Integrating environmental considerations in digital health technology assessment and procurement: Stakeholders’ perspectives

Abstract

Background

Digital health technologies (DHTs) are promoted as means to reduce the environmental impact of healthcare systems. However, a growing literature is shedding light on the highly polluting nature of the digital industry and how it exacerbates health inequalities. Thus, the environmental footprint of DHTs should be considered when assessing their overall value to healthcare systems. The objectives of this article are to: (1) explore stakeholders’ perspectives on integrating the environmental impacts of DHTs in assessment and procurement practices; (2) identify the factors enabling or constraining the operationalisation of such a change; and (3) encourage a constructive dialogue on how environmental issues fit within healthcare systems’ push for more DHTs.

Methods

Semi-structured interviews were conducted with 29 stakeholders involved in DHTs in a large Canadian academic healthcare centre. Data were collected and analysed through a mixed deductive-inductive process using a framework derived from diffusion of innovations theories.

Results

The integration of the environmental impact of DHTs in assessment and procurement is contingent upon key micro-meso-macrosystemic factors that either enable or constrain changes in practices and processes. Innovation (micro) factors include stakeholders’ recognition of the environmental issue and the extent to which it is feasible for them to address the environmental impact of DHTs. Organisational (meso) factors include the organisation's culture, leadership, policies, and practices, as well as the expertise and professional skillsets available. Finally, external (macro) factors include political and regulatory (e.g., national strategy, laws, standards, norms), economic (e.g., business models, public procurement), and professional and scientific factors (e.g., evidence, methodologies, clinical guidelines).

Conclusion

Considering the environmental impact of DHTs depends on micro-meso-macrosystemic factors involving a variety of stakeholders and levels of governance, sometimes with divergent or even antagonistic objectives and expectations. It highlights the importance of better understanding the complexity inherent in the environmental shift in healthcare.

Introduction

To tackle climate change, many governments are implementing strategies and actions to reduce their carbon footprint.^1–5 In this context, public procurement is seen as an important lever to accelerate the green transition because it can, for instance, introduce environmental norms and standards in public tenders.^6–10 Public procurement is the process followed by a public service to acquire goods, services, works, and other supplies, usually by means of a contractual arrangement after public competition. ¹¹

Given their contribution of approximately 4.6% to global greenhouse gas (GHG) emissions, healthcare systems are expected to reduce their environmental impact to better align with their health mission.^12–16 However, the green transition faces many challenges related to the unique nature of healthcare, as stringent safety, efficacy, and efficiency standards must be met to protect patients and deliver quality care. The design of eco-friendly health technologies that meet public procurement requirements is especially challenging, as health innovators fear that additional environmental criteria could compromise the quality and safety of care for patients and increase costs for healthcare systems.^17,18 For example, using less-resistant eco-friendly materials instead of plastic to make prostheses or using washable devices instead of disposable ones as they may increase the risk of nosocomial infection. ¹⁷ In this context, digital health technologies (DHTs) (Box 1) are frequently promoted as means to reduce the environmental impact of healthcare systems without compromising the quality and financial sustainability of services. For instance, virtual consultations reduce GHG emissions related to the travel of healthcare professionals and/or the transfer of patients and their families from one healthcare site to another.^19–22

Box 1:

Definition of digital health technologies.^23,24

The World Health Organization (WHO) defines DHTs as “the field of knowledge and practice associated with the development and use of digital technologies to improve health (…). It includes a wider range of smart and connected devices (…) such as the Internet of Things, advanced computing, big data analytics, artificial intelligence (AI) (…) and robotics”. ²³ Certain DHTs are considered as medical devices (e.g., Software as a Medical Device). A medical device is defined as “any instrument, apparatus, appliance, software, material, or other article, whether used alone or in combination, together with any accessories, including the software intended by its manufacturer to be used (…) for human beings for the purpose of diagnosis, prevention, monitoring, treatment or alleviation of disease, etc.” ²⁴

Though DHTs may cut travel emissions, especially for patients who can only access services by car, they are far from being carbon neutral. A growing literature is shedding light on the highly polluting and energy-intensive nature of the digital industry and how it exacerbates health inequalities.^21,25–28 The environmental impact of DHTs results from high energy consumption (e.g., computing power, algorithm training), the material composition of the technology (e.g., rare-earth minerals, plastics), and the infrastructures that support their use (e.g., data servers, cable networks).^28–31 Currently, digital technologies account for 4% of global GHGs emissions.^31–33 For instance, the emission rate of certain AI technologies is comparable to the emission rate associated to the entire lifecycle of a car.^30,34 If current trends continue, GHGs from digital technologies are likely to account for more than half of the transportation sector's emissions by 2040. ³⁵ Thus, given their central role in strategies to improve access, quality, and efficiency of healthcare worldwide, the environmental footprint of DHTs must be included when calculating healthcare systems’ GHGs emissions. According to Thompson, ²⁹ if their environmental impact is not considered, these technologies will continue to contribute significantly to climate change, with negative consequences for the health of individuals and communities.

Because integrating environmental considerations in the assessment and procurement of DHTs is an emerging concern, it is important to better understand the practical implications for healthcare systems. Thus, the objectives of this article are to: (1) explore stakeholders’ perspectives on integrating the environmental impacts of DHTs in assessment and procurement practices; (2) identify the factors enabling or constraining the operationalisation of such a change; and (3) encourage a constructive dialogue on how environmental issues fit within healthcare systems’ push for more DHTs, as well as within our increasingly digitalised societies.

