Health Privacy and Political Attitudes

Transcription

1 Health Privacy and Political Attitudes Concerning the Electronic Patient Record in the Netherlands Master Thesis Author Evert Mouw Student number S Institute Leiden University Department Political Science First supervisor Prof. Dr. Joyce Outshoorn Second supervisor Dr. Frits G.J. Meijerink Wordcount Ca incl. attachments

2 Abstract The storage of personal health information in nationwide electronic patient record (EPR) has led to a heated political debate in the Netherlands. The link between the political attitudes of individuals and the value given to health privacy is largely unknown. This study introduces the Health Privacy and Political Attitudes Survey. Results: most political attitudes are not a good predictor of ideas on health privacy, but authoritarian attitudes strengthen the trust in (government regulated) electronic records. Weak evidence suggests that women are more enthusiastic about the EPR. Earlier research from the Rathenau Institute (2009) indicated that higher educated persons are less enthusiastic about the EPR; this study weakens their finding. Abbreviations BSAS EHR EPR HPPAS MC MD MP NHS NKO WVS British Social Attitudes Survey Electronic Health Record Electronic Patient Record Health Privacy and Political Attitudes Survey Medisch Contact (a Dutch medical magazine) Medical Doctor Medical Practitioner National Health Service (UK) Nationaal KiezersOnderzoek (Dutch Parliamentary Election Studies) World Values Survey Front cover illustration I composed the front picture using these original pictures: database icon by Mauricio Piacentini, public domain folder (blue) by the Open Clip Art Library, public domain medical imaging by Nevit Dilmen, GNU Free Documentation License no person by an unknown author, GNU Free Documentation License physician icon by Aha-Soft, free or non-commercial use 2

3 Acknowledgements Prof. Dr. Joyce Outshoorn for guiding me while writing this thesis. Dr. Frits G.J. Meijerink for reviewing the draft and final versions of the thesis. Drs. Armèn Hakhverdian for helping with the draft research proposal. Mireille Schaap, my girlfriend and medical student, for proofreading my thesis. Martin Tückermann, MA, historian, for many interesting discussions. The Leiden Institute for Advanced Computer Science for some good courses. All errors are my own. Writing Conventions This thesis is written using the British spelling. I use random capitalisations in headings if I feel it to improve readability. References are given using the Harvard system, APSR style (American Political Science Review). For dates, the ISO-8601 international standard is often used (for example, means February 4, 2009). For more information about this standard, please refer to: time_format.htm About the Author Evert Mouw is a graduate student of political science in Leiden and of medical informatics in Amsterdam. He also is a Microsoft certified systems engineer (MCSE). Personal website Thesis website post@evert.net 3

4 List of Tables Table 1: Public popularity of health databases by country Table 2: Count of privacy in the 2006 platforms Table 3: gender and age Table 4: religion Table 5: political parties, absolute numbers Table 6: political attitude correlation Table 7: age Table 8: gender Table 9: gender and party Table 10: education Table 11: multiple regression (linear) for DV principle Table 12: Encoding of the party variables in the SPSS database Table 13: Dutch political parties in parliament after the elections of List of Figures Figure 1: Political map in The Floodgates of Anarchy (Christie and Meltzer, 1970) Figure 2: Google Trends: electronic health record Figure 3: Google Trends: electronic medical record Figure 4: Google Trends: ele[c k]tronisch patientendossier Figure 5: Sigmund, a Dutch comic by Peter de Wit Figure 6: Privacy Barometer, party ranking of Figure 7: Age pyramid of the respondents, output from SPSS Figure 8: Age pyramid for the population at large (CBS, 2010) Figure 9: Religious groups (CBS, 2009) Figure 10: gender difference Figure 11: gender and political positioning

5 Contents 1. Foreword 7 2. Introduction New technology, new politics Reading Guide Previous work Literature review Political space (attitude mapping) Privacy Health Privacy Opting in, opting out EPRs and State Building Trusting the Government The current state of health databases It started around Popularity and use Storage and accidents Corporate initiatives Introduction in the Netherlands Privacy and politics in the Netherlands Dutch election programs Privacy Count Analysis Privacy Barometer Research question Summary Theory and expectations Conceptualisations Expected attitudes based on political dimensions Left vs right Libertarian vs authoritarian Materialism vs postmaterialism The role of education, age, gender, and religion The relation between health privacy and other privacy issues Research methods and materials: The Health Privacy and Political Attitudes Survey A quantitative approach Design of the survey The required number of participants Viral marketing Results Demographic Analysis Sample size and gender Survey running time Age distribution Age and gender Education Religion Evaluation of the viral marketing and randomness Summary of the Demographic Analysis 41 5

6 10.2. Statistical Analysis Political party dimensions Political attitude axes of the BSAS Age and principle Gender and principle Gender and party Religion and principle ID card and principle Education and principle Summary of the Correlations Found Multivariate linear regression model Discussion of the Statistical Analysis Comments Report About the survey format Against the restricted use of health information Reasons for being against the EPR Ideas on the implementation of the EPR Analysis of the comments Conclusions References Books and Articles Materials News Media Attachments Att: Questionnaire Att: datafile addendum Att: party encoding in the data file Att: SQL queries Att: Dutch elections of Att: EPR mentioning per Platform Att: all comments Comments received from the survey free text input Comments received by Att: The Medisch Contact survey 76 6

7 1. Foreword in the past no government had the power to keep its citizens under constant surveillance. With the development of television, and the technical advance which made it possible to receive and transmit simultaneously on the same instrument, private life came to an end. Nineteen Eighty-Four (Orwell, 1949) Today, I made a walk through the woods and heathland of the Veluwe, a region in the Netherlands where my father s family has lived for centuries. The region is known for its beautiful nature and old-fashioned people. Only two generations ago, these people were independent, lived on their own, and knew nothing about extensive regulations, digital networks en electronic databases. In many respects, I belong to this region living and thinking in ways that many would regard as out-dated. After the walk, I switched on my computer. I have multiple computers at multiple locations, earned my Microsoft systems engineer certification years ago, and I also explored the world of arcane command line use with Linux. Being in information technology is not just about technical skills. The geek world has its own literature, culture, and politics: from the Hitchhiker s Guide to the Galaxy to the formation of new identities. The network has replaced the city. Digital property is not bound to space or time. The flow of information can cross boundaries that are taken for granted by the old world. This new culture and technology is guiding the main culture to new roads. Access and availability of information is becoming limitless. The individual thus will have access to a limitless choice of ideas, lifestyles and art. That changes the relation between society and the individual. Society will be less able to guide the individual to some best option, because the individual wants choice and wants to choose himself. Those who are first to discover the new information networks are more likely to reject the authoritative tendencies of modern society and modern government. The people of the current postmodern technoculture are often not understood by mainstream society, for they are too far ahead of the others. I am a part of that culture, too. So where does that leave me? On one foot I am standing in a world too old to be compatible with modern culture, and on the other foot I am standing in the world of the artificial, too far ahead of the mass to be compatible with modern culture. But what is, then, modern culture? I like to define it as the nation-state with a fat government. It did not exist a few hundred years back, but now it does and it taxes and regulates the not-so-free citizens as if they were part of a machine. Its properties are bound to space and time. And, as such, it has difficulty coping with 7

8 the new digital networks that are definitely not bound to space and time. The masses and the governments want to tighten their grip on the new developments. Instead of doing nothing and becoming irrelevant, some of them now try to guide and control the new developments. They limit free speech on the internet and they create massive databases and espionage networks to spy on their own citizens in order to protect and care for their citizens. It is a bit like the Orwellian Big Brother, but in some respects, it is becoming even worse. I will use medical records, the subject of this thesis, as an example. The electronic patient record (EPR) will improve health care and will make patients less dependent on one hospital or one doctor. It was developed by universities and private organisations, but in the Netherlands, it is now regulated by the government. The digital records make patients independent of space: everywhere in the Kingdom of the Netherlands, their records will be available and medical care can be arranged. No time is needed to transfer the records, and they will be available 24 hours a day, 7 days a week (independent of time). The sciences of the artificial will penetrate and change our world like never before. New possibilities are emerging, and new risks. Many computer scientists and programmers love the new possibilities, but at the same time urge the public to take measures to secure their privacy and data. The problems begin to emerge when you think about the meaning of ownership of data. Who owns the data? The creator of the data, the owner of the physical storage where it is to be stored, the system administrator who controls the computer storage, the person that is described by the data, or some higher regulating body? It is hard to define, but for sure, when you can control both the storage of, and the access to your data, then nobody can dispute your ownership. For example: you have an encrypted memory stick with some nude photographs of your girlfriend. As long as you do not give the password to others, those photographs are yours. But them to some friends, and you lose all control. Or store them online and you might think that you are the only one that has access, but who knows who can have access to your online photographs, your online s, your online tweets, and the list goes on and on. Who knows who has access to your online medical records you have to trust the administrators of the database. You have no control over the storage media, nor do you have control over the access mechanisms. Things would be different if all your medical records would be encrypted and only you would have the private key, but such is not the case. I am using my own server. I like to store my on my own server. Maybe I am like that guy in Vernor Vinge s book Rainbows End that used an old computer, so that he could have control over his own hardware and software. In the end, even his computer was taken 8

9 over. Digital information is of such nature; it is very hard to protect. Digital information is very easy to replicate, according to Icke (2009), astrophysicist in Leiden. He dislikes the EPR with a passion, because he fears that the information will not be safe, and very hard to delete. He would prefer to carry his medical records on his own memory stick. Mind though, that I understand why the networked databases approach was chosen. Each hospital will administer their own databases, and a real central database for the medical records will not exist. Instead, a national directory will be implemented, so that every doctor can search for medical records and access them, wherever they are. This decentralised, distributed database system is more robust and more secure. It has its advantages over the idea to carry your records on a memory stick. Such memory sticks are easily forgotten, stolen or lost, and while the data can be encrypted, such loss could be very dangerous for someone who needs medical help. Databases also help to improve health care because they make epidemiological research easier. But for legal, practical and other reasons, patients will often not be able to change or delete information about them. For example, a diagnosis made by a doctor should not be changed by the patient, because the doctor is held responsible to create a professional diagnosis report, and the patient record is the place where his report is stored. So much private, personal information about you in some database that you cannot control may be bad enough, but it is only the beginning. Just imagine if all those different databases, all your and phone communications, your credit card and banking transactions, your online video, music and book orders, your calendar, all your movements in the public transport or on the road, your fingerprint, the people you have contact with, and finally your medical records are all linked together. Yes, that is right, linked together to get one big picture of all that is you. That is possible with today s technology and databases in the Netherlands. New, advanced computer programs will be able to make an estimate of your role and activities in the social networks you are part of, and will be able to make better predictions on your behaviour than you yourself would be able to. And all that data is collected on you, without you being in control. That is our situation in the Netherlands, anno domini Of course, no institution has the legal authority to combine all those data sources and to access all that information about you. Maybe the secret service, but who knows. One of the nice properties of digitised data is that it can be stored indefinitely, replicated without limits, and sure you feel fine about your data being stored indefinitely, without being in control, without knowing for absolutely sure how that data will be used in the future. 9

