, prospective, nonrandomized, observational, unblinded study has limitations, which makes comparisons between the anesthetic groups and between countries uncertain. We did not prescribe anesthetic choice, and we limited our convenience sample to 30 patients having one operation type in Australia. We aimed to provide a life cycle assessment of three anesthetic approaches to a total knee replacement, but we caution comparison between the three groups. A prospective study powered appropriately would be a considerable undertaking and of limited benefit given the initial hypothesis posed by this study.

We acknowledge anesthetic practice variability, particularly choice of anesthetic gases with high global warming potential.[11] Use of desflurane and nitrous oxide in our small study could skew group results markedly (e.g., greater than100 kg carbon dioxide equivalent emissions for either nitrous oxide or desflurane use).[13] We chose to exclude the one patient receiving nitrous oxide as the relative carbon dioxide equivalent emissions from using nitrous oxide compared with sevoflurane/total intravenous anesthesia/spinal anesthesia are very high, making intergroup comparison difficult.

Comparisons between the amount of equipment/drugs/gases are influenced by the duration of the operation. Many items have greater use in the first hour (induction, drug administration, spinal anesthesia) than for subsequent hours. Nevertheless, other environmental effects are more closely dependent upon duration (electricity for the air warmer and scavenging), carbon dioxide absorbent use, and oxygen use.

We excluded orthopedic surgery and all operating room heating/ventilation/air conditioning carbon dioxide equivalent emissions, focusing solely upon anesthesia. Anesthetic breathing circuits were changed weekly,[28,39] a practice common in Australia,[40] Germany,[41] and elsewhere. Reusable laryngoscope blades, handles, face masks, and surgical gowns were used.[15] We averaged the carbon dioxide equivalent emissions for all 20 drugs studied by Parvatker et al.,[31] using this average for unstudied drugs (cefazolin, paracetamol, and tranexamic acid).[31] Drugs given in relatively large quantities (cefazolin) dominated the pharmaceutical carbon dioxide equivalent emissions. Cardboard/paper was routinely separated preoperatively.

Avoiding the use of desflurane and nitrous oxide is only the beginning of actions that anesthetists can undertake to reduce their workplace carbon footprint. The fuel efficiency of the average U.S. car is 0.40 kg carbon dioxide equivalent emissions/mile, so in our study, the average anesthetic carbon contribution (17 kg carbon dioxide equivalent emissions) is like driving 42 miles (without desflurane or nitrous oxide). Several activities can safely reduce the anesthetist's carbon footprint. For spinal anesthesia, reducing O2 flows from 10 l to 6 l/min reduces driving by 1 mile/h. For general anesthesia, reducing fresh gas flow with sevoflurane by 1 l/min saves 3 miles/h. Replacing 1 l/min fresh gas flow sevoflurane with total intravenous anesthesia saves another 3 miles/h. Using the minimum plastic and glass use will reduce the carbon dioxide equivalent emissions 1 kg carbon dioxide equivalent emissions/h, equaling saving 3 miles/h. Converting from Australia's electricity mix to Europe's for spinal procedures will save 2 kg carbon dioxide equivalent emissions, equaling 5 miles/h. When combining these mentioned carbon sparing activities, you have halved the miles driven for the 3-h anesthetic.

Decreasing the carbon footprint of some activities is challenging; a minimum of pharmaceuticals and equipment are required. Further, anesthesiologists cannot change the carbon intensity of electricity, although we can advocate.[13] The use of renewable energy decreases the carbon dioxide equivalent emissions associated with cleaning reusable equipment, with promising plans locally for Victorian electricity generation.[42] For the European Union/United Kingdom/U.S. anesthesiologist, moving from single-use to reusable anesthetic equipment right now will have financial and environmental benefits.[15] Our study quantifies carbon dioxide equivalent emissions of individual areas of anesthesia practice. We encourage cognizance of one's carbon footprint, emphasizing that instigating multiple, seemingly small changes in our workplace patterns is the best path to low carbon anesthesia.