Theoretical and conceptual background

This study stems from a larger research project investigating the conditions that enable or constrain the integration of the Responsible Innovation in Health (RIH) domains of Silva et al. ³⁶ into DHTs assessment and procurement practices. The RIH framework brings together five value domains: (1) population health value; (2) health system value; (3) economic value; (4) organisational value; and (5) environmental value. ³⁶

In this study, we focus on the environmental value domain where eco-responsibility is defined as “a product, process or method that reduces the negative impacts of an innovation” at key lifecycle stages, including raw material sourcing (e.g., recycled materials), manufacturing (e.g., waste reduction), distribution (e.g., packaging), use (e.g., energy efficiency), and disposal (e.g., remanufacturing). ³⁶ In this regard, we consider the inclusion of a technology's environmental impact in a decision-making model historically focused on clinical effectiveness and cost as an innovation. By innovation, we refer to new working methods and routines that aim to improve results, efficiency, profitability, relevance, or stakeholders’ experiences. ³⁷ It also refers to ideas, practices, or goals that are considered new by an individual or group. ³⁸ As with medical devices in general, decision-making around DHTs is shrouded in uncertainty, since strong evidence on effectiveness, relevance, quality, efficiency, and safety is often lacking. Consequently, decision-making conflicts are not uncommon in a context where no single choice is identified as optimal by all stakeholders.^39,40

As with any innovative practice, taking the environmental impact of DHTs into account may be in tension, competition, or even contradiction with one or more existing practices, cultures, and/or powers. This can lead to confrontations between different conceptions, priorities, or perspectives.^41,42 Therefore, stakeholder adhesion will depend, among other things, on the level of clarity provided about the innovation's impact on roles and practices. ³⁷ Addressing the environmental impact of DHTs may require changes at the individual (e.g., knowledge, perceptions, attitudes, skills), organisational (e.g., processes, rules, responsibilities, stakeholder dynamics, organisational culture and routines), and healthcare system levels (e.g., laws, directives, norms and standards, financing and business models, market dynamics). Because the changes associated with this new practice can be seen as a gain, a threat, or a loss depending on one's role and understanding, various stakeholders may engage in promotion and argumentation – to convince and justify their views – and negotiations – to find compromises and gain support.^43–46 Moreover, this innovation must be contextualised by looking at the specific dynamics, interactions, and interdependencies between stakeholders and levels of healthcare governance (e.g., organisational preparedness, healthcare system governance, power issues in healthcare organisations; specificity of public vs. private healthcare organisations and systems; context of so-called developed vs. developing countries).^37,41,46 In this regard, considering the environmental impact of DHTs may require significant changes and adjustments in the functioning, cultures, dynamics, rules and laws, practices, and operations of an organisation and/or a healthcare system.^{37,42,45,47,48} The introduction of such a change also implies anticipating potential risks and unintended consequences.^37,42

The role of public procurement in stimulating and disseminating green technologies has been widely studied.^49–55 In the healthcare sector, however, few studies have addressed the micro-meso-macrosystemic factors and conditions involved in the successful integration of environmental concerns into device and technology evaluation and procurement practices.^49,56–62 This contrasts with the literature on innovation, which has already shown that underestimating the complexity and plurality that characterise healthcare organisations and systems leads to difficulties and failures of change.^37,63,64 In this respect, because decision-makers and innovation promoters still largely hold a rational, linear, and predictable view of change, the importance of understanding how stakeholders perceive, negotiate, and interpret change in regard to their contexts, realities, and interests is underestimated.^{37,41,65–67} Thus, in order to better address the complexity and dynamics inherent in the integration of change in healthcare organisations and systems, we adopted a framework derived from Greenhalgh et al.'s ³⁷ model of innovation diffusion. The framework consists of three dimensions: (1) the innovation's attributes; (2) the internal system; and (3) the external system (see Table 1 for details).^37,42,46 Greenhalgh et al.'s model is one of the most comprehensive frameworks for studying micro-meso-macrosystemic factors impacting innovation diffusion in healthcare organisations and systems. It was developed based on an exhaustive systematic review covering several fields (e.g., rural sociology, medical sociology, management and organisational science, health promotion, evidence-based medicine). It has been widely used to study complex changes and innovations in healthcare organisations and systems.^{42,63,68–75} By adopting a multidisciplinary social scientific approach to research on environmental issues in healthcare, the use of this framework allows for a better understanding of the complexity inherent in the green transition within healthcare systems.

Table 1.

The dimensions of the integrative analytical framework adopted for the study.^37,42,46

Innovation attributes (micro factors)
Observability and relative advantage	The innovation is observable and translates (or entails) clear advantages and benefits, especially in the performance of responsibilities or tasks.
Compatibility	The innovation is compatible with perceived values, norms, and needs.
Complexity	The innovation is perceived by key stakeholders as simple to use or implement, and is easily adopted.
Risk/issues	The innovation does not involve a high degree of uncertainty about the outcomes, and is not perceived as risky or problematic.
Knowledge	The knowledge required to use the innovation is codified and transferable from one context to another.

Internal system (meso factors)
Implementation support	Support is available, in the form of training and assistance, to increase the capacity of target individuals or groups to act.
Skills and competencies	The skills and competencies required to implement and use the innovation are adequate.
Funding, resources, and dedicated time	Financial and human resources adequately support the implementation of the innovation, its use, and evaluation.
Tension for change	The innovation takes place in an environment where stakeholders perceive the current situation as needing improvements.
Stakeholders receptive to change	Stakeholders are responsive to the innovation, and therefore can more easily commit to the proposed change and adapt to incorporate it.

