Strategies and tactics to reduce the impact of healthcare on climate change: systematic review

Abstract

Objective

To review the international literature and assess the ways healthcare systems are mitigating and can mitigate their carbon footprint, which is currently estimated to be more than 4.4% of global emissions.

Design

Systematic review of empirical studies and grey literature to examine how healthcare services and institutions are limiting their greenhouse gas (GHG) emissions.

Data sources

Eight databases and authoritative reports were searched from inception dates to November 2023.

Eligibility criteria for selecting studies

Teams of investigators screened relevant publications against the inclusion criteria (eg, in English; discussed impact of healthcare systems on climate change), applying four quality appraisal tools, and results are reported in accordance with PRISMA (preferred reporting items for systematic reviews and meta-analyses).

Results

Of 33 737 publications identified, 32 998 (97.8%) were excluded after title and abstract screening; 536 (72.5%) of the remaining publications were excluded after full text review. Two additional papers were identified, screened, and included through backward citation tracking. The 205 included studies applied empirical (n=88, 42.9%), review (n=60, 29.3%), narrative descriptive (n=53, 25.9%), and multiple (n=4, 2.0%) methods. More than half of the publications (51.5%) addressed the macro level of the healthcare system. Nine themes were identified using inductive analysis: changing clinical and surgical practices (n=107); enacting policies and governance (n=97); managing physical waste (n=83); changing organisational behaviour (n=76); actions of individuals and groups (eg, advocacy, community involvement; n=74); minimising travel and transportation (n=70); using tools for measuring GHG emissions (n=70); reducing emissions related to infrastructure (n=63); and decarbonising the supply chain (n=48).

Conclusions

Publications presented various strategies and tactics to reduce GHG emissions. These included changing clinical and surgical practices; using policies such as benchmarking and reporting at a facility level, and financial levers to reduce emissions from procurement; reducing physical waste; changing organisational culture through workforce training; supporting education on the benefits of decarbonisation; and involving patients in care planning. Numerous tools and frameworks were presented for measuring GHG emissions, but implementation and evaluation of the sustainability of initiatives were largely missing. At the macro level, decarbonisation approaches focused on energy grid emissions, infrastructure efficiency, and reducing supply chain emissions, including those from agriculture and supply of food products. Decarbonisation mechanisms at the micro and meso system levels ranged from reducing low value care, to choosing lower GHG options (eg, anaesthetic gases, rescue inhalers), to reducing travel. Based on these strategies and tactics, this study provides a framework to support the decarbonisation of healthcare systems.

Systematic review registration

PROSPERO: CRD42022383719.

Introduction

The direct and indirect human health impacts of climate change have been well documented over the past two decades. The 2015 Paris Agreement and subsequent research have painstakingly established the association between planetary and human health.^{1 2 3 4} However, until recently, less attention has been directed towards the impact healthcare systems have on climate change. This issue is gaining momentum, spurred on by the implementation of the Paris Agreement and the introduction of the UK National Health Service (NHS) net zero carbon strategy.^{5 6}

In essence, healthcare systems occupy a special place; although they are on the frontlines in dealing with climate induced demand for healthcare, they are also major emitters.^{7 8} Healthcare systems must therefore address a predicament: maintenance of high quality care with the resilience and capacity to respond to escalating climate induced demands; and mitigation of their own, substantial contributions to the climate crisis. The action required to manage the complexities of the environmental and societal costs of delivering health services is not trivial. Healthcare systems, notably hospital facilities, are energy intensive, high consumption organisations that produce considerable quantities of waste.^{8 9 10 11} Although there is an irrefutable duty of care to patients and a fundamental commitment to do no harm at the point of delivery, healthcare systems have until recently remained largely unrecognised contributors to the climate crisis.¹² However, healthcare systems are well positioned as environmental stewards to get their own house in order and promote benefits—lowering the collective carbon footprint while simultaneously improving long term health by reducing low value care.

That said, healthcare systems currently lag behind other service sectors in reducing carbon emissions.^{6 13} To be compliant with the Paris Agreement and contribute to the Intergovernmental Panel on Climate Change’s target of limiting global warming to 1.5°C above preindustrial levels, healthcare must play its part, aiming for net zero carbon emissions by 2050.^{13 14 15} There is a need to act in a swift and decisive manner because humanity has a limited window to achieve significant progress in reducing carbon emissions or irreversible changes will exceed the boundaries of the planet.¹⁶

Most estimates place average global emissions from healthcare at about 4.4%, with a country range of 4-8.5%.¹⁵ As confronting as these numbers are, it is possible they will get worse before they get better.¹⁷ More frequent large scale events stimulated by climate change (eg, bushfires, floods, cyclones, heatwaves, and other weather sequelae) will require more carbon inducing care, placing additional, often overwhelming loads on overstretched healthcare systems grappling to contend with the current high burden of chronic diseases, non-communicable diseases (such as cardiovascular diseases), communicable diseases, constant introduction of new interventions and technologies as we discover new ways to manage disease, and ageing populations.^{10 11 18 19} With slow progress on reduction strategies, increased demand on healthcare systems in turn generates further environmental impacts in a vicious cycle (fig 1).

Fig 1.

Positive feedback loop reflecting current association between the healthcare system and climate change. GHG=greenhouse gas

Against that backdrop, we sought to examine the quality and quantity of global evidence on ways in which healthcare systems contribute to climate change and the proposed and implemented ways of reducing the effects of healthcare systems on the climate. Table 1 presents our key terms.

Table 1.

Glossary of terms

Term	Definition
Climate change	Change in global climate patterns and events, particularly in response to human activity resulting in increased carbon dioxide (CO2) and other greenhouse gas (GHG) emissions produced primarily through use of fossil fuels.20
Healthcare system	Organisations, people, infrastructure, and resources involved in promoting, restoring, or maintaining health for individuals, groups, populations, and communities21; this includes but is not limited to delivery of healthcare by public and private health services and also healthcare quality, governance, the health workforce, and medical technologies and products.
Low value care	Care that is of little or no benefit to the patient, such as unnecessary tests or procedures; according to some accounts, currently estimated at 30% of care.22
Micro level of the system	Healthcare system on frontlines where clinicians interface with patients, involving healthcare delivery such as at a clinic, general practice, or hospital department through to an entire hospital or provider facility.23 24
Meso level of the system	Healthcare system at regional or network level; this can range from a health district or hospital network to a provincial or state jurisdiction. Examples include carbon footprint of an entire United States hospital network; an NHS England region; the New South Wales healthcare system25; or an emergency medical services network.26
Macro level of the system	Whole health sector, eg, a national system or collection of national healthcare systems in a World Health Organization region. Examples include emissions attributed to surgery reported at the national level27; carbon budget of the entire Chinese healthcare system11; total carbon footprint of asthma exacerbation across the UK28; or WHO’s Eastern Mediterranean region.
Carbon footprint	GHGs attributed directly or indirectly to an entity, individual, or product, which is expressed as carbon dioxide equivalents (CO2e)29; measurement of total GHG emissions (including CO2, methane, and nitrous oxide) along supply chains11; estimates of level of emissions in CO2e for which a country, business, institution, individual, or another entity is responsible.30
Decarbonisation	Process of reducing CO2 and other GHG emissions from a specific activity, process, or procedure.31
CO2 equivalent	Unit used for comparing different GHGs (eg, methane) based on global warming potential (GWP) through conversion to equivalent amounts of CO2 with the same GWP—usually expressed as million metric tonnes of carbon dioxide (MMTCDE). CO2e for a gas is calculated by multiplying tonnes of the gas by its GWP,32 eg, methane has a GWP of 21, therefore one million metric tonnes of methane is equivalent to 21 MMTCDE. Essentially, CO2e are a measure of GHGs created by an identifiable set of activities; this includes emissions under the control of an emitter and indirect GHG generation from supplied electricity.33
Carbon neutral	Balancing CO2e released into the atmosphere by removing corresponding amount of CO2.34
Net zero	Point at which there is a balance between GHG equivalent emissions and removal of GHG equivalent emissions.34 Balance can be achieved at the product, individual, country, region, or global level.30

