Evading risk through depoliticization: the case of
the Dutch electronic patient record
Tim ten Ham 5871565
Research Master Social Sciences Master thesis
Supervisor: Christian Bröer Second Reader: Patrick Brown 18 August 2014
Evading risk through depoliticization: the case of the Dutch
electronic patient record
‘Freedom’s just another word for nothing left to lose’ (Kris Kristofferson, Me and Bobby McGee, 1969)
Abstract
This study examines how a governance organisation employs various governmental techniques in order to 1) successfully implement an information technology that is generally perceived as controversial and illegitimate and 2) manage institutional risk. More specifically, I have investigated how policy makers and implementing agency personnel deal with uncertainty in the concrete practices of the implementation of this information technology in Dutch health care. On the basis of semi-structured interviews with twenty policy makers and/or implementers and twelve interviews with first-line health care professionals, observations of information meetings and a variety of documents I traced three governmental techniques that need to allay anxiety regarding this technology: seduction, emotion management and relieving in combination with strict surveillance of health care professionals. Even though their and citizens‟ voluntary collaboration is formally specified it is demonstrated that these strategies compromise health care professionals‟ and citizens freedom of choice. As such, freedom is redeployed as a technical instrument in the achievement of governmental purposes and objectives. It is concluded that by evading a risk framework in the communication to health care professionals and the wider public the information technology is strongly depoliticized. Depoliticization is seen as the most effective way wherein which the technology is to be adopted and institutional risk is managed, but at a certain point it might unwittingly sustain the uncertainty that is to be mitigated.
Contents
Preface 4
1. Introduction: the implementation of a controversial information technology 5 2. Governmentality as a response to institutional risk 7
2.1 Policy is implementation 7
2.2 Evading risk communication as a governmental technology 7
2.3 Implementing framing and feeling rules 9
3. Methodology: interpretive policy analysis 10
3.1 The challenge of accessing local knowledge 10
3.2 Analysis 13
4. Rethinking implementation strategies 14
5. Implementation strategies disentangled 15
5.1 Seducing health care professionals 16
5.1.1 Economic seduction 16
5.1.2 Rhetoric seduction 17
5.2 Emotion management 19
5.2.1 Removing fear and anxiety 19
5.2.2 The black list and the employment of embarrassment 21 5.3 Relieving individual health care professionals 22
6. GPs and pharmacists: coping with ambivalence 24
7. The role of citizens in the implementation process 27
8. Conclusion and discussion 31
References 34
Appendix I: Timeline and overview of the data collected 36 Appendix II: Map of Regionale Samenwerkingsorganisaties in the Netherlands 45
Appendix III: Coding scheme 46
Preface
This thesis is the capstone of my study sociology that I started back in 2008. At the time, it would never have occurred to me doing a research master and least of all that I would study the implementation of the Dutch electronic patient record. Yet this is what happened. Conducting my own research, over a period of time that lasted more than a year, appeared to be a process that is simultaneously frustrating and inspiring. It was now and then quite hard to find the right balance between creativity and pragmatism that is needed in order to develop a suitable research design and to ask interesting questions that can be answered with the means at hand. Also, it appeared to be quite hard to make sense of the data that have been collected over the course of several months, to find patterns based on rigorous analysis and to construct a story that makes sense with the help of theoretical and empirical insights. I hope that I have succeeded in the job of developing a cohesive, systemic and clear scientific argument and as such have rightfully applied the knowledge that I have acquired over the years.
As my research on the electronic patient record has started out of pragmatic considerations I could not have imagined how societally and politically sensitive this issue was. Quite apart from the fact that I now realise what the societal relevance of social research means, it gave me the feeling that I really discovered „things‟ and researched a case that was worth investigating. Even though the topic has been subject to heated debate in the past, it is not my intention to position myself in this debate. I have neither a background in computer and information sciences nor in health care, so I would like to emphasise that I cannot and will not say anything about the desirability of the electronic patient record itself. It was my aim to neutrally map the strategies of a governance organisation, but with the help of a sociological theory that critically assesses how power is exercised over individuals and collectives in advanced liberal societies.
I am grateful to Olav Velthuis, as he suggested me to examine the electronic patient record as an empirical topic for my thesis project. Without our conversation this trajectory would never have started at all. Also, I would like to thank Guido van „t Noordende, as he could explain to me what the electronic patient record is from a technical perspective. Next, I would also like to thank Pita Spruijt and Anne Knol as they involved me in their research process. The knowledge and experience that I have acquired during my work as a research-assistant have definitely aided me in writing a better thesis. Subsequently, I am also grateful to John Grin. As my internship supervisor he also took the time and pleasure in advising me about the thesis project. Finally, I am indebted to Patrick Brown, who has given me some fruitful recommendations on a theoretical level and Christian Bröer, who has aided me in bringing an indispensable analytic focus in my argument.
1. Introduction: the implementation of a controversial information technology
Formally, general practitioners (GPs) and pharmacists are free to make use of an emergent information technology in the Dutch health care sector. Nowadays, this information technology is called LSP (Landelijk Schakelpunt), which is a new name for the national electronic patient record (EPD1). Interestingly, one of the GPs that I visited had established a connection with LSP, but during the conversation I discovered that he actually was a fervent opponent of the technology: „and to tell you one more thing‟, he firmly continued his tirade to me, „I do not collaborate with LSP. On principle.‟ I was rather surprised that he rejected LSP on principle, as I clearly saw a pass reader on his desk and even an LSP-advertisement poster in his waiting room, an indication that his clinic was connected with the new infrastructure. „Is that true?‟ he faltered. When I explained to him what the pass reader on his desk and the poster meant his confusion rapidly increased: „You say: I‟m doing LSP. I say: oh, I thought I was obliged to do this.‟ „You don‟t need to do anything‟, I said. „No?‟ the GP answered. „No‟, I confirmed. „Okay, that‟s what I thought‟, the GP muttered.
The EPD enables the exchange of medical information between various health care professionals throughout the country. In April 2011 the Dutch Senate rejected unanimously the legislative proposal that would legally oblige first-line health care professionals, GPs and pharmacists, but also hospitals to make use of this EPD. According to the Senators on duty, this information technology was so large in scope that it made medical information too easily accessible for unauthorised individuals. This could precipitate unacceptable privacy risks for citizens. Also, there were heated discussions among physicians about the desirability of the technology. Remarkably, even though one of the traditional institutions of our representative democracy blocked further implementation of the EPD it is exactly this information technology that is currently being implemented in the Netherlands and with which this GP is confronted.