External system (macro factors)
Socio-political environment	The socio-political environment raises awareness of the concrete or potential benefits of the innovation.
Political or professional guidelines and decisions	Directives, guidelines, political and/or professional decisions, and cross-level synergies enable – or directly or indirectly support it – to increase the innovation's chance of success.
Science and evidence	Available evidence contributes to the consideration of innovation and its actual or potential impact.
Support from other stakeholders	The innovation is supported by other stakeholders outside of the healthcare system.

Methods

We conducted a qualitative study in a large healthcare organisation in Quebec, Canada (referred to as “the City hospital”), between March and July 2021(Box 2). ⁷⁶ A leader in digital innovation, the City hospital's senior management had already initiated reflections and actions to integrate DHTs into healthcare delivery in a responsible way, including environmental concerns. It is expected that the results of this study will contribute to stimulating reflection and inform decision-making on the subject within the City hospital, and elsewhere.

Box 2:

Description of the City hospital.

The City hospital is one of the largest academic hospitals in the province of Quebec and the entire country (Canada). It offers specialised and sub-specialised services to adult patients. It employs over 14,000 people, including over 1000 physicians, 4000 nurses, and over 2000 other healthcare professionals. It treats around 500,000 patients annually. ⁷⁷ It also houses one of the largest medical research centres in Canada, with more than 2300 staff and an academic mission to produce and disseminate knowledge and research results. It has been ranked by the U.S-based magazine Newsweek as Canada's most innovative hospital and one of the world's top 250 Best Smart Hospitals for 2021. ⁷⁸ The City hospital is positioning itself as a leader in tackling global warming. It is committed to reducing its greenhouse gas emissions by 55% by 2030, and aims to become carbon neutral by 2040, without compromising the quality of healthcare. Canada's healthcare system generates around 4.6% of the country's total greenhouse gas emissions (detailed data for Quebec and the City hospital are not available).^15,79

Recruitment

We selected the interviewees through purposive sampling. We aimed for a broad internal diversification of profiles and opinions within the City hospital. ⁸⁰ We identified key informants based on their involvement and roles in the assessment, procurement, implementation, and/or use of DHTs: clinicians, health technology assessment (HTA) specialists, middle and senior managers, technology providers, researchers, jurists/lawyers, and patients. To identify potential participants, the City hospital's senior management first provided a list of individuals who work in and are concerned by DHTs (e.g., research, HTA, procurement, project management, use -clinicians and patients-, legal questions) as well as DHT companies with whom it collaborates. Then, to mitigate organisational bias and to ensure a wide range of perspectives, we conducted internet searches to identify other potential participants and technology suppliers not on the initial list (e.g., through reports, documents, projects, conferences and symposiums). Finally, respondents also suggested other potential participants at the end of the interview, including patients, procurement specialists, clinicians, and project managers. ⁸¹

Recruitment was conducted by the first (HA) and third authors (PL). The team then discussed to ensure that the profiles, expertise, and perspectives of potential participants met the expectations of the study and would allow a broad spectrum of perspectives to emerge. Such an approach was essential to avoid having a homogeneous “sample” of respondents that risked conveying a “formatted” institutional discourse. Furthermore, the authors had no particular interest or involvement (personal or professional) within the City hospital that could affect their work, from data collection to publication of results. The lead author (HA), who piloted the entire study (recruitment, data collection and analysis, publication of results), had no previous contact with most respondents. The previous contacts he has had with four respondents were in an academic research context and do not present any potential conflicts of interest or bias. In this respect, HA was not involved in the initial recruitment of the four interviewees. Their names were on the list provided by the hospital's senior management. We retained them after the team judged that their contribution was essential to the project. HA then contacted them via the formal procedure used to invite all potential participants, as required by the City hospital Research Ethics Committee.

We sent personalised e-mail invitations to potential participants with an explanation of the project and the reason for the request. Of the 42 invitations sent, 29 individuals agreed to be interviewed. Potential interviewees who could not be reached or participate in the study were actively involved in the management of the COVID-19 pandemic (e.g., clinicians, policy-makers, managers, women running technology companies). Table 2 summarises participants’ characteristics. Most interviewees had several professional experiences and/or roles related to the topic.

Table 2.

Summary of the study participants’ characteristics.

	Gender, n
Type of participants (category of respondents)	Female	Male	Total n
Decision-makers/senior managers	2	1	3
Mid-level managers	3	3	6
Clinicians/managers	1	1	2
Clinicians/computer scientists/researchers	2	3	5
Procurement and contracting specialists	2	0	2
Jurists and/or lawyers	1	0	1
Health technology assessment specialists	1	0	1
Patients	1	3	4
Technology providers	0	5	5
Total	13	16	29

Data collection

The first author (HA) conducted semi-structured interviews in French (27) and in English (2) on the Zoom™ videoconferencing platform between March and July 2021. Due to COVID-19, they were mainly conducted from home and it was ensured that no other individual was present during the interviews. They lasted between 30 and 90 min, were audio recorded with the permission of the participants, and transcribed by HA. Prior to the interview, each respondent received a consent form summarising the project's objectives and implications. HA obtained verbal and written consents for each participant. Interview questions were guided by the RIH framework of Silva et al. ³⁶ HA took notes during and after the interviews and used them to help contextualise the subsequent analyses. He first tested the qualitative interview guide with two respondents prior to the start of the study. The guide then evolved in light of responses and emerging information. One interview was conducted with each respondent. The interviews, along with our in-depth knowledge of the field acquired through extensive involvement in various DHTs and innovation research projects, allowed for triangulation of sources and data. ⁸²

Data analysis

HA transcribed, coded, and analysed the data using Dedoose™ software. He performed the first round of analysis and developed a preliminary coding framework. In the second round, the framework was iteratively challenged, discussed, and refined by the second (LR) and third authors (PL). We conducted a deductive-inductive thematic analysis. Our deductive analysis was guided by our analytical framework (Table 1), while our inductive analysis aimed to capture emerging themes not covered by the framework. After agreeing on the different identified themes, we concluded that none required the addition of a new framework dimension as all identified themes fitted within our analytical framework. We reached data saturation for the observations and themes reported in the findings. We did not return transcripts to participants for comment and/or correction, as the answers were clear and did not require any particular clarification. No feedback was requested on the findings.