Methods

Search strategy and selection criteria

In this review, we assessed the quality and quantity of evidence on ways in which healthcare systems contribute to climate change and the approaches, models, and tools available to decarbonise healthcare systems. The review was prospectively registered on PROSPERO (CRD42022383719) and reported in accordance with the preferred reporting items for systematic reviews and meta-analyses (PRISMA) guidelines (supplementary materials 1 and 2).³⁵

The search strategy was designed by the review team in conjunction with a research librarian, and run across eight electronic databases. Seven of these databases were searched from their respective inception dates to November 2023: Business Source Premier, CINAHL, Cochrane Reviews, Embase, Health Business (EBSCO), Medline, and Web of Science. Scopus was searched from 1990 to November 2023. Supplementary material 3 gives more details on the search terms.

Publications were eligible if they were full text articles in peer reviewed journals. All study designs (eg, case studies, reviews) were included. Reports from authoritative international agencies (eg, World Health Organization, World Bank, United Nations, and Organisation for Economic Co-operation and Development) or government reports at a federal or national level were also eligible if assessed to be of suitable quality. A primary focus on the effects of human healthcare systems on climate change was required. Articles were excluded if they were not available in English, were not published as full papers, or did not have a primary focus on healthcare systems’ effects on or contribution to climate change. Supplementary material 4 presents full inclusion and exclusion criteria.

Records identified in the database search were imported into Rayyan, a screening software tool,³⁶ and duplicates removed. For inter-rater reliability, pilot screening of titles and abstracts was conducted on a set of 200 randomly selected publications and any resulting discrepancies were resolved by the group. In our original review protocol, it was planned that six investigators would screen records in pairs. However, given the large volume of literature identified in the search, 15 investigators screened records in pairs. The full set of titles and abstracts retrieved from searches were then independently double screened for inclusion by investigators working in teams of two or three (EL, LE, EM, GD, CLS, AC, KB-C, SW, GF, RP, LP, HRA, SS, CR, AP). The full texts of the remaining publications were then duplicate screened against the inclusion criteria to ensure consistency in the final set of included texts. Supplementary material 5 presents reasons for exclusion during the full text review. Regular meetings took place to resolve any disagreements by consensus at each step of the process, with JB available as arbiter.

Data analysis

Data extraction and quality appraisal were conducted independently by investigators, and verified during synthesis processes. Four quality appraisal tools were applied to assess the quality of included publications: Joanna Briggs Institute critical appraisal checklist for systematic reviews and research synthesis³⁷; Mixed-Methods Appraisal Tool (MMAT)³⁸; Scale for the quality Assessment of Narrative Reviews (SANRA)³⁹; and Authority, Accuracy, Coverage, Objectivity, Date, Significance (AACODS) tool for evaluation and critical appraisal of grey literature.⁴⁰ After initial appraisal, the quality of each article was assigned to one of four categories: critically low, low, moderate, high. Articles determined to be of critically low quality were excluded from analysis; supplementary materials 6-10 provide further details.

A purpose built data extraction workbook was developed in Microsoft Excel. Data extraction variables included publication details (author, title, year, country of publication), healthcare system level (micro, meso, macro), healthcare system sector (eg, primary care, surgery, mental health), and measurement of climate impact (eg, carbon footprint). A thematic analysis of the extracted data was performed using an inductive approach to identify themes emerging from the literature.⁴¹ One reviewer (SW) first familiarised themselves with all extracted data, and then created a set of six initial broad themes and 24 subthemes using inductive thematic analysis.⁴¹ A subgroup of five reviewers (CLS, KBC, GD, SW, EL) discussed and revised the initial coding of themes to 12 themes. The whole authorship team conducted a review of these broad themes for content and accuracy, and the final nine themes were then reviewed by the senior author (JB). Once the nine broad themes had been developed, a subgroup of reviewers (CLS, KBC, GD, SW, EL, EM) met to code the data within each broad theme into appropriate subthemes, again using an inductive thematic analysis. Two broad categories were also developed to frame the themes as avenues of enacting change (overarching strategies), or specific approaches for reducing GHG emissions in healthcare systems (decarbonisation tactics). The whole authorship team reviewed and approved all categories and subthemes, which were finalised by the senior author (JB). Supplementary materials 11-13 provide details.

Patient and public involvement

This research did not focus on any specific patient population. Although no patients were directly involved in setting the research question or conducting the review, we ensure we have healthcare consumers’ input, expertise and advice on the progress of our research at the intersection of climate change, health, and healthcare, which feeds into and shapes all our work in this area. Healthcare consumers are an integral part of our results dissemination strategy.

Results

Study selection and quality assessment

Electronic database searches identified 33 737 publications, 15 793 of which were duplicates, leaving 17 944 publications for title and abstract screening; 739 studies met the inclusion criteria for full text screening. Of these, 28 were excluded because they could not be retrieved, leaving 711 texts to be screened (fig 2). Two additional papers were identified, screened, and included through backward citation tracking.

Fig 2.

Preferred reporting items for systematic reviews and meta-analyses (PRISMA) diagram

Most of the articles in this review were appraised as high (n=112, 54.6%) or moderate (n=80, 39.0%) quality. Only 13 articles (6.3%) were identified as lower quality but were included because they provided valuable information for the review. Seven studies were considered ineligible based on critically low quality appraisal ratings (supplementary materials 6-10). After full text screening and quality appraisal, 205 publications were included (fig 2).

Study characteristics

Types of studies

Included publications were categorised by study design (table 2): empirical studies; reviews including narrative, scoping, and systematic; non-empirical “narrative descriptive” publications (eg, case studies, reports); and those applying multiple methods. Studies were also categorised based on the level of the healthcare system to which they pertained: micro, meso, or macro (table 2; see table 1 for full definitions).

Table 2.