10 So, our government has a huge responsibility, one it might not be able to handle. I feel that such a responsibility should be felt deeply within the government. They should not create databases when it is not absolutely necessary. The database containing the fingerprints of all Dutch citizens does not provide much confidence. Nor does the plan to store all public transport usage information of individuals for seven years. The EPR is no exception to this: the web application that would enable patients to see which doctors and institutions did access their records is not yet ready, but despite that fact, the government wants to roll out the EPR as soon as possible. All data communications for the EPR are secured and encrypted, but the data stores are not. With yesterday s technology, it is possible to create additional levels of security that make it much, much harder to steal medical records or to gain unauthorised access. Those technologies should be used for such sensitive data as medical records. Fairly good is not good enough. It should be excellent, because it should be secure even in the technological world of tomorrow. And yes, I understand that doctors want to be able to access the medical records in cases such as emergencies or when the patient is not conscious and cannot give access himself. But giving them the possibility to just gain access to any record by default is wrong. I like the EPR and its potential possibilities. But I would rather control the data myself, or at least see a more secure implementation with encrypted storage and biometric access controls. Governments are not aware of their moral responsibility in an increasingly technological world which they fail to understand. I am not sure how corporations will compare with projects such as Google Health and Microsoft Health Vault. It often comes down to usability for the masses, what makes corporate designs similar to governmental designs. We might need a completely new, open, secure design for sharing and controlling this kind of personal data. 10

11 2. Introduction 2.1. New technology, new politics Your medical records and your genetic information are yours, and are confidential. Both are related to your health and life expectancy. A breach of confidentiality would infringe your privacy, but a lack of information availability could endanger your life. Governments are involved in regulating the emerging Electronic Patient Records (EPR) and genetic databases. That makes the regulation of health privacy a political issue. Those who opt out, are they critical of government control and proponents of individual liberty? We wonder which kind of people they are, and how they think about politics. Which political attitudes identify the opt-outers? And, when trying to answer this question, what do we exactly mean when we talk about political attitudes? The EPR, also linked to health privacy, is a hot political issue at this moment. The social relevance of the political attitudes of opt-outers is easy to grab. Politicians, civilians and voters are trying to find a balance between government control, health, and privacy. A better understanding of the relationships between political preferences, privacy needs, and centralised health information can improve public management of health information. The scientific relevance touches a few different areas. Most public opinion research and discussion on the EPR and genetic databases has been done by medical researchers and computer scientists. The political aspects of the public preferences are barely touched. We know that most people favour the EPR, but is there a linkage to political attitudes? For example, do left-wing voters, on average, prefer the EPR more? This question is not yet answered by political scientists because the introduction of the EPR is such a recent development. Similar genetic issues, though, have been included in some surveys. The question is probably part of the bigger quest to understand the relations between political attitudes and the priority given to health privacy. Existing theories that explain the interconnections between political attitudes and health privacy issues are mostly characterised by their lack of existence. This research tries to explore this terra incognita. 11

12 2.2. Reading Guide Not much, if any, scientific literature on the relationship between political attitudes and health privacy exists. Why this is so is explored in the next section. This lack of knowledge means also that a theory must be developed from scratch. Luckily, some foundations for such a theory are present. Concepts such as privacy, political attitudes, and EPRs, are well covered by scientific literature. That means that a new theory should use already accepted insights and concepts. It is usual to first present an introduction and an extensive literature review to explore the current state of scientific knowledge. But in this case, the current knowledge is mostly void. The literature review acts more as an overview of the building blocks of my theory it gives an idea of current insights on privacy, political attitudes, and the role of government. As such, it acts more as a background and reference section. If you are already familiar with theories on privacy and political attitudes, you should skip the literature review. If you have knowledge of current developments of electronic health records, then you should skip the chapter on the current state of health databases. And for those that know how political parties score on privacy, the chapter on privacy and politics in the Netherlands can be skipped. 3. Previous work Health privacy is a subclass of privacy in general, but it is quite new and the concept needs special treatment. In short, as personal information on your health becomes digitised, new privacy issues emerge. You can read more about it in the subsection Health Privacy on page 18. Data on health privacy related to political attitudes is scarce because electronic medical records are so very recent. Only very few nations have implemented EPRs. So, the subject is too recent to be subjected to social research, and the number of countries is too small to have attracted many researchers and their surveys. This double-edged sword has eliminated the possibility of any datasets that contains both data on political attitudes and health privacy. Sometimes, it is even difficult to find datasets that contain data on both political attitudes and traditional privacy issues. For example, the World Values Survey does not include privacy issues, although some other surveys do. The surveys designed by medical researchers, information scientists and public opinion experts only measure the popularity of medical records and how people feel about a number of 12

13 practical issues; they do not care about political attitudes. Examples of such surveys are done by Pollara and Earnscliffe (2003), Rynning (2007), Public Opinion Strategies (2005), Westin (2008), UMR Research (2008), Maurice de Hond (2009), and TNS NIPO (2009). And mostly all surveys end theories designed by social scientists do not recognize health privacy as a separate concept. Examples are Westin (2003), Margulis (2003), and the NKO (2006). Those that do mention health privacy as a separate concept, such as Hustead and Goldman (2002), Solove (2002), Cross (2006), Tanne (2008), and Castle and DeBusk (2008), do not explore the relation with political attitudes. I could only find one weak exception: the British Social Attitudes Survey. The BSAS is a well-known and widely used data source. The inclusion of health privacy questions might be inspired by the fact that the UK is one of the first countries to use electronic health databases. Still, while a lot of questions related to political attitudes are available, the number of questions related to health privacy is very low only one question. In fact, no questions on medical records are included, just only one question on genetic databases can be found. Also, the number of years in which health privacy questions were included is very small. This means that the BSAS is not very useful as a data source. Not a survey, but a study nonetheless, was done in The Netherlands by the Rathenau Instituut (2009), ordered by the Upper House. The study was qualitative, so the results are not representative for the Dutch population. One interesting finding of the study is that higher educated persons are more likely to object to the inclusion of their personal data in the electronic medical record. This lack of useable data introduces a necessity to collect new data. Then a new theory on the relationship between health privacy and political attitudes can be tested. This means that a new study must be designed. The building blocks of such a study are concepts such as privacy and political attitude. Therefore, I first present a literature review of these concepts. 4. Literature review This research deals with a number of subjects that are already well described in the social sciences. Literature exists on political multi-axial positioning or political space, and also on privacy. Useful literature for this research, in which concepts and subjects used in this research are explored, is summarized in the subsections below. Also mentioned are important surveys and data sets relevant to this research. 13

14 4.1. Political space (attitude mapping) Different positions in the political space can be represented using a simple right-left onedimensional model, but many social scientists did not find this satisfactory. Early in 1939, Leonard W. Ferguson identified three factors (dimensions): Religionism, Humanitarianism, and Nationalism. Hans Eysenck published his book Sense and Nonsense in Psychology in He found two factors, which he named Radicalism (R-factor) and Tender- Mindedness (T-factor). Oddly, rotating the graph of Eysenck 45 degrees renders the same factors of Religionism and Humanitarianism of Ferguson. Other two-axial models followed suit, like those of Milton Rokeach, David Nolan, Jeff Greenberg & Eva Jonas, Jerry Pournelle, and Ronald Inglehart. An eight-axial model was proposed by Patrick Mitchell. An example of a two-axial map is given in Figure 1. Figure 1: Political map in The Floodgates of Anarchy (Christie and Meltzer, 1970) The model of Inglehart became very well-known and often used for its application by the World Values Survey. His two-dimensional model can be used to map values of individuals, and also to compare the values of populations in different countries. On the vertical scale, values of religion, authority, and tradition dominate. The bottom is represented by traditionalist positions such as patriotism. The top is occupied by secularism and non-traditionalism. The horizontal scale measures the transition from industrialism to post-industrialism. The left side represents industrialism and survivalism, and the need for physical security. The right side 14

15 represents the open self-expression and post-materialism that results from the wealth in postindustrial societies. The right side is associated with trust and tolerance. The political spectrum or political space can be graphed quite easily when a two-axial model is used. The multiple two-dimensional models have given rise to a number of political spectrum graphs. The Inglehart-Welzel Cultural Map of the World is famous it groups cultures together using the two dimensions that are measured by the WVS. In The Netherlands, a map well known by the public is the two-axial model used by the Electoral Compass (Dutch: Kieskompas). The academic director is André Krouwel, a political scientist. The two axes are the conservative progressive axis and the left right axis (Kieskompas, 2010). The model is very useful because the map is often updated to reflect actual party positions. The British Social Attitudes Survey lists four different dimensions under its topic attitude and values scales. These dimensions show some similarity to the other multi-axial models mentioned. The dimensions are: Left right scale Libertarian authoritarian scale Postmaterialist materialist scale Welfarist individualist scale 15

16 4.2. Privacy Holvast (2009) provides an overview of the history of privacy. The abstract of his chapter History of Privacy in the multi-author book The Future of Identity in the Information Society provides an excellent introduction into the concept of privacy. I could not improve the wording, so I replicate the abstract in its original form: Discussion on privacy issues is as old as mankind. Starting with the protection of one s body and home, it soon evolved in the direction of controlling one s personal information. In 1891, the American lawyers Samuel Warren and Louis Brandeis described the right to privacy in a famous article: it is the right to be let alone. In 1967 a new milestone was reached with the publication of Alan Westin s Privacy and Freedom when he defined privacy in terms of selfdetermination: privacy is the claim of individuals, groups, or institutions to determine for themselves when, how, and to what extent information about them is communicated to others. History of privacy makes clear that there is a strong relationship between privacy and the development of technology. The modern discussion started with the use of cameras and went on to include the development and use of computers in an information society in which personal data on every individual is collected and stored. Not only is it a great concern that privacy is eroding but also that we are entering a surveillance society. This loss of privacy seems to be even more the case since the protection of privacy is strongly dependent upon the political will to protect it. Since 9/11, however, this political will world-wide is oriented more toward the effective and efficient use of technology in the battle against criminality and terrorism than it is toward protecting privacy. Therefore it is time to re-evaluate the use of technology and the protection of privacy. It is not only privacy that is at stake but above all democracy. Westin (2003) also defines privacy as the claim of an individual, social group, or association to determine what information about himself/itself should be known to others. He writes that privacy is frequently determined by the individual s power and social status. The rich can withdraw from society when they wish; the lower classes cannot. Ironically, though, the rich, the famous, and the politically powerful are also the people whose privacy is most assaulted by the media, political rivals, government investigators, and the like. He distinguishes four aspects of personal privacy needs: family life, education, social class, and psychological makeup. As a result, the individual s needs are constantly changing. 16