The study was approved by the City hospital Research Ethics Committee (Number: Comité d’éthique de la recherche-City hospital: 20.399). The research team is composed of researchers (3 PhD) and physician-researchers (2 PhD/MD, 1 MD/MPH), including two women (2 PhD) and four men (1 PhD, 2 PhD/MD, 1 MD/MPH). The team has extensive experience in qualitative and quantitative research in digital health, innovation, HTA, medicine, public health, and health services and policy.

Findings

We identified seven themes related to our analytical framework – innovation attributes, internal system, external system – that captured factors enabling or constraining the integration of the environmental impact of DHTs in assessment and procurement practices (Table 3 summarises the results). Table 4 provides illustrative citations for each theme (translated from French to English when needed by the authors).

Table 3.

Key enabling and constraining factors.

Enablers

Constrainers

Healthcare systems’ role in taking care of patients affected by climate change and the need to green care practices; Education, training, and support programs for clinicians to better understand environmental requirements; Leveraging and coordinating existing eco-friendly initiatives; Integrating an environmental policy in the hospital's strategic plan; Adopting international norms, standards, and certifications; Companies will invest in green technologies if the government requires meeting environmental standards, regulatory and legal frameworks are in place, and funding and procurement models are available; Environmental criteria must be communicated to companies during innovation design and development stages; Professional medical federations and colleges must legitimise greener care delivery.

DHTs’ use in reducing healthcare systems’ immediate environmental impact (e.g., virtual consultations); Limited number of suppliers for highly specialised medical technologies Clinicians’ familiarity with and preference for specific medical technologies; Lack of fit with current technological infrastructure and ongoing supplier contracts; Lack of tools and frameworks to operationalise green care policies; Lack of HTA expertise and skillset to examine complex DHTs lifecycles; Lack of developers, designers, and engineers trained in eco-friendly computing practices; Ministry-level public tendering processes aiming to procure cheapest technologies; Aversion to spend more on green technologies in post-pandemic budgets; Corporate strategies and market dynamics that expend beyond healthcare systems; Current lack of solid evidence and methodologies to measure the environmental impact of DHTs.

Table 4.

Illustrative respondent citations per theme.

Themes	Illustrative citations
Recognition of the environmental issue	“Honestly, for me, it's still a theoretical debate. Right now, it's not a high priority for me. My number one priority [is to have AI technologies that can detect the disease] (…). I know it's very narrow (…). But I’m more concerned about the technical aspects of making things work, than the environment.” (Researcher 1) “If you have an AI model, there is definitely an electrical cost to running the algorithm. But it's so minimal, even the training costs, compared to, say, predicting the response to a prescribed treatment for a patient. Imagine a wrong treatment prescribed to a patient for three months, the environmental footprint can be terrible too, not to mention the health problems. I don’t see AI in healthcare as having a negative footprint.” (Technology Provider 1)
Feasibility of addressing the environmental impact of DHTs	“The technology must meet the clinical needs (…). Because it's so specific, what we’re asking for, that there are few companies that can do it. So, we have to ask ourselves an existential question: Do we give up on this because it doesn’t meet certain environmental criteria? What do we do to meet the needs of the patients? It's like benefit/risk management, it's always a balancing act.” (Manager 1) “No matter how much we want to include plenty of clauses, if awareness is not raised among clinicians and their management, it's useless. They’re already resistant to change. They often have an idea in mind before we even go to tender (…). It has to come from the clinicians and the field. For example, the operating room can raise concerns about environmental issues (…). There's awareness and education to be done (…). Afterwards, we could translate this into contracts.” (Manager 2).
Organisational culture, policies, practices, and infrastructure	“I feel like it's really a culture change. It goes beyond just a technology that you’re going to implement (…). They’re trying to introduce that mindset: paperless City hospital and things like that (…). It's part of their development strategy. But in practice, it's not necessarily easy. It's something that takes a lot of time and it's even worse (…). There's a certain movement and a culture that's being put in place.” (Manager 3) “We’re starting to include it more and more. We have a policy, which is good. Increasingly, we’re questioning this dimension in procurement, for example. We ask the question systematically (…). People are aware of the issue within the organisation (…). We include it in ‘population health’ for the moment. It's not a separate dimension (…). We have questions more on the lifecycle. But energy consumption of DHTs, we don’t have it in our framework.” (Clinician-Manager 1) “It becomes very complex because you’re getting into the internal policies of the organisation. We would need ISO standards or other standards and certifications that could support that (…). When you make it a standard, you also force suppliers to align themselves with it. They will arrange for tenders because they have an interest in developing this. So, you’re going to make them aware of it and you’re going to achieve your objectives (…). It's difficult. You have to be able to frame it all.” (Manager 2)
Expertise and skills	“At the HTA level, they don’t look at that dimension (…). I think it's a whole other area of technology. It must be completely managed and devoted to a well-defined resource (i.e., a person who is an expert in the field). Because it has to consider the whole chain of the technology: What does it generate as an environmental or economic impact? etc. (…). I think that it's out of the HTA team's mission (…). It’d be really great if they could do it. But it's a whole world.” (Manager 4)
Political factors	“We need to make it a standard. Today, it's more of an exception. Because the problem is that the Public Procurement Authority monitors everything. If there's the slightest mistake, they’ll slap us on the fingers. Also, these standards aren’t necessarily known among suppliers. So, we can’t really demand it, unless there's a contractual clause (…). We can demand it, but we’ll be sued for undermining competition (…). We’ll lose.” (Manager 5) “You really have to look at the regulatory side (…). You also need suppliers to be competitive. Because we agree, in the end, it's the price that decides. The Ministry isn’t going to want to pay huge prices. We’ve already lost enough money during the pandemic.” (Manager 6)
Economic factors	“[Names of several large medical technology and device companies] have their PSRs (Pharmaceutical Sales Representatives) who talk with the minister's office, and on a regular basis (…). Then, I imagine they talk to hospital management as well. These large companies have salespeople who spend their time in hospitals.” (Technology Provider 2)
Professional and scientific factors	“[Professional federations and colleges] have influence, for example, in guiding change in certain areas, clinical or not. Yes, they have a role to play, precisely to mitigate the clinicians’ concerns.” (Manager 7) “It's not straightforward, because we’re in healthcare. We can’t start gambling with people's lives or the quality of patient care.” (Manager 1)

