Over the last two decades, large-scale digital health programs and services, such as electronic health records (EHRs), have proliferated with equal access, improved quality, and resource optimization as wider health policy goals. National eHealth policy programs have gained attention not only for their benefits, but also for several unintended consequences and failed expectations. Given the mixed, complex results of these programs, questions are increasingly being asked about the executive power of political-governance strategies and how large-scale digitalization programs are governed to achieve the health policy goals of quality improvement and equal access with the necessary digital transformations.

In this study, based on a qualitative systematic review, we take a closer look at how governance has been implemented and considered in the realm of digitalization in health care by scholars in health care and the health sciences. The objective of the review is to arrive at an informed, recognizable conceptualization and consideration of governance. By summarizing existing and emerging models and the ways they are considered, including suggestions for improvement, we analyze the features of each model and how they possibly interfere with each other. In turn, we hope to inform the discussion on improving governance to obtain eHealth policy goals.

In a 2013 editorial in the British Medical Journal, the English national program for information technology (IT) was discussed 10 years after its inception [1]. This program promised to revolutionize care in the English National Health Services (NHS) and was originally scheduled to run for 2 years and 9 months starting in April 2003 [2].

The editorial summarized the results of the national program for IT and discussed its flaws and benefits. We draw attention to the following conclusion on governance: “There have been substantial improvements in the technical knowledge base underpinning information systems in the NHS, in organizational capacity to introduce any new IT system, and in information governance processes and procedures. Ten years on, only a handful of hospitals can be described as paperless, and most communication between NHS organizations still occurs by snail mail, fax, or patient messenger” [1].

Governmental opponents of the program described it as a huge disaster, but in the final benefits statement released by the Department of Health, its proponents largely described it as a success. They predicted that by 2022, financial benefits would be £10.69 bn (US $13.86 bn), outweighing the cost of £9.78 bn (US $10.09 bn).

Similar accounts of the contested, intended, and unintended results of eHealth programs were described in the United States in 2019: “The US health system has recently achieved widespread adoption of EHR systems, primarily driven by financial incentives provided by the Meaningful Use (MU) program” [3]. Although successful in promoting EHR adoption and use, this program and other contributing factors also produced important unintended consequences, such as failed expectations, EHR market saturation, innovation vacuum, physician burnout, and data obfuscation, with far-reaching implications for the United States health system [4]. To avoid unintended consequences, these authors proposed improved governance, including efforts from diverse players such as health care providers, administrators, health information technology (HIT) vendors, policymakers, informatics researchers, funding agencies, and outside developers. The authors also argued for the promotion of new business models, collaboration between academic medical centers and informatics research departments, and improved methods for evaluating HIT.

These accounts of mixed, complex results, challenges to the realistic calculation of benefits and costs, and implementations of intentional changes constitute the background of this paper. Perhaps unintended consequences could have been minimized through stronger, better governance of these processes in England and the United States.

At the outset, we understand health care governance as a political process regarding national eHealth goals. National health care services differ between countries and regions, and they are broadly categorized into 2 systems: single-payer and multi-payer systems [5]. National authorities are responsible for governing their respective national health care services, including eHealth governance strategies that are often top-down through legal regulation and reimbursement schemes, where resources are allocated to prioritized service providers. However, political governance can be driven by evidence-based medical imperatives, which are bottom-up strategies [6]. National eHealth policy adjustment in such cases occurs according to new medical evidence.

Governance is extensively discussed in social and political sciences and is one of the most common social science terms [6]. One prominent definition is “an interactive process through which society and the economy are steered toward collectively negotiated objectives” [6]. Political goals are normally the result of compromises obtained through collective negotiations between stakeholders. “Public governance should thus be considered as composite and mixed with inbuilt tensions between competing concerns, actors, ethics, resources, and time horizons” [6,7].

This paper considers eHealth governance as a dynamic process in which development, decisions, implementation, evaluation, and adjustments overlap and interact. The development grows out of conscious and unconscious interdependencies. To realize deliberate policies in such interdependent situations, collective negotiations might be increasingly important. In Norway, the word “co-management” captures this process, which blurs the strict distinction between bottom-up and top-down governance.

One group of researchers has emphasized the need for more empirical studies in the field of governance to strengthen both conceptual and empirical knowledge [8].

We report on the results of a systematic, qualitative review of how eHealth governance has been conducted and experienced. We investigate how governance has been addressed and has played out in eHealth programs as described in the health sciences literature. We do not aspire to compare different enactments of governance, but instead, perform a descriptive analysis by conceptualizing the characteristics of governance and the experiences presented in the cited papers. We also discuss how different models interfere with each other and summarize suggestions for improvements.

