Evaluating the Environmental Impact of Clinical Research: A Full Life Cycle Analysis of a French Academic Randomised Clinical Trial

ABSTRACT

Climate change poses the greatest threat to human health in the 21st century. The healthcare sector contributes approximately 5% of global greenhouse gas emissions and has a significant environmental impact. Although clinical trials are crucial for identifying effective and safe treatments and preventing disease, their environmental impact is poorly documented. Our study aimed to assess the environmental impact of a publicly funded, academic clinical trial by adapting life cycle assessment (LCA) methodology to clinical research.

We performed a retrospective, simplified, full LCA using the EF 3.0 methodology on a prospective, double‐blind, randomised controlled neurosurgery trial. The trial included 202 patients at 18 university hospitals throughout France. To identify hotspots of interest, 16 impact indicators and their combination into a single score were evaluated. The results showed that climate change (or greenhouse gas emissions) was the most important indicator, accounting for almost 30% of the single score. Greenhouse gas emissions were estimated at 31.6 t of carbon dioxide equivalent. The next most important were resource use of fossils (24%), resource use of minerals and metals (12%), and particulate matter emissions (8%). The main hotspots identified were patient transport and travel by clinical research assistants for source data verification.

In conclusion, by using a full LCA approach, our study confirms that conducting a clinical trial has a substantial environmental impact, particularly with regard to greenhouse gas emissions. The main hotspots identified were related to patient transport and clinical research assistants' travel.

Trial Registration: The SUCRE study (Treatment of Chronic Subdural Hematoma by Corticosteroids: A Prospective Randomised Study)—clinicaltrials.gov identifier: NCT02650609.

Abbreviations

CRA

clinical research assistant

CTU

clinical trial unit

EF

environmental footprint

GHG

greenhouse gas

ISO

International Organization for Standardization

LCA

life cycle assessment

LCI

life cycle inventory

LCIA

life cycle impact assessment

NCT

National Clinical Trial

NIHR

National Institute for Health and Care Research

PEF

product environmental footprint

Pts

points

tCO₂e

tonnes of carbon dioxide equivalent

1. Introduction

Climate change is the leading global health threat of the 21st century. Modelling by the Intergovernmental Panel on Climate Change suggests that average global temperatures could rise by up to 4.4°C by 2100. This would pose an ‘existential threat’ to humanity if current greenhouse gas (GHG) emission trends continue. In any case, the average temperature is expected to rise by 2°C by 2050 because of the existing carbon emissions and the lack of adequate adaptation and actions of socio‐economic stakeholders worldwide [1]. The healthcare sector has a substantial environmental impact. It contributes approximately 5% of global GHG emissions and generates significant waste, making it a major contributor to the overall environmental footprint of modern societies [2]. In this context, the health system should prepare for two critical shifts: (1) evolving care demand, driven by environmental changes (e.g., rising heatwaves and air pollution) altering population health patterns, and (2) transforming care delivery, adapting to the double carbon constraint: reducing carbon intensity by cutting emissions from healthcare operations (energy, transport and waste) and substituting non‐renewables (fossil‐fuel‐dependent resources such as plastics and chemicals) with sustainable alternatives.

Although clinical trials are essential for identifying effective and safe treatments and preventing disease, they also have a significant environmental impact. In 2021, Adshead et al. [3] estimated that the 350 000 clinical trials registered on ClinicalTrials.gov emitted 27.5 million tonnes of carbon dioxide equivalent (tCO₂e), based on the results of a few pilot studies [4, 5, 6]. More recently, a few studies have examined carbon footprints in clinical trials in both publicly funded academic [7, 8] and pharmaceutical industry [9, 10, 11] settings. Two reviews of the literature were recently published, listing all the studies that have estimated the carbon emissions associated with conducting clinical trials [12, 13]. These reviews demonstrate that the carbon emissions reported in these studies can vary significantly (from 17.6 to 2498 tCO₂e). Similarly, the differing scope of what is considered to be the clinical trial, the use of different emission factors and variability in the definition and presentation of the activities with the highest emissions within a trial (‘hotspots’) make relevant comparisons between trials difficult. However, certain hotspots do appear to recur, such as research team or patient transport, although the proportion of the total emissions they represent varies greatly between trials.