Innovation attributes (micro factors)

Recognition of the environmental issue

Though all interviewees acknowledged the challenges of climate change and agreed that actions should be taken to quickly address this issue, only five respondents believed that the environmental impact of DHTs should be an assessment and/or procurement criterion. While the majority of the respondents (25/29) underscored the role of healthcare systems in taking care of patients affected by climate change and the need to green care practices, two technology providers argued that questioning the environmental impact of DHTs was irrelevant. They put forward the benefits of DHTs in reducing the environmental footprint of the healthcare system and argued that, for example, though AI technologies are energy-intensive, their carbon impact is lesser than other therapies (e.g., cancer care). They believed that DHTs are necessary to make healthcare systems more eco-responsible.

Feasibility of addressing the environmental impact of DHTs

Respondents raised several issues constraining the feasibility of assessing the environmental impacts of DHTs. A manager emphasised how highly specialised medical fields require very specific technologies that are developed by a limited number of suppliers. Because the City hospital must acquire the technology that best meets patients’ clinical needs, adding environmental considerations to highly specialised DHTs makes it much more difficult to find technology that addresses all of these concerns. Another manager pointed to the reluctance of clinicians who often request specific products and technologies. Given their autonomy of practice, it is challenging to impose eco-friendly alternatives. The latter explained that if the environmental requirement were to come from the clinicians themselves, the City hospital would likely integrate it into its operations, assessments, and procurement practices.

Internal system (meso factors)

Organisational culture, policies, practices, and infrastructure

In addition to feasibility challenges, respondents discussed organisational constrainers and proposed enablers for the City hospital. For a manager, an organisational culture change is required to address an ensemble of environmental concerns that go beyond the environmental impact of DHTs, including clinical practices, food services, and waste management. They further explained how such change is time-intensive and involves strong awareness-raising for all stakeholders – clinical, technological, and administrative. For this respondent, clinicians, given their influential position within the organisation, require more education and training to better understand the environmental implications of their practices and the technologies they use, as well as adequate support to address these issues. In the same vein, to enable such a change, the respondent argued that the City hospital also needs to better leverage and coordinate the small eco-friendly initiatives already implemented in the hospital.

For other respondents, the City hospital has already undertaken steps towards a greener organisational culture through an environmental policy included in its 2019–2023 strategic plan that covers energy and water consumption, smart buildings, recycling, composting, and procurement. A senior manager argued that this policy made the City hospital an environmental pioneer in the Quebec healthcare system, as mechanisms and actions have been implemented to raise awareness among the staff. This point was supported by another manager who reported that environmental issues are increasingly integrated into policies under the population health umbrella. However, the respondent specified that the environment is not a dimension of its own and environmental impacts of DHTs are not part of the actions implemented to date.

Though the City hospital integrates environmental issues in internal policies, a manager pointed out that their practical implementations remain challenging. One interviewee explained how the organisation's current technological infrastructure limits the types of technologies that can be integrated into its system. A newer and greener technology may not fit with older and more energy-intensive systems. Furthermore, integrating greener technology would involve breaking existing contracts with some suppliers, which could result in significant financial penalties for the organisation. It would also require a continuous updating and/or replacing of obsolete technologies and equipment on a regular basis to ensure patient safety, which the organisation does not necessarily have the capacity or means to do.

According to a manager, to foster greener practices, the City hospital should adopt international standards (e.g., The International Organization for Standardization: ISO) with external and independent certification processes and mechanisms as technology providers must also comply with these standards and certifications in a transparent manner. The same respondent also explained how the City hospital is lacking tools and frameworks (e.g., rules, guides, protocols) to better operationalise green practices.

Expertise and skills

Respondents identified other operational constrainers pertaining to the specific expertise and advanced skillset required for HTA teams to properly examine the entire lifecycle of DHTs and analyse the economic impact of the pollution generated by these technologies throughout their production, use, and disposal. A manager explained how the City hospital does not currently have this expertise in-house and would require additional resources to this end. For another respondent, the national regulatory agencies in charge of approving and certifying technologies for reimbursement and use in the healthcare system should be responsible for this extremely complex task, not the City hospital. A technology provider further explained how companies lack developers, designers, and engineers trained in eco-friendly computing practices to meet such environmental requirements and that this expertise is not easily available. This is partly because most computer sciences, information technology, and system design training programs do not integrate environmental issues into their curricula.