17 After giving a historical overview of privacy developments, he concludes his article with the comment that privacy is a quality of life topic that will exert a major influence on the quality of civic life in the 21 st century. Privacy can mean a lot of different things in the literature. Solove (2002) mentions six general headings: 1. The right to be let alone (from Samuel Warren and Louis Brandeis). 2. Limited access to the self - the ability to shield oneself from unwanted access by others. 3. Secrecy - the concealment of certain matters from others. 4. Control over personal information - the ability to exercise control over information about oneself. 5. Personhood - the protection of one s personality, individuality, and dignity. 6. Intimacy - control over one s intimate relationships or aspects of life. Woven into these six headings is physical privacy, or the ability to shield your body from unwanted access. In the American constitution, the right of the people to be secure in their person is guaranteed in the fourth amendment. The Dutch constitution is more explicit in article 11, where the invioliability (or; integrity of the body is guaranteed) of the human body is protected 1. Current research on bodily integrity and the relationship between the body and citizenship is primarily done in women s studies. For example, Outshoorn (2008) opens her working paper with European states vary widely in the extent to which they grant control to women over their own bodies and in guaranteeing the basic human right of integrity of the body and remarks that the issue is still a central concern for women in general today. Margulis (2003) gives an overview of the concept of privacy and current theories about privacy in his article Privacy as a Social Issue and Behavioral Concept. He describes what privacy is and what privacy does, the costs of losing it and the socio-political aspects of privacy. A summary of his article follows: Defining privacy is tricky because the concept is elastic and a wide variety of definitions exist. Privacy can be seen positive as beneficial for individuals and society, but some authors warn for negative consequences such as the protection of criminals and liars. Margulis (2003) combines two different theories of privacy, and he finds many overlapping characteristics. The limited-access approaches discuss how individuals and groups control or regulate access to themselves. Both theories describe our need for privacy as a continuing dynamic of changing internal and external conditions, to which we respond by regulating privacy. Privacy has 1 Dutch: Ieder heeft, behoudens bij of krachtens de wet te stellen beperkingen, recht op onaantastbaarheid van zijn lichaam. 17

18 universal characteristics but the forms that privacy can take is probably culturally specific. The attempts to regulate privacy may be unsuccessful. Privacy benefits on the socio-political level include opportunities for political expression, political choice, and participation in associations such as family and religion. At the psychological level, privacy is important for social interaction and a healthy self-definition. Losing privacy can result in stigma. A stigmatized individual can be the target of negative stereotypes and discrimination. (Margulis, 2003) The social importance of privacy has diverse aspects. Democracy is supported by, and supports, privacy. Not only individuals have a need for privacy; groups and organisations can also have a need to keep some of their information private. Balancing the interests of individual privacy against public goods (such as law enforcement) often results in less individual privacy. This may be caused by the framing of individual privacy, while whole groups and even the whole society benefits from privacy rights. (Margulis, 2003) The World Values Survey does not cover privacy issues directly: The WVS is a survey that has been conducted since 1981 in four waves throughout approximately 80 nations. While surprisingly excluding questions that deal with privacy and surveillance, we are able to take questions about trust, governance, authority, relationships and gender from the WVS. This allows us to see how countries scores on these values relate to their attitudes toward privacy and surveillance. We will be able to hypothesize whether or not different cultural values elicit different attitudes toward privacy. (Zureik, Stalker and Smith, 2006, p.12) Their research is finished, but the summary report (The Surveillance Project, 2006) does not mention the promised correlation between the WVS data and their findings. This will be done in their book (Zureik, Stalker, Smith, Lyon and Chan, 2010), which is forthcoming. The British Social Attitudes Survey does include multiple privacy questions. For example, the questionnaire contains the likelihood of genetic information being used in getting insurance. This means that this data set allows correlating political preferences to privacy opinions, including health privacy Health Privacy An important subclass of privacy is health privacy. This is sometimes recognised in law: the USA federal government has issued the Model State Public Health Privacy Act. Gostin, Hodge and Valdiserri (2001) explain the model act, and comment that existing state laws are 18

19 inconsistent and inadequate. They also stress the importance of health privacy for the quality of public health. Health privacy mostly concerns medical records and genetic information. Margulis (2003) discusses genetic privacy and mentions the overall weakness in legislative protection of health-information privacy. He calls the genetic/medical privacy a major policy battleground. An example can be found in the Dutch constitution (see 4.2). The physical inviolability of the human body is made explicit, but in the constitution, it is not explicitly extended to (medical) information about the human body. The professional secrecy for medical practitioners is arranged in other (secondary) laws, such as the medical code of practice. This forces medical professionals to protect the medical records of their patients, but health privacy as a fundamental basic right vis-à-vis the government is not made explicit. A famous case in the USA is Jacobson v. Commonwealth of Massachusetts (1905). The Supreme Court decided that the government can enforce vaccination if it is necessary for public health. In such cases, when a germ is affecting a large part of the population, the common interest (herd immunity) is held more important than individual bodily integrity. But in another case, Zucht v. King (1922), the Supreme Court ruled that a public school student could not be forced to have a vaccination because no danger to public health was present. More on these and other cases are described by Curry (2002). Castle and DeBusk (2008) write about nutritional genomics and the ethic difficulties which face registered dieticians. Electronic health records often contain genetic information, but US federal privacy protection is only minimal. Health privacy concerns not only affect isolated individuals. Genomic disorders often affect whole families. Medical records can reveal family implications, like increased risk of Alzheimer s disease to other family members. Another problem noted by Castle and DeBusk (2008) is the problem of unsolicited information: after a genetic test for some problem x, the doctor might see test results for some other illnesses y. The patient might profit from pro-active treatment for y, but maybe he/she does not want to know about the problem. Patients who can read their own medical records might face such unsolicited information. An example of a group that advocates health privacy is the American based Health Privacy Project 2. Hustead and Goldman (2002) discern four components of medical information

20 privacy protection: 1) access, 2) use, 3) disclosure, and 4) storage and security. They also stress the importance of non-discriminatory use and adequate legislation. Goldman and Hudson (2000) recommend both self-regulatory policies and better legislation to engender public trust and confidence in both traditional health care as well as e-health activities. Hustead and Goldman (2002), Solove (2002), Cross (2006), Tanne (2008), and others also stress the importance of new legislation. The emerging technological possibilities must be covered by law. Castle and DeBusk (2008) advise strict controls around electronic health records to ensure that RDs can make the most of the benefits of electronic record management, while gaining the client s trust that private health records will remain private Opting in, opting out Difference of opinion exists about opting in or opting out. Privacy proponents argue that an opt-in system offers the best privacy safeguards, while those who focus on the best patient care possible like opt-out systems best. An example of this discussion is given by Watson and Halamka (2006). In the editorial of the same issue of the BMJ journal, Norheim (2006) agrees with opting out: It s ethically sound for patients to opt out. He defends the notion of soft paternalism because people do not always make the choices that are best for them. However, the Royal College of General Practitioners strongly recommend an opt-in system (Cross, 2006) EPRs and State Building Freeman (2002, p.764) writes that information policy in health care can be thought of as a kind of state building. He bases his conclusion on his analysis of current trends in information technology, health care and expanding government administrative territory. By defining and regulating the medical information infrastructure, the government increases the role of the state. Such collectivisation could increase the power of the state over medical practitioners and patients. But it should be noted that government is not a single, focused agent (Freeman, 2002, p.765). The internal contradictions of government may be well illustrated by the findings of Miller and Tucker (2009). They found that in the United States, electronic medical records are sooner implemented in states that lack privacy regulations: privacy protection may inhibit adoption if hospitals cannot benefit from easily exchanging patient information. According to their study, 20

21 the yearly adoption rate of electronic medical records is 11% lower than in those states that have state privacy laws Trusting the Government Is trust in the government related to political attitudes or individual political ideology? Miller (1974) researched the American trust in their government. He concluded that the American voters are polarised, so that a Democratic policy will create distrust with the Republican voters, and vice versa. A policy that would try to find a middle ground would create moderate distrust with both groups. In such a situation, some distrust with the Government is unavoidable. His conclusion was attacked by Citrin (1974), who makes a distinction between trust in the political system (governmental system) and the current policy of the current political party in office. This kind of discussion is typical for the topic of trust and government. Trust is hard to define, and even harder to measure. Dalton and Klingemann (2007, p.353) say that political trust when viewed from the individual level is a puzzle. Furthermore, political trust is usually even more randomly distributed between social groups (p.356). In general, we just do not know much yet about the factors that influence trust in the government. Trust in others might be an altogether other variable than trust in the government. Killerby (2005), using data from the World Values Survey and the Worldwide Governance Research Indicators Dataset, concludes that trust in government does not have a statistically significant correlation with the level of social trust (trust between people). Also, the level of trust in the government is not necessarily indicative of the quality of governance. Is one s willingness to let the state organise medical records dependent on one s trust in the government? Maybe it is, but the ambiguity of the concept political trust and the different but related trust in the governmental system and the actual politicians who make up the government do little to make the idea verifiable. Trust in the government is a related but too vague a concept to be of much use. 5. The current state of health databases The unified concept of health privacy includes both genetic privacy issues and medical records issues. Therefore, I include both genetic databases and medical records in the more general term health databases. How popular and widespread are health databases in our society? Which problems and issues do practitioners and societies face? 21

22 5.1. It started around 2006 General public interest in electronic health records can be shown to have started in In that year, many people started to submit queries to Google on that subject. This can be visualised using Google Trends 3. Figure 2 shows the results from the search query electronic health record (taken from the internet at ). Figure 2: Google Trends: electronic health record A similar query, now for the phrase electronic medical record, shows that some people already searched online for such information since 2004, but this phrase seems to get a bit out of vogue. The spike in 2006 corresponds with the start of online searches for the electronic health record : Figure 3: Google Trends: electronic medical record Because this research is conducted in the Netherlands, I have added the queries in the Dutch language 4. They are called elektronisch patientendossier (in red) and electronisch The correct spelling is patiënt but I choose to search for patient because this includes the former and also catches common spelling errors and computer encoding problems. 22

23 patientendossier (in blue). Both queries mean the same thing, there is just a small spelling variation possible. A single spike at the end of 2008 indicates the start of the public discussion on EHRs in the Netherlands: Figure 4: Google Trends: ele[c k]tronisch patientendossier This, of course, creates just a first impression. Health information was digitised long before 2006, as is shown in the next subsection Popularity and use In the year 2000, Iceland sold the medical and genealogy records of its 275,000 citizens to a private medical research company (CNN, Jerrold Kessell, 2000). Iceland has a homogeneous population and a detailed history of family linkages. DeCode, a private medical research company, will use the information to undertake unique genetic research. Many physicians criticised the selling of sensitive information and were concerned about the relation between medic professionals and their patients. Another concern is the level of control that is given to a private company. Not only has the Icelandic government handed over the control of genetic information, it also makes it hard or even impossible for third parties to access the information. Such acts could evoke public resistance, and gaining public approval by means of a referendum would seem appropriate. But no such referendum was held, and the Icelandic population is not very concerned: only five percent has opted out. Genomic information is often stored in medical records. Current public usage of Electronic Patient Records (EPRs) is limited to only very few nations. The United Kingdom was one of the first to implement a nationwide EPR system: the project of the National Health Service started in Another early bird is Sweden: the Swedish National Patient Summary was launched in Sweden has outsourced the implementation to two IT vendors, InterSystems 23

24 and Tieto. The Canadian province Alberta has a large-scale functioning network, but a nationwide system does not yet exist. The USA is still struggling to implement electronic records. Some other nations are preparing for the implementation of EPRs. Public opinion polls measuring the willingness to allow personal health information to be stored in electronic databases have been held in many countries. Because the polls did not use the same questions, the results are not comparable. The general impression, though, is that health databases are popular with the public. It is hardly researched how support for an EPR is linked to political allegiance. One of the few surveys that questioned both EPR support and political attitudes came to this conclusion: Support for the creation of this type of network or exchange is strong and consistent regardless of political affiliation, age, education, or socio-economic status (Public Opinion Strategies, 2005). In the American two-party system, 72% of the Republicans favoured the EPR, and 71% of de Democrats likewise. Table 1: Public popularity of health databases by country Country Year Perc. Support in favour of Source Canada % genetic research database Pollara and Earnscliffe (2003) Sweden % shared, national HER Rynning (2007) USA % health information network Public Opinion Strategies (2005) USA % electronic PHR Westin (2008) Australia % individual HER UMR Research (2008) Netherlands % electronic EPR Maurice de Hond (2009) Netherlands % electronic EPR TNS NIPO (2009) Some medics are less enthusiastic than the general population. Nearly 31% of the Dutch MDs have already objected to the use of their medical information in electronic health records, while an additional 25% considers doing so (Katzenbauer, 2009). The main arguments of the opponents are based on a lack of confidence in the privacy, security and maturity of the EPR. They also regard the EPR as a threat to their professional secrecy. Dutch citizens are not enthusiastic at all, according to Maurice de Hond (2009). But a similar research by TNS NIPO (2009) found the opposite. DeVrijePsych (2009) explained this partly by the fact that the survey of TNS NIPO was held before the results of Katzenbauer (2009) became public. The survey of Maurice de Hond was held after the publication of many doubts expressed by MDs. 24

