Benefits realization in digital transformation: the translation from policy to practice in health care

Abstract

Purpose

While digital transformation holds immense promise, organizations often fail to realize its benefits. This study aims to address how policies for digital transformation benefits realization are translated into practice.

Design/methodology/approach

The authors apply a qualitative, comparative case study of two large, public-sector health care organizations in Sweden. Through document and interview data, the authors analyze the process of translation.

Findings

The study finds that practice variation is primarily caused by two types of decoupling: policy-practice and means-ends. Contrary to previous studies, coercion in policy compliance is not found to decrease practice variation.

Research limitations/implications

The limitations primarily stem from the empirical selection of two large public health-care organizations in Sweden, affecting the study’s generalizability. Reducing practice variation is more effectively achieved through goal alignment than coercion, leading to implications for the design of governance and control.

Practical implications

Policymakers should, instead of focusing on control-related compliance, work to align organizational objectives and policies to decrease practice variation for successful benefits realization.

Social implications

The study contributes to better benefits realization of digital transformation initiatives in health care. As such, the authors contribute to a better functioning and more transformative health care in times of increased demand and decreased supply of health-care services.

Originality/value

The study challenges conventional wisdom by identifying that coercion is less effective than goal alignment in reducing practice variation, thereby enhancing the understanding of policy implementation dynamics in health-care settings.

Citation

Isik, L., Nilsson, C., Magnusson, J. and Koutsikouri, D. (2024), "Benefits realization in digital transformation: the translation from policy to practice in health care", Transforming Government: People, Process and Policy, Vol. 18 No. 2, pp. 303-317. https://doi.org/10.1108/TG-11-2023-0177

1. Introduction

Digital transformation comprises a challenging but potentially rewarding endeavor for private and public sector organizations alike, resulting in significant investments currently being directed to digital transformation initiatives (Eom and Lee, 2022). While associated with significant potential benefits (Pekkola et al., 2022), research has also identified difficulties in attaining said benefits (Serra and Kunc, 2015).

The public sector is multifaceted and inherently complex, characterized by a distinct governance structure (Arundel et al., 2019) that diverges significantly from the market-driven principles (Williamson, 2008) of the private sector. Public organizations operate within a framework defined by legal regulations, political directives, societal norms, citizens accessibility and values. This sector is marked by the involvement of numerous stakeholders (Arundel et al., 2019; Jonathan, 2020), underscoring its unique operational ethos. In the realm of health care, digital transformation encapsulates a spectrum of initiatives aimed at augmenting operational efficiency among health-care providers, fostering patient-centered care models and addressing organizational, managerial and socioeconomic challenges inherent to the sector (Kraus et al., 2021). Furthermore, in health care, digital transformation must navigate the added complexity of managing highly sensitive data, necessitating advanced security and compliance measures to protect patient information (Jonathan, 2020).

Organizations engage in digital transformation in aspiration of certain benefits. These benefits vary from pure rationalization (i.e. reduced headcount through automation) to increased quality of services and products and changes in organizational culture (Pekkola et al., 2022). As argued by Wessel et al. (2021), digital transformation, however, impacts not only in the form of continuous improvements but fundamentally changes the very identity of the organization. From this perspective, the benefits of digital transformation are diverse and varying in nature. Digital transformation is fundamentally linked to benefits realization (Holgeid et al., 2021), as its success is measured by how well technological changes are converted into concrete, measurable benefits for the organization.

Investments in digital transformation increasingly demand effective realization of benefits (Zwikael et al., 2018), but projects successful in these terms can still fail to realize benefits (Chih and Zwikael, 2015; Zwikael and Smyrk, 2012). This gap has led to new models incorporating cultural and social factors (Nielsen and Persson, 2017). Despite efforts and focus on benefits management over the last decade (Breese et al., 2015), many digital transformation projects still fail to achieve expected benefits. The challenge lies in the complex dynamics of translating policy goals into effective actions (Prior et al., 2014). Albeit extensive effort and focus on benefits management in the last decade (Breese et al., 2015), projects today still fail to realize the predicted benefits. The challenge lies in the complex dynamics of translating policy objectives into incremental actions (Prior et al., 2014) that effectively realize the predicted benefits.