Since November 2012 private parties, united in the Vereniging van Zorgaanbieders
voor Zorgcommunicatie (VZVZ), a new cooperation between the professional bodies of
several health care associations, are now in charge of the implementation of the EPD. They have chosen a new strategy in order to make the EPD successfully operative in Dutch health care. The name has changed from EPD into LSP and since legally enforcing participation and implementation has failed, VZVZ has introduced a market strategy in the hope that health care professionals now voluntarily collaborate. However, as my encounter with the GP in this section demonstrates, health care professionals‟ freedom of choice seems to be highly questionable. In my view, the situation of this GP cannot be viewed in isolation. Although there has been and still is considerable reserve, and to a lesser extent, even strong opposition to the introduction of the EPD in health care it is interesting and relevant to find out how the technology becomes implemented nonetheless.
In more sociological terms this thesis aims to explore how power is exerted over individuals and collectives within „networked‟ forms of policymaking under radical uncertainty (Hajer & Wagenaar, 2003: 5). This empirical case is particularly interesting as
an air of illegitimacy, controversy and ambiguity surrounds LSP, mainly due to its contested political history and its unanimous rejection by a democratic institution. Hence, more specifically I am interested in how policy implementers deal with illegitimacy and ambiguity in implementation practices. How do they deal with the radical uncertainty regarding the eventual successful adoption of LSP in health care? After all, implementing a controversial technology requires collective learning from mistakes made on earlier, adopting new roles and strategies by policymakers and implementing agency personnel and finding new ways to create support from health care professionals and citizens.
As can be read the research question is almost entirely empirically driven. Theoretically, I consider the implementation process of LSP as a case of governmentality (Foucault, 1982) in response to the management of institutional risk (Rothstein, 2006). Governmental techniques, or implementation strategies as I call them, involve sanctioning, disciplining and normalizing health care professionals during the implementation process. These strategies are seen as the most suitable way to cope with the political vulnerability of the governance organisations involved. This is, briefly stated, how I understand the governance of this information technology. LSP-stakeholders, consisting of a wide range of public and private actors2 have come to a new way of governing, settled in a legally non-binding covenant agreement, in which they have formulated objectives for the period 2013-2016. The assumption behind the covenant is that enlightened self-interest of its participants – a decrease of costs in health care – and societal interests – qualitatively better and more efficient health care – will ultimately meet. This approach transfers political decision-making and implementation of policy from the traditional political realm, the classical-modernist institutions of representative democracy, to new professional arenas and market players (Bovens, 2005: 122; Hanssen, 2009: 120).
The argument of this thesis is based on interviews with policymakers and policy implementers or regional project managers3, health care professionals, observations of information meetings and a variety of documents. These methods provided me with on the spot information about the strategies that policymakers, implementers and their organisations employ in order to attain their policy goals. In the next section I dissect the governmentality perspective and present the concepts that enable me to interpret the implementation strategies that I have encountered. Subsequently in chapter three I provide the reader with an overview of how I have come to my findings: here I shed light on how I collected and analysed the data. These data are the body of the thesis and present the main argument. Chapter four discusses my respondents‟ reflexivity concerning the uncertainty with which the implementation has to do deal and the political vulnerability of the LSP-project. In chapter five I present the implementation strategies that are employed by policy implementers with the aim to induce health care professionals to connect to LSP. In chapter six I discuss how GPs and pharmacists deal with the ambivalence that is deliberately maintained in the practices of the implementation process. In chapter seven the
2 The most notable covenant partners are: Vereniging van Zorgaanbieders voor Zorgcommunicatie (VZVZ), Landelijke Huisartsen Vereniging (LHV), Vereniging Huisartsenposten Nederland (VHN), Nederlandse Vereniging van Ziekenhuizen (NVZ), Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie
(KNMP), Nederlandse Patiënten en Consumenten Federatie (NPCF) en Zorgverzekeraars Nederland (ZN). 3 From now on I call them policy implementers or regional (LSP) project managers interchangeably.
role of citizens in the implementation process, and more particularly in the opt-in procedure, is discussed. Finally, in chapter eight there is space for some concluding remarks.
2. Governmentality as a response to institutional risk
2.1 Policy is implementation
As can be read in the introduction I tend to use the concepts LSP-policy and implementation of LSP rather loose. Before I outline how the notion of governmentality can enhance our understanding of the deployment of a new information technology in health care, I briefly touch on the subtle difference between policy and implementation to bring a more precise analytic focus in my argument. Even though policy and implementation are in reality closely corresponded with one another they need to be clearly separated analytically according to Pressman & Wildavsky (1973). Roughly said, whereas policies imply theories, implementation entails empirics. Pressman & Wildavsky (1973: xxii) perceive policy as a hypothesis „containing initial conditions and predicted consequences. If X is done at time t1, then Y will result at time t2.‟ Policies are based on assumptions, and assumptions are necessarily hypothetical. For example, it is assumed that if health care professionals rely on a new information technology such as LSP the efficacy and quality of health care increases. LSP can thus be seen as a means or solution to achieve a particular end. Implementation, on the other hand, is seen as „a process of interaction between the setting of goals and actions geared to achieving them‟ (ibid.).
Thus, implementation is not a programmed process, but a process of interaction (Lipsky, 1993; Tops, 2002: 175) – interaction between implementation- and regional project managers and health care professionals. These implementers constantly need to anticipate on problems, relations and positions as they encounter them in concrete situations. Hence, LSP-policy emerges in the concrete practices of the implementation process. In this study, the analytic and empirical focus lies here. VZVZ has formulated very clear objectives in their Business plan 2013-2016. These objectives very concretely refer to a particular number of health care professionals that have their computer systems connected to LSP and a particular number of citizens that have given opt-in: permission to have (elements of) their medical records shared between health care professionals with whom they have a treatment relationship. In other words, VZVZ operates from an implementation-paradigm that is guided by a focus on agreements concerning performance (Tops, 2002: 179). If I use the term implementation I simply refer to the process wherein which action is undertaken by LSP project leaders to stimulate GPs, pharmacists and citizens to collaborate with LSP in order to obtain desired results as formulated in the
Business plan 2013-2016.