Life cycle assessment (LCA) is a standardised method of assessing the potential environmental impacts of products and services, taking into account all processes and environmental factors throughout their life cycles. Our work aimed to adapt the LCA method to a Phase 3 academic clinical trial to evaluate its environmental impacts. This work had two objectives: to conduct a thorough evaluation of the environmental impacts of a publicly funded, academic clinical trial (not just carbon emissions) for the first time and to pinpoint the primary factors contributing to these impacts to explore potential mitigation strategies.

2. Methods

2.1. Life Cycle Analysis

When we started working on this project, there was no publicly available guidance on how to evaluate the environmental impact of a clinical trial. Therefore, we used the product environmental footprint (PEF) method, developed by the European Commission, which provides harmonised rules for conducting a reliable and transparent assessment of the environmental impacts of products and organisations [14, 15, 16]. The PEF method relies on the LCA, also known as life cycle analysis, an iterative methodology for assessing environmental impacts associated with all the stages of the life cycle of a commercial product, process or service. The International Organization for Standardization (ISO) provides standards for LCA in ISO 14040 and 14044. These standards describe the four main phases of an LCA:

1. Definition of goal and scope: Main methodological choices are made in this phase as definition of the functional unit, identification of the system boundary (activities and stakeholders), main assumptions and limitations.

2. Life cycle inventory (LCI): Involves the data collection and the calculation procedure for the quantification of inputs and outputs of the studied system; these concern energy, raw materials and other physical inputs, products and waste, and emissions to air/water/soil.

3. Impact assessment: LCI results are associated with environmental impact categories and indicators through the life cycle impact assessment (LCIA) method, which first classifies emissions into impact categories and then characterises them into common units.

4. Interpretation: Results are interpreted, and the most relevant impact categories, life cycle stages, processes and elementary flows are identified (hotspots).

2.2. Transposition of the LCA to Clinical Trial Process

2.2.1. Definition of Goal and Scope: Life Cycle Mapping

We chose to assess the environmental impact of the SUCRE study (Treatment of Chronic Subdural Hematoma by Corticosteroids: A Prospective Randomised Study—NCT02650609), sponsored by Rennes University Hospital (France). This study was selected because it was representative of a complete, Phase 3 French publicly funded, academic‐sponsored clinical trial with available tracking data. This study was a prospective, double‐blind, randomised controlled trial comparing methylprednisolone with placebo in the treatment of chronic subdural haematoma without clinical and/or radiological signs of severity. Characteristics of the protocol have been published elsewhere [17]. The study included 202 patients in 18 university hospitals in France. The time frame was 9 years from funding application (2014) to submission for publication of the study results (2023).

To assess the LCA of any product, process or service, the entire system must first be established and understood. To achieve this, a series of information‐gathering meetings were held with the clinical trial project team. These meetings enabled life cycle mapping of the clinical trial and definition of the LCA boundaries. The different process steps were defined as follows: study design, regulatory procedures, logistical set‐up, conduct of the study, and closing and archiving (see Figure 1a). Eventually, the functional unit was defined as ‘Carrying out a multi‐centre, randomized academic trial with drugs’. The core trial activities (see Table 1) and the stakeholders (see Figure 1a) were subsequently defined.

FIGURE 1.

(a) Life cycle mapping and life cycle inventory of a clinical trial. (b) EF 3.0 life cycle impact assessment method.

TABLE 1.

Core trial activities included in the LCA.

Core trial activity	Exclusions/assumptions
Research personnel activity: Utility consumption
Data collection and exchange (digital and paper): Equipment (computers and screens, printers and paper), data storage, Wi‐Fi use, email exchanges, web requests and paper printing	Excluded: Storage back‐ups
Transport: On‐site set‐up, monitoring and close‐out visits, patient transport to and from the clinical site, and transport for presentation of results	Excluded: Team commuting
Drug and placebo production and distribution	Assumption: Proxy for drug composition and packaging weight
Consultations: Utility consumption, medical devices and consumables	Excluded: Electricity consumption in the consultation room, manufacture of CT scans and radiation emitted by CT scans
Waste (including drug waste)

Note: For the following, the activity data from the investigating centres were extrapolated from the data collected for the Rennes centre, where appropriate in proportion to the number of patients included: patient transport, investigating team activity, number of CT scans performed.