Digital transformation initiatives are often cross-functional, involve stakeholders from an ecosystem of actors, and it often takes time before benefits become visible (Doherty et al., 2012), which makes benefits realization even more challenging. While we see a clear pattern of isomorphism in the policies for benefits realization among organizations (Villadsen, 2011), we also see significant practice variations (Røvik, 2016).

While most organizations have explicit policies in place to assure benefits realizations (Serra and Kunc, 2015), there are reportedly significant variations between policy and actual practice (Prior et al., 2014). According to the sociology of translation (Latour, 1987), ideas such as benefits realization should not be assumed to mechanistically diffuse from policy to practice. Instead, the translation involved between policy and practice is a complex interaction between sender and receiver, associated with a repackaging and adaptation by the receiver (Czarniawska and Sevón, 1996).

Despite the wealth of digital transformation literature, there is still a gap in the research regarding the nexus between digital transformation and benefits realization, with few studies bridging this gap. This study seeks to bridge this gap by applying translation theory to analyze not just the enactment but, more importantly, the realization of benefits from digital transformation within the public sector. Our focus extends beyond the implementation of digital strategies to scrutinize how these strategies translate into tangible benefits, examining the often complex divergence between policy intent and actual outcomes in practice.

Drawing upon this brief rationale, our research question is:

RQ1.

How is digital transformation benefits realization distorted in the translation from policy to practice?

This study is operationalized through a comparative case study of two large public-sector health-care organizations in Sweden. Through a combination of content analysis of policies and interviews with key stakeholders in the organizations, we use the phased model of translation proposed by Pedersen (2007) to explore the translation from policy to practice in the two organizations. In doing so, we contribute new insights to the study of policy-practice translation in public sector organizations.

2. Previous research and theoretical framing

2.1 Benefits realization of digital transformation

Vial (2019) conceptualizes digital transformation as a broader process that goes beyond mere technological upgrades, enhancing an entity by inducing significant changes through information, computing, communication and connectivity technologies. Within this expansive transformational process, digital transformation projects are not the ultimate objective but rather serve as pivotal catalysts or foundational stages. These projects facilitate transformative shifts, laying the technological groundwork for comprehensive organizational change, thus acting as crucial enablers rather than equivalents of digital transformation.

Kraus et al. (2021) conducted a thorough literature review of 340 articles on digital transformation in health care. The results from the overview were divided into five categories:

1. Operational efficiency in health-care providers;

2. Patient-centered approaches;

3. Organizational factors and managerial implications;

4. Workforce practices; and

5. Socioeconomic aspects.

Among these articles, none had an explicit focus on how the benefits of digital transformation projects in health care are realized. This indicates a need for further exploration of the subject.

Benefits management, a discipline since the 1980s, provides insights into translating the strategic goals of digital transformation projects into actionable steps toward digital transformation (Serra and Kunc, 2015). Numerous sources, including academics, authors, consultants, stakeholder groups and certification organizations (Breese et al., 2015), have contributed to the vast amount of literature on benefits management. Despite the wealth of knowledge, there are still significant challenges in realizing the benefits of digital transformation (Marnewick and Marnewick, 2022).

Albeit extensive efforts and focus on benefits management in the last decade (Breese et al., 2015), projects still struggle to achieve expected benefits. This is evident in a study on digital transformation projects from 1994 to 2002 where only 26% of the projects achieved expected benefits (Doherty et al., 2012). This trend persists, as highlighted in Marnewick and Marnewick (2022), which reference studies by Standish Group and others indicating only 30% of digital transformation projects succeed in realizing the aspired benefits. These failed projects create significant financial, personnel and legitimacy losses for organizations (Marnewick and Marnewick, 2022).