Overview of key characteristics of included publications (n=205)

Classification	Publications, n (%)
Study designs
Empirical	88 (42.9)
Review	60 (29.3)
Narrative descriptive	53 (25.9)
Multiple methods	4 (2.0)
Healthcare system level*
Micro (eg, frontline clinician level, including healthcare delivery at a clinic or department level to whole hospital or facility level)	83 (35.3)
Meso (regional or network level, eg, health district, hospital network)	31 (13.2)
Macro (whole healthcare system level, including national or global systems)	121 (51.5)

^*If applicable, publications were categorised into more than one category; therefore, the healthcare system level total does not add up to 100%.

Geographical focus

High income countries were over-represented in the included publications (n=101, 69.2%, excluding reviews). The most commonly discussed country was the United States (n=27, 13.2%), followed by the UK (n=21, 10.2%), Australia (n=17, 8.3%), and Canada (n=14, 6.8%). Figure 3 depicts the study setting by country. Many publications (n=30, 14.6%) focused on the macro level without discussing a specific country or region.^{6 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70}

Fig 3.

Setting of included empirical publications by country. Some publications refer to more than one country. Countries were counted for all empirical publications

Publication dates

Most included records were published in recent years (fig 4), with 85.4% (n=175) from 2019 onwards, indicating the growing interest in this topic.

Fig 4.

Included publications by year

Topics covered

A wide range of healthcare related topics were discussed, including surgery (n=36)^{24 27 51 54 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102}; internal medicine specialties (ie, renal services, gynaecological services; n=18)^{23 57 75 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117}; digital health (n=15)^{59 66 71 118 119 120 121 122 123 124 125 126 127 128 129}; primary care (n=15)^{53 56 63 75 130 131 132 133 134 135 136 137 138 139 140}; hospitals (n=14)^{82 92 141 142 143 144 145 146 147 148 149 150 151 152}; supply chains (eg, procurement; n=10)^{42 44 153 154 155 156 157 158 159 160}; oncology (n=8)^{48 58 161 162 163 164 165 166}; pharmaceuticals (n=7)^{45 47 167 168 169 170 171}; diagnostics (n=6)^{52 172 173 174 175 176}; chronic condition treatment (n=5)^{28 61 177 178 179}; workforce (n=5)^{180 181 182 183 184}; pathology services (n=5)^{75 174 185 186 187}; dentistry (n=4)^{60 188 189 190}; critical care (n=3)^{90 191 192}; clinical trials (n=2)^{67 193}; emergency medical services (n=2)^{26 194}; mental health services (n=2)^{195 196}; outpatient care (n=2)^{197 198}; screening (n=2)^{199 200}; travel related to doctors’ medical education (n=1)²⁰¹; and palliative care (n=1).²⁰² Fifty papers examined the system as a whole, taking a macro perspective.^{6 7 11 19 25 43 46 49 50 55 62 64 65 68 69 70 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236}

Scope

Twenty nine (14.1%) publications specifically used terminology that refers to GHG emissions according to three emission scopes defined by the Greenhouse Gas Protocol²³⁷ (direct emissions, indirect emissions from energy use, indirect supply chain emissions).^{7 27 44 45 47 48 50 55 65 68 69 81 85 125 142 152 166 172 184 193 197 200 202 207 209 219 224 229 232} Table 3 presents definitions of the three scopes.

Table 3.

Three scopes of carbon footprint

Scope	Definition
1	Direct emissions from healthcare facilities and healthcare owned vehicles.
2	Indirect emissions from purchased energy sources, including electricity grids, and energy used for steam, cooling, and heating.
3	Indirect emissions from healthcare supply chain, including emissions related to transportation, production, and disposal of medical goods and services.

Source: Definitions derived from Greenhouse Gas Protocol.²³⁷

Measurement of healthcare emissions

Forty eight publications developed or applied specific models or tools to measure healthcare emissions; some publications used several models or tools (table 4). Models included life cycle assessment (LCA; n=27),^{60 76 84 104 107 111 119 122 128 130 141 142 155 162 173 174 175 179 186 190 193 200 209 212 213 223 231} economic input-output LCA (n=2),^{197 212} LCA multiregion input-output models (n=1),²⁵ environmentally extended input-output models (n=6),^{7 11 24 77 213 232} environmentally extended multiregional input-output model (n=4),^{43 44 231 232} and the Bilan Carbone model (n=1).¹⁶⁶ Tools used in the literature included carbon calculators (n=6),^{52 101 115 121 165 176} Carbon Trust recommendations (n=1),²⁶ the Eyefficiency tool (n=1),⁸⁹ the Pollard model (n=1),¹⁸⁸ the Ringelmann smoke chart (n=1),¹⁵⁷ and the US Environmental Protection Agency’s waste reduction model (n=1).¹⁴⁴ Other studies (n=16) used publicly available datasets (eg, UK Department of Business, Energy, and Industrial Strategy conversion factor data; environmental impact data provided by the German Federal Environment Agency), or previously published literature to quantify GHG emissions in their facilities.^{74 83 94 99 110 113 123 126 129 148 161 162 163 168 178 202}

Table 4.

Tools for measuring healthcare emissions from empirical studies

Method	Description	References
Model*
Life cycle assessment (LCA)	Method used to assess environmental impacts of product or service throughout its entire life cycle, from extraction of raw materials to final disposal of product or service, with goal of identifying stages of life cycle that have greatest environmental impact, and to suggest ways to reduce environmental impact of product or service. In practice, boundaries are placed on sources of emissions to facilitate analysis. Studies also used other LCA related models, such as life cycle impact assessments (LCIA), life cycle costs (LCC), and organisational LCA (O-LCA).	60, 76, 84, 104, 107, 111, 119, 122, 128, 130, 141, 142, 155, 162, 173-175, 179, 186, 190, 193, 200, 209, 212, 213, 223, 231
Bilan Carbone	Method of carbon footprint analysis that accounts for all sources of GHG emissions, such as carbon dioxide, methane, nitrous oxide, and other GHGs, from direct and indirect sources.	166
Economic input-output LCA (EIOLCA)	Monetary flows across economic sectors are compiled as part of input-output models. These input-output models are then linked to sector specific environmental data, which form part of a life cycle inventory for evaluating carbon impact of a system.	197, 212
LCA multiregion input-output (LCA-MRIO)	Model that considers economic and environmental interactions between different regions or countries to assess environmental impact of product or service. Applying input-output analysis, interrelations between different economic sectors and their environmental impacts are assessed to estimate total environmental impact of product or service across entire supply chain.	25
Environmentally extended input output (EEIO)	Method that combines economic input-output data of industrial sectors with environmental information to provide comprehensive picture of environmental impacts of economic activities in a country. Examples include national EEIOs, such as the United States EEIO.	7, 11, 24, 77, 213, 232
Environmentally extended multiregional input-output (EE-MRIO)	Model that tracks the flow of economic expenditure and associated carbon emissions between regions and countries. Carbon emission calculations are derived from economic activity.	43, 44, 231, 232
Tools
Carbon calculator (online software)	Software tools that enable individuals, businesses, and organisations to calculate their carbon footprint generated over given period of time. Software uses data inputs provided by user and then applies calculations based on established methods to provide estimated total carbon footprint; this includes calculators developed by US EPA.	52, 101, 115, 121, 165, 176
Carbon Trust recommendations	Data collection tool based on recommendations from the Carbon Trust used to assess energy consumption of emergency medical services systems. The tool uses reports on energy consumption associated with ambulances, vehicles, on site energy usage, and aviation fuel to determine overall carbon footprint.	26
Eyefficiency tool (environmental LCA)	Sustainability benchmarking tool used to estimate GHG emissions and efficiency of surgical processes using data from consumption, travel, and waste. Process involves determining scope, measuring emission sources, and applying emissions factors to determine overall GHG emissions footprint.	89
Pollard model (decision support tool)	Using detailed care pathway approach, this model simulates carbon footprint of service reconfigurations and accurately calculates bottom line implications and associated resource consequences of changes in health sector.	188
Ringelmann smoke chart	Graphical tool used to visually assess density of smoke from combustion source.	157
US EPA’s waste reduction model (WARM V15)	Tool developed by US EPA that provides high level estimates of GHG emissions produced by materials and material management practices; allows comparison of practices for industry and researchers.	144