External system (macro factors)

Political factors

Respondents strongly agreed that the current political context is not conducive to integrating environmental considerations into DHTs assessment and procurement. A clinician-manager stressed the contrast between the growing public awareness of environmental and climate change issues and the orientation of governmental public policies and actions. Six interviewees highlighted the fact that the public tendering process of the Ministry of Health aims to procure the cheapest technologies. Two of them argued that the short-term cost-oriented vision is a major constraint for organisations seeking to implement green technologies because they are often more expensive than their counterparts. For a senior manager, if the City hospital's and the society's desire for greener technologies is not translated into coherent governmental policies, directives, and actions in favour of the environment, it will be very difficult, if not impossible, to initiate the necessary changes in the field. Another manager provided a concrete example of this constrainer. She explained how the City hospital, in accordance with its internal environmental policies, could provide a preferential margin to an eco-friendly technology supplier in the case of a tender where two suppliers score equally on the other evaluation criteria. However, in this scenario, the loosing supplier would be in their legal right to sue the City hospital through the Public Procurement Authority (i.e., the public contracts oversight body in Quebec).

Economic factors

Closely intertwined with the political constraining factors described above are the economic factors brought forward by interviewees, and which relate to the COVID-19 pandemic, market demand, and broader market dynamics.

For a manager, because the COVID-19 pandemic made visible the impacts of climate change and the urgency to act, one might have expected the government to develop policy and implement necessary actions towards a more sustainable healthcare system. However, the manager further explained that the pandemic demanded such colossal public spending over this period, the government's priority was then to cut expenses rather than invest the significant amounts required to procure green technologies, which remain perceived as more expensive than non-eco-friendly alternatives.

For a technology provider, green technologies are more expensive because they reflect the costs incurred by companies to meet environmental standards. Given the magnitude of the investments that may be required, companies are often reluctant, or simply refuse, to invest in the development of green technologies, especially if there is uncertainty about the availability of a real market demand and a return on investment. This respondent further clarified that technology providers would be willing to invest to comply with environmental standards if: (1) the government requires meeting these standards; (2) adequate regulatory and legal frameworks were in place; and (3) funding and procurement models were available. Relatedly, another technology provider argued that the addition of environmental criteria would need to be communicated to the industry ahead of time because such criteria must be addressed early on during the technology design and development stages. As such, suppliers should not be presented with a fait accompli once production and commercialisation have begun and only informed when tenders are made public.

While the government has a role to play in creating market demand for green technologies, a technology provider warned that the disruptive nature of such a change for companies should not be underestimated. According to a manager, well established companies are reluctant to change their advantageous and profitable business models. Their financial and human resources hold significant influence over political decision-making and over the content of the laws and regulations being adopted. Corporate strategies and market dynamics that largely extend beyond the healthcare system level are thus important to consider. It is expected that established companies will react to a (re)shaping of public markets by trying to continue to capture these markets or by ensuring that they benefit from the new opportunities offered by eco-friendly DHTs.

Professional and scientific factors

Anchored in the medical perspective, respondents discussed a professional enabler and a scientific constrainer. A manager reported that clinicians require clear medical protocols and practice guidelines as well as clear medico-legal liability frameworks to green their clinical practices. Towards this end, professional federations and colleges could play an essential role in implementing credited training programs and in producing guidelines and directives supporting the use of green medical technologies and practices. By legitimising greener care delivery, professional bodies could help reduce clinicians’ concerns over risks that may arise with more eco-friendly technologies. Though clear protocols and guidelines are necessary to get clinicians on board, a clinical manager explained how the current lack of evidence and the difficulty of measuring the environmental impact of green DHTs (e.g., opaque global production chains, lack of established calculation methods) are important constrainers. Because the clinical field is largely rooted in the “evidence-based medicine” paradigm, regulatory bodies, decision-makers, and clinicians require solid scientific evidence to change their practices. The respondent warned that, without quantifiable data, it will be difficult to change clinical practices.

Discussion

Contributions of the study and implications for practice and policy

Our findings point to the complexity of greening DHTs in a high-income country, given the factors and conditions needed to overcome practice-oriented obstacles at different levels of healthcare system governance. In this sense, our study shows that a moral imperative is insufficient to convince stakeholders to embrace such a transition. As reported by a body of literature on the diffusion of innovations, our study points to the importance of considering the multiple changes required to green healthcare: individual working practices and methods, organisational structures and processes, as well as policies, governance, and regulatory and legal frameworks.^37,63,83

More specifically, our study clarifies the perspectives of multiple stakeholders and elucidates key micro-meso-macrosystemic factors required to green DHTs in a developed country's public healthcare system: (1) innovation attributes that include stakeholders’ recognition of climate change and whether they can address the environmental impact of DHTs; (2) internal system factors that revolve around the organisation's culture, policies, practices, expertise, and professional skillsets; and (3) external system factors that include political, economic, professional, and scientific issues. Below we discuss the different factors in relation to the existing literature on the topic.