We have restricted our search to PubMed because digital health is a popular emerging topic, and it is of interest to learn how governance is addressed and considered in health sciences. On the basis of political and social sciences, we contribute to a cross-fertilization of scientific traditions by communicating with practitioners, medical internet researchers, and political health authorities.

How do the papers included in the qualitative review define and consider governance?

The objective is to arrive at informed, recognizable conceptualizations and considerations of governance models of eHealth, as presented and discussed in the scientific literature. In turn, we hope our results will help inform the discussion on how to govern such processes to obtain collectively negotiated objectives.

The remainder of the paper contains an account of the methods used, quantitative and qualitative results, and a discussion section in which we distinguish between and elaborate on models based on our constructs from interpreting the empirical results of the cited papers. We discuss suggestions for governance improvements, present our conclusions, and point to areas for further research. We also discuss the limitations of this review. Due to the coronavirus pandemic, we have provided a postscript that connects our models and conclusions to possible (eHealth) governance in the COVID-19 era.

A qualitative systematic review is a method for integrating or comparing with the findings from qualitative studies. It looks for themes or constructs that lie in or across individual qualitative studies and may employ selective or purposive sampling. This type of review produces a narrative synthesis with a thematic analysis and includes interpretive, conceptual models. The accumulated knowledge resulting from this process may lead to the development of a new theory, an overarching narrative, a wider generalization, or an interpretative translation. The goal is not aggregative in the sense of adding studies together, as with a meta-analysis; rather, it is interpretative and broadens the understanding of a particular phenomenon [9].

Our search was performed in November 2018 using the previous version of PubMed. PubMed was updated in the spring of 2020.

We searched “All fields” with the following search criteria:

Search (“governing reforms” OR (telemedicine governance) OR (governance “national telemedicine programme”) OR (governance “national ehealth programme”) OR (governance “national ehealth programme”) OR (governance innovation ICT) OR (“innovative procurement” health*) OR (“innovative procurement” health) OR (“whole system demonstrator programme” lessons) OR (Sundhedsplatformen) OR (Governance “regional EHR implementation” Denmark)).

The search was restricted to scientific papers, including systematic reviews published in peer-reviewed journals. A systematic review was defined as an overview with an explicit question and a methods section with a clear description of the search strategy and methods used to produce the review. The review was also expected to report on and analyze empirical data. As many papers were retrieved (220), we included only reviews and papers published from 2010 onwards in the final review for pragmatic reasons.

The governance of general health reforms, innovation programs, public reforms, and innovation programs was excluded, as were conference papers, dissertations, proceedings, and irretrievable papers.

Articles were stored in the free software Rayyan, where we inserted the selection criteria and read the abstracts [10]. On the basis of the inclusion and exclusion criteria, two reviewers independently screened the list of the titles and abstracts from the literature searches and identified potentially relevant studies, the full-texts of which were then retrieved. We resolved any disagreement through dialogue between the two reviewers based on the selection criteria. The citations were then exported to EndNote.

Data collection was carried out using a web-based data-extraction form created by the authors in Google Docs, which is enclosed in Appendix 1. The 2 authors collected data based on full-text papers. The following quality domains were assessed to identify, include, and critically appraise the studies: explicit accounts of data-collection methods, explicit accounts of methods used to analyze the findings, and overall assessments of the qualities of the papers and reviews.

We analyzed the papers from a grounded perspective by interpreting the qualitative data in the included papers. Grounded theory is a form of empiricism that emphasizes inductive reasoning and hypothesis generation, in contrast to the hypothetic co-deductive model of the positivist scientific method [11]. Our interpreted models can be used as working hypotheses for further empirical development and research.

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses diagram below (Figure 1) presents the quantitative results [12].

In the background section, we briefly indicated 2 broad governance strategies: political governance, which normally depicts top-down processes and medical governance, which normally depicts bottom-up processes. On reviewing the 11 papers, we found it difficult to subsume all of them under either political or medical strategies.

Therefore, we constructed four governance models to categorize and conceptualize our findings. We chose to construct models because “Models seek to simplify phenomena as an aid to conceptualisation and explanation” [13]. They are also tools to realize a narrative synthesis, which includes thematic analysis and interpretive, conceptual models.

The 4 models are: political governance, medical governance, internet and global business governance, and self-governance. We briefly describe the models below before sorting the different papers into them.

The following 4 papers align with the political-governance model: Atalag addresses a single content model for eHealth interoperability and secondary use in New Zealand [14]; Park and Atalag address the national approach to health care ICT standardization and focus on progress in New Zealand [15]; Kierkegaard addresses interoperability after EHR deployment and describes persistent challenges and how the Danish government applied a combination of bottom-up and top-down strategies to realize full-scale EHR implementation [16]; de Riel et al [17] address success factors for implementing and sustaining a mature electronic medical record, iSante, in a low-resource setting in Haiti (Table 1).