Models are general method used to measure GHG emissions at a certain level, whereas tools were developed to measure emissions using one or more models.

GHG=greenhouse gas; US EPA=United States Environmental Protection Agency.

^*Some publications used several models.

Key themes emerging from the literature

From the thematic analysis, nine themes emerged across the included publications, which were then grouped into one of two categories: overarching strategies (table 5) or decarbonisation tactics (table 6). The most frequently discussed overarching strategy was the need for effective policies and governance surrounding healthcare sustainability (n=97) and the most common decarbonisation tactic was changing clinical and surgical practices (n=107). Papers discussing numerous themes or subthemes were included in several categories.

Table 5.

Publications in the overarching strategies category, grouped into themes

Theme	Examples	Subtheme and definition		References
Enacting policies and governance (n=97)	Existing sustainability targets include the UK National Health Service’s net zero by 2040 policy,224 and the Paris Agreement to achieve net zero emissions by 2050.58 The Healthier Hospital Initiative was organised collaboratively with Health Care Without Harm, Practice Greenhealth, and the Centre for Health Design, and has since engaged over 1000 hospitals to improve healthcare sustainability in the United States.212 Tax incentives for investment into renewable energy are available through the US Inflation Reduction Act 2022.224	Priority setting and leadership (n=51)	Policy making and priority setting from governments and regulatory bodies relating to climate change in healthcare systems	6, 7, 11, 19, 43, 46, 50, 51, 56, 57, 59, 63-65, 67, 70, 76, 79, 95, 96, 104, 105, 109, 111, 114, 115, 120, 125, 127, 128, 130, 133, 138, 143, 152, 154, 166, 170, 184, 188, 206, 209, 210, 216-218, 220, 221, 231, 232, 234
		Require auditing, evaluation of processes, reporting of progress, and benchmarking across facilities (n=27)	Mandatory reporting of sustainability processes based on standardised metrics	6, 7, 23, 45, 55, 58, 61, 68, 70, 73, 88, 124, 136, 142, 151, 159, 168, 173, 179, 205, 207, 208, 212, 213, 216, 219, 221
		Financial policy (including procurement; n=21)	Incentivisation of sustainable practice through payment models	6, 42-44, 48, 49, 70, 73, 89, 96, 123, 133, 138, 143, 162, 171, 180, 200, 221, 224, 233
		Multisectoral approach including dedicated organisations (n=21)	Collaboration between stakeholders from different sectors and industries to instigate change	7, 23, 56, 69, 73, 89, 96, 103, 112, 117, 162, 191, 192, 201, 203, 205, 210, 217, 220, 221, 233
		Selecting targeted emissions reductions including required use of tools (n=15)	Setting quantifiable and mandatory targets for reducing emissions with required use of standardised tools and metrics	7, 11, 19, 25, 45-47, 58, 70, 71, 89, 136, 193, 218, 219
Changing organisational behaviour (n=76)	Establishment of an Environmental Sustainability Special Interest Group within an Australian renal clinical network. This group surveyed dialysis facilities, and found only 21% had formal sustainability education within their organisation.105 After surveying 11 hospitals in Turkey regarding their sustainability practices, recommendations for organisational change were identified including creating an “environment friendly team” to lead sustainability initiatives, and providing training to all hospital staff related to sustainable practice.80	Education and training (n=59)	Formal education and training programmes or informal knowledge sharing for professional healthcare clinicians and staff	24, 25, 48-50, 57, 63, 65, 68, 69, 73, 75, 77, 79, 81, 83, 86, 91, 99, 102, 103, 110, 112, 114, 116, 127, 130, 133, 135, 137, 138, 142, 143, 150-152, 165, 166, 171, 172, 174, 178-180, 185, 187, 192, 195, 196, 202, 205-207, 210, 212, 220, 228, 233, 235
		Leadership and empowerment (n=26)	Leadership groups and committees supporting decarbonisation actions and empowering change within healthcare organisations	27, 46, 58, 63, 70, 71, 77, 80, 93, 105, 127, 133, 135, 136, 152, 154, 165, 181, 182, 185, 191, 196, 199, 216, 221, 234
Actions of individuals and groups (n=74)	Medical organisations specifically promoting sustainable or low carbon practice, eg, the British Thoracic Society promoting low carbon inhaled therapies.139 Healthcare professionals championing sustainability efforts within their institutions, and also connecting with local communities to share ideas and advocate for change.151 Encouraging patients to work towards improving their health and the health of those around them to reduce health risks and unnecessary use of health services.236	Advocacy (n=45)	Open promotion and systemic advocacy towards addressing climate change in healthcare systems	7, 48-51, 53, 58, 61, 64-67, 71, 81, 85, 86, 88-90, 96, 98, 103, 107, 114, 116, 117, 127, 133-135, 139, 142, 151, 171, 180, 181, 183, 184, 192, 195, 196, 216, 218, 232, 233
		Community involvement and education (n=21)	Educating, empowering, and uniting communities to take action on climate change issues related to the healthcare system	43, 53, 56, 81, 90, 103, 112, 130, 133, 151, 160, 168, 192, 210, 212, 218, 221, 229, 232, 233, 236
		Individual responsibility (n=15)	Individuals, including patients and health professionals, acting independently to minimise their carbon footprint	57, 59, 63, 73, 81, 89, 91, 96, 103, 143, 160, 162, 192, 220, 233
		Benefits of reducing GHG emissions (n=15)	The flow-on effects of reducing GHG emissions, such as reduced disease burden and cost savings	19, 28, 50, 69, 96, 102, 107, 108, 112, 137, 177, 212, 215, 219, 226
Using tools for measuring and monitoring GHG emissions (n=70)	The ENERGY STAR Portfolio Manager is a standardised benchmarking tool for the energy efficiency of buildings. Its development and use has enabled measurement and reporting of over 50 000 metric tons of GHG savings.207 Tools such as life cycle analysis should be used to measure and monitor GHG emissions, and could be incorporated into health technology assessment and metrics for measuring quality of care.58 59 114 164 227 (More detailed examples of tools and methods are in table 4.)	Using tools to improve decision making on procurement and services (n=43)	Using established tools and databases to quantify GHG emissions and inform sustainable improvements	6, 43, 47, 58, 59, 62, 66, 68, 73, 75, 77, 79, 84, 88, 96, 104, 105, 107, 109, 112, 124, 137, 139, 142, 146, 149, 154, 164, 192, 200, 205, 208, 211, 212, 216-218, 224, 225, 227, 230, 231, 234
		Develop tools to assess GHG emissions from different levels (n=25)	Developing new tools that quantify GHG emissions from several sources, and in different contexts	7, 11, 19, 26, 43, 44, 69, 103, 124, 125, 146, 171, 177, 188, 193, 194, 207, 209, 211, 219, 221, 222, 225, 227, 228
		Standardise platforms for reporting or decision making (n=17)	Consolidation of standard and systematic ways of reporting GHG emissions across health and other industries	7, 26, 45, 47, 48, 51, 61, 71, 107, 114, 148, 154, 163, 177, 207, 211, 219