2.2.2. Life cycle inventory

A data collection worksheet (Data S1) was produced so that activity data could be collected in a standardised way. Identified stakeholders were asked to collect data related to their activities on the worksheet. Only activities undertaken for the clinical trial, over and above routine care, were included. Data were gathered from the trial documentation, site investigator file, sponsor's trial‐specific databases and tracking files, mailboxes and electronic case report forms application. Where primary data for the trial were not available, we used two different databases: the NegaOctet database (data specific to France) for processes such as sending an email or performing an online search and the EcoInvent 3.9 database (data specific to Europe) for energy mix and transportation (car, train and flight). For licensing reasons, emission factors cannot be shared here. The life cycle inventory is presented in Data S2.

2.2.3. Impact Assessment

A full LCA using the PEF methodology and the EF characterisation methodology determines 16 indicators designed to quantify various environmental impacts associated with the process throughout its entire life cycle. These impact indicators are determined based on activity data collected in the LCI, combined with emission factors from databases (characterisation). The 16 indicators can be categorised into several key areas (see Table 2) and grouped into a single score. The calculation of the single score involves normalisation and weighting according to the following procedure (see Figure 1b):

TABLE 2.

Impact categories included in PEF and details of the methods and indicators used to assess them [18].

Impact category	Description	Indicator and unit of measure
Acidification	Acidification from air, water and soil emissions mainly due to combustion processes in electricity generation, heating and transport	Accumulated exceedance (AE): The potential impact of substances contributing to acidification is converted to the equivalent of moles of hydron (a general name for a cationic form of atomic hydrogen, mol H+ eq).
Climate change	Increase in the average global temperature resulting from greenhouse gas (GHG) emissions	Radiative forcing as global warming potential (GWP100): The global warming potential of all GHG emissions is measured in kilograms of carbon dioxide equivalent (kgCO2e); namely, all GHGs are compared to the amount of the global warming potential of 1 kg of CO2.
Ecotoxicity, freshwater	Impact of toxic substances on freshwater ecosystems, which may damage individual species as well as the functioning of the ecosystem	Comparative toxic unit for ecosystems (CTUe) based on the USEtox model: The CTUe unit expresses the estimated potentially affected fraction of species integrated over time and the volume of the freshwater compartment per unit of mass of the chemical emitted.
Eutrophication, freshwater	Eutrophication and potential impact on ecosystems caused by nitrogen and phosphorous emissions mainly due to fertilisers, combustion and sewage systems	Fraction of nutrients reaching freshwater end compartment: The potential impact of substances contributing to freshwater eutrophication is converted to the equivalent of kilograms of phosphorus (kg P eq).
Eutrophication, marine		Fraction of nutrients reaching marine end compartment: The potential impact of substances contributing to marine eutrophication is converted to the equivalent of kilograms of nitrogen (kg N eq).
Eutrophication, terrestrial		Accumulated exceedance (AE): The potential impact of substances contributing to terrestrial eutrophication is converted to the equivalent of moles of nitrogen (mol N eq).
Human toxicity, carcinogenic effects	Impact on human health caused by absorbing substances through the air, water and soil. Direct effects of products on humans are not measured	Comparative toxic unit for humans (CTUh) based on the USEtox model: Expresses the estimated increase in morbidity (the number of disease cases) in the total human population per unit of mass of the chemical emitted.
Human toxicity, noncarcinogenic effects		Comparative toxic unit for humans (CTUh) based on the USEtox model: Expresses the estimated increase in morbidity (the number of disease cases) in the total human population per unit of mass of the chemical emitted.
Ionising radiation	Impact of exposure to ionising radiations on human health	Human exposure efficiency relative to U‐235: The potential impact on human health of different ionising radiations is converted to the equivalent of kilobecquerels of uranium 235 (kBq U235 eq).
Land use	Transformation and use of land for agriculture, roads, housing, mining or other purposes. The impact can include loss of species, organic matter, soil, filtration capacity and permeability	Soil quality index: Representing the aggregated impact of land use on biotic production, erosion resistance, mechanical filtration and groundwater replenishment (dimensionless—Pt).
Ozone depletion	Depletion of the stratospheric ozone layer protecting from hazardous ultraviolet radiation	Ozone depletion potential (ODP): Expressed in kilograms of trichlorofluoromethane (CFC‐11) equivalent, a gas known for its harmful effect on the ozone layer. The potential impacts of all relevant substances for ozone depletion are converted to their equivalent of kilograms of trichlorofluoromethane (also called Freon‐11 and R‐11); hence, the unit of measurement is in kilograms of CFC‐11 equivalent (kg CFC‐11 eq).
Particulate matter emissions	Impact on human health caused by particulate matter (PM) emissions and its precursors	Impact on human health: Measured as the change in mortality due to PM emissions, expressed as disease incidence per kilogram of PM2.5 emitted.
Photochemical ozone formation	Potential of harmful tropospheric ozone formation from air emissions	Tropospheric ozone concentration increase: The potential impact of substances contributing to photochemical ozone formation is converted into the equivalent of kilograms of nonmethane volatile organic compounds (kg NMVOC eq).
Resource use of fossils	Depletion of non‐renewable resources and deprivation for future generations	Abiotic resource depletion—Fossil fuels (ADP‐fossil): The amount of materials contributing to resource use of fossils is converted into megajoules (MJ).
Resource use of minerals and metals		Abiotic resource depletion (ADP ultimate reserves): The amount of materials contributing to resource depletion is converted into equivalents of kilograms of antimony (kg Sb eq).
Water use	Depletion of available water depending on local water scarcity and water needs for human activities and ecosystem integrity	Weighted user deprivation potential: The potential impact is expressed in cubic metres (m3) of water use related to the local scarcity of water.