In the political-governance model, eHealth is governed top-down through a national eHealth strategy. The papers included in this model describe national top-down strategies on technical infrastructure, legal frameworks, and institutional structures, such as national advisory boards or a multi-editorial expert group [15,16] (Table 1).

Atalag describes the national top-down strategy of New Zealand as a “middle-out transitional approach to achieving semantic interoperability in eHealth” [14]. Kierkegaard describes how the Danish authorities combined top-down and bottom-up approaches to realize national eHealth goals: “Changes in the organizational setup and redistribution of responsibilities between the Danish regions and the state play a pivotal role in producing viable and coherent solutions in a timely manner” [16].

The following 4 papers align with this model: Wade et al [18] presented a qualitative study of ethical, medico-legal, and clinical governance matters in Australian telehealth services; Sutton considers a Picture Archive Communication System and diagnostic imaging service delivery from the UK perspective [20]; Bagot et al [21] performed a qualitative analysis comparing with the experience of Australia and the United Kingdom in integrating acute stroke telemedicine consultations into specialists’ “usual practice”; and Crocker et al [19] address patient safety and image transfer between referring hospitals and neuroscience centers, exploring possible methods of improvement.

The bottom-up strategy may grow out of local needs [18,20], and sometimes it is picked up and transformed into a national top-down strategy. For instance, Crocker et al [19] claim that “Part of the remit of the National Programme for Information Technology in the NHS (NPfIT) was to improve neuroscience teleradiology.” The researchers wanted to evaluate whether their experiences were part of the national top-down strategy; so the paper describes how the medical society experienced their recommendations presented in an NHS report from 2004 being “largely ignored” in the national program. National top-down strategies do not necessarily align with bottom-up strategies [19].

Wade et al [18] state that, from a medical point of view (bottom-up), concerns about data security in telemedicine need to be governed through national top-down guidelines.

The following 2 papers align with this model: Mackey et al [23] addressed health domains for sale and argued for the need for global political health governance on the internet, and Mackey et al [23] discussed a call for a moratorium on the top-level domain “generic health” to prevent the commercialization and exclusive control of web-based health information [22].

These authors argue that the national jurisdiction of privacy and data protection is largely ignored and challenges existing international governmental organizations, such as the World Health Organization and European Union [7]. According to the authors, the business interests of large actors and market mechanisms overrule both political and medical governance.

Mackey et al [22] claim that the processes of the Internet Corporation for Assigned Names and Numbers (ICANN) “appear to favour business interests and generation of profits over the future integrity of the Health Internet, failing to make any tangible commitments to protect public health or enforce norms as would be found in a responsible global governance Framework.”

However, they also state that “Direct-to-consumer advertising of prescription products is not allowed in the vast majority of countries other than the United States and New Zealand, and it may be unlawful for pharmaceutical manufacturers to engage in multijurisdictional web-based advertising that could occur through future gTLDs (generic top-level domain name)” [23]. There is a lack of global governance in health that may challenge human rights in public health [23].

We have included one paper in this category by Williams et al [24], who discussed governance in terms of individual agency as opposed to political governance. They also envision the opposite of individual agency: the ways in which digitalization and the use of technological affordances change individuals and identities. The paper describes a complex model, a dynamic dualism between the governance of individuals by eHealth versus the self-governance of eHealth.

We have presented the included papers under 4 governance models: political governance, which is normally limited to nation-state jurisdiction, international, evidence-based medicine and health care guidelines, large-scale industries governed by business goals, and self-governance by consumers.

Table 1 lists the 11 papers by author, year published, country in which the governance model is discussed, and model or strategy. Some papers do not refer to a specific country because they describe global models or models related to the consumers’ use of the IoT products and services.

Table 2 lists the included papers by author, year published, country in which the characteristics, experiences and challenges are described, and the suggested improvements related to a specific governance strategy. Three papers do not relate to a specific country but describe aspects related to global governance and self-governance models.

This section focuses on the features and considerations of the different models, the ways in which they interfere with each other, and suggestions for improved governance.

Scholars in the political-governance category articulate the need for more documentation of the successful, locally-led governance of donor-funded systems, including any capacity-building for local responsible entities and joint system design, planning, and implementation [17]. Experiences with political strategies point to a demand for ongoing efforts to improve them, with stronger political executive power building on medical relevance to achieve wider goals.

In the medical-governance category, scholars point out the need for continued monitoring of a more robust image-routing system across the NHS and audits of its functionality by the Connecting for Health Safety Team and the National Patient Safety Authority to combine bottom-up and top-down strategies [19].

In support of this argument, Sutton claims that “There is much hope that despite the current economic climate, the established IT programmes in the UK are able to continue to facilitate the development of new solutions. In 2010, the Department of Health in England adopted the ‘Quality, Innovation, Productivity and Protection’ agenda to maximize the benefits of existing IT systems and promote improved patient care and productivity” [20].