Overarching strategies definition: mechanisms for healthcare system improvement relating to climate change.

GHG=greenhouse gas.

Table 6.

Publications in the decarbonisation tactics category, grouped into themes

Theme	Examples	Subtheme and definition		References
Changing clinical and surgical practices (n=107)	In 2019, an Australian tertiary hospital reported replacing desflurane with low flow sevoflurane, which resulted in carbon emission savings of 270 474 kg carbon dioxide equivalent.83 Wide-awake, local anaesthesia, no tourniquet (WALANT) hand surgery has been found to be more environmentally friendly than traditional hand surgery with sedation owing to reduced waste and use of equipment.78	Replacing identified high GHG practices with lower GHG options (n=80)	Using sustainable or low GHG options when applicable, including healthcare products and processes	11, 23, 27, 28, 43, 48, 54, 57, 61, 63, 68, 71, 73-75, 77-79, 81-87, 90, 91, 93, 94, 96-98, 100, 102, 103, 106, 107, 112-114, 120, 126, 128, 129, 133, 135, 138, 139, 148, 149, 152, 159, 161, 163, 165-168, 170, 171, 173, 174, 177-179, 183, 189, 190, 192, 196, 214, 217, 224, 227, 229, 230, 232-234, 236
		Reducing the amount of low value care (n=41)	Reduction of unnecessary or inefficient tests, procedures, and products	6, 7, 24, 28, 42, 51, 52, 55, 66, 68, 69, 75, 77, 79, 93, 97, 99, 101, 109, 116, 130, 142, 149, 151, 152, 169, 174, 175, 177, 179, 185, 186, 190-192, 205, 212, 214, 224, 229, 233
		Reducing water usage (n=6)	Reduction of water used in clinical and surgical practice	23, 25, 50, 108, 143, 189
Managing physical waste (n=83)	By incorporating the 5R principles (reduce, reuse, recycle, rethink, and research), hospitals can reduce the amount of waste generated in operating theatres.88 Three primary hip operational rooms at a large NHS hospital in the UK were audited and revealed nearly 15% of clinical waste disposal contained clean recyclable items such as cardboard and plastics.95	Waste minimisation approaches (n=71)	Methods related to reducing, reusing, recycling, and repurposing waste	11, 25, 43, 47-51, 54, 55, 57, 60, 61, 63, 67, 72, 73, 77, 78, 80, 82, 84, 85, 88, 91, 95-98, 100-102, 104, 105, 107, 110, 112, 116, 135, 138, 142-144, 149, 151, 152, 154, 155, 159, 160, 169, 171, 177, 183, 189-192, 198, 206, 211, 217-219, 222, 224, 228, 229, 232-234
		Waste disposal (n=29)	Choosing appropriate and low carbon methods of segregating and disposing of healthcare waste	50, 57, 73, 77, 80-82, 86, 88, 89, 93, 95, 107, 114, 143, 144, 150, 151, 153, 157, 160, 177, 183, 199, 206, 210, 223, 232, 236
Minimising travel and transportation (n=70)	Fractionated treatments of radiotherapy might potentially reduce distance travelled by patients because a substantial portion of carbon footprint produced by radiotherapy is from patient travel.161 A prospective randomised control trial investigating follow-up care through digital health app video consultations compared with in-person clinic consultations concluded that telemedicine decreases 11.248 kg of travel related GHG emissions per patient (based on car estimates).94	Using models of care that reduce travel (n=41)	Use of telemedicine, integrated models, or more local services that minimise patient travel	6, 24, 43, 47, 48, 51, 59, 63, 71, 73, 74, 79, 84, 94-96, 105, 109, 119, 121, 122, 124-126, 128, 132, 135, 148, 161, 163, 165, 171, 172, 176, 188, 191, 195, 196, 217, 228, 229
		Access to alternative forms of transportation (n=25)	Promotion and access to shared or active transport options for patients and clinicians	7, 47, 55, 69, 79, 105, 114, 121, 131, 137, 140, 142, 152, 161, 166, 174, 176, 190, 192, 197, 202, 224, 226, 232, 233
		Streamlining medical services (n=14)	Organising medical services and consultations to maximise efficiency and minimise patient travel	61, 79, 86, 93, 127, 131, 132, 148, 156, 161, 166, 176, 188, 193
		Reducing professionals’ travel for conferences and education (n=13)	Reduction of business travel for health professionals and other stakeholders	43, 47, 51, 63, 85, 117, 120, 171, 191, 195, 201, 224, 228
Strengthening infrastructure (n=63)	Carbon emissions produced by diagnostic imaging can be reduced by limiting unnecessary imaging, optimising use of equipment on needs basis (ie, turning off scanners when not in use) and considering standby power usage when obtaining new equipment.175 By adding piping and insulating boiler plant equipment, upgrading lighting and building controls, St Joseph Hospital in Canada was able to avoid 15% of overall GHG emissions and captured this using energy benchmarking.207	Energy efficient buildings (n=40)	Prioritising energy efficiency in the design and infrastructure of healthcare facilities	47, 50, 55, 57, 61, 63, 69, 72, 73, 80, 81, 92, 93, 97, 98, 109, 112, 135, 141-143, 146, 147, 152, 160, 171, 192-194, 207, 210-212, 218, 221, 224, 229, 232, 234, 235
		Energy conservation processes (n=34)	Conserving energy through optimising equipment use and efficiency	26, 27, 43, 57-59, 61, 63, 73, 84, 88, 96, 97, 108, 114, 117, 118, 135, 143, 145, 151, 152, 161, 174, 175, 184, 194, 195, 202, 206, 221, 226, 228, 229
Decarbonising the supply chain (n=48)	Decreasing procurement emissions requires understanding sustainability and the lifecycles a product can go through, in addition to developing a relation with suppliers.89 A large tertiary hospital based in Victoria, Australia has produced a consensus of priorities actions for procurement, including a commitment from suppliers to achieve carbon neutrality, incorporating sustainability clauses in procurement contracts, implementation of reusable equipment when possible and policy changes so that companies provide life cycle assessment of products before purchasing.142	Sustainable production and procurement (n=43)	Considering sustainability in the manufacturing and procurement of healthcare products	7, 25, 42-44, 47, 50, 54-56, 59, 65, 73, 84, 85, 98, 104, 105, 110, 112, 114, 127, 136, 139, 142, 143, 151, 152, 158, 171, 175, 183, 193, 197, 210, 212, 215, 218, 224, 229, 232, 234, 235
		Decreasing energy grid emissions (n=10)	Reduce reliance on non-renewable forms of energy grid supply	7, 26, 50, 87, 89, 158, 194, 204, 212, 232

Decarbonisation tactics definition: tactics for reducing greenhouse gas emissions within healthcare systems.