Despite the wealth of knowledge available, there are still significant hurdles to realizing the benefits of digital transformation. Several factors contribute to this challenge, such as difficulties in integrating intangible benefits in the projects, difficulties in specifying measures of benefits, lack of leadership/governance and lack of focus on organizational changes and factors (Doherty et al., 2012). Previous research has also highlighted the need for more complexity and uncertainty factors to be incorporated in benefits management models (Breese et al., 2015).

2.2 The sociology of translation

The sociology of translation aims to increase our understanding of the “travel of ideas” as introduced by Latour (1987). Ideas are perceived as manifestations, able to travel through time and space, but reconstructed during the adoption process (Czarniawska and Sevón, 1996). The notion of translation opposes the more instrumental approach to the diffusion of innovation (Rogers et al., 2014), adding nuance to how practice may diverge from established policies.

Translation has received ample scholarly attention (Zilber, 2006), primarily within two key research streams, as noted by Nielsen et al. (2022). The first and most prominent stream investigates the adoption of new ideas at the organizational level of analysis. In this context, scholars like Røvik (2016), examining knowledge transfer (Hultin et al., 2021) and exploring lean management principles, use translation as a framework to theorize how ideas are adopted by organizations and how ideas change. The second research stream has studied the role of creators of ideas such as thought leaders, business schools and consultants on the broader organizational level of analysis (Czarniawska and Joerges, 1996).

Lindberg and Erlingsdottir (2005) perceive the process of translation as a progression from dis-embedding to packing (for the idea to be able to travel), followed by an unpacking and a re-embedding after the travel in the new context. The receiver carries out the unpacking and re-embedding, which entails reinterpreting of the original idea so that it aligns with the new context. In a study of the translation of eco-taxation in Scandinavia, Pedersen (2007) identifies translation as comprised of two major phases. During the first phase, the idea is institutionalized by the sender into a set of instructions, and in the second phase, the receiver cherry-picks instructions that best fit with the intent of use.

We depart from the premise that translation should not be viewed in isolation, but rather it should be understood as part of a broader, more dynamic translation process. In line with prior research on the travel of ideas across organizational and geographical boundaries (Czarniawska and Joerges, 1996; Wedlin and Sahlin, 2017), we acknowledge that our starting point—the policy—may actually serve as a waystation in a more extensive process of translation. Crucially, the formation of the policy itself can be influenced by a myriad of factors, including experiences, civil servant expertise, consultancy input and trends in both research and practice. Here, our focus is to enhance our understanding of how localized adaptation within our organization could potentially propagate further. For example, health care might disseminate its tailored version of the policy as a best practice, thereby extending the translation process. As a result, our study echoes the view that the movement of ideas is multidirectional rather than a linear translation process from origin to destination (Nielsen et al., 2014).

3. Method

We conducted a qualitative, comparative case study of two large public health-care organizations in Sweden during the spring of 2023 (Yin, 2014). The rationale for choosing health-care organizations was one of convenience, given that two of the involved researchers had health-care organizations as their primary research focus and part-time employment, i.e. assuring unrestricted access. The organizations operate under a governance structure that incorporates both political and administrative machineries.

Data were collected through the analysis of policy documents, encompassing a total of 11 documents for Organization A and 11 for Organization B. This included approximately 1,000 pages, covering budgets, policies, strategies and guidelines, see Table 1. These documents, free from sensitive content, adhere to the public access principle, ensuring their availability to all residents under the law of governing transparency.

In the analysis of policy documents, we used a targeted search strategy including key terms “digital,” “transformation,” “benefit,” “effect,” “efficiency,” “benefit realization,” “value,” “close-to-home care” and “innovation.”

In addition, 23 semi-structured interviews with four executives, nine middle managers, five unit managers and five staff in core organizational roles (i.e. nurses, strategists and project managers) were conducted, see Table 2. The rationale for choosing to interview both managers and core organizational roles was that we found that individuals at different levels were able to provide specific details and different nuances of practice in the two organizations.