2.2 Evading risk communication as a governmental technology
The Minister of Public Health has pointed out that citizens and health care professionals are to be informed about the risks of electronically storing and exchanging medical
information4, but, as will be demonstrated later, it seems that VZVZ has clearly opted for a different strategy. After all, it seems to evade risk communication to the public altogether. The only way wherein which information is provided to the public is by means of brochures and leaflets that can be found in health care facilities and the VZVZ website. How can this strategy be comprehended? Rothstein (2006: 216) points out that governance organisations are increasingly confronted with scrutiny and accountability. Failure has always been part of governance, but due to increased complexities, such as the participation of a wide variety of actors with divergent interests in the current covenant structure, potential failure „create institutional risks that can threaten the legitimacy of governance organisations and their practices in managing societal risks‟ (ibid.). Whereas VZVZ has been established to implement a solution for a societal risk – health threats related to medication errors5– they must, in order to secure their own survival, reflexively manage the negative externalities of governance itself. In other words, they continuously need to manage their own vulnerable position as a governance organisation that is responsible for the implementation of an information technology that is generally being perceived as controversial and illegitimate. This must be realised if we aim to understand the organisation‟s behaviour and strategies.
In order to understand VZVZ‟s response to institutional risks the „Risk Government Model‟ developed by Wardman (2008) is clarifying. He bases this model on three imperatives for engaging in risk communication. First, the normative imperative is guided by the believe that „risk communication should raise people‟s awareness of risk and enable citizens to be party to individual risk decisions that personally affect them and their communities‟ (Wardman, 2008: 1621). Hence, the communication of risks are informed by two democratic ideals: an obligation felt by government agencies or business organisations to inform the wider public, while the public must also have an opportunity to be involved in how risks are communicated and managed. Second, the instrumental imperative construes risk communication as a resource that may be employed by an organization to help them accomplish their objectives. Third, the substantive imperative „advocates that risk communication should be utilized to generate improvements in understanding and the quality of knowledge available when making risk related decisions‟ (Wardman, 2008: 1622). This model seems to me plausible, as a core element of VZVZ‟s communication strategy is to keep citizens substantively disengaged, while evading communication about LSP in terms of risk to the wider public is instrumentally utilised in a wish to smooth the path of a new and controversial technology to the health care sector.
The Risk Government Model is in correspondence with the notion of „governmentality‟ (Foucault, 1982). In short, this notion sees political rule in advanced liberal societies not as direct command and control over its people, but rather assumes and retains their individual freedom. In fact, it sees freedom as one of its key resources. Hence, effective political authority seems more effective if individuals and collectives are guided
through their freedom (Bröckling, Krasmann & Lemke, 2010: 49). Freedom of choice can
also be observed during the implementation of LSP as health care professionals and
4 Vergaderjaar 2011-2012. Kamerstuk 27529, no. 109. 5
According to the 2006 HARM-report there are annually 41.000 hospital admissions related to medication errors, with estimated costs of 85 million euros a year.
citizens are formally free in their decision to work with or to be included in LSP. I demonstrate in the next section how this freedom of choice gets shape in the implementation process under influences of several technologies of government. These technologies are „the complex of mundane programmes, calculations, techniques, apparatuses, documents and procedures through which authorities seek to embody and give effect to [political rationalities and] governmental ambitions‟ (Rose & Miller, 1992: 175).
Together with political rationalities, the discursive fields wherein which power is exercised and the ways wherein which authorities morally justify their power, governmental technologies establish a connection between the activities and lives of individuals and groups with the aspirations of authorities (ibid.: 176). In other words, by investigating the implementation strategies of LSP-stakeholders we see that the modern state does not coerce but „acts in ways that try to direct free will‟ because in Foucault‟s (in: Bröckling, Krasmann & Lemke, 2010: 31) words: „it incites, it induces, it seduces, it makes easier or more difficult; it releases or contrives, makes more probable or less, but is always a way of acting upon one or more subjects by virtue of their acting or being
capable of action. A set of actions upon other actions [emphasis added]‟. After all, the
target group nonetheless retain certain action prospects as they are provided with freedom of choice, albeit under influences of incentives. This is how I would like to interpret the implementation strategies I have mapped: transferring a new technology that was once a project instigated and implemented by the state to market actors and professional bodies should not be interpreted as a diminishing of state power, but as a restructuring of governmental techniques (Bröckling, Krasmann & Lemcke, 2010: 40).
2.3 Implementing framing and feeling rules
In order to understand the current implementation process of LSP it is particularly interesting to take into account health care professionals‟ perspectives as it might shed light on how cooperation is brought about. Following Yanow (2000: 18), as clients of LSP they should not be conceived of as „„passive targets‟ of „policy missiles‟, but as active constructors of meaning as they „read‟ legislative language and agency objects and acts‟. These perspectives yield insights into how they give meaning to LSP. Following Nijhof (2003) and Bröer (2006) I assume that people give meaning to social phenomena and by doing so they employ existing frameworks. These existing frameworks are called discourses: collectively shared ways of interpreting certain phenomena (Nijhof, 2003: 20). Bröer (2006: 55) developed an analytical model that elucidates how people might cope with policy discourses, as it is hypothesised that these resonate in their perceptions and subjectivity. The way health care professionals cope with the policy discourse might affect their decision to cooperate with LSP or not. His analytical model, which he calls discursive resonance, presumes three different ways wherein which people can perceive social phenomena.
First, people might reproduce the dominant policy discourse if they mimic its assumptions and meanings or definitions of the situation. This is called consonance. Second, dissonance exists if people‟s definition of the situation partly corresponds and is partly incongruent with the dominant discourse. Deviation of the dominant discourse goes hand in hand with an argumentative conflict as people might hold two views or beliefs that
are actually inconsistent with each other. We will see this clearly in the answers of some health care professionals in the subsequent sections. Finally, people are conceived of as autonomous if their views or beliefs are not affected by the dominant discourse, but by other discourses or different ways of thinking or feeling about the subject matter (Bröer, 2006). Employing this model has a very clear advantage: it does not reduce individual health care professionals to simple proponents and opponents, or structure the debate in progressive versus conservative, but it rather sheds light on how individual health care professionals reproduce and struggle with a macro-sociological conflict (Ellis, Barry & Robinson, 2005).