GHG=greenhouse gas.

Overarching strategies

Policy and governance level initiatives

The need for policies and incentives to promote, guide, monitor, or evaluate changes at the country, state, and facility level was discussed in 97 publications^{6 7 11 19 23 25 42 43 44 45 46 47 48 49 50 51 55 56 57 58 59 61 63 64 65 67 68 69 70 71 73 76 79 88 89 95 96 103 104 105 109 111 112 114 115 117 120 123 124 125 127 128 130 133 136 138 142 143 151 152 154 159 162 166 168 170 171 173 179 180 184 188 191 192 193 200 201 203 205 206 207 208 209 210 212 213 216 217 218 219 220 221 224 231 232 233 234}; 39.2% of the publications discussing this theme were empirical studies. The most common strategy focused on priority setting and leadership from national and international bodies (n=51), such as integrating sustainability into policy or setting long term sustainability goals across organisations.^{6 7 11 19 43 46 50 51 56 57 59 63 64 65 67 70 76 79 95 96 104 105 109 111 114 115 120 125 127 128 130 133 138 143 152 154 166 170 184 188 206 209 210 216 217 218 220 221 231 232 234} This was followed by publications (n=27) that suggested mandating evaluations, benchmarking, and reporting of the carbon footprint for healthcare organisations to strengthen the evidence base for future decision making.^{6 7 23 45 55 58 61 68 70 73 88 124 136 142 151 159 168 173 179 205 207 208 212 213 216 219 221} An equal number of publications (n=21 each) highlighted using financial policy and incentives for sustainable healthcare processes^{6 42 43 44 48 49 70 73 89 96 123 133 138 143 162 171 180 200 221 224 233} and the need for a multisectoral approach to mitigate the carbon footprint of healthcare,^{7 23 56 69 73 89 96 103 112 117 162 191 192 201 203 205 210 217 220 221 233} such as collaborations between large healthcare organisations. Fifteen publications recommended setting targeted emissions reductions.^{7 11 19 25 45 46 47 58 70 71 89 136 193 218 219}

Organisational behaviour change

Organisational change to develop a future ready workforce was dealt with in 76 publications.^{24 25 27 46 48 49 50 57 58 63 65 68 69 70 71 73 75 77 79 80 81 83 86 91 93 99 102 103 105 110 112 114 116 127 130 133 135 136 137 138 142 143 150 151 152 154 165 166 171 172 174 178 179 180 181 182 185 187 191 192 195 196 199 202 205 206 207 210 212 216 220 221 228 233 234 235} Roughly 30.3% of the publications discussing this theme were empirical studies. Ongoing education and training within healthcare organisations and tertiary institutions was identified in 59 publications as a way of ensuring the workforce is aware of and equipped for sustainability challenges.^{24 25 48 49 50 57 63 65 68 69 73 75 77 79 81 83 86 91 99 102 103 110 112 114 116 127 130 133 135 137 138 142 143 150 151 152 165 166 171 172 174 178 179 180 185 187 192 195 196 202 205 206 207 210 212 220 228 233 235} Publications discussing organisational leadership focused on implementing multidisciplinary teams, “green champions,” and other organisational role models to manage sustainability at different levels of the healthcare system and empower employee decision making (n=26).^{27 46 58 63 70 71 77 80 93 105 127 133 135 136 152 154 165 181 182 185 191 196 199 216 221 234}

Individual and group action

The literature outlined several important roles for individuals and groups in reducing healthcare’s carbon footprint (n=74)^{7 19 28 43 48 49 50 51 53 56 57 58 59 61 63 64 65 66 67 69 71 73 81 85 86 88 89 90 91 96 98 102 103 107 108 112 114 116 117 127 130 133 134 135 137 139 142 143 151 160 162 168 171 177 180 181 183 184 192 195 196 210 212 215 216 218 219 220 221 226 229 232 233 236}; 20.3% of the publications discussing this theme were empirical studies. Promoting change through advocacy, lobbying, and open support of sustainability initiatives was the focus of 45 publications,^{7 48 49 50 51 53 58 61 64 65 66 67 71 81 85 86 88 89 90 96 98 103 107 114 116 117 127 133 134 135 139 142 151 171 180 181 183 184 192 195 196 216 218 232 233} including the need for community involvement and education when enhancing the sustainability of healthcare practices (n=21).^{43 53 56 81 90 103 112 130 133 151 160 168 192 210 212 218 221 229 232 233 236} Several publications emphasised individual responsibility as a key factor in making change (n=15), particularly through engaging with low carbon healthcare and living an active lifestyle.^{57 59 63 73 81 89 91 96 103 143 160 162 192 220 233} Additionally, 15 articles identified benefits of reducing GHGs in the health system, which included financial savings and reducing disease burden.^{19 28 50 69 96 102 107 108 112 137 177 212 215 219 226}

Using tools for measuring and monitoring GHG emissions

The impetus to develop, use, and standardise tools for measuring and monitoring GHG emissions was a prominent theme (n=70 publications).^{6 7 11 19 26 43 44 45 47 48 51 58 59 61 62 66 68 69 71 73 75 77 79 84 88 96 103 104 105 107 109 112 114 124 125 137 139 142 146 148 149 154 163 164 171 177 188 192 193 194 200 205 207 208 209 211 212 216 217 218 219 221 222 224 225 227 228 230 231 234} Nearly 37.1% of the publications discussing this theme were empirical studies. Evaluative tools, along with existing models such as LCA, can be used to improve decision making on procurement and services, and was a feature of 43 publications.^{6 43 47 58 59 62 66 68 73 75 77 79 84 88 96 104 105 107 109 112 124 137 139 142 146 149 154 164 192 200 205 208 211 212 216 217 218 224 225 227 230 231 234} Developing tools to measure GHG emissions and monitor the movement of materials was discussed in 25 publications.^{7 11 19 26 43 44 69 103 124 125 146 171 177 188 193 194 207 209 211 219 221 222 225 227 228} Several publications (n=17) emphasised the need for standardised platforms for reporting and comparing the sustainability of products or processes to enable healthcare organisations to make informed decisions about their current practices.^{7 26 45 47 48 51 61 71 107 114 148 154 163 177 207 211 219}