The interviews, conducted in Swedish, consisted of 14 open-ended questions, focusing on digital transformation projects and their benefits realization in practice. Each interview lasted approximately 60 min was recorded, transcribed and translated into English. Structured questions were also included to capture interviewees perceptions of key success factors and challenges. Two questions used a numeric scale from 0 to 5 to assess the organization’s ability to realize digital transformation benefits and the level of organizational emphasis on these benefits. In addition, respondents evaluated the proportion of management projects that successfully achieved their predetermined objectives.

We used a thematic analysis approach (Braun and Clarke, 2006) to code and analyze documents and interviews, focusing on benefits realization perceptions in policy and practice and the necessary conditions for translation. Initial data coding and sorting were performed using ATLAS.ti, with collaborative discussions to refine the coding scheme. One researcher conducted interviews/material coding, consulting the second researcher for code verification. Our process model (Figure 1) was developed in three steps. First, we analyzed the translation process in both organizations using the initial coding scheme. Second, recurring patterns (e.g. legitimacy, accountability, vested interest) were identified. Drawing from Pedersen’s (2007) study on eco-taxation within institutional frameworks, we created a process model incorporating sociological isomorphism. Our model includes “Coercive” and “Legitimate” stages, reflecting institutional influences per Pedersen. Through abductive analysis of organizations A and B, we identified “adopted,” “adapted,” and “variation” stages, extending existing theories to public health-care policy-to-practice translation. Further details are in the discussion chapter.

4. Results

4.1 Policy

4.1.1 Organization A.

The 2023 budget delineates an economic climate that is more strained than it has been since the early 1990s. It emphasizes the need for prioritizing and allocating resources where they yield the highest benefit for residents. The budget states that all corporations and administrations are to implement operational changes that enhance efficiency so that funds can be redirected from administration and bureaucracy to core activities. The organizations finances will be scrutinized to ensure sustainable economic development. Examples of expected initiatives include “coordination, process automation, or other digital transformation efforts” (A-D2). The council allocates additional funds during the year for this, emphasizing the importance of vigilance regarding the outcomes of the initiatives, “Before deciding on the financing of a measure, the expected effects should be described, which should then be followed up.” (A-D2)

Organization A’s Innovation and Digital Transformation Policy outlines its direction and intentions for digital transformation and innovation. While it does not explicitly mention benefits realization, it emphasizes deploying digital solutions for sustained benefits. The policy mandates compliance for decision-makers in innovation and digital transformation. In addition, there is an IT and digital transformation strategy that begins by defining the perspectives digital transformation should support within the organization. This strategy emphasizes that digital transformation enhances citizen welfare and aligns with the council’s budget goals. The term benefits realization is not used. However, it is stated that the organization “in order to increase the ability to realize benefits from digital transformation must focus on […] ”(A-D7). For the strategy, a less stringent approach is applicable, emphasizing the use of “should be taken into account” instead of the policy’s formulation, which implies its more authoritative nature. Under the heading “Patient Benefit in Focus” (A-D3) in the overview of the organization’s long-term efforts, the institution outlines its sustained commitment to digital transformation, emphasizing patient benefits: “This transformation allows employees to dedicate less time and effort to manual tasks, enabling them to focus more on patient care” (A-D3). Associated with this is a roadmap that lists needs, which are sorted and prioritized within a timeline. There is no mention of benefits realization.

Organization A’s project model, recommended for IT and digital transformation, supports the benefit realization phase as it includes a set of templates that also support post-project work. Importantly, the application of this model is not mandatory within the organization but is widely adopted. The model encompasses a definition of benefits, stating that in a project, benefits are defined as positive outcomes perceived by the stakeholders resulting from a change.

4.1.2 Organization B.

Organization B faces a strained financial situation due to factors such as high inflation and increased utility costs. Therefore, the organization has set the main objectives for 2023: resource management, efficiency improvement, increased productivity and reduced net cost. The administrative level in Organization B relies on the detailed budget document to achieve its various goals, and monitoring is implemented down to the lowest level through a specific analytics solution. This document describes the council’s goals from various documents and the administrative goals required to achieve them. There is a clear sense of urgency conveyed in the document, where the conditions for the administration’s planning are outlined in terms of staffing crisis, economic crisis and communication crisis.