Furthermore, policies do not only implement „framing rules – „rules according to which we ascribe meanings to situations‟ but also „feeling rules‟ (Hochschild, 1979: 566). Hochschild (ibid.) defines feeling rules as „guidelines for the assessment of fits and misfits between feeling and situation‟. According to Bröer & Duyvendak (2009: 340) besides framing rules, policy measures also contain feeling rules: rules that are implicated in concrete interventions and that attempt to steer health care professionals‟ emotions. Making use of some implementation strategies while neglecting others can be more successful if it „resembles pre-existing mindsets‟ (ibid.). If a particular intervention or way of approaching „strikes a responsive chord‟ in health care professionals it may affect their perceptions and their willingness to participate and may remove feelings of sceptic caution. In light of the analytical framework that I have introduced I can formulate the research question on a more theoretical level: how does a governance organisation employ various governmental techniques in order to 1) successfully implement an information technology that is generally being perceived as controversial and illegitimate and 2) manage institutional risk?
3. Methodology: interpretive policy analysis
3.1 The challenge of accessing local knowledge
This study is based on an interpretive approach to policy analysis (Yanow, 1993; 2000). Such an approach focuses on the meaning that policies have for a wide variety of policy-relevant publics. Also, it enables the delineation of „how‟ a policy means by identifying the practices through which policy meanings are communicated and which audiences are targeted (Yanow, 2000: 8). In order to detect how the implementation of LSP occurred in practice I decided to comply with this interpretive presupposition, as it would give me access to various „communities of meaning‟ that Yanow (2000: 10) distinguishes: policymakers, implementing agency personnel and affected clients, which all are part of the implementation process of LSP. With the aim to gain access to the words and actions of these policy-relevant actors one „interpretive‟ method takes central stage: the analysis of semi-structured interviews. Observations and the contents of documents are used as complementary sources of information. I elaborate on these methods in the subsequent part of this chapter.
Due to the multitude of actors and my initial ignorance of their existence I was confused where to start and which organisation to approach at the outset. An extra
difficulty was the secrecy with which this project was surrounded, which urged me not only to approach my respondents carefully and neutrally, but which gave me also immediately substantial insights into the strategies of the organisations involved. As it is my aim to shed light on the strategies that are employed in order to make LSP operative, I decided to focus my attention on the actors that I perceived as the driving force behind the implementation process, namely the central organisation (VZVZ) that is in charge of the implementation in the Netherlands and various smaller organisations that facilitate the implementation regionally, which are chiefly Regionale Samenwerkings Organisaties (RSOs). I came across these RSOs somewhat late when I found out that regionalisation is a key element of the implementation. This made it necessary to collect data in various regions in the Netherlands and conduct interviews with several LSP project managers who were actively involved in facilitating connection and opt-in in their specific region. Finally in January 2014, when I visited the first RSO in Eindhoven, I received a map of the Netherlands with these regions, the names of the organisations and the names of the people who were employed by these organisations. This map was an important opportunity for me to approach more, and more relevant, respondents for my study6.
Sometimes I contacted the director of the organisation, but I would preferably come into contact with LSP project leaders as they could give me more insight into the micro-strategies that they concretely employed in their daily work. After all, this is an important part of their job. LSP project leaders have more direct contact with individual health care professionals than directors. Directors are more involved on an administrative level, as most of them were for instance member of VZVZ‟s Region Council. But also speaking with directors was interesting as these were mostly people who were involved with the EPD-project for many years and they could tell me more about the political sensitivity of LSP and relatedly, the vulnerable position of VZVZ and the way they tried to manage this precarious position in administrative and implementation practices. I have also included people in my sample who were employed by smaller, regional pharmacist associations and freelancers with their own IT-company whom were hired by RSOs or regional associations of walk-in clinics if they performed a similar job compared to the other LSP project managers. I finalised my sample with two employers of the professional bodies of GPs (Landelijke Huisartsen Vereniging and Nederlands Huisartsen Genootschap respectively) and a GP and a pharmacist who were (in)formally involved with the implementation for VZVZ and who played a so-called ambassador-role. As I will demonstrate later in my thesis these ambassadors are key players in the implementation process.
Besides conducting interviews with policy implementers, also some observations have been acquired at information meetings7. In sum I have visited four meetings and one connection test organised by a walk-in clinic. Even though the latter was mainly a technical affair it also shed some light on the fluid, informal and decentralised relations within this policy network. Furthermore, I visited one meeting that was only attended by regional project managers and that was organised by VZVZ, one presentation that was given by VZVZ on the mart for health care and information technologies (mainly attended
6 This map can be found in the appendices. 7
A more detailed, chronological overview of the data that has been collected (interviews, observations and documents) can be traced back in the appendices.
by suppliers and consultants), an information meeting that was organised by a RSO and which was mainly attended by health care professionals and managers from second-line care, and a LSP workshop organised by VZVZ and which was attended by medical assistants from GPs. Because I visited these meetings I not only detected how policy implementers dealt with questions from health care professionals and how they conveyed their message to their target groups, but I also noticed how they collectively reflected about the current situation and which strategies would be most effective to attain their policy goals.
Most national and regional information meetings that were especially meant for health care professionals from first-line care and which were substantially larger as several dozens of them were invited, were unfortunately organised in the period before the start of my thesis project (between January 2012-September 2013). The aim of these meetings was to inform health care professionals about the benefits of LSP and to facilitate connection. During my research project this connection was already done for the majority of first-line health care professionals and the focus was more but not solely on the opt-in procedure. But this also depended on the region. Besides I also examined various policy documents and what Yanow (2000: 16) calls policy artefacts: brochures, leaflets and opt-in forms that are used in GP-clinics and pharmacies with the aim to inform citizens about the importance of exchanging medical information. These documents however, were not the main focus of the analysis, but chiefly provided me with some useful background information. Most policy documents could be found on the websites of VZVZ and the RSOs that I visited and were the formal basis of LSP policy.