25 5.3. Storage and accidents Privacy issues became even more urgent after November 2007, when the UK government lost two discs with Child Benefit data containing name, address, date of birth, National Insurance number and, where relevant, bank details of 25 million people (BBC news, 2007). While these were no medical records, it still illustrates the privacy and security concerns for centralised databases. Only a year later, the USA would face an even worse privacy catastrophe. In November 2008, the medical records of millions of patients were stolen from Express Scripts, a medical benefits management company (New York Times, John Markoff, 2008). In some cases, the threat of identity theft can be nearly as damaging as the actual theft of medical records. In 2009, hackers claimed to have accessed 8 million patient records and 35 million prescriptions from the US state Virginia s prescription drug database (Krebs and Kumar, 2009). The authorities decided to shut down parts of the computer system, although prescriptions could still be filled. The Dutch decentralised data storage could be the best way to protect patient data. McGilchrist, Sullivan and Kalra (2007) argue that accidents such as the loss of sensitive personal data of 25 million citizens in the UK in 2007 would not occur so easily when patient data is stored in multiple institutions, with multiple levels of security. A single, centralised database is also a single point of failure and a single target for attackers. Another reason for decentralised information systems is to prevent a rigid centralised approach that may conflict with local or specialised needs. Kmietowicz (2007) writes that MPs are confused about which details should be included in the two medical databases that were operated in the UK. He quotes Kevin Barron; MP, member of the Labour party, and chairman of the cross party health committee: A highly centralised approach to the NHS IT programme has stiffed local activity, causing frustration and resentment. Relaxing central control will make local trust and strategic health authorities feel more engaged in the project. The chip card used to get access to the Dutch EPR system was hacked in early 2009 by Erik Westhovens, a security specialist (de Winter, 2009). According to him, it was easy to hack the chip card it took him just a few hours and some basic equipment because it uses an undisclosed algorithm, while many argue that security algorithms should be open to enhance public testing. The minister of health reacted by announcing a newer chip (Novem, 2009). 25

26 5.4. Corporate initiatives A recent development is the offering of personal medical record storage by companies such as Microsoft and Google. Microsoft HealthVault, as of still in beta phase 5 but launched in October 2007, promotes 6 its service as follows: Make more informed health decisions for you and your family. Store your health information in one convenient place and share it with others you trust. You control your health information. You decide who can share it, and what they can share. We always ask for consent before allowing another person or Web site to access health information. A major recent success for HealthVault is the adoption by the Mayo Clinic, an international renowned medical practice (Timmer, 2009). Google Health, launched in February 2008, offers a similar service, with a similar privacy policy 7 : Google stores your information securely and privately. We will never sell your data. You are in control. You choose what you want to share and what you want to keep private. A successful pilot project was carried out at the Cleveland Clinic in Ohio. Other medical centres were impressed: This is truly a patient-controlled health record, and that s a very significant step in the drive toward a more consumer-oriented system of health care, said Dr. John D. Halamka, chief information officer of the Harvard Medical School (New York Times, Steve Lohr, 2008). Kidd (2008) discusses the two EPR systems developed by the British government, and compares them with the corporate alternatives. They seem to be rather optimistic about the corporate offering: But perhaps this whole development by the NHS is all too little, too late. Is the NHS summary care record a 20 th century healthcare solution being overtaken by 21 st century technology and increased sophistication in the use of the internet in the community? Given the choice of having governments create and exert a degree of control over your internet based personal health record, and being able to do it yourself with a little help from Microsoft or Google, which would you choose? 5 Beta phase: when software is feature complete, but has not yet reached production quality. That means that the software still needs some wide-scale testing, so beta software is often released to the public with a warning that it is still beta and a request for early adaptors to fill in bug reports

27 Steinbrook (2008) also describes the rise of these company owned health record services. He ends his article with some caution: If concerns about privacy, security, and commercial exploitation can be allayed, this nascent enterprise should have a smoother birth. Another slightly cautious overview is given by Tanne (2008). She enumerates some risks such as the lack of legal regulations and the problems that could occur when an EPR contains erroneous information Introduction in the Netherlands The Ministry of Health ordered research on electronic patient records in 1997 or maybe even earlier. In 1997, a memorandum was written by the ministry with some early ideas on the EPR and on a communication network (Ministry of Health, 1997). These ideas were partly inspired by European projects, such as the CEN norm ENV Medical Informatics, an experimental standard, also from Serious thinking on the EPR began in 2005, after the Ministry of Health (2005) presented a blueprint for the development and implementation. This blueprint was no longer based on one big centralised database such as currently implemented in the UK, but it contained designs for a central directory and communications system to exchange and access medical records. Also mentioned were other countries working on an EPR, such as the USA, the UK, Sweden, Germany and Canada. After the presentation of this blueprint, discussions on the EPR were present in the Lower House year after year. The nationwide EPR was announced going live on At this date, the Parliament had not yet approved the plans, but the medical profession could then already test the technology and the procedures, and citizens received a letter informing them of the plans and explaining them how to opt out if they would not participate. The last few months (as of ) the Upper House has voiced many sceptical questions and approval is not yet certain. The recent fall of the cabinet (government) on could have some influence. Decision making about sensitive issues was halted until after the elections and a new cabinet. A number of issues are marked by the parliament as sensitive or controversial 8. Although the EPR is not one of them, the Second Chamber has had no time to debate the issue, so it will be handled by the next parliament and cabinet. Still, a new cabinet is likely to have fresh ideas about how to continue with the EPR, especially after the many questions raised by the public and the Upper House

28 6. Privacy and politics in the Netherlands Privacy International is a London based watchdog organisation. Government and corporate intrusions on privacy are recorded and compared, and an international privacy ranking of countries is published on their website: In their ranking, both the Netherlands and the United Kingdom (both have introduced an EPR) score low on privacy. Germany and Italy are doing better, and Greece has the best privacy score of the European Union members, according to Privacy International (2007). Individuals who care about their health privacy are likely to have a preference for parties that upheld privacy. But which political parties care about privacy in the Netherlands? 6.1. Dutch election programs Political interest for EPRs can be measured in different ways, like counting how often politicians touch the subject in the media. To get a quick first impression, it is easier to study the most recent political party election programs. Here follows a short lexical analysis of the Dutch programs for the parliamentary elections of Note that the analysis is done very roughly, because the only objective here is to get an impression of the recent state of affairs. For each political party, I show 1) how much attention is given to privacy issues, and 2) if any attention to electronic health records is given. Figure 5: Sigmund, a Dutch comic by Peter de Wit 28

29 Privacy Count For each political party which made it to the parliament of the Netherlands after the elections of , I have counted the frequency of the word privacy in its platform 9. When a proposal with negative consequences for privacy included the word privacy, then I counted it as a negative number. The source of the election is the study of Leiden political scientists (Leiden University, 2006). The results are shown in Table 2: Count of privacy in the 2006 platforms. A list of party descriptions, seat numbers, specific texts and other details can be found in the attachment Att: Dutch elections of 2006 on page 69. Table 2: Count of privacy in the 2006 platforms party + - rank (abbr) CDA PVV VVD PvdD SGP CU GL PvdA SP D Of these parties, the following parties did mention and favour the EPR: VVD, D66, and SGP. For such a privacy-aware party as D66, this is striking. It comes as a surprise that in the parliamentary debate of , Koşer Kaya (D66) proposed 10 to delay the EPR until patient privacy would be guaranteed. It is surprising she departs from the party programme but on the other hand, this is what voters would expect from D Analysis The political interest in privacy issues and electronic health records at the moment of the parliamentary elections of 2006 in the Netherlands is not very high, but a number of parties show some attention to the subject. At first look, most low-ranking parties have a Christian and/or right-wing ideology, while most high-ranking parties have a secular and/or left-wing ideology. 9 Count executed on using Adobe Reader. 10 Dutch: Handelingen , nr. 43, Tweede Kamer, pag

30 6.2. Privacy Barometer A far more thorough measurement of how much attention is given to privacy by political parties is done by the Dutch organisation Privacy Barometer (2010). The graph below is shown on their website and is based on how parties actually act or vote on more than 37 laws, proposals and discussion items. The CDA, again, scores very low, while GL and SP score fairly well. D66, a party which has a reputation of upholding privacy, is only doing less badly than most other parties. Of the four lowest-ranking parties, three are Christian. The fourth is the PvdA, a social democratic party. 7. Research question Figure 6: Privacy Barometer, party ranking of The preceding paragraphs have given a first impression about how political parties think about privacy and EPRs. But the subject of this study is not the political party. It is the individual. Do someone s political attitudes predict how that individual thinks about the EPR? This is the main topic of this study. It is also a subject that is not yet researched (see 3, Previous work). For that reason, I have designed a new study. The central question of the study involves the individual and his/her political attitudes. A careful wording and further elaboration of this question is needed. The main question of this research is whether the importance given to health privacy issues such as health records, if regulated by the government, is correlated to political attitudes. For 30

31 example, do individuals who favour an authoritarian government also, in general, care less about their health privacy? But how do we specify political attitude and how do we measure a preference for health privacy? The individual political attitude can be determined in two ways. It can be 1) measured by party allegiance, that is, voting behaviour. Although this measures real-world political behaviour, it is a poor method to measure one s political attitudes: often, no political party does exist that exactly mirrors one s political attitudes. This is especially true for two-party systems. A more appropriate method is 2) the mapping to a position in the political space. This political space can be represented using one or more dimensions (axes). This is explained in more detail on page 14. Thus, to satisfy both methods to measure individual political attitude, two subquestions needed answering. First, which political parties are voted for by those concerned about health privacy issues? And second, is there a correlation between individual political-ideological variables (political position) and the health privacy preference of the individual? These correlations might be dependent on culture and national political and health traditions. This research was held in the Netherlands. The introduction suggested that the health privacy is just a subcase of privacy issues in general. Such an assumption might prove false. Health privacy could, in theory, be unrelated to other privacy preferences. To test this, a third subquestion was asked. Is there a linkage between individual health privacy preference and between the individual stance towards other privacy issues? Summary Main question: health privacy preference political attitude Sub question 1: health privacy preference political party allegiance Sub question 2: health privacy preference political position / space / scales Sub question 3: other privacy preference health privacy preference 31