Analysis reveals that goals related to the digital transformation of health care are included in the administrative goals, which are also followed up on, but nothing is mentioned about benefit realization or impact assessment. In parallel, the organization has developed a digital transformation policy that is based on the vision of “The Good Life.” The digital transformation policy outlines four main processes and emphasizes the importance of administrative simplification, intraorganizational collaboration and the role of each management team in driving the necessary changes to enable digital transformation and realize its effects within the organization.

The digital transformation strategy provides a vision for how health care interacts with patients and residents through digital technologies. The driving force for digital transformation within health care is described as stemming from patient and organizational needs, technical possibilities and trends. Organization B has a policy document that acknowledges “benefit realization” as a political area of priority established by the council.

Organization B has established procedural documents based on a common project model. One of the documents (cost benefit analysis) describes a methodology to be used for calculation, benefit realization, monitoring and allocation of responsibilities to ensure benefit realization. However, the respondents are not aware that this particular document exists. There are three distinct templates for projects: large/complex, medium and small. Only in large/complex projects is there a requirement to specify outcome goals, benefit realization and attach a cost-benefit calculation, detailing when monitoring should occur and who is responsible for realizing the benefits.

On the administrative level within Organization B, there is a divergence from the model due to lack of knowledge of the existence of a model for benefit realization. This administrative level uses its own proprietary benefit calculation method. All investments exceeding €50,000 are expected to comply with a process formalizing from request to decision and realization. Several respondents express that this specific process functions well and that the specified benefits are realized.

4.2 Practice

4.2.1 Organization A.

The organization has commenced its efforts to improve its ability to realize the benefits of digital transformation initiatives. Discussions are actively taking place within the organization’s management. According to one participant, the finance director provided a distinctive approach, emphasizing the necessity of a thorough cost-benefit analysis: “You need to do a benefit calculation on this. We cannot expect it to just happen and turn out well” (A9, top management “TMT” member). In response to this clarion call, the organization has custom-developed its own cost-benefit analysis framework, tailored to its needs while avoiding undue complexity. The aim is to equip the organization with a tool for prioritizing projects.

The new approach has increased the organization’s focus on benefits realization:

I also perceive that we are discussing benefit realization […], thanks to our initiation of conversations on benefit calculations. We have, in a way, sown this seed, and I believe we have genuinely begun to think, but it is somewhat dependent on the individual. (A9, TMT member)

Interviewees agree that the organization prioritizes benefits realization, although many criticize the absence of systematic approach. Or, as one interviewee expresses it:

We have a model […] we identify the effects on the operations and also the person responsible for realizing those effects. But my personal experience is that we follow up on it to a rather limited extent. (A10, heads of care unit management)

Another participant expresses skepticism about the effectiveness of the current approach:

Some people probably think that we have such a model, and maybe we do. But personally, I don’t think it works particularly well. We often end up with unclear benefits realization, which is regrettable, and that’s something we need to improve upon. (A2, TMT member)

The responses collectively underscore a common criticism, that is, the inadequacy or even total absence of follow-up after a project’s completion. Despite this, several respondents accentuated the significance of establishing a functional follow-up system. The neglect of such a system was attributed to various reasons, including the challenge of maintaining momentum post-project and the difficulty in identifying suitable measures to adequately assess benefits.

While the project model is partially adopted, 70% of the respondents are unfamiliar with it. Even if the project model outlines explicit responsibilities for benefit realization, it lacks clarity, leading to uncertainty about who is responsible. One Division Head expresses this ambiguity:

No, we don’t have that. No, in that case, it would fall on the VO-chief, mm. We don’t have specific, like, people who […] no. But I perceive that I have a responsibility in that matter myself.