Stakeholders’ perspectives on green digital health technologies

Scholarly work has shed light on how an individual's attitude towards and knowledge of an innovation affect if and how they will adopt the innovation.^37,63,83 In our study, though respondents were aware of climate change issues and agreed that the healthcare system should implement green actions, their perspectives diverged on whether DHT assessment and procurement decisions should consider the environmental impact. Several respondents argued that considering the environmental impact is unjustified as they either prioritised the quality, effectiveness, and safety of DHTs over broader environmental concerns or argued that, on the contrary, DHTs can help green the healthcare system. Both perspectives are supported in the literature. Previous work has demonstrated how health concerns are often dichotomised from and prioritised over environmental concerns and a growing body of literature reports on the digital industry's efforts to tackle climate change.^17,84–88

The literature on DHTs also argues that, in addition to reducing the carbon footprint of healthcare (e.g., virtual care eliminating transportation-related pollution), DHTs offer substantial financial benefits while improving access to services for patients.^89,90 Yet, this view may underestimate the environmental harms caused by the digital industry as a whole.^{28–31,89,91} Diffusion of innovations scholars have underscored that, to facilitate the adoption of an innovation, it must be compatible with stakeholders’ values, norms, and needs, and its relative advantage and benefits must be very clear. ³⁷ In the case of our study, the innovation is the consideration of the environmental impact of DHTs in assessment and procurement decisions. At present, stakeholders are not exposed to the relative advantages and benefits of green DHTs and the latter may not align with their healthcare priorities as they are mainly involved in cutting-edge DHTs (e.g., AI) that promise to improve access, quality, and efficiency, while reducing (at least theoretically) healthcare's environmental footprint.^{22,89,91–99} Consequently, this context makes it difficult for stakeholders to consider the potential negative environmental impacts of a DHT that improves patient care. For example, clinicians will not adopt a green DHT if they are not convinced that the technology will reduce the environmental footprint of their practice without sacrificing quality, efficacy, and safety. That said, individual commitment to green DHTs is insufficient if healthcare organisations are not prepared to adopt them.

Organisational preparedness

According to Rajagopalan et al., ¹⁰⁰ the success of an organisation's green transition largely depends on whether its structures and processes support the green vision and strategy by promoting a greener work culture and scaffolding employees’ green actions. In line with our study results, organisational leadership, a strong green strategy, adequate resources, and in-house capabilities and expertise are recognised as major challenges to procure green DHTs.^61,100–104 Consequently, considerable resources and investments may be required to integrate the environmental impact of DHTs in assessment and procurement decisions,^61,105 thereby leaving the organisation unprepared for this transition.

This is the case for the City hospital, and many other healthcare organisations in Quebec, because of a series of austerity-led reform measures in recent years. These reforms have largely compromised the ability of Quebec's healthcare organisations to innovate and engage in transformative initiatives that require significant financial, material, and human resources and investments. Nevertheless, one of the most established ways of driving the ecological transition in DHTs is to prioritise green procurement as part of the organisation's overall green strategy.^100,101,105 This requires organisations to abandon purchasing practices based on the initial cost of the technology. ⁶¹ The organisation will not be able to do this unless it can demonstrate that, with at least the same level of efficiency, quality, and safety, green DHTs that require a higher purchasing cost are more cost-effective than cheaper standard technologies.^61,106 Because the green digital market is nascent, data supporting these types of cost-benefit analyses are lacking. This, in turn, adds a layer of difficulty for public healthcare organisations aiming to purchase green DHTs as the Ministry of Health closely scrutinises their spending.

Another factor affecting public healthcare organisations’ preparedness for a green transition is the lack of in-house expertise on eco-responsible healthcare, including a shortage of specialists in environmental impact assessment within HTA teams (e.g., economic evaluation; efficacy, quality, and safety of DHTs), eco-responsible procurement specialists, and sustainable development managers.^61,105 In the same vein, training and awareness-raising among the organisation's staff, particularly clinical staff, on the environmental impacts of healthcare in general is important.^37,100 As reported in this study, if clinicians are reluctant or refuse to use green technologies, there is little chance that the organisation will acquire them. In this respect, clinical leadership within the organisation is essential to help clinical teams and management staff change their perspectives on green DHTs.^37,100

However, procurement of green DHTs remains challenging because of macrosystemic obstacles. As observed in our study, though the City hospital has internal environmental strategies and policies, its status as a public healthcare organisation makes it subject to national governance, laws, and regulations that limits its capacity for action.

Macrosystemic preparedness

The macrosystemic environment (e.g., socio-political, institutional, regulatory, economic, and governance) plays an important role in the green transition of a healthcare organisation.^16,107,108 The latter requires the support from and a synergy with healthcare system policies and orientations, and other concerned stakeholders (e.g., market and industry, research and innovation, professional colleges and federations, insurances, patients, civil society).^{16,61,100,109,110} Below we discuss five spheres of macrosystemic influence.

First, because public procurement accounts for 15–30% of gross national product worldwide, the government could oblige technology suppliers to develop and market green DHTs by encouraging healthcare organisations to purchase them.^61,100,111 With its purchasing power, the government plays a crucial role to generate demand for specialised green products that companies will not develop if the economic model (e.g., purchases, reimbursement) does not encourage them to invest in this field, or if it reduces profit margins.^112–115

Second, considering the environmental impact of DHTs will require trade-offs between product costs and environmental criteria as green DHTs are more expensive to produce. ¹¹⁶ Such a choice is difficult in a context where decision-makers are focused on controlling rising healthcare costs.^113,115 That said, several governments have already introduced policies to support green procurement (e.g., UK, Italy, Belgium).^61,116–119 For its part, Quebec has a sustainable development strategy for the province but lacks a clear green procurement framework.^117,120 To remedy this situation, several specialists have called on the Quebec government to: (1) capitalise on its “strike force through its purchasing power” and revise its public procurement policies and framework to foster green procurement and (2) require environmental impact assessments to ensure that the “Sustainable Development Act” is actually implemented. ¹²⁰ On this last point, the literature reports that to accelerate the green transition, governments should require healthcare organisations and industry players to render their carbon footprint data public.^100,113