32 8. Theory and expectations Health records touch multiple privacy dimensions: secrecy, control and intimacy (Solove, 2002). Secrecy will be compromised if unauthorized access to the health record occurs. Control will be difficult when medics, not patients, control the information in the patient record (Kidd, 2008). And often, patient records will contain sensitive information about the intimate life of patients. The individual weighs these privacy concerns against the possible private and public health advantages to be gained Conceptualisations A discussion of these concepts can be found in the Literature Review on page 13. This summary is intended to offer a clear understanding of the concepts as used hereafter. Privacy: The claim of an individual, social group, or association to determine what information about himself/itself should be known to others (Westin, 2003). Health privacy: A subclass of privacy concerning medical or genomic information. Other privacy: All privacy not concerning medical or genomic information. Health privacy preference: The preference given to health privacy, measured by asking if one likes or opposes the electronic health record. Other privacy preference: The preference given to other privacy issues. Political position: A multi-axial position in the political space Political party allegiance: The political party one votes for Expected attitudes based on political dimensions One s weight given to privacy is partly dependent on one s trust in the government and others and is thus related to one s political attitudes. A political attitude cannot simply be described by the words left, right or centre. The complexity of the differences between parties and ideologies is not well represented by such a simple left-right dimension. A multiaxial representation of the political spectrum is more appropriate. I explain my expectations using three of the four dimensions used by the BSAS (see 4.1 on page 14). I have skipped the welfarism vs individualism axis because in the Netherlands, that overlaps too much with the left vs right axis. Please note that my theory is highly speculative, because no previous research in the field of health privacy and political attitudes is known to me. Here follows my theoretical expectations, mostly based on common sense and intuition. This research tries to falsify these expectations. 32

33 Left vs right I do not expect much difference here. Although right-wing voters are more critical towards governmental projects, left-wing voters might be opposed because they are more concerned about privacy issues. Probably, those voters near the centre will have more trust in de government, while ultra-right and ultra-left voters will have a more critical attitude and thus be more critical about electronic records on civilians in general. But it is hard to tell, because good predictors for trust in the government do not exist (see page 21) Libertarian vs authoritarian It is easy to grasp that state-driven projects appeal more to authoritarian voters, who trust the government, than to libertarian voters. I expect a big difference between these two opposites in respect to their enthusiasm for electronic health records regulated by the government. The case of EHRs administered by private companies might be different for libertarians. They usually are fond of economic freedom and technological advancement, so when EHRs are controlled by private organisations, they might like it Materialism vs postmaterialism I expected that those who value material goods and health would be less concerned about privacy issues at least, I expect them to be more ready to sacrifice some privacy if that would result in better health care. But postmaterialists could be more concerned with health in general. So, although I expect materialists to show more support for EHRs than postmaterialists, I do not expect this division to run very deep The role of education, age, gender, and religion Earlier research done by the Rathenau Instituut (2009) established that lower educated persons do not object to the EPR, while higher educated persons are more likely to object. There is no reason to expect different results from this study. No earlier work is known to have found a relation of health privacy with gender or age. I do not expect to find such relations, although it is always worthwhile to test against background variables. Religion might have a correlation with a preference for authoritarian government and, in that way, increase the enthusiasm for medical records. Research from the Privacy Barometer (2010) showed that Christian parties score low on privacy in The Netherlands. 33

34 8.4. The relation between health privacy and other privacy issues The term health privacy already has the implicit meaning that it is a special kind of privacy, not to be confused with other forms of privacy. It is reasonable to expect some correlation between one s health privacy concerns and one s concerns about other privacy issues; but the correlation might not be as strong as one would assume. Health privacy as a standalone concept is further explored in 4.3 on page Research methods and materials: The Health Privacy and Political Attitudes Survey 9.1. A quantitative approach Choosing between a qualitative and a quantitative approach is dependent on the objectives of the study. This research aims to correlate individual political attitudes to individual health preferences. It tries to determine whether a measurable correlation exists between two variables in a population. Such measurements call for a quantitative approach Design of the survey The lack of a solid data source inspired me to create my own survey, one which includes questions on both political attitudes and on health privacy. This is not easy to do, because the time and funds for this research are very limited. But a small online survey is doable. Which means that the number of questions should be low, so that participants do not quit before completing the questionnaire. Because of the practical restraints, I only targeted Dutch speakers in The Netherlands. I name this survey the Health Privacy and Political Attitudes Survey (HPPAS). The big things I want to cover are 1) health privacy, 2) general privacy, 3) political attitudes, and 4) some personal background variables such as age, education, occupation, and sex. The general privacy must not be overlooked. It needs to be contrasted with the specific health privacy attitudes. Note that one might care not much about one s health privacy, but still prefer private organisations over the government to regulate one s health records. So any inquiry into one s 34

35 health privacy preference must include both one s willingness to share health information with government controlled databases and with private controlled databases. I have tried not to invent my own questions. As much as possible, I want this survey to be comparable with existing surveys. Besides not wanting to re-invent the wheel, using the question wordings from well-known surveys enables future comparative use of this survey. How to measure political attitudes is a well-established practice. The BSAS (British Social Attitudes Survey) is a well-known survey from the UK with a large questionnaire and big data sets. To keep my research compatible with well-known surveys such as the BSAS, I used a subset of their questions. I selected a few questions for each political dimension that appear in recent editions of the BSAS. The questions taken from, or strongly related to, this survey are marked with BSAS with the year and question number from the user guides from which I copied them. The BSAS questions are adapted to the Netherlands: where the word Britain appears, it is replaced by the Dutch word Nederland. The BSAS was designed for a two-party system, while the Netherlands have a multi-party system. So I included some questions from a Dutch survey as well. Besides, because my survey was held in the Netherlands, it was important for my survey to be comparable to an important Dutch survey. One of the most important election surveys is the NKO, the Nationaal Kiezersonderzoek, which is Dutch for Dutch Parliamentary Election Studies. The NKO, though, has only a small questionnaire compared with the BSAS, so the pool of questions that can be used is smaller. Notwithstanding, the NKO question on euthanasia might be useful to determine if health-related self-determination extends to health privacy. Health privacy questions should include a few of the questions the Dutch medical journal Medisch Contact which held its own survey targeted at its medical audience (Katzenbauer, 2009). These specific questions on the Dutch EPR allow for a comparison with earlier research on the popularity of the Dutch EPR project. The questionnaire itself is included in the attachments The required number of participants Normally, around 1000 participants are required to get statistically significant results. A number of ca. 200 ~ 300 participants can be sufficient for smaller research projects with somewhat lower significance requirements. This research is mainly trying to answer a is there 35

36 question, and not a how much exactly question. That justifies the lower significance requirement, and in turn, the lower number of participants (a smaller sample) Viral marketing Using the internet for a survey makes it possible to use viral marketing 11. Participants were offered the opportunity to invite a few friends by . Viral marketing has the potential to reach many different persons and groups using the principle of six degrees of separation (Milgram, 1967). This means that the chance of only reaching like-minded persons, which would derandomise the selection of participants, is minimal. To remember who was already invited, I generated unique IDs from the addresses. So, I did not store the addresses without permission, but only stored a Whirlpool hash 12. Such a hash cannot be reduced to the address itself, but nonetheless is unique for a specific address. Closing the survey had to be done with some care. Some people might have unread invitations in their mailbox. So, a week before the survey was closed, no new invitations were sent. Participants did not have the possibility to send invitations during that last week. The survey was located online at: 11 An informal introduction to viral marketing can be found at

37 10. Results Demographic Analysis Does the sample offer a good representation if the population? This is evaluated in this demographic analysis Sample size and gender Of the 218 respondents, 124 (57%) are male and 94 (43%) are female Survey running time The survey started on and ended on , so the running time was exactly one month. Most responses were recorded during the first two weeks of the survey. This is likely to reduce the chance that new political developments or media coverage has had much influence on the survey outcome Age distribution The average age of the respondents was 36. That is only slightly below the average age of the Dutch population, which was 38 years in 2003 (NIDI, 2003). But a visual examination of the age pyramids for my survey and for the population at large reveals some differences. In my survey, people aged 30 to 50 are somewhat underrepresented, while older people and especially young adults are somewhat overrepresented. 37

38 Figure 7: Age pyramid of the respondents, output from SPSS Figure 8: Age pyramid for the population at large (CBS, 2010) 38

39 Age and gender The age pyramid already showed that the female respondents are, on average, younger. This is confirmed by the following query result 13 : Table 3: gender and age gender age f 32.3 m 39.5 To determine the significance of this difference between the gender groups, a Student s independent samples t-test is performed. Levene's test significance = 0,002 < 0,05 so equal variances not assumed. In other words, the variability in age is of significant difference between the sexes (which is not good). The 2-tailed significance of the t-test is zero, so there is a significant difference of the variable principle between males (mean = 32,3; sd = 13,0) and females (mean = 39,5; sd = 15,9) ( t(215) = 3,693; p = 0) Education My survey selection method has led to a high number of higher educated respondents. Many invitations were sent to students and members of student associations. How does this affect this study? In the Dutch general population of 2008, 25% has a higher degree 14 (CBS 2007). In my sample, it is 138/218*100% = 63%. That is a very substantial difference with possible consequences for the conclusions that can be drawn from this research. The study results are more reliable for the higher educated parts of the population. On the brighter side: 63% is closer to 50% than 25% is to 50%. This means that my survey data is well fit to detect differences between the higher educated and the rest. 13 See SQL attachment, formula 1 14 In the Netherlands: HBO and WO 39

40 Religion This an overview of the number of different religious groups in the sample 15. Table 4: religion religion count % 0 not religious believer Roman Catholic Reformed Presbyterian Protestant Islam other Alas, no Moslems are in this sample, while in The Netherlands, they constitute 5% of the population (CBS, 2009). In the sample, a larger part is not religious or not a member of a religious organisation (56% + 16% = 72%) than in the Dutch population (42%). Figure 9: Religious groups (CBS, 2009) Evaluation of the viral marketing and randomness The viral marketing has worked less well then I had hoped for. Many respondents are close to me, studying in Leiden or close to my family or former work circles. The sample is still very 15 See SQL attachment, formula 2 40

41 diverse, but the selection method could not prevent this lack of randomisation. Results from this study are thus primarily applicable to the groups that were preferred by the selection method Summary of the Demographic Analysis The sample is skewed towards high educated, atheist, young males, often studying in Leiden. Still, all age groups are well represented, as are women Statistical Analysis The most important dependent variable (DV) is principle: the question whether, in principle, one opposes (-1), is neutral (0) towards, or favours (1) the EPR. The theory identifies a number of independent variables (IVs) that could affect principle and those variables are analysed. Also, I try to find empirically eventual correlations by using a correlation matrix. The variables of influence are combined in a general linear model to compare the strength of influence of the various variables Political party dimensions A correlation between party affiliation and health privacy hits the core of this study. The results of the survey are given in the table below. Table 5: political parties, absolute numbers party votes principle objected longterm PVV 6 17% 83% 33% abstain 3 33% 33% 33% D % 16% 70% PvdA 29 62% 26% 71% SP 20 63% 18% 80% n/a (not active) 19 63% 11% 76% PvdD 7 64% 14% 79% VVD 22 66% 11% 84% SGP 7 71% 14% 64% GL 26 73% 13% 85% CDA 28 73% 9% 84% CU 12 75% 17% 71% not 13 96% 0% 100% other 1 100% 0% 100% Using the method of Kieskompas (2010), one can create an ordinal ordering of the Dutch political parties on two scales: [left right] and [progressive conservative]. I have recoded the original party variable to two other variables 16 : party_leftright and party_conservative. 16 See 15.3 Att: party encoding in the data file 41