The importance of creating benefit and value for patients in work repeatedly surfaced as a compelling theme: “The goal when working here is of course for the patients to somehow have it a bit better, […] that is after all the benefit.” (A5, heads of care unit management)

4.2.2 Organization B.

Organization B is confronting an economic scenario unprecedented in its recent history. As delineated in the preceding section, the local level possesses a document termed “detailed budget,” which explicitly stipulates that the organization is grappling with crises on multiple fronts: financial, communicational and human resources. The studied organization has a structure that has been established for the realization of benefits, with a chief executive responsible for benefit realization at the central level and implementation executives at the local level.

As noted, there is an aspiration to involve different levels and functions within the organization in benefit realization. Both central and local-internal projects are initiated throughout the year. There is a consensus among the respondents that the realization of benefits is of significant importance. The top management of the studied administrative organization prioritizes benefit realization work, and there is a formal structure in place to manage it:

We may be one of the administrations that have worked most purposefully and systematically with both benefit realization […], so for me, it is fundamental when deciding to do something new that it really creates increased value; otherwise, one should, of course, refrain. (B5, TMT member)

The administration uses a benefit calculation method developed in-house. It’s used for investments larger than €50,000, targeting the ex ante to ex post process. Several respondents highlight the model as successful. For the initiatives where the benefit calculation model is used, the administration’s top management serves as the stakeholder and follows up to ensure that realization has occurred. However, the same process is not applied for digital transformation initiatives that fall below €50,000.

There is a variation in the approach to benefit realization depending on the type of project or investment, contingent on diverging knowledge among coworkers as to the existence of the benefits realization model per se. Several respondents express that, according to the project model, the responsibility for benefit realization lies with the receiving organization or project initiator ex post. There is an understanding that benefit realization requires systematic follow-up and evaluation to determine whether the project has resulted in benefits. However, the respondents identify clear challenges in benefit realization ex post:

We have a very poor track record in the [organization], particularly in large IT projects […] not in any of the projects I have looked at, have the benefits been realized as stated. (B1, TMT member)

5. Discussion

Our study addresses the research question of how benefits realization is distorted in the translation from policy to practice. Our results show that both organizations have isomorphic policies for benefits realization, but their actual practices vary significantly. Through analyzing the translation processes of the two organizations, we have identified patterns in translation, resulting in a process model for how policy is translated into practice (Figure 1). We use the process model of translation to analyze and contrast how policy is translated into practice within the two case organizations.

In Organization A, policy documents suggest the importance of value derivation from digital investments. Yet only the framework of the project model explicitly addresses benefits realization. This project model, designed to shepherd initiatives from inception to realization, has been predominantly shaped by the central IT department. While other departments within Organization A have made sporadic and piecemeal contributions to its development, their influence has been comparatively minor. The project model is not coercive but rather serves as a voluntary guide, offering a collection of templates and instructions for the organization. This voluntariness has led to different levels of adoption across the organization; some parts of the organization use the model extensively, others selectively adopt parts of it and a few rely on entirely different models.

While the general idea of realizing value from digital investments is widely accepted and considered legitimate, its explicit manifestation through the project model is less uniformly endorsed. This is in line with Pedersen’s (2007) notion of “cherry-picking instructions” to better suit the local context. In Organization A, this has resulted in the development of local benefit calculators, tailor-made to meet specific organizational entities’ unique needs. Despite a central policy directive, the translation of policy into practice exhibits nuances and is susceptible to local adaptations (Czarniawska and Joerges, 1996; Pedersen, 2007), resulting in high variation in practice. Figure 2 contains an overview of the translation process of Organization A.

Organization B has an explicit policy for benefit realization. The document does not cover the needs from practice and instead refers to key performance indicators (KPIs) for benefit realization, i.e. is more control oriented (Uzunca et al., 2022). Even though there is an explicit document sanctioned by the political council which could be expected to be replicated (Røvik, 2016) due to coercion (Pedersen, 2007), it instead causes division and disagreement in practice. Because of this polarization (Nielsen et al., 2022), the idea is not deemed legitimate. Parts of the document are used at the local level, “hooking on the idea” (Pedersen, 2007) about the organizational structure and modifying (Røvik, 2016) it to their own context. The other part of the document lacks clarity on account of focusing exclusively on KPIs. Therefore, the organization acts according to Røvik’s (2016) radical translation process by designing a bespoke model that matches the practice needs. This translation resembles a game of “Chinese Whispers,” in which the original policy is reformulated and adapted to align with the core operations at the local level. The meaning of the policy significantly changes in this process, resulting in a practice that differs significantly from policy. The policy is therefore partially translated, and there is a high variation in practice (Figure 3).