Sampling health care professionals was different. Ideally, I should have selected my respondents on the basis of thorough theoretical sampling, which was unfortunately not possible. This was mainly due to practical concerns. It was particularly hard to approach GPs, as this group is very busy with their daily work and who were in many instances not willing to be interviewed. So I was induced to approach these GPs whom I knew indirectly through my own personal networks. I tried to include as much variation in attitudes in my sample, but I think it is highly unlikely that I have captured this variety completely. My approach to pharmacists was slightly different. I simply called several pharmacies with the question whether it was possible to ask them a few questions about LSP. Most of them agreed. After a couple of interviews I noticed the similarity of the pharmacists I interviewed. Therefore, I decided to include one pharmacist who had completely different ideas about the subject matter. Because most health care professionals have been approached by a convenience sample (Bryman, 2008: 183) the perspectives of health care professionals should be treated as a point of departure form where one could possibly examine more closely the perspectives of health care professionals and their interpretation of the implementation process in a subsequent study.
Table 1. Overview of the data collected.8
3.2 Analysis
The interviews are the core elements of my data. The observations and documents that I have collected provide me with some contextual information I used in order to strengthen my argument. I asked my respondents broad questions about their exact function, their and their organisation‟s aims and the things they wanted to accomplish during their employment. Very often I simply asked how the implementation was going. Also I was particularly curious in which way they approached health care professionals in the process, what they viewed as the most effective strategies, what they experienced as difficulties, whether they were satisfied about the organisation of the implementation, regionally and nationally, and of course what their prospects were for future developments. These were all questions that elicited long and detailed narratives, emotions and effusions of which I also was surprised. In short, my respondents were willing to talk and discuss these politically and societally sensitive matters with me in depth. The interviews were coded and analysed with Atlas.ti version 6.2. I started coding my first interviews while I was still very busy with arranging other interviews.
I devoted careful attention to implementation techniques, which gave me a clear directive what the data meant in light of my, still broadly formulated, research question. I coded interesting chunks of text and defined my codes simultaneously as I was still unknown with the precise techniques that were employed. Besides taking notice of techniques I also decided to label other seemingly important features of the implementation process such as the negative side-effects of the working procedures that were pursued. Also I seriously took into account respondents‟ reflections about their own and others‟ decisions and actions which I considered significant. This resulted in a descriptive coding scheme that can be traced back in the appendices. The same goes for the interviews that I have conducted with health care professionals. Especially, I took notice of doubts, misunderstandings, disagreements and conflicts and tried to categorize them as such. In order to elevate my description of the implementation process to a more conceptual level, which is called axial coding (Charmaz, 2006: 60-63), I printed all codes,
8 Apart from two interviews, all have been conducted on the work places of the respondents. Most interviews lasted between 1 and 2 hours.
Data N
Interviews
LSP project managers 12
Policy makers (directors RSOs + two policy makers from national umbrella organisations GP)
6
Health care professionals (including two ambassadors)
14
Observations 5
Documents/policy artefacts 15
put them on the ground and tried to cluster them together in order to find patterns. By doing this I tried to transcend the descriptive elements of my study and to translate my data to a more conceptual or etic perspective (Eriksen, 2001: 36; Bernard & Ryan, 2010: 110). In sum, I created 155 codes, of which 113 refer to answers given by policymakers and implementers, whereas 42 codes refer to answers given by health care professionals. On a descriptive level I have discerned 26 different implementation strategies. The most important strategies are discussed in the subsequent parts of the thesis, the remainders are displayed in the appendices. In total, I have coded 1620 text fragments in 32 interviews.
4. Rethinking implementation strategies
When the Senate rejected LSP in 2011, this seemed to signal the end of this incipient technology. At least it formally ended the involvement of the state. Some reports have been published that attempted to account for the failure of the adoption of LSP (NSOB, 2012; Pluut, 2010). It is beyond the scope of this thesis to discuss the failures that have been made in the political decision-making process prior to the crucial rejection of the Senate in 2011. However, the state‟s failure to implement the technology successfully has served as an important learning moment for current policymakers and implementers. The rejection of the Senate has made them more reflexive about current strategies and their potential success. This reflexivity is based on the awareness that eventual adoption and success of LSP is still highly uncertain. After all, if health care professionals and citizens seem not to participate, for whatever reason, LSP is destined to perish. This forces VZVZ in a highly vulnerable position, as they aim to implement and communicate about a technology that is generally being perceived as controversial and illegitimate by health care professionals and citizens. In the terminology of Rothstein (2006: 216), VZVZ is not only preoccupied by managing a societal threat, but mainly by managing institutional risk – i.e. their own vulnerability. VZVZ is faced with threats to the legitimacy of current methods of governance and ultimately, their very existence.
This became clear during the interviews with policymakers and implementers, but also in various policy documents. For example, in the Business plan 2013-2016 VZVZ reports the results from a risk analysis (pp. 81-84). They map the risks which might hinder a successful deployment of LSP. I interpret this as a clear indication of an increased awareness about the possibility of failure, due to „conflicts and puzzles of governing‟, such as „inherent uncertainties, fragmented organizational sections, constrained resources, ungovernable actors and unintended consequences‟ (Rothstein, 2006: 217). The potential failure of LSP, combined with its controversial and illegitimate status, thus create institutional risks with which VZVZ has to deal during the implementation process. Simultaneously, this clarifies what VZVZ understands as risk: not so much risks which could be precipitated by the deployment of an emergent technology, but rather the risk that certain policy goals could not be attained and the system will be aborted. This also indicates what the focus is of the covenant: maximising connections of health care professionals and citizens‟ opt-in, as this is the only way that can ensure LSP‟s survival.
The core challenge that policymakers and implementers face is how they must cope with potential resistance from health care professionals, politicians and the wider public, as they are aware of the fact that resistance may undermine the attainment of their policy goals. They know that they must treat the second chance that they have received very carefully, or otherwise they “flush ten years of hard work and 300 million euros of tax money through the toilet”, as one of my respondents replied. Attention, in whatever from, to LSP may result in discussion, scepticism, unrest and, in the worst case, objection from opponents. This has to be prevented by all means. I cannot stress enough that all actors who are involved with the implementation are very well aware of this. This RSO-director clearly explained this to me:
“I think LSP has much to do with politics and national tendencies and policies, it has to be too careful now and then. From this construction it cannot be innovative […] Regionally, we can take some risks only if one takes for example privacy very seriously, what we continuously do, but something might go wrong. The national organisation cannot afford this.”