Normalisation: The environmental impacts calculated for each indicator are normalised by dividing them by a normalisation factor corresponding to the total impact of a reference (e.g., annual global emissions for climate change). This step allows impacts from different indicators to be compared on a common basis (e.g., % of annual global impact).

Weighting: Once the results have been normalised, weighting factors are applied to assign relative importance to each impact indicator. The weighting factors are generally determined by panels of experts, social studies or environmental policies and reflect societal priorities for each impact. Applying these factors aggregates all the standardised impacts into a single environmental score. For normalisation and weighting factors, see Data S3.

The single score is expressed in points (Pts), a unit based on the impact of an average European per year in 2010, and depends on the characterisation method. The lower the score, the lower the impact on the environment.

The impact characterisation method used was EF 3.0 in SimaPro 3.5 software. By entering the LCIs of all the inputs and outputs from the SUCRE study, we were able to determine emission factors from 16 indicators and a single score for this clinical trial.

2.3. Uncertainty and Sensitivity Analyses

Uncertainty was estimated using the Monte Carlo simulations on the impact characterisation method and a data quality analysis (see Data S4) in accordance with the data quality requirements of the PEF repository.

Sensitivity analyses (see Data S4) were performed on (1) the average distance of patient transport to investigating centres other than Rennes (extrapolated data) and (2) a mitigation in clinical research assistant (CRA) transport by replacing air travel with train travel.

3. Results

The SUCRE study's single score was estimated at 2.88 points. The main indicator impacting the single score was climate change (28.5%), with total GHG emissions estimated at 31.6 tCO₂e, or 156 kgCO₂e per patient included (Table 3). The other main impact factors were resource use of fossils (23.5%), resource use of minerals and metals (12.2%) and particulate matter emissions (8.3%).

TABLE 3.

Environmental impacts of the SUCRE study (see Section 2 and Data S3 for detailed methodology).