Our findings accentuate the importance of a “modifying mode” in policy-practice translation (see Røvik), i.e. that an essential element of the originating context is retained and modified to function in the receiver’s context. While both organizations had established policies, they displayed variance in terms of translation to practice. The foundational principles were retained in both cases but with adaptations suited to each public health-care organization’s unique local setting. Notably, only specific components of the primary policy, like the benefits calculations and the assessment of potential benefits were deeply integrated into practice. Such selective adaptation underscores the inherent challenges in wholeheartedly adopting wide-ranging policies and the significant strategic modifications organizational entities undertake to remain contextually relevant and feasible (Lappi et al., 2019). In other words, local conditions dictate which parts of the policy are incorporated, leading to customizations that address the local context’s distinct challenges and opportunities.

The identified discrepancies between policy and practice in our findings can be explained through the concept of policy-practice decoupling (Bromley and Powell, 2012) in situations when the value of a policy is not evident and the policy competes with the resources intended for the core mission. Since patient benefit is the primary mission in both studied public health-care organizations, any policy that is not directly aligned with patient benefit will lead to significant practice variations. This observation aligns with the arguments presented by Kooiman and Jentoft (2009), who claim that such decoupling occurs due to an inherent hierarchy of values and norms in organizations. In other words, when there is a misalignment between policy and practice due to competing values and norms, these policies will be deprioritized and result in increased practice variation. This opens a call for a more value-aligned approach to policy formulation and execution, reflecting Kooiman and Jentoft (2009) emphasis on the role of values and norms in governance.

Drawing on Meyer and Rowan (1977) work on decoupling, organizations create and sustain gaps between formal policies and practice to avoid tradeoffs between external legitimacy and internal flexibility. Our empirical observations confirm that all interviewees place a high value on the realization of benefits, yet that there is significant variation in practice. As noted by Bromley and Powell (2012), the original idea of decoupling as a gap between policy and practice should be complemented by that of a gap between means and ends. Means-ends decoupling becomes particularly relevant for organizations that experience increased organizational structural complexity, perpetual reform, and where resources are diverted away from core goals. This type of setting corresponds well to that of health care, and Orton and Weick (1990) see it as typically prevalent in organizations involved in the production of complex public goods.

In relation to benefits realization of digital transformation, we see strong tendencies for not only policy-practice decoupling but also means-ends decoupling. This is particularly evident in the emphasis on KPIs for reporting in Organization B, which, rather than facilitating actual benefit realization, is a clear sign of means-ends decoupling, where the primary objective could be seen as symbolic implementation of the means decoupled from the ends aligning with Bromley and Powel’s concept. Even digital transformation per se could be seen as fitting the bill of means-ends decoupling, where digital transformation has been shown to signal modernity and function as a means for attaining increased legitimacy in the eyes of external stakeholders (Bromley and Powell, 2012).

5.1 Contribution to research

Our study offers two main contributions to research. First, we find that practice variation is caused by two types of decoupling: policy-practice and means-ends. When policy fails to align with underlying core values/objectives or when the policy instruments themselves become geared toward means decoupled from ends, this will increase the likelihood of policy-practice distortion, supporting previous findings from e.g. Park et al. (2017) and Tushnet (1995). Second, we find that coercion in the translation process from policy to practice did not decrease distortion nor result in a decrease in practice variation. This has previously been acknowledged in the global diffusion of policies (Dobbin et al., 2007), as well as in the organizational surveillance literature (Sewell and Barker, 2006). This contrasts with previous findings from the study of management accounting and public administration, where bureaucratic control is coercive in essence (Brown, 2007). More recent developments focusing on the substitution of control with transparency (Bol et al., 2016) may hold interesting future findings that will further nuance these negative externalities of coercion. In summary, the study reveals that the observed practice variation stems from two distinct forms of decoupling: policy-practice and means-ends. In contrast to prior research, the study does not identify a decrease in practice variation resulting from coercion in policy compliance.