So the potential outburst of objection and with this, potential failure of the adoption of LSP, cause this radical uncertainty that covenant partners need to handle in the implementation process. This became also very clear at an information meeting that I visited. LSP project managers from the whole country were invited with the aim to be instructed about effective strategies to evade this objection and to implement LSP successfully. The main message was that top-down enforcement of commitment was perceived as counterproductive, but that health care professionals have to be made enthusiastic within the context of high-quality health care. In the next section I would like to delve deeper into the current implementation strategies that are employed by LSP project leaders in order to create commitment from health care professionals.
5. Implementation strategies disentangled
I have visited many regions in the country where various RSOs are present. These are small organisations, mostly with only a few employers, have the legal status of a foundation and are sponsored by health care organisations from the region, among others hospitals, regional GP and pharmacist associations, other health care facilities and sometimes patient organisations or even the municipality. All RSOs share the same objective and function on the basis of a similar assumption: the efficiency and quality of health care increases if the communication between health care professionals improves with the help of innovative information technologies. There are about fifteen RSOs in the Netherlands, as these organisations fit in with the regional organisation and demand of health care. Most of these RSOs take part in a „task force‟ or regional steering committees, wherein which also other interested parties participate. Generally, project leaders employed by RSOs, or in some cases project leaders from walk-in clinics, have the most
important function regarding the stimulation of connecting health care professionals and opt-in by citizens.
Therefore, these were for me crucial respondents as they could give me detailed and on the spot information about the strategies that they and their organisations employ in order to reach their policy goals. Generally, across the regions I heard similar stories about the organisation and the concrete practices of the implementation. Nearly all regions organised information meetings just before or after the private restart of LSP in November 2012. Newsletter are sent to them, brochures and flyers are used and much information can be found on websites of VZVZ, RSOs and the professionals bodies of GPs, pharmacists and hospitals. However, these are not necessarily governmental techniques. There must be a clear relationship between technologies of domination and technologies of the self, between authorities ambitions‟ and health care professionals‟ actions (Rose & Miller, 1992; Lemke, 2001). This chapter must shed some light on this relationship. I argue that health care professionals are not simply neutrally informed about what LSP enables, but that VZVZ and RSOs will try everything in their powers to make them work with LSP.
5.1 Seducing health care professionals
5.1.1 Economic seduction
Health care professionals who intend to establish a connection between their computer systems and LSP must invest some costs in order to realise this. Mainly, they need to put much effort in this process in terms of time, which is already a scarce resource for most health care professionals. Therefore, they are financially compensated by VZVZ. GPs and pharmacists are entitled to receive two subsidies: a compensation to realise connection (aansluitvergoeding) and a compensation for opt-in (structurele vergoeding) if they include at least a certain percentage of their patient population9. However, health care professionals cannot claim these subsidies without taking into consideration some requirements. There is a strict deadline for the aansluitvergoeding, which increases the pressure to establish a connection. Also, the structurele vergoeding cannot simply be received. In order to receive their financial compensations in 2013 GPs and pharmacists must have included at least 30% and 10% respectively of their patient population before the end of the year. Before the end of 2014 both groups must have included at least 40% of their patients to be eligible for this subsidy. However, the intention behind this strategy may seem clear, as this project manager explains: “we agreed that before 31 December 2013 they must have a minimal amount of opt-in as a prerequisite to receive the structural compensation, so this gave a boost, but it was intended as an incentive.” A project manager from another region adds: “I have devoted much attention to the monetary compensations in the newsletters. […] Well, for pharmacists it is a drive to do something. So gradually, pharmacists were progressing in terms of connection.”
For most health care professionals the financial compensation was indeed interpreted as a trigger to establish a connection to LSP, albeit from a pragmatic perspective: “Well, financial support has been helpful. If it would have been very
expensive and in the case we had not received any financial support then it would have left hanging. So it has forced us to make a decision. It gives a certain urgency, a kind of deadline on the decision-making.” This decision-making process is confirmed by another GP: “It was a kind of lure, to persuade GPs to do this. […] And this has been the driving force for the internal discussion, as this is not a primary need from our perspective.” Even the GP whom we met in the introduction and who appeared to be against LSP has accepted the subsidy, but he nevertheless remained stuck in an argumentative struggle:
“It seems as if too much money has been invested, too much monetary interests and these people don‟t appear to represent the interests of patients. That‟s just my feeling and particularly with the lures, you know, that‟s just wrong […] as they say: it will be obliged in the end, we‟re just lured and that doesn‟t feel proper.”
It is important to realise that a connected GP or pharmacist is not automatically a GP or pharmacist who actually makes use of the information technology. As one director of a RSO lamented: “Yes, they are connected, but that does not mean that they employ it. These are two different things. They have reaped the subsidies, let me put it that way.” Hence, financially seducing health care professionals seems to be a necessary condition for having them making a connection to LSP, but it appears to be far from sufficient for them to employ it in their daily work. Much, much more has to be done than this.
5.1.2 Rhetoric seduction
When I asked the spokesperson from VZVZ what he considered as their most important task regarding LSP he responded convincingly: “Use by health care professionals. The most important thing is eventually, and I call it health gains by means of efficiency in the exchange of data. That is our most important aim”. This is not a surprising answer, as the active use by health care professionals is the major reason why the private restart has been made possible altogether. The covenant partners in general and VZVZ in particular need to undertake action in order to make LSP successfully operative, as this is not going to be initiated by health care professionals or citizens themselves. When I asked my respondents in the very same interview how they imagined accomplishing this the spokesperson responded: “We tell things to care providers. And that is objective, that is neutral […] it is actually descriptive. […] We are seen as neutral.” Later in the interview, the implementation-manager initially confirmed his answer, but tended to nuance this point of view somewhat: “Well, I‟m very objective. I point out what it [LSP] enables and it is up to them if they want to connect. Obviously, there is some sales talk in it, but I try to convey it as objectively as possible.” This is a contradiction in terms. How can one tell an objective story to health care professionals if you want them to think, feel and do something
differently? Sales talk is by definition not neutral.