Impact category	Total	Points	Contribution to the single score (%)
Climate change	31 605 kgCO2e	8.22E‐01	28.5
Resource use of fossils	529 654 MJ	6.78E‐01	23.5
Resource use of minerals and metals	0.3 kg Sb eq	3.51E‐01	12.2
Particulate matter emissions	1.6 × 10−3 disease incidence	2.39E‐01	8.3
Photochemical ozone formation	139 kg NMVOC eq	1.63E‐01	5.7
Acidification	117 kg mol H+ eq	1.30E‐01	4.5
Ecotoxicity, freshwater	248 071 CTUe	1.12E‐01	3.9
Ionising radiation	7397 kBq U235 eq	8.78E‐02	3.0
Eutrophication, freshwater	4.1 kg P eq	7.19E‐02	2.5
Eutrophication, terrestrial	340 mol N eq	7.13E‐02	2.5
Eutrophication, marine	32.4 kg N eq	4.92E‐02	1.7
Human toxicity, noncarcinogenic effects	4.8 × 10−4 CTUh	3.83E‐02	1.3
Human toxicity, carcinogenic effects	2.4 × 10−5 CTUh	3.00E‐02	1.0
Water use	3382 m3 deprived	2.51E‐02	0.9
Land use	153 596 Pt	1.49E‐02	0.5
Ozone depletion	7 × 10−4 kg CFC‐11 eq	8.19E‐04	0.0
Total (single score)		2.88	100

The ‘conduct of the study’ phase was the main contributor to the environmental impact, and this was true for all indicators analysed (between 95% and 100%). Within this phase, the major hotspots identified were patient transport for study consultations, CRA transport, and study drugs and exams (see Figure 2).

FIGURE 2.

Impact of research activities on the following: (a) climate change, (b) resource use of fossils, (c) resource use of minerals and metals, and (d) particulate matter emissions. CRAs: clinical research assistants.

3.1. Uncertainty and Sensitivity Analyses

The part of values containing uncertainty data was estimated at 70% with the Monte Carlo simulations on the characterisation method. The two most significant activities (patients and CRA travels) had a data quality ratio greater than 2, meaning that the data quality is considered very good (Data S4).

The sensitivity analysis made on the transports of patients from other centres than Rennes showed that by reducing the distance by 10%, a maximum reduction of 5% in the single score was observed (Data S4). Furthermore, this does not alter the contribution of the stages to the impact on the indicators studied (data not shown). In the same way, the sensitivity analysis performed on the mode of transport for CRA travels showed that the climate impact of the trial was reduced by 3% if air travel was reduced by 35% (6000 km) and replaced by train travel.

4. Discussion

In recent years, environmental sustainability in healthcare, including clinical research, has become an increasingly important issue considering the global threat to the climate. As major players in clinical research, university hospitals need to examine the sustainability of their practices in the context of clinical trials, as well as implementing eco‐design for routine care.

To the best of our knowledge, this is the first study to report the results of a full screening LCA of a clinical trial. Previous studies have only focused on GHG emissions from clinical trials using various methodologies, without questioning the relevance of using GHG emissions as the sole environmental indicator. Our results show that climate change (driven by GHG emissions) is a major contributor to the overall environmental impact of clinical trials. However, notable impacts were also identified in resource use of fossils, resource use of minerals and metals and particulate matter emissions. Importantly, the activities responsible for the highest climate impacts, notably patient and clinical research associate (CRA) transportation, also contributed substantially to other categories. This strong overlap suggests that targeting these hotspots through eco‐design strategies could achieve concomitant reductions across multiple environmental indicators. Based on these findings, we also believe that climate impact could serve as a pragmatic surrogate endpoint to estimate overall environmental impacts of clinical trials, especially when the same activities drive results across several categories. As suggested by Griffiths et al. [7], carbon‐based assessments may allow simplified and more accessible evaluations, avoiding the need for resource‐intensive, full LCAs when performing routine or large‐scale assessments.

Nevertheless, comprehensive LCAs remain valuable in specific settings where interventions may differentially affect other environmental dimensions, providing a more complete understanding of environmental trade‐offs. The carbon footprint of the SUCRE study was evaluated at 31.6 tCO₂e (corresponding to the climate change indicator), corresponding to 156 kgCO₂e per patient included. This is roughly equivalent to the GHG emissions of three French citizens for 1 year, or nine return flights from Paris to New York. In comparison with other healthcare interventions, the carbon footprint of a patient enrolled in the SUCRE study approximates that of a cataract operation [19] or 3–4 haemodialysis sessions [20].