5.2 Contribution to practice

In addition to the contribution to research, we offer one main contribution to practice and policy. First, our study identifies that the primary mechanism for decreasing practice variation for policy compliance lies not in coercion but in goal alignment. Policymakers should assure that the rationale for policies is explicit in terms of linking these to the overarching mission of the organization. This insight has consequences for the design of governance and control, where coercive mechanisms may be replaced with alignment mechanisms through, e.g. enhanced transparency and line-of-sight.

5.3 Limitations

Our study has two main limitations, both related to generalizability. First, we have chosen two large (not small) public (not private) health-care organizations in Sweden (not elsewhere). As noted by Bannister (2007), comparing and transferring insights across institutional environments is laden with problems. However, we contend that our findings can offer valuable insights into the translation process from policy to practice in other national health-care systems. Second, our study uses translation theory as its analytical lens. While other theories could offer different perspectives on the complex process of translating policy into practice, such as the resource-based view (Oberländer et al., 2021), we argue that translation theory provides a robust conceptual framework for analyzing the trajectory from policy to practice in health-care settings.

5.4 Future research

We identify two distinct avenues for future research from our study. First, an interesting extension to our research would be to conduct comparative case studies that examine the differences and similarities of policy-practice translation between diverse national health-care settings. Second, this study consistently emphasizes the alignment of policy documents with the core mission and patient benefit goals. We believe that focusing on the impact of such policies on patient outcomes could add a meaningful layer of analysis. This would further our understanding of policy-practice alignment and means-end decoupling.

Figures

Figure 1.

Process model for policy-practice translation

Figure 2.

Translation process of organization A

Figure 3.

Translation process of organization B

Table 1.

Overview and number of policy documents by type

Type	Description	Org A	Org B
Budgets	The budget is the main policy document and applies to all activities and is superior to other documents. The budget sets goals, appropriations, performance requirements, assignments and investment plans as well as ownership directives.	1. Budget 2023, 289 pages	1. Budget 2023, 62 pages 2. Budget 2022, 62 pages
Policies	Policies contains guiding principles for how councils and companies should conduct their daily operations within a specific area to achieve the goals of the council.	1. Policy for innovation and digitalization, 8 pages	1. Vision – the good life, 20 pages 2. Digitalization policy, 3 pages 3. Policy for integrated information management with other health care, 3 pages
Strategies	In a strategy, the long-term direction for councils or companies is established to achieve the goals of the council, i.e. how the operations should be developed. Through the strategies, the organization makes choices and identifies courses of action.	1. Life science strategi 16 pages 2. IT and digitalization, 12 pages 3. Innovation, 7 pages 4. Research, education and development, 20 pages 5. Goals and strategic direction for primary care, 156 pages	1. Strategy for the transformation of health care, 11 pages 2. Strategy for residents and health care, 43 pages 3. Plan for the national strategy for life science, 17 pages
Guidelines	In guidelines, the council establishes rules and requirements for how councils or companies should work within a specific area or with a certain type of issue to achieve the goals of the council. Guidelines usually concretize a policy.	1. Implementation plan for close-to-home care, 40 pages 2. Project portfolio, 39 pages 3. Digitalization roadmap, 27 pages 4. Digitalization roadmap, 25 pages	1. Health-care service offerings, 19 p 2. Operational plan, 39 pages 3. Benefit realization and monitoring plan, and model for impact assessment, 7 pages.

Source: Authors’ own creation

Table 2.

Overview of interviewees

Position	Organization A	Organization B
Executives	2	2
Middle managers	4	5
Unit managers	2	3
Staff in core organizational roles (nurses, strategists, project managers)	2	3

Source: Authors’ own creation