Hence, health care professionals are continually „informed‟ about what LSP enables. I have written „informed‟ within quotations-marks as I argue that VZVZ‟s implementation-managers and regional project managers actually do much more than providing neutral information about how LSP works. LSP policymakers and project managers seduce, induce and persuade health care professionals of the benefits of the system. They think of „seducing strategies‟ as one of my respondents candidly clarified. In
fact, creating awareness among health care professionals concerning the usefulness and necessity of electronically exchanging medical information via the LSP-infrastructure is being done, so they acquire the realisation that LSP is the solution for their problems. Apparently, most GPs and pharmacists do not see this this added value on the basis of their daily work. In short, a need to use LSP is to be generated by means of the dissemination of „reliable‟ information. It is assumed that health care professionals are raised out of their state of oblivion when they are confronted with the benefits of LSP by means of such an awareness-raising campaign. Ellis, Barry & Robinson (2005) have observed a similar presumption in the case of the placement of offshore wind farms in Northern Ireland. This quote might illustrate my point:
“If we want to bring about a change then we need to take a few steps before we make it. This means that people first need to know that it [LSP] exists, subsequently they must understand what it is and eventually they must make use of this thing that we offer. So we need to invest much in the first trajectory, we need to explain what is does and why it happens. If we have achieved that knowledge then it becomes acceptable to do something with it, to accept that change actually.” (Spokesperson VZVZ).
Also on the regional level, project managers have adopted the strategy of seducing and influencing health care professionals:
“But the weakness of an organisation as this is – I mean, I cannot oblige it. I can facilitate, but I cannot say to a pharmacy: I want you to opt-in sixty per cent of your population within three months. That is not possible. So I am dependent on the effort of my pharmacists – so the level on which I act is mindset. So to incite them to do it, and to remove possible obstacles that might hinder them” [emphasis added].
This regional project manager vividly demonstrates her working strategies. She is aware of her leeway for reasonable action and the possibilities and impossibilities that her position entails.
If we return briefly to the notion of governmentality we realise that the market model, instigated by VZVZ and the covenant partners, enables the employment of governmental practices. This is where governing and modes of thought come together in practice. Not for nothing Foucault (1982) defined governmentality as the „conduct of conduct‟. By seducing them with sales talk and rhetoric means, implementation- and project mangers structure and shape the possible actions of health care professionals (Lemke, 2001). Noticeably, governing from the market model is not a way to force people to do what the governor wants, but authorities are continuously busy with organizing and creating the conditions under which individual health care professionals can make use of their freedom to cooperate with LSP. During the implementation process, they must acquire the idea that the technology is useful and necessary for their daily work, and this must encourage them to establish a connection to it.
5.2 Emotion management
5.2.1 Removing fear and anxiety
Gradual and almost invisible, organisational change is declared as the most effective way to bring LSP into being. Health care professionals are told that connection to LSP only brings about a minor change in their working processes. Hence, one of the reforms that have been implemented since the private restart is placing so-called „regional fences‟ into the LSP-infrastructure. This means that electronic exchange of medical information between GPs and pharmacists is only possible within designated regions, but hospitals can retrieve medical information that has been made available by health care professionals in other regions. I was told by VZVZ that the regionalisation of the LSP-infrastructure was a requirement from the Senate to continue the implementation at all, but this was refuted by another respondent, who was employed by one of the professional bodies of GPs:
“No, you already pointed this out. It [VZVZ] is a private enterprise. There are no requirements, there have been eruptions and motions in the Senate. So it is a kind of feeling, a kind of self-censorship. You see this in Russia as well and other scary countries. VZVZ needs to ensure that they are dropped off the radar, so they cannot be made accountable for this […] At a meeting two, two-and-a-half years ago there has been determined that regionalisation does not improve the level of security and is not desirable and necessary from the perspective of health care professionals. We all agreed on this. Yet another decision has been made. These things happen.”
This remark induced me to interpret this intervention as a strategy used by VZVZ and the covenant partners in order to increase support from health care professionals. Not coincidentally, some of the project managers that I have spoken used the metaphor of a „flywheel‟, in order to clarify how they have organised the implementation: “If one wants to get it working throughout the country then one needs to turn a very large flywheel. The first turns are then very heavy, while if one approaches it regionally then one has much smaller flywheels which are far more easily to turn.” Also, the information that can be exchanged between health care professionals is limited and this also goes for the type of health care professionals who can retrieve external medical information.
I interpret these (provisional) interventions as a desired sluggishness of the implementation process. The regionalisation and the restricted availability of the functionalities that LSP offers can plausibly be perceived as a desired deceleration of the implementation, as the awareness of potential distrust and a lack of support are still strong among the covenant partners. A LSP project manager explains this: “I think we must not try too much. We must have a very firm idea of what our aims are, but we must do it gradually. If we go too quick, we lose many adherents, which I consider not reasonable.” This quote demonstrates the uncertainty and the potential objection with which policymakers and regional project managers could be confronted with in the implementation process. After all, critics of LSP argue that the technology is too extensive; it enables the exchange of too much medical information between too many health care professionals. In order to decrease objection the functionalities of LSP are decreased as well. We see here how organisational interventions are geared towards the management of feelings (Bröer & Duyvendak, 2009). A small LSP must give health care professionals and
particularly GPs the feeling that electronically exchanging medical information occurs in a secure and confidential environment.
VZVZ is aware of the fact that over the years LSP has achieved a negative image and has firmly established the objective to allay any distrust in this new care infrastructure, as the spokesperson clearly says: “The biggest fear was in fact that the system was national. […] To remove that fear, it was argued: we want you to do it on a small-scale. Just like OZIS.” OZIS is an older electronic exchange system that functions regionally, and is mainly used by pharmacists for many years. Therefore, the covenant partners have decided to eliminate OZIS gradually as this would be a very strong incentive for switching to LSP, particularly for pharmacists. If this is done, they are more or less forced to use LSP, because they cannot perform their profession properly without a complete overview of the medication history of their patients – which is after all provided by information technologies. However, delimiting LSP‟s functionalities has been done with the aim to generate support for GPs, whereas pharmacists, who are used to automation in their daily working processes, long for progress in the availability and the exchange in data, as this pharmacist explains:
“And then it [LSP] doesn‟t do anymore than OZIS. So they want to make sure that it is something better, while the other says: it should be nationally operating, but that has been the political discussion, that was a sensitive issue. But if you want to have a system that functions properly then you cannot get away with a few fences.”