Decarbonisation tactics

Changing clinical and surgical practices

Changing clinical and surgical practices as a solution for reducing healthcare’s carbon footprint was advocated in 107 publications^{6 7 11 23 24 25 27 28 42 43 48 50 51 52 54 55 57 61 63 66 68 69 71 73 74 75 77 78 79 81 82 83 84 85 86 87 90 91 93 94 96 97 98 99 100 101 102 103 106 107 108 109 112 113 114 116 120 126 128 129 130 133 135 138 139 142 143 148 149 151 152 159 161 163 165 166 167 168 169 170 171 173 174 175 177 178 179 183 185 186 189 190 191 192 196 205 212 214 217 224 227 229 230 232 233 234 236}; 37.4% of the publications discussing this theme were empirical studies. This solution included replacing high GHG practices with lower GHG options (n=80),^{11 23 27 28 43 48 54 57 61 63 68 71 73 74 75 77 78 79 81 82 83 84 85 86 87 90 91 93 94 96 97 98 100 102 103 106 107 112 113 114 120 126 128 129 133 135 138 139 148 149 152 159 161 163 165 166 167 168 170 171 173 174 177 178 179 183 189 190 192 196 214 217 224 227 229 230 232 233 234 236} such as choosing anaesthetic gas types and systems with lower carbon footprints.^{54 71 73 81 83 85 86 96 100 229} This advocation was also seen in primary care settings (n=6).^{63 130 133 135 138 139} As a way of minimising healthcare usage, and thus reducing healthcare’s carbon footprint, 41 publications reinforced the need to assess clinical practices for low value care.^{6 7 24 28 42 51 52 55 66 68 69 75 77 79 93 97 99 101 109 116 130 142 149 151 152 169 174 175 177 179 185 186 190 191 192 205 212 214 224 229 233} This included identifying and reducing unnecessary processes or procedures, minimising drug overprescription, or encouraging a preventative care approach to reduce the need for health services. Reducing water usage during procedures, disinfection and sterilisations, laundry, recycling water, and practising water efficiency were also suggested in six publications.^{23 25 50 108 143 189}

Managing physical waste

Dealing with waste was a focus of 83 publications^{11 25 43 47 48 49 50 51 54 55 57 60 61 63 67 72 73 77 78 80 81 82 84 85 86 88 89 91 93 95 96 97 98 100 101 102 104 105 107 110 112 114 116 135 138 142 143 144 149 150 151 152 153 154 155 157 159 160 169 171 177 183 189 190 191 192 198 199 206 210 211 217 218 219 222 223 224 228 229 232 233 234 236}; 32.5% of the publications discussing this theme were empirical studies. This theme included waste minimisation methods (n=71),^{11 25 43 47 48 49 50 51 54 55 57 60 61 63 67 72 73 77 78 80 82 84 85 88 91 95 96 97 98 100 101 102 104 105 107 110 112 116 135 138 142 143 144 149 151 152 154 155 159 160 169 171 177 183 189 190 191 192 198 206 211 217 218 219 222 224 228 229 232 233 234} such as reducing, reusing, and recycling medical equipment, and using reusable alternatives over disposable items. Several publications highlighted the need for effective segregation and disposal of waste (n=29),^{50 57 73 77 80 81 82 86 88 89 93 95 107 114 143 144 150 151 153 157 160 177 183 199 206 210 223 232 236} including choosing disposal methods such as incineration that help minimise carbon emissions.

Minimising travel and transportation

Patient or clinical travel and transportation was emphasised as an important and addressable source of carbon emissions in 70 publications.^{6 7 24 43 47 48 51 55 59 61 63 69 71 73 74 79 84 85 86 93 94 95 96 105 109 114 117 119 120 121 122 124 125 126 127 128 131 132 135 137 140 142 148 152 156 161 163 165 166 171 172 174 176 188 190 191 192 193 195 196 197 201 202 217 224 226 228 229 232 233} Some 42.9% of the publications discussing this theme were empirical studies. Using carbon efficient models of care was seen as a method to reduce emissions (n=41), for example, telemedicine or home based treatment options.^{6 24 43 47 48 51 59 63 71 73 74 79 84 94 95 96 105 109 119 121 122 124 125 126 128 132 135 148 161 163 165 171 172 176 188 191 195 196 217 228 229} A further 25 publications discussed alternative forms of transportation (eg, car sharing, buses, trains, and active travel such as cycling) for patients and clinicians.^{7 47 55 69 79 105 114 121 131 137 140 142 152 161 166 174 176 190 192 197 202 224 226 232 233} Streamlining medical services, such as emergency service coordination, redesigning schedules for clinicians to enhance ride share and closing sites with lower use, was discussed in 14 publications.^{61 79 86 93 127 131 132 148 156 161 166 176 188 193} Several papers (n=13) also highlighted reducing professionals’ travel (eg, clinicians, researchers, industry partners) for conferences and education as a decarbonisation method.^{43 47 51 63 85 117 120 171 191 195 201 224 228}

This theme had the most implementation evaluations. These studies often highlighted reductions in carbon emissions resulting from teleconferencing or telemedicine,^{122 163 176} and they also tended to be of high quality.

Strengthening infrastructure

Sixty three papers recommended improvements to healthcare infrastructure^{26 27 43 47 50 55 57 58 59 61 63 69 72 73 80 81 84 88 92 93 96 97 98 108 109 112 114 117 118 135 141 142 143 145 146 147 151 152 160 161 171 174 175 184 192 193 194 195 202 206 207 210 211 212 218 221 224 226 228 229 232 234 235}; 33.3% of the publications discussing this theme were empirical studies. Improving the energy efficiency of healthcare facilities such as hospitals was a feature of 40 publications.^{47 50 55 57 61 63 69 72 73 80 81 92 93 97 98 109 112 135 141 142 143 146 147 152 160 171 192 193 194 207 210 211 212 218 221 224 229 232 234 235} For example, consideration of design features such as insulation and passive shading options can help reduce reliance on electrical heating and cooling, and therefore reduce GHG emissions. Thirty four publications advocated for using energy conservation measures—such as turning off appliances when not in use—to help minimise carbon emissions at a facility level.^{26 27 43 57 58 59 61 63 73 84 88 96 97 108 114 117 118 135 143 145 151 152 161 174 175 184 194 195 202 206 221 226 228 229}

Decarbonising healthcare supply chains

Forty eight papers proposed decarbonising the supply chains as a way of reducing GHG emissions.^{7 25 26 42 43 44 47 50 54 55 56 59 65 73 84 85 87 89 98 104 105 110 112 114 127 136 139 142 143 151 152 158 171 175 183 193 194 197 204 210 212 215 218 224 229 232 234 235} Roughly 39.6% of the publications discussing this theme were empirical studies. Numerous publications (n=43) focused on ways to reduce the carbon footprint by deploying sustainable production and procurement mechanisms within supply chains, such as using sustainable or local sources of food supply, or using supply chains with low emissions.^{7 25 42 43 44 47 50 54 55 56 59 65 73 84 85 98 104 105 110 112 114 127 136 139 142 143 151 152 158 171 175 183 193 197 210 212 215 218 224 229 232 234 235} Decarbonising the sources of energy for healthcare by using renewable sources rather than fossil fuel sources and monitoring energy grid consumption was noted in 10 publications.^{7 26 50 87 89 158 194 204 212 232}

Discussion

Key areas for policy and practice

From our findings we developed a rich picture of the strategies and tactics for reducing the emissions produced by healthcare systems (fig 5). This schematic figure is not an exhaustive example, but rather illustrates how these strategies and tactics to reduce emissions produced by healthcare systems are interconnected and affect flow within and outside of the healthcare system. Authorisation to proceed starts at the top, with effective policies, governance, and high level leadership. Measurement of GHGs through recognised and increasingly sophisticated tools is necessary, and now more pronounced, as is the use of tools and frameworks to report on progress, and make decisions about alternative, greener options. Tools and frameworks for use at micro, meso, and macro levels are now more widely available. Several options for individuals and groups to orchestrate change and improvement, typically targeted at organisations such as hospitals, care facilities for older people, and general practices, have been generated.