42 To correlate the party dimensions to all three health privacy variables, I use a Spearman correlation method because all variables are not metric, bur ordinal measures. All correlations found are very weak (coefficients 0,01 to 0,07) and not statistically significant (significance 2- tailed 0,386 for principle vs party_leftright and 0,873 for principle vs party_conservative). Only the health variables have a strong correlation (coefficients 0,59 to 0,76) between them, which is reassuring because all three of them should measure some aspect of the health privacy attitude. Another, more simple, method to dimension the Dutch political space is by dividing the parties on the Christian secular axis. The Christian parties are CDA, CU and SGP, and the secular parties are the others. A t-test (independent samples) comparing these groups gives these results: Levene s significance is 0,159 > 0,05 so equal variances assumed. The significance (2-tailed) p = 0,091 > 0,05 ( t(180) = -1,699) so there is no statistical significant difference. For principle, Christian or secular party preference does not matter Political attitude axes of the BSAS As explained in the theory, I test against three axes of the BSAS political attitudes scales. The results of the Spearman correlations with principle are given in Table 6. Table 6: political attitude correlation quantity variable ρ (rho) significance n left-right levelling -0,068 0, left-right management -0,120 0, authoritarian criminality 0,094 0, authoritarian law 0,203 0, postmaterialist euthanasia 0,128 0, Of these results, the only significant correlation is between principle and law (survey question: The law should always be obeyed). The significance (2-tailed) is 0,003 < 0,01 and the correlation coefficient r = 0,

43 Age and principle I have calculated the average age of my respondents using SQL queries on the database 17 : Table 7: age principle age It may look like those who, in principle, oppose the EPR (principle = -1) are on average somewhat older than those that favour it (principle = 1). To test the possible linkage, I use a Spearman correlation because principle is an ordinal variable. The Spearman correlation disproves the linkage, because the significance (2-tailed) is 0,382 > 0,05 (with correlation coefficient -0,59). So, in this sample, there is no correlation between principle and age Gender and principle Does gender make a difference for the answer to the first question of the survey? It looks like men are, in principle, more often opposed to the EPR. Table 8: gender in principle female male oppose 19% 27% neutral 18% 22% favour 63% 51% A Student s t-test for independent samples is performed to measure the difference between the sexes. Levene's significance 0,146 > 0,05 so we may assume equal variances. The t-test significance (2-tailed) gives 0,077 which means that the 0,202 ± 0,114 mean difference between males (mean = 0,44; sd = 0,797) and females (mean = 0,23; sd = 0,856) is very likely a structural one ( t(216) = 1,779; p = 0,077). The means of both genders are within each other s standard deviations, so the difference between the genders is not very spectacular, to say the least. 17 See SQL attachment, formula 3 43

44 Figure 10: gender difference Confounders such as the younger average age of women in this sample and the relation between age and principle could explain at least a part of this outcome. This study is mainly set up to research the relation between political attitudes and health privacy. A possible confounder could be gender differences in political attitudes, resulting in gender differences for the health privacy variable principle. Therefore, I explored the correlation of gender and party preference Gender and party Which party is more popular with women and which party more popular with men? Because more men than women entered the sample, I divided the number of votes for a party by some gender by the total number of that gender. Dividing the female part by the male part, and then checking if the result is greater or lesser than 1, determines whether a party is more popular with females (0, Boolean false) or with males (1, Boolean true). This database query 18 results in Table See SQL attachment, formula 4 44

45 Table 9: gender and party party 0 = feminine, 1 = masculine sum(total) (other) - 36 CDA 1 28 CU 1 12 D GL 0 26 PvdA 0 29 PvdD 0 7 PVV 1 6 SGP 1 7 SP 0 20 VVD 0 22 Please note that these values are valid for this sample only. In the general population, the VVD has the biggest gender gap with proportionally more males voting VVD than females, although the PVV might be a contender (Outshoorn, 2010, personal communication). For this sample, a political attitudes Kieskompas (2010) map can be drawn: Figure 11: gender and political positioning This visualises that in this sample, women tend to favour progressive parties, while men have more sympathy for conservative parties. 45

46 Religion and principle Does religion have a correlation with health privacy? To compare the non-religious group against the religious group, an independent samples t-test is performed. I have included the believer, but not member of a religious organisation in the religious group. Result: Levene s test variance significance = 0,250 > 0,05 so equal variances assumed. The 2-tailed significance is 0,203 > 0,05, so there is no significant difference between non-religious (mean = 0,44; sd = 0,802) and religious (mean = 0,28; sd = 0,847) subgroup from this sample in respect to health privacy ( t(216) = 1,276; p = 0,203). In other words, non-religious and religious people in this sample do not show statistically relevant differences in respect to the principle variable ID card and principle To test the relation between health privacy (principle) and other privacy, in this case one s attitude for ID cards (idcard), I have performed a Spearman s correlation test. Rho = 0,277 (p = 0,000) so there is a significant correlation, albeit not very strong. (The correlation coefficient for a positive correlation is between 0 (no correlation) and 1 (full correlation).) Education and principle Table 10: education in principle higher educated other educated oppose 27% 19% neutral 17% 25% favour 56% 56% From the table above follows those higher educated persons, on average, are more likely to disagree with EPRs. To prove this, a Student s independent samples t-test for is performed for principle against education level. Levene s significance is 0,063 > 0,05 so equal variances are assumed. The t-test significance (2-tailed) gives 0,470 so for this sample, the difference between higher educated (mean = 0,29; sd = 0,865) and the others (mean = 0,38; sd = 0,786) is not statistically significant ( t(216) = -0,724; p = 0,470). This sample proves no statistically relevant difference between higher educated persons and others in respect to the EPR. I have also tested against two other variables for health privacy: objection and longterm. Both variables cannot assume equal variances, and longterm also gives no statistically significant result. The objection variable does not really test a health privacy attitude but measures if a person has undertaken the action to fill in objection forms. In this case, education 46

47 does matter: with a t-test significance (2-tailed) of 0,000 it gives a mean difference of 0,337 ± 0,076 ( t(197) = 4,4; p = 0,000). Higher educated persons, in this sample, are more eager 19 to fill in an objection form against the EPR Summary of the Correlations Found This is a short overview of the correlations found in the sample until up here. Males are on average older than females, which could affect the other correlations that have an age component. + women - principle + women + progressive + education + objection + law + principle Multivariate linear regression model To further investigate these correlations and their influence on health privacy (principle), I perform a multivariate linear regression. The suspect variables are law and gender, and I have added education, age, party_leftright and party_conservative to the mix because these variables were discussed earlier. Using the enter method, a significant model emerged: F (6) = 2,448; p = 0,027 < 0,05. This model explains only 4,6% of the variance (adjusted R 2 = 0,046). The only significant factor was law. Although gender appears to have some influence, it is not significant. See Table 11. Table 11: multiple regression (linear) for DV principle Variable B SE B ß Sig. (p) (Constant) 0,661 0,345-0,057 law 0,183 0,064 0,216 0,005 gender -0,238 0,127-0,142 0,063 education -0,021 0,049-0,032 0,672 age 0,000 0,004 0,008 0,916 party_leftright -0,025 0,029-0,088 0,385 party_conservative 0,018 0,035 0,054 0, Discussion of the Statistical Analysis Dutch political parties have not politicised the EPR topic. Public discussions and disagreements were minimal. This might explain why no relation between the individual 19 On a scale from -1 to 1, that is 0,337 / 2 * 100% 16%. 47

48 preference for the EPR (principle) and his/her political party preference could be found, not even using multiple groupings (left vs. right, conservative vs. progressive, Christian vs. secular). Or it can be the other way round: the EPR preference divides all parties. Whatever the case, in the Netherlands, party preferences are not measurably connected to health privacy. In my theory, I expected that on the libertarian authoritarian BSAS scale, libertarians would have less preference for a government-managed EPR than authoritarians. This has been proven by the correlation between principle and law. Respect for the law, and authority and regulation in general, assumes that rules and regulations will be valued and followed. Much fear and some real risks surrounding the EPR are based on the breach of rules and security. More trust in the law decreases the fear for such risks. This correlation between trust in the EPR and trust in authoritarian government is therefore understandable. On two other BSAS scales, the left right scale and the materialism postmaterialism scale, I did not expect to find much difference in respect to health privacy, and indeed I could not find evidence of such differences. The lack of correlation goes also for age. Older people have as much trust in new developments as young people do, at least when it concerns digitisation of medical records. This fits with my theoretical expectation. Women are more likely to support the EPD than men. My theoretical expectation was to find no difference. The evidence for this finding is indeed weak, as the multivariate regression model could not prove (nor disprove) a significant relation. Still, it is hard to explain this gender difference in the sample, which is very real. I consulted Outshoorn (2010), professor in women s studies. She did not know of similar findings in other research. She suggested that women go to the doctor more often, for multiple reasons such as female reproduction and to accompany children and elderly. This could increase their familiarity and trust with medical registries. But in my sample, women were on average young, so their experience with doctors is likely more limited. This is negated by another property of the women in my sample: they were also highly educated. Such people, known as the worried well, are suspected to make more doctor visits. Do women have more trust in EPRs because they tend to vote more progressively? That is hard to say, and progressive parties are known to have more respect for privacy issues (see 4 on page 13). The fact that women from this sample tend to be more progressive than men is not unlike the same phenomenon in the population at large, as explained by Vollebergh, Iedema 48

49 and Meeus (1999). The results from this sample reflect earlier research in this respect, which adds confidence in the sample quality. Although a correlation between religion and health privacy was expected, this research could not prove one. Assumptions about the Christian trust in the government might be mistaken. Maybe this sample contained not enough religious individuals to provide enough power to reveal the correlations. Another explanation for the lack of correlation could be that religion is a big umbrella term, under which many worldviews and attitudes manifest themselves. The somewhat weak correlation between health privacy and other privacy was expected in my theory. Both concepts are related, but at the same time they differ enough to have a strong weakening effect on the correlation. Higher educated persons are more likely to fill in an objection form against the EPR, which is understandable because better educated individuals are more experienced with forms. This finding is also consistent with earlier work from the Rathenau Instituut (2009). But in relation to their in principle preference for the EPR, I could find no difference with the lower educated. This contrasts with the suggestion from the Rathenau Instituut (2009) that higher educated persons are less likely to support the EPR. It also diverges from my theoretical assumption that no differences would be found. My research was quantitative, while the Rathenau research was qualitative they already mentioned that their results could not be extrapolated to the Dutch population at large. Therefore, I feel that my finding that the level of education has little to do with one s preference for the EPR is holding firm. 49

50 12. Comments Report Respondents were able to enter a free text comment at the end of the survey. The comments can be read in the comments attachment (in Dutch, and edited to remove personal data). Some respondents did not fill in the comments, but sent me an with comments. Here, some of the comments and s of interest are presented in English and analysed. When multiple respondents have sent a comment with the same meaning, I have condensed it to one comment About the survey format The question on daily occupation (variable: occupation) must be changed. Currently I am both working and caring (caring profession). Another respondent commented that he was both working and studying. I think my question format could indeed be improved to catch cases such as this one. For the question on religion, more choices should be possible for Islam, such as Sunnites and Shiites. I did not do so, because 1) I wanted to follow the NKO question format, and 2) because Moslems only constitute 5% of the Dutch population. I would prefer to not only say which party I voted back in 2006, but also which party I favour right now. That can be quite a difference. That is true. I borrowed this question format from the NKO, and this historical way of collecting data enables comparisons with earlier surveys and with the party platforms of 2006, but it makes comparisons with the current Kieskompas less solid. It is partly compensated for by the other questions that deal with the political attitude scales. Still, I would recommend to ask respondents for their current party preference when I would have to repeat this survey. I am missing some important questions in the survey, like the handling of confidential information and the rights of consumers. Some questions lack an I won t say option, like if you do not want to say which political party you voted for, or which religion you belong to. Although I borrowed the question formats from well-known and well-designed surveys, I think that this comment is correct. Such a won t say option should be offered. There were some (technical ) errors, so I could see the address of another respondent. Some commenters also complained about the technical error that led to an bombardment of their mailboxes. Fully my fault, and I apologise. 50