Third, though the current regulatory and political context in Quebec is constraining the green transition in healthcare, some organisations have already implemented internal environmental initiatives and strategies thanks to strong organisational leadership. For example, the Integrated Health and Social Services Centre in Laval, a large health and social care organisation, developed a clear strategy and objectives for its green transition, assessed its total GHGs emissions, and is in the process of reviewing “the use of single-use supplies [and ] over-packaging”, adopting a “responsible procurement policy and a supplier code of conduct”, and “training super-users in responsible procurement”.^121,122 The literature on the diffusion of innovations has demonstrated that if one organisation successfully adopts a change, comparable organisations will be more likely to adopt it as well. ³⁷ As such, if the government were to recognise this organisation's environmental practices as the norm through supportive policies and directives, other organisations could more successfully take on a green transition.

Fourth, guided by the “evidence-based medicine” paradigm, healthcare systems will only adopt green DHTs if there is robust scientific evidence on their quality, safety, and efficacy, and on their contribution to healthcare objectives.^{21,89,97–99,123} Kaack et al. ³⁰ emphasised the importance of assessing the environmental impact of DHTs at a systemic level, rather than focusing on their marginal effects. This is important to avoid their rebound effect, which occurs when a technology provides real benefits for individuals, but generates other negative impacts (e.g., societal and environmental) at the population scale.^30,124 For Hensher, ¹²⁵ environmental considerations should be integrated into the economic evaluation of health technologies in order to convince decision-makers and other stakeholders. To this end, several HTA agencies are adapting their assessment methodologies and guidelines to better integrate environmental sustainability (e.g., Canadian Agency for Drugs and Technologies in Health, National Institute for Health and Care Excellence, International Network of Agencies for Health Technology Assessment).^{109,113,114,118,126} However, such a change is complicated by the current lack of: scientific consensus on the approaches used to integrate environmental impacts in HTA as standardised methods for quantifying and qualifying these impacts have yet to be developed^109,115,127; environmental data available for the entire lifecycle of the technology^{109,113,114,125,128}; multidisciplinary expertise to carry out data analyses ¹⁰⁹ ; and commitment from the industry to publish their data.^109,129

Finally, healthcare professionals require a clear positioning from their professional colleges and federations to adopt green DHTs. Medical colleges and federations exert great influence by training professionals, producing clinical practice guidelines, and clarifying medico-legal responsibilities.^16,130 As such, their disengagement is recognised as an obstacle to the integration of health innovations.^16,37,42,63 In Quebec, their “conservatism” on the environmental question contrasts with healthcare professionals’ growing awareness of healthcare's environmental impact.^{109,131–133} According to Khan et al., ¹³⁴ incorporating sustainable health principles into clinical guidelines and protocols produced by medical colleges and federations can motivate clinicians to integrate environmental considerations into their practices (e.g., choice of technology use). The lack of clinical guidelines and protocols to support and encourage such a change is recognised as a major obstacle (e.g., quality, medico-legal liability). ¹³⁴

Strengths and limitations

Given its exploratory nature, the study does not claim generalisability. It rather aims to foster discussion and reflection on the micro-meso-macrosystemic implications of addressing environmental issues in DHTs. Therefore, the transferability of our results must be contextualised, as healthcare organisations and systems differ widely across jurisdictions and countries. For example, the City hospital operates in a publicly funded healthcare system, which makes it different from privately funded healthcare organisations and systems. Furthermore, our study took place in a developed and established healthcare system. Thus, the level of stakeholder recognition and interest in environmental issues could be very different in healthcare systems in low- and medium-income countries.

Several potential interviewees could not be reached or were unable to participate in the study due to their active involvement in managing the pandemic (e.g., decision-makers, managers, clinicians, women running technology companies). Nonetheless, the participants’ rich and diversified experiences and expertise provided a substantial body of data that allowed us to generate a deep understanding of the topic under study. In addition, even with ongoing communication and critical discussions with PL and LR to ensure the solidity, credibility, and reliability of the entire research process, the fact that the first author (HA) piloted the study from start to end (e.g., recruitment, collection, analysis) could constitute a limitation. However, the team members’ extensive experience in qualitative research, which allowed us to adopt a reflexive stance towards the participants and the data, facilitated a deeper and more nuanced immersion into the data. In this regard, the adoption of a rigorous research approach, informed by theory and guided by sound qualitative methods, contributes to the solidity, credibility, and reliability of our results.

Conclusion

Digital technologies are expected to play a central role in twenty-first century healthcare systems. However, their contribution to improving care and services for populations should not come at the expense of the environment. Because there is growing agreement that DHTs contribute to pollution and climate change, sustainable healthcare systems must consider the environmental and social impacts along the entire lifecycle when assessing the added value of these technologies for patients and the population. As reported in this study, the achievement of such a change depends on multiple micro-meso-macrosystemic factors that involve diverse stakeholders and governance levels with, sometimes, diverging interests, perspectives, expectations, and objectives. Based on an innovation diffusion theory framework, our results provide an original contribution to the scientific literature, highlighting the importance of paying more attention to the complexity and particularity inherent in healthcare organisations and systems. It emphasises the fact that the green transition in healthcare, far from being a simple technical issue and/or “intrinsically justified”, cannot be achieved without enabling conditions and factors at different levels of healthcare system governance. Our results also provide a useful empirical basis for future studies on the integration of environmental concerns into the development, assessment, procurement, and use of DHTs in healthcare organisations and systems.