Fig 5.

Strategies and tactics for reducing healthcare system greenhouse gas (GHG) emissions according to healthcare system level: micro (eg, frontline clinician, including healthcare delivery at a clinic, department, whole hospital, or facility), meso (regional or network level, eg, health district, hospital network), macro (whole healthcare system, including national or global systems)

On decarbonisation specifically, attention was focused on modifying medical, surgical, and other clinical practices, reducing physical waste, and minimising transportation costs or eliminating travel altogether through new clinical models such as virtual care, telehealth, or telemedicine.^{24 59 63 81 94 97 120 121 126 165 172 196} When enacting decarbonisation tactics that affect patient care, it is vital that lower carbon alternatives provide equal or improved patient outcomes compared with current practices and be acceptable to and supported by patients.¹⁰⁵ Other specific activities for reducing carbon emissions and greening healthcare, directly or in those services indirectly controlled by healthcare systems, include building new or remediating existing healthcare infrastructure and reducing the carbon footprint along healthcare supply chains. Given the large contribution of supply chain emissions and the high potential for reducing emissions,²³⁸ it is surprising that supply chain decarbonisation approaches, including those related to food supply, were less frequently discussed than others. Another surprising lack of attention was discussion at the intersection of universal health coverage and healthcare system decarbonisation because reducing the need for care through primary prevention would have a direct impact on carbon emissions by lowering demand.^{49 140}

A way forward

Overall, a framework of useful activities to substantially reduce the carbon budget of healthcare is now available, and represented in figure 5. Although this framework provides a platform to assess these activities, determining which elements would be most effective, and under what conditions, will partly depend on the characteristics of the individual healthcare system.⁹⁶ Barriers to implementing sustainable practice into healthcare systems might also exist, including patients accepting changes, cost and funding mechanisms, attitudes towards change, and workforce capacity in an already overburdened system. As these decarbonisation approaches are put into practice, any barriers and potential enablers should be explored further.

The included literature was mainly from high income countries. However, we found some examples of lower-middle income country healthcare systems and organisations taking the lead in advancing the overarching strategies identified, such as developing tools to measure emissions (eg, the Aga Khan Health Services (AKHS) freely available tool),⁴⁴ and the decarbonisation tactics, such as reducing physical waste and emissions from waste disposal (eg, Snigdha and colleagues)¹⁵⁸ and improving the environmental sustainability of clinical practices.^{111 156 199} Some of the decarbonisation tactics might be more challenging to implement in lower-middle income country healthcare systems, such as strengthening infrastructure.⁹⁶

Regardless of the setting, having an overarching set of mechanisms and approaches might be useful and highlights the need for a multipronged approach, including how we measure emissions and progress, especially given the range of measuring tools that have emerged. The actions include effective governance and supportive policies, appropriate financing, and strengthening infrastructure and service delivery.²⁰ Collaborative, international, multisectoral leadership is needed to learn from progress elsewhere and to localise and prioritise strategies and support direct actions.

Achieving a step change reduction in emissions is a socially and ethically responsible imperative. In principle, it is feasible and beneficial to roll out these strategies and tactics at scale to achieve systemic change—this is central to the mission of healthcare providers: to do no or less harm, to alleviate suffering, and to improve the health status of the population. Healthcare can set an example for other industries and sectors as responsible environmental stewards. This review suggests that healthcare systems and professionals have recognised their duty and are embracing the necessary changes, but there is a long way to go.

Strengths and limitations

The comprehensive analysis of 18 years of studies, frameworks, and tools that assess and quantify the carbon footprint of healthcare is a key strength of the paper, as is the development of a new framework. Previous reviews did not appear to focus sufficiently on mitigation strategies or use the systematic review methodology.

Our review was limited to papers and authoritative reports written in English, and so some papers might have been missed. However, eight databases were searched, therefore the likelihood of new themes being identified is low. Most of the included literature discussed healthcare systems in high income countries, which is a well known bias. This bias can create challenges when extrapolating identified strategies to low-middle income countries. However, the included papers revealed that healthcare systems in low-middle income countries are undertaking similar strategies to high income countries. A review specifically focused on healthcare systems in low-middle income countries, and the ethical implications of climate change and decarbonisation tactics, is warranted. Inductive approaches, as in this study, allow flexibility and deeper understanding of the data, but introduce issues of inter-rater reliability and bias. Our review used 19 investigators for screening, extraction, synthesis, and interpretation to minimise bias.

Conclusions

We reviewed the literature to generate strategies and tactics for reducing the impact of healthcare systems on climate change. Implementing these strategies should enable substantial progress towards reducing healthcare’s carbon footprint and support a greener health sector powered by renewable sources. However, healthcare systems would need to adapt these approaches to their context and needs to maximise their effects. Healthcare can lead the way in shifting to net zero by 2050, and by doing so we can reduce the burden of patients who would otherwise need more care as a consequence of a warming world.

What is already known on the topic

• The carbon footprint of healthcare systems has been estimated at about 4.4% of global emissions, but comprehensive reviews investigating mitigation are lacking

• One review took a global approach when examining the environmental impact of healthcare systems; however, this study was not peer reviewed and data were limited to one year

• Another study discussed the sustainability of healthcare at a global level, however it was not conducted as a systematic review, and the methods used to evaluate and collate data were unclear

What this study adds

• This review includes 18 years of studies, frameworks, and tools assessing the carbon footprint of healthcare systems, and the steps taken to measure and reduce these impacts

• Overarching strategies and specific tactics, models, and tools were identified that could be used to decarbonise healthcare systems, aiming to reach net zero emissions by 2050

Web extra.

Extra material supplied by authors

Web appendix 1: Supplementary material 1-4

Web appendix 2: Supplementary material 5-13

Ethics statements

Ethical approval

No individual participant data are included in this manuscript. All data are aggregated from existing publications. Ethics approval was therefore not required. We attest that we have obtained appropriate permissions and paid any required fees for use of copyright protected materials.

Data availability statement

All datasets generated and analysed for this study, including the search strategy, list of the included and excluded studies, data extracted, analysis plans, and quality assessment are available in the article or supplementary material and upon request from the corresponding author. No individual participant data were used for this study.