51 The discussion on the EPR has two aspects. One is the value for care, and I think an EPR can improve healthcare. The other is the big brother is watching you discussion, and because I worry about that, I dislike the EPR. The survey does not provide an opportunity to clearly distinguish the two aspects. Actually, I have one question about who should have control over the EPR (the control variable), but I agree that the first question ( principle ) could offer more choice, like I favour the EPR if no government control or involvement Against the restricted use of health information Some people would favour the use of genetic information, but those must fear that such use will be forbidden in the future. This is quite interesting, and reminds me that one should never make assumptions. I copied this question from the BSAS, and I copied also the implicit assumption that the use of genetic information for things such as insurance is always perceived as a bad thing. This comment proves that this needs not to be true Reasons for being against the EPR Medical records can be out-dated or biased. This can result in not getting the right treatment. The pay as you drive ( kilometerkastjes ), the public transport chip, fingerprints in the passport I worry about my freedom. Too many errors are possible resulting from EPR use. An assertive ( mondige ) patient can stand up for himself. Problem is, many patients are not able (physically or mentally) to stand up for themselves. For me, an important requisite for each digital database with personal data is a very rigid and secure access policy. The government has, in this respect, not a good reputation. This covers other databases such as voting computers and public transport chip cards as well. A lady of 66 years old (name known to author) comments that she works in a sector where the electronic child record will be used. She has experienced that the data is being used by unauthorized parties, such as insurers. Parents are hard pressed to deliver information about their children, under penalty of withdrawing funds they have a right to. She also has had a bad experience with the general practitioner of her old mother who sent a complete medical record (paper-based) to her using postal services. Understandably, this lady is very critical towards electronic records and the government. 51

52 Ideas on the implementation of the EPR I favour the EPR, but only when local (not nation-wide). I favour the EPR, but it should be accessible only for people with a medical professional secrecy. It may never be used by investigation services. Data should not be stored in databases, but be carried with the patient using data cards or memory sticks. In that way, access control is very easy. And implementation is less costly than the (nationwide) EPR. This is true in many cases, but on the other hand, many medical memory sticks may be lost, stolen, or otherwise compromised. Additional biometric security, such as fingerprint authentication, can help, but still decentralised data storage is now always more safe. Other problems are the juridical obligation to medical providers to create a record of their activities (medical records must be created anyway) and the impossibility to use decentralized medical records for secondary use such as epidemiological research. From a nurse: It is hard to delete or change data from the EPR after it is inserted. Data inserted can often not be corrected by the person who inserted the data. The current EPR is too limited. Additional, often very important, paper based data cannot be added. Confidential data, only meant to be known to the patient and the direct medical staff treating the patient, cannot be inserted into the EPR Analysis of the comments A number of problems with the design of the survey were given. I do not believe that the design of the survey was bad. Difference of opinion and insight in the design of surveys is to be expected for any survey. But some comments were worth considering for a possible future health privacy survey. They also revealed a few weaknesses in well-known big surveys such as the BSAS and the NKO. Overall, I do not think that the design weaknesses mentioned in the comments are disturbing enough to raise serious concern about the validity of the results from this survey. The number of comments with concerned remarks about the privacy and lack of data security of the EPR is much higher than the number of comments positive on the EPR. This contrasts with the mean value of the principle variable: most of the respondents favour the EPR. People who like to share their opinion using the free text input are often opponents of the EPR. This could mean that the opponents are more likely to read and comment on developments in health privacy. If that is true, than it does not bode very well on the EPR: those who are more concerned are less enthusiastic. 52

53 In many cases, the critics are not fundamentally against the idea and benefits of digitising the medical records, but are worried about privacy and data security. Those worries are backed up with very real examples of how things can go wrong. Others point to the possibility of wrong data in patient records and the barriers to change such wrong data. 53

54 13. Conclusions Health privacy issues in the Netherlands transcendent political attitudes. One s position in the political space is not a good predictor for one s opinion on the electronic patient record (EPR). A somewhat weak but significant correlation between voting for authoritarian parties and liking the EPR was found. People who like authoritative government are more often a proponent of an EPR regulated by the government. A less significant relation was found between being female and favouring the EPR. This study was explorative in the sense that it was the first study that tried to link health privacy to political attitudes and the size and selectiveness of the sample was of relatively low quality. To get more definitive results, a larger and more random sample must be taken. A future study can focus on fewer variables; especially one s preferences for the EPR and for authoritative government are worth investigating. A future study would also do well to read (About the survey format) to improve the questionnaire. Those who oppose the EPR were more motivated to add free text comments or to send s expressing their concerns. These opponents did include knowledgeable people; I recommend that the government not only should improve its information services, but also should pay attention to possible privacy pitfalls of the current EPR implementation. For leaders of parties that are libertarian and focused on privacy, this study adds even more incentive for such caution. Voters of such parties value their privacy more. In this respect, it is remarkable that people from my sample that voted Green-Left do not take a more critical stance towards the EPR. GL is one of the most pro-privacy parties (together with the SP and PvdD). All political leaders should avoid taking risks with EPRs. The fact that both privacy and health evoke much emotion, and are both important for the well-being of their voters, are enough reason to be careful. Trust in technological advancement should be accompanied with a deep understanding of the technology itself and the possible consequences of a bad design. 54

55 14. References Books and Articles Castle, David, and Ruth DeBusk The Electronic Health Record, Genetic Information, and Patient Privacy. Journal of the American Dietetic Association 108 (8): CBS Statistics Netherlands Religie aan het begin van de 21ste eeuw. Kengetal E-16. Christie, Stuart, and Albert Meltzer The Floodgates of Anarchy. London: sphere. Diagrams by Bill Hicks. Citrin, Jack Comment: The Political Relevance of Trust in Government. The American Political Science Review 68: Cross, Michael GPs leader sets conditions for electronic care records. British Medical Journal 332 (3): Curry, Lynne The human body on trial: a handbook with cases, laws, and documents. Hackett. Dalton, Russell J., and Hans-Dieter Klingemann Oxford handbook of political behavior. Oxford: Oxford University Press chapter 18: Social and Political Trust, pp Everts, Philip Democracy and Military force. Basingstoke: Pagrave Macmillan chapter 4 Theory Formation and Empirical Knowledge - The Concept of Public Opinion, pp Fitzgerald, Thomas Rethinking Public Opinion. The New Atlantis pp Freeman, Richard The health care state in the information age. Public Administration 80: Goldman, Janlori, and Zoe Hudson Virtually Exposed: Privacy And E-Health. Health Affairs 19: Gostin, Lawrence O., James G. Hodge, and Ronald O. Valdiserri Informational Privacy and the Public s Health: The Model State Public Health Privacy Act. American Journal of Public Health 91: Holvast, Jan The Future of Identity in the Information Society. Boston: Springer chapter History of Privacy, pp Hustead, Joanne, and Janlori Goldman Genetics and Privacy. American Journal of Law & Medicine 28: Katzenbauer, Maartje Te vroeg voor landelijk EPD. Medisch Contact 64 (5): Dutch. Kidd, Michael Personal electronic health records: MySpace or HealthSpace? British Medical Journal 336 (5): PMID: Killerby, Paul Trust Me, I m From the Government: The Complex Relationship between Trust in Government and Quality of Governance. Social Policy Journal of New Zealand (7). 55

56 Kmietowicz, Zosia MPs dismayed at confusion about electronic patient records. British Medical Journal 335 (9):581. PMID: Leiden University department of Polical Science, ed Verkiezing van de Tweede Kamer der Staten Generaal 22 november Verkiezingsprogramma s. Amsterdam: Rozenberg Publishers. Dutch. Margulis, Stephen T Privacy as a Social Issue and Behavioral Concept. Journal of Social Issues 59: McGilchrist, Mark, Frank Sullivan, and Dipak Kalra Assuring the confidentiality of shared electronic health records. British Medical Journal 335 (12): PMID: Milgram, Stanley The Small World Problem. Psychology Today 2: On the six degrees of separation principle. Miller, Amalia R., and Catherine Tucker Privacy Protection and Technology Diffusion: The Case of Electronic Medical Records. NET Institute Working Paper No , February Miller, Arthur H Political Issues and Trust in Government: The American Political Science Review 68: NIDI De bevolkingsontwikkeling in een notendop. Demos (9). Norheim, Ole Frithjof Soft paternalism and the ethics of shared electronic patient records. British Medical Journal 333 (7):2 3. PMID: Orwell, George Nineteen Eighty-Four ed. Outshoorn, Joyce Women s Movements and Bodily Integrity: towards a dynamic institutionalist approach. FEMCIT IP WP5 Working Paper No. 1. Paper prepared for the Workshop Gender, Politics and Institutions: Towards a Feminist Institutionalism? Rennes. FIRST DRAFT. Pagliari, Claudia, Don Detmer, and Peter Singleton Potential of electronic personal health records. British Medical Journal 335 (8): PMID: Rynning, Elisabeth Public Trust and Privacy in Shared Electronic Health Records. European Journal of Health Law 14: Editorial. Solove, Daniel J Conceptualizing Privacy. California Law Review 90 (7): Steinbrook, Robert Personally Controlled Online Health Data - The Next Big Thing in Medical Care? The New England Journal of Medicine 358 (4): Tanne, Janice Hopkins Fears over security as Google launches free electronic health records service for patients. British Medical Journal 336 (5):1207. PMID: Vollebergh, W. A. M., J. Iedema, and W. Meeus The Emerging Gender Gap: Cultural and Economic Conservatism in the Netherlands Political Psychology 20 (6):

57 Watson, Nigel, and John Halamka Patients should have to opt out of national electronic care records. British Medical Journal 333 (7): Westin, Alan F Social and Political Dimensions of Privacy. Journal of Social Issues 59: Zureik, Elia, Lynda Harling Stalker, Emily Smith, David Lyon, and Yolande Chan Privacy, Surveillance and the Globalization of Personal Information: International Comparisons. Kingston: McGill-Queen s University Press. Forthcoming Materials CBS Statistics Netherlands Retrieved from the internet on de Hond, Maurice Invoering van elektronisch kinddossier. Retrieved from the internet on Dutch. de Hond, Maurice Veel Nederlanders negatief over elektronisch patiënten dossier. Retrieved from the internet on Dutch. DeVrijePsych EPD. Retrieved from the internet on Dutch. Icke, Vincent Jan Veltkamp Symposium. Panacee and MFLS. Speaker [Conference]. Academiegebouw Leiden Inglehart, Ronald. N.d. The Inglehart Values Map. Retrieved from the internet on Kieskompas Electoral Compass. The academic director is André Krouwel. Ministry of Health Nota informatievoorziening in de zorg. Dutch. Ministry of Health ICT in de zorg - Van Elektronisch Medicatie Dossier naar Elektronisch Patiënten Dossier - Plan van aanpak, 1 maart Dutch. NKO Stichting Nationaal Kiezersonderzoek Dutch Parliamentary Electoral Studies website.. Outshoorn, Joyce to Evert Mouw. Personal communication, Pollara, and Earnscliffe Public opinion research into genetic privacy issues. BioPortal. Privacy Barometer De actuele stand van de politiek over privacy. Dutch. Privacy International The 2007 International Privacy Ranking. Watchdog organisation. Public Opinion Strategies Attitudes of Americans Regarding Personal Health Records and Nationwide Electronic Health Information Exchange. The Markle Foundation. 57