However, and this is where policymakers and implementers are very well aware of, is that LSP is actually a so-called „growth model‟. Delimiting its functionalities can easily be perceived as a means to a desired end, as this project manager states: “Its national character will be foisted on through the back door. I am sure about that. And the need will grow. […] I mean, this cannot be stopped. I think they [VZVZ] know, but I presume they think we do it harmlessly, regionally and then we ramp it up.”
However, the strategy of decreasing the functionalities might have severe drawbacks for a successful implementation as it can inadvertently diminish support for LSP, as we could see in the critique that most pharmacists expressed. Viewed in this light, it is worth quoting one of the RSO-directors extensively:
“You know, it is as if one constructs a railway which brings about protests in a neighbourhood as these people do not want a railway in their backyard. But you tell these people: only one train a week will ride by. Whereupon the neighbourhood replies: if it is so little, we agree. But we already know that this railway is only profitable if twenty trains a day can make use of it. If you have started to tell the neighbourhood that only one train a week will make use of it then it becomes quite hard to make them accept twenty trains a day. Well, this is also what VZVZ – and I understand it, but eventually it is so stupid […] How do we connect mental health professionals? How do we arrange digital exchange concerning care for the elderly? […] If I would tell our GPs and pharmacists, who support electronic information-exchange, that mental health professionals would be included in LSP, they drop out massively. [emphasis added]”
This indicates the vulnerable position of LSP and once more underwrites its controversial state. Also it clarifies how fragile health care professionals‟ support for the system is, even from health care professionals who seem to advocate the technology.
5.2.2 The black list and the employment of embarrassment
LSP project managers do not have an easy job. Their target population is mostly divided and large, as this group consists of several hundreds of pharmacists and GPs. All these health care professionals must be informed and stimulated to cooperate with LSP – one by one. I elaborate on the individual approach in the next subsection, as I first want to delve deeper into alternative approaches that are used in order to create more support for LSP and increase the „level of acceptance‟ as some of LSP project managers seemed to formulate it. Another respondent, who is employed by one of the professional bodies of GPs mentioned the existence of a noticeable language and culture gap between policymakers and health care professionals. The language and the ideas of policymakers are not congruent with the experience of health care professionals, “policymakers talk about LSP as „the‟ solution”, as he stated. He referred to information meetings that he visited and where he observed a significant difference in the reaction of health care professionals when managers or their own health care colleagues had the floor: “they [health care professionals] get the support from the whole room and they are within five minutes much further than a VZVZ-employer as they know what goes on in the field, what situations one might encounter.”
It seems that most policymakers and implementers are very well aware of the fact that health care professionals accept a fundamental change in their working processes seemingly more easily when colleagues appear to be enthusiastic. Therefore, VZVZ and RSOs deploy several ambassadors who are firm believers in this new technology and who can motivate their colleagues to get down to LSP. On websites and in newsletters prominent GPs from the region are interviewed and who basically tell so-called „success stories‟. One LSP project leader confirmed this strategy: “I noticed eventually that they [GPs and pharmacists] make each other enthusiastic. That is the crux of the matter.” This is a crucial transformation of implementation strategies. Rather than using formal constraints, social control is used as a device to increase the level of participation. Enthusiastic health care professionals are deployed to address their colleagues if they remain behind with opt-in or their computer systems have not yet established a connection. These opt-individuals are key actors in the implementation process. I have also spoken with an ambassador, who really believed in the importance of electronically exchanging information with LSP for the profession of GPs and he very clearly explained what this social control entailed:
“In three years you can find a list for patients in the walk-in clinic where one can read that it is not possible to retrieve information from these, say, ten GPs. You must not subject them to penalties. That is what I always stress: please don‟t do this. […] But this [the list with deviant GPs, TtH] is only in three years. Some of the opponents are afraid for this, that they are forced. Not by health care insurers, because that was really stupid that this insurer obliged it, and we were very angry about what he did. No, eventually the professional body will clean itself.”
A couple of months later I visited an information meeting organised by one of the RSOs. Even though most visitors worked in hospitals rather than first-line care this meeting was interesting nevertheless. One of the speakers was the director of the walk-in clinic from that region and he ended his presentation with a firm statement that was in correspondence with what the GP-ambassador pointed out: “We intend to create a situation wherein which it is embarrassing if a GP or hospital is not connected to LSP by the end of 2015 [emphasis added]”.
Rather than imposing cooperation from the state – by means of legislation – on health care professionals, LSP is currently implemented from a market strategy that lets health care professionals free to cooperate or not. But what we see is that the pressure to make use of LSP comes from different sources. The social control that is exerted by other health care professionals is a very powerful and effective constraint to enforce cooperation eventually, as GPs are sanctioned in a different way. In other words, deploying health care ambassadors and intensifying social control can feasibly be interpreted as a governmental technology, a „procedural device through which individuals and collectives shape the behaviour of each other or themselves‟ (Bröckling, Krasmann & Lemke, 2010: 48). Health care professionals are governed according to what Foucault (1982) calls: self-regulation, or „technologies of the self‟. As the state cannot enforce participation, it is now a matter of cultivating their preferences combining with stimulating social control among health care professionals:
R: “Indeed, they [health care professionals] must see the added value. And they must want to do something with this and that is a process, they must want it together. This can be traced in Nijmegen en Twente where this has gradually started; they address each other when information is not used.
I: Health care professionals address each other?
R: Health care professionals, mutually. Or if your record isn‟t put in order, hey listen colleague; I cannot work with this information. Well, but that is – then you are very far in the process, and most regions are not so far.”
Power is exercised over individual health care professionals as they are confronted with dominant forms of subjectification: the construction of how human beings think, feel and act (Foucault, 1982: 216), but most importantly, as one project manager strikingly stated: “they must not have the feeling that they are forced.” The actions of project managers act on the possible actions of individual health care professionals. As they retain an action perspective, they retain a feeling of freedom. Precisely as this freedom is a prerequisite to implement LSP in the first place, the target group accepts the implementation.
5.3 Relieving individual health care professionals
Even though policy implementers attempt to stimulate cooperation with LSP by means of deploying enthusiastic GPs and pharmacists and by fostering social control among them, they do not stop here. VZVZ and RSOs position themselves as organisations of which their main task is to provide support for health care professionals. If health care professionals have made the decision to connect their computer systems to LSP the role of VZVZ and their regional counterparts suddenly starts to change. They transform from seducers to