Comparing the results obtained with other studies in the field remains challenging because of the substantial variability, primarily due to differences in methodology and the domains of emissions assessed. The recent reviews [12, 13] show consistent identification of major hotspots across academic and pharmaceutical industry studies (study team facilities, site monitor visits and trial management meetings). Absolute carbon emissions were generally higher in pharmaceutical industry‐sponsored trials than both the national and international academic trial results, reflecting additional activities such as drug manufacturing, packaging, distribution and laboratory sample logistics. The level of intervention of a study thereby appears to be a key determinant of its environmental impact. Nevertheless, to put our results into perspective, we can look at the carbon emissions of national academic studies. Lyle et al. [5] report a mean of 78.4 tCO₂e (range 42.1–112.7) in 12 pragmatic randomised controlled trials, corresponding to a mean of 306.2 kgCO₂e (80.0–883.7) per patient included. More recently, Griffiths et al. reported a total of 89 tCO₂e and 45 kgCO₂e per patient for the PRIMETIME study [7] and an average of 63.3 tCO₂e (18–109) and 208 kgCO₂e (8.5–725) per patient for seven national academic studies [8]. By way of comparison, the total carbon emissions reported for the SUCRE study (31.6 tCO₂e total and 156 kgCO₂e per patient) are slightly lower, but methodological differences make direct comparisons challenging. Earlier assessments were based on the GHG reporting protocol [4, 5, 6], whereas more recent works have used a monocriteria life cycle approach [9, 10, 11] or a methodology adapted specifically for clinical trials [7, 8]. Standardisation, methodological harmonisation such as that developed by Griffiths et al. [7] and adaptation to various contexts (e.g., academic/pharmaceutical industry and national/international) will be essential to enable reliable comparison and foster collaboration across the research community.

Most previous studies reported clinical trial unit (CTU) emissions, trial‐specific patient assessments, meetings and travel by study personnel as major hotspots for carbon emissions [6, 7, 8, 9, 10]. Our results clearly confirm that trial‐specific patient assessments have the greatest impact, particularly with regard to patient and trial staff transport; this major effect is also evident in the other three main indicators (resource use of fossil fuels, resource use of minerals and metals and particulate matter emissions). However, it should be noted that CTU emissions were not identified as being hotspots of the SUCRE study, which may be due to the analytical scope excluding commuting. Moreover, as Mackillop et al. previously highlighted [9], energy in France remains relatively low‐carbon, which may contribute to the limited contribution of French CTU premises in carbon emissions.

4.1. Limitations

The work presented here began in early 2023, when there was no specific, recognised methodology for assessing the environmental impact of a clinical trial. We therefore opted for a generic LCA methodology, which we adapted to reflect the successive phases of a clinical trial's life cycle. To ensure the feasibility of this first ‘screening’ exploratory LCA, several assumptions were necessary, and certain activities were excluded from the assessment. In particular, commuting was not included in the LCA methodology, unlike the carbon footprint methodology, as it is not considered to be solely associated with the clinical research study (see Section 2). However, based on the results obtained, it is unlikely that the excluded data would have altered the respective contributions of the activities to the various indicators.

LCA remains a relatively recent methodology, and its precision strongly depends on the quality and completeness of underlying databases, which are frequently updated. Data on healthcare, pharmaceuticals and web practices still require improvement. Additionally, access to some databases is restricted: EcoInvent and NegaOctet, for instance, are private, pay‐per‐use databases. HealthcareLCA, a database specific to healthcare, is one of the few open‐access resources available, containing 7000 emission factors for medical devices, drugs, medical interventions, companies and services. However, the calculation methodologies for these data are not standardised, and some do not fully adhere to LCA principles. Therefore, information from this database must be used cautiously, with careful consideration of the methodologies employed to estimate the environmental impact.

Except for digital emissions factors derived from the NegaOctet database, emissions factors used in this study were European and not country‐specific, primarily sourced from the EcoInvent database. Of these, only the electricity mix seems likely to vary greatly between countries. However, given the very small proportion of electricity‐related activities in our study, changing the electricity mix would likely have little effect on our conclusions.