In a paper concerning ethical and legal matters in Australian and UK telehealth, the authors reported that Conducting organizational case studies would give a deeper understanding of the matters identified, particularly those of governance and system change” [18]. In their comparison of the integration of acute stroke consultations and specialists’ usual practice, the authors suggest that “Future research might investigate the transferability of UK and Australian experiences to broader European, Asian and American networks” [21]. In general, medical culture is similar across national borders.

The medical society is global and agrees on evidence-based medicine being the core driver of medical diagnoses and treatment procedures. However, integrating telemedicine into care requires juggling international medical culture and local cultural variations in medical delivery. Medical-governance strategies need to embed evidence-based medicine and telemedicine into different organizational set-ups. Incremental “juggling” is considered useful for solving organizational challenges.

Regarding internet and global business governance, the authors of the two reviewed papers suggest that “Focusing on the public good can be a first and crucial step to ensure an accurate, reliable, and evidence-based online presence for health for this generation and the next” [23]. They recommend that the “Internet community needs to be vigilant to ensure the reliability and trustworthiness of health information online, and take immediate action to secure the future integrity and proper governance of this important namespace for the health internet” [22]. The authors hope to encourage ICANN to appoint the World Health Organization and a multitude of governmental and health nongovernment organizations as sponsors. This presupposes that the World Health Organization has executive governance power with the ability to overrule the global internet market and the political governance within each member state. They express a need for political governance to dominate global business strategies.

The papers propose solutions for “future e-Health governance approaches to ensure the appropriate management of health could be accomplished by requesting ICANN to recategorize health as a sponsored gTLD and proactively appoint WHO as its sponsors” [23]. They suggest that the World Health Organization should develop policies to ensure accountability and transparency in gTLD operations that meet the best interests of the global health community and enforce eligibility rules regarding all future health registrants.

They propose that the World Health Organization’s possible appointment as a gTLD sponsor should be governed by a diverse, globally representative board of health stakeholders in partnership with responsible internet service providers. “This governance mechanism can have representation and be organized into subject-specific advisory panels to review and recommend content to be included on for health” [23].

In the self-governance category, consciousness and data activism is important, which resonates with an argument from Alan Peterson, who calls for “algorithmic accountability’ which highlights the potential for citizens to create alternative futures—ones oriented to fulfilling human needs rather than techno-utopian visions” [31].

By limiting our inclusion criteria to 2010 onwards, we might have missed important papers that could enhance the review. Qualitative reviews are not all-embracing, and the results are colored by the search criteria. Nevertheless, our paper provides valuable insights into various governance strategies at play in the realm of eHealth and valuable suggestions for further work to improve governance and scientific knowledge.

The current coronavirus pandemic might provide an excellent illustration of our findings and main conclusions.

To govern and control the pandemic, collective, transparent efforts are in demand: political decisions followed by financial incitements, firm global medical knowledge and advice, international business actors’ financial and ethical decisions, and personal decisions by citizens to adhere to the advice. Following a process when the collectively negotiated goal to fight the pandemic is at stake, puts strong demands on each of these “nodes” in dynamic co-governance.

We have identified 4 different governance models linked to national eHealth programs, national health ICT infrastructures, regional and local professional networks, Health Internet businesses, consumer-driven self-management solutions, and virtual health services beyond national health services.

The political model depicts governance through various management strategies to influence wider health political goals and innovation. The model also enacts eHealth governance by establishing financing schemes, infrastructures, standards, and laws. Stronger leadership, by involving stakeholders in national health IT boards, is suggested to achieve goals. The medical-governance model enacts governance within and by eHealth communities grounded in evidence and trust. The authors of papers addressing this model argue that evidence-based medicine should be the basis for the development of political-governance strategies. The internet and global business model put international business actors in dominant roles in eHealth governance, challenges the jurisdictions of nation-states, and limits the influence of international health actors, such as the World Health Organization. The authors of papers that address this model envision a dystopian future, especially for the governance of health data and services on the internet, and they argue for action by global health authorities. In the self-governance model, governance is characterized by shifting, dynamic tensions: governance and control of the individual by their submission to eHealth versus self-governance within eHealth by empowered, conscious citizens.

On the basis of this review, we conclude that to achieve health policy goals in large-scale eHealth policy programs, collective negotiations between nation-states, global policy actors, medical and self-governance actors, and global business and industry actors are essential. Digital technology affordances present opportunities for both benefit and harm concerning the realization of health policy goals, a dynamic that future studies should scrutinize. Technological affordances and both optimistic and pessimistic views deserve serious consideration.

According to our findings, further research is needed to produce knowledge about