58 Rathenau Instituut Presentatie focusgroepen EPD. Dutch. Expertmeeting EPD - Eerste Kamer op The Surveillance Project Global Privacy of Data - International Survey, Summary report. Ipsos Reid. TNS NIPO Meerderheid Nederlanders heeft vertrouwen in elektronisch patiëntendossier. Retrieved from the internet on Dutch. UMR Research National E-Health Transition Authority Quantitative Survey Report. egovernment Resource Centre. Westin, Alan Americans Overwhelmingly Believe Electronic Personal Health Records Could Improve Their Health. The Markle Foundation. Zureik, Elia, Lynda Harling Stalker, and Emily Smith Background Paper for the Globalization of Personal Data Project International Survey on Privacy and Surveillance. The Globalization of Personal Data Project, Queen s University News Media ANP. 2008a. Klink krijgt bezwaren tegen patiëntendossier. nu.nl. Dutch. Published on ANP. 2008b. Meer bezwaren tegen patiëntendossier. nu.nl. Dutch. Published on BBC news UK s families put on fraud alert. bbc.co.uk. Published on CNN, Jerrold Kessell Iceland sells its medical records, pitting privacy against greater good. CNN. Published on Computer idee Elektronisch patiëntendossier. Computer idee (1): Dutch. de Winter, Brenno EPD-pas gekraakt om ongelijk Klink te bewijzen. Webwereld. Dutch. Published on Hubers, Fred Elektronisch patientendossier - Hoe veilig zijn je medische gegevens? Computer easy (1):12. Dutch. Krebs, Brian, and Anita Kumar Hackers Want Millions For Data on Prescriptions. The Washington Post. Published on New York Times, John Markoff F. B. I. Looks Into a Threat to Reveal Patient Data. The New York Times. Published on New York Times, Steve Lohr Google Health Begins Its Preseason at Cleveland Clinic. The New York Times. Published on

59 Novum Gekraaktechip EPD wordt vervangen. nu.nl. Dutch. Published on Nu.nl, Wieland van Dijk Hackers stelen miljoenen patiëntendossiers. nu.nl. Dutch. Published on Timmer, John Microsoft s HealthVault scores a big win with Mayo Clinic. Ars Technica. Published on

60 15. Attachments Att: Questionnaire The English translation of the Dutch questions is given at the left side; on the right side are the Dutch wordings. The origin of each question is given below the specific question: a note is placed after the question, in [ square brackets ]. Each question leads to a variable in the database. The internal database name is given in {accolades}. The first two questions are taken from Medisch Contact and are marked with MC. Questions that I have introduced myself are marked with EM. (English) Introduction The Netherlands have decided to introduce an electronic health record (EPR). Such an EPR contains all your medical data. Medical doctors can help you faster because they can access your data with a computer. But some people worry about their privacy. This survey tries to find out if a link exists between the need for (medical) privacy and political attitudes. Do opponents of the EPR vote right? Or do they vote left? Of maybe do they vote for the centre? This survey has 20 questions. I hope you will participate! On the EPR 1. In principle, I favour the introduction of the EPR oppose the introduction of the EPR stand neutral against the introduction of the EPR [ MC (replaced as MP by I ) ] {principle} 2. The objection form against the EPR I have filled in I have not filled in I yet consider to fill in (Dutch) Inleiding Er is besloten om een elektronisch patiëntendossier (EPD) in te voeren in Nederland. Zo'n EPD bevat dan al je medische gegevens. Doktoren kunnen je daarmee sneller helpen omdat ze met de computer bij jouw gegevens kunnen. Sommige mensen maken zich echter zorgen om hun privacy. Dit onderzoek probeert te achterhalen of er een link is tussen die behoefte aan (medische) privacy en politieke voorkeur. Stemmen tegenstanders van het EPD links? Of stemmen ze juist rechts? Of stemmen ze op het midden? Hieronder staan 20 vragen. Ik hoop dat je mee wilt doen! Over het EPD 1. Ik ben in principe voor invoering van het EPD tegen invoering van het EPD neutraal over invoering van het EPD [ MC (replaced as MP by I ) ] {principle} 2. Het bezwaarformulier EPD heb ik wel ingevuld heb ik niet ingevuld overweeg ik alsnog in te vullen [ MC ] {objection} 60

61 [ MC ] {objection} 3. Electronic patient records are on long term useful harmful make no difference [ EM ] {longterm} 4. Electronic patient records can best be managed by medics and hospitals private organisations, such as Microsoft Vault or Google Health, where you choose yourself the government with national rules and organisation don t know [ EM ] {control} On other privacy 5. Every adult in Britain should have to carry an identity card agree strongly agree neither agree nor disagree disagree disagree strongly [ BSAS and d: Compulsory identity cards ] {idcard} 6. How likely or unlikely do you think it is within the next 25 years that genetic information will be used to judge a persons suitability for getting a) health or life insurance? very likely quite likely not very likely not at all likely can t choose {gen-insurance} b) a job they have applied for? very likely quite likely not very likely not at all likely can t choose {gen-job} c) credit at the bank? 3. Elektronische patiëntendossiers zijn op lange termijn nuttig schadelijk maken geen verschil [ EM ] {longterm} 4. Elektronische patiëntendossiers kunnen het beste beheerd worden door doktoren en ziekenhuizen private organisaties, zoals Microsoft Vault of Google Health, waar je zelf voor kan kiezen de overheid met landelijke regels en organisatie geen mening [ EM ] {control} Over overige privacy 5. Iedere volwassene in Nederland zou een identiteitsbewijs bij zich moeten hebben zeer eens eens neutraal oneens zeer oneens [ BSAS and d: Compulsory identity cards ] {idcard} 6. Hoe waarschijnlijk denk je dat in de komende 25 jaar genetische informatie gebruikt wordt om een iemands geschiktheid te bepalen voor het krijgen van a) ziektekosten- of levensverzekering? heel waarschijnlijk waarschijnlijk weinig waarschijnlijk zeer onwaarschijnlijk geen mening {gen-insurance} b) een baan? heel waarschijnlijk waarschijnlijk weinig waarschijnlijk zeer onwaarschijnlijk geen mening {gen-job} c) een lening van de bank? heel waarschijnlijk 61

62 very likely quite likely not very likely not at all likely can t choose {gen-loan} [ BSAS a,b,c: Likelihood of genetic information being used ] On political attitudes 7. Some people feel that the differences in income in our country should increase. Others think that it should decrease. Others have a moderate opinion. Where would you place yourself on a line from 1 to 7, with 1 meaning more income difference and 7 less difference? a number between 1 and 7 can t choose [NKO st wave 23 / BSAS a: Left-right scale ] {levelling} 8. Management will always try to get the better of employees if it gets the chance agree strongly agree neither agree nor disagree disagree disagree strongly [ BSAS e: Left-right scale ] {management} 9. Opinions differ on how the government fights crime. Where would you place yourself on a line from 1 to 7, with 1 meaning the government is too harsh and 7 meaning the government should be more punitive? a number between 1 and 7 can t choose [ NKO st wave 25 / BSAS a: Libertarian-authoritarian scale ] {criminality} 10. The law should always be obeyed, even if a particular law is wrong waarschijnlijk weinig waarschijnlijk zeer onwaarschijnlijk geen mening {gen-loan} [ BSAS a,b,c: Likelihood of genetic information being used ] Over politieke attitudes 7. Sommige mensen vinden dat de verschillen in inkomens in ons land groter moeten worden. Anderen dat ze kleiner moeten worden. Natuurlijk zijn er ook mensen met een mening die daar tussenin ligt. Waar zou u uzelf plaatsen op een lijn van 1 tot en met 7, waarbij de 1 betekent dat inkomensverschillen groter moeten worden en de 7 dat ze kleiner moeten worden? een getal tussen 1 en 7 geen mening [ NKO st wave 23 / BSAS a: Left-right scale ] {levelling} 8. Het management zal altijd proberen van werknemers te profiteren als ze de kans krijgen zeer eens eens neutraal oneens zeer oneens [ BSAS e: Left-right scale ] {management} 9. Over de manier waarop de overheid optreedt tegen criminaliteit wordt verschillend gedacht. Waar zou u uzelf plaatsen op een lijn van 1 tot en met 7, waarbij de 1 betekent dat de overheid te hard optreedt en de 7 dat de overheid harder moet optreden? een getal tussen 1 en 7 geen mening [ NKO st wave 25 / BSAS a: Libertarian-authoritarian scale ] {criminality} 10. De wet moet altijd nageleefd worden, zelfs als een bepaalde wet fout is zeer eens 62

63 agree strongly agree neither agree nor disagree disagree disagree strongly [ BSAS e: Libertarianauthoritarian scale ] {law} 11. Looking at the list below, please tick the box next to the one thing you think should be Britains highest priority, the most important thing it should do. Britain should maintain order in the nation give people more say in government decisions fight rising prices protect freedom of speech can t choose [ BSAS a: Post-materialist scale] {priority} 12. Some people say that euthanasia should always be forbidden. Others say that euthanasia should be possible on patient request. Of course, some have an opinion in between. Where would you place yourself on a line from 1 to 7, with 1 meaning the euthanasia should always be forbidden and 7 meaning euthanasia should be allowed? a number between 1 and 7 can t choose [ NKO st wave 22: Postmaterialist scale? ] {euthanasia} 13. If you could vote right now, which party would you choose? Not voted / am entitled to vote Not voted / not entitled to vote Abstention CDA CU D66 GL PvdA PvdD PVV eens neutraal oneens zeer oneens [ BSAS e: Libertarianauthoritarian scale ] {law} 11. Kijk naar de onderstaande lijst. Kies wat volgens jou de belangrijkste prioriteit zou moeten hebben in Nederland, dat dus wat als belangrijkste gedaan moet worden. Nederland moet de orde handhaven mensen een stem geven in overheidsbesluiten stijgende prijzen tegengaan de vrijheid van meningsuiting beschermen geen mening [ BSAS a: Post-materialist scale] {priority} 12. Sommige mensen vinden dat euthanasie altijd verboden moet zijn. Anderen vinden dat euthanasie mogelijk moet zijn als de patiënt daarom vraagt. Natuurlijk zijn er ook mensen met een mening die daar tussenin ligt. Waar zou u uzelf plaatsen op een lijn van 1 tot en met 7, waarbij de 1 betekent dat euthanasie moet worden verboden en 7 dat euthanasie mogelijk moet zijn? een getal tussen 1 en 7 geen mening [ NKO st wave 22: Postmaterialist scale? ] {euthanasia} 13. De vorige verkiezingen voor de Tweede Kamer zijn in 2006 gehouden. Wat heeft u toen gestemd? Niet gestemd / wel stemgerechtigd Niet gestemd / niet stemgerechtigd Blanco CDA