Finally, the choice of the SUCRE study, an interventional drug study, limits its transposability to other study designs (particularly non‐interventional studies). However, the choice of this study enabled the assessment of processes presumed to carry substantial environmental impact, such as the production and distribution of drugs and placebos, the monitoring of data from a multicentre trial and performing examinations specific to the trial. Varying the types of studies assessed should now be encouraged so that recommendations can be made depending on the type of study and so that relevant good practice can be shared.

4.2. Perspectives

Given the predominance of travel in the environmental impact of clinical trials, several emerging practices at the international level may offer effective mitigation strategies, such as decentralised clinical trials and the use of remote monitoring. However, alongside technical and organisational solutions, it is important to remember that the most effective way to reduce emissions is to minimise research waste, which results from research that is conducted but has no impact, for example, because it lacks methodological rigour or unnecessarily duplicates a previous study [21, 22]. Echoing Altman [23], we advocate that ‘less research, better research and research for the right reasons’ also meet the goal of reducing the environmental impact of clinical research. It is in this spirit of minimisation that researchers should design and conduct clinical trials. In this context, the UK National Institute for Health and Care Research (NIHR) Carbon Reduction Guidelines highlight areas where sound research design can reduce waste without compromising scientific validity or methodological rigour [24]. Beyond methodological refinement, the broader adoption of environmental management systems in clinical research is required. Stakeholders across the research ecosystem—from sponsors and investigators to patients—need to be engaged in this transition. Therefore, qualitative studies need to be conducted to understand the drivers and barriers to adopting environmental approaches in clinical trials, with the aim of shaping a new, sustainable clinical research paradigm.

Ultimately, integrating sustainability across all levels of clinical research is imperative. Efforts should extend beyond individual studies to encompass funding allocation processes [25] and the criteria used to evaluate research performance [26, 27]. Only a systemic paradigm shift—embedding environmental considerations as a core dimension of research design, conduct and evaluation—will enable the true eco‐design of clinical research.

5. Conclusions

Through the first complete LCA of a clinical trial, our study has shown that a French publicly funded academic clinical trial has a substantial environmental impact and confirmed the interest in the use of GHG emissions as an indicator of this impact. Should immediate mitigation efforts focus on patient transport and study staff travel, which are identified as hotspots, a global paradigm shift will be necessary to achieve eco‐design in clinical research.

Author Contributions

Claire Fougerou‐Leurent wrote the manuscript. Claire Fougerou‐Leurent, Bruno Laviolle and Louise Forteau designed the research. Alain Renault, Catherine Mouchel, Chloé Rousseau, Enora Marion, Sabrina Cochennec, Anne Ganivet, Marie‐Laure Gervais and Loïc Fin performed the research. Claire Fougerou‐Leurent, Bruno Laviolle, Margot Brion, Louise Forteau and Maëlle Perrot‐Loyer analysed the data.

Ethics Statement

The LCA carried out in this work was a retrospective analysis of data collected during the SUCRE study. The data used were clinical trial documentation, activity data from the sponsor's staff, and interviews with trial staff and study site personnel. No patients participating in the trial were involved in the LCA, and no personal information from the participants was collected or shared.

Conflicts of Interest

All authors have completed the ICMJE uniform disclosure form at http://www.icmje.org/disclosure‐of‐interest/ and declare financial support from CIC Inserm 1414 for the submitted work; Louise Forteau and Maëlle Perrot‐Loyer were employees of the Environmental Consulting Firm O2M Lab at the time of the study, and Alain Renault is an employee of the University of Rennes. All other authors were employees of the Rennes University Hospital at the time of the study. The authors declare no financial relationships with any organisations that might have an interest in the submitted work in the previous 3 years; no other relationships or activities could appear to have influenced the submitted work.

Funding

The authors have nothing to report.

Consent

The data collection template (Data S1) is reproduced with permission from the Environmental Consulting Firm O2M Lab.

Supporting information

Data S1: fcp70086‐sup‐0001‐Supplementary_Material1.pdf.

Data S2: fcp70086‐sup‐0002‐Supplementary_Material2.docx.

Data S3: fcp70086‐sup‐0003‐Supplementary_Material3.docx.

Data S4: fcp70086‐sup‐0004‐Supplementary_Material4.docx.

Data Availability Statement

Data are available on request from the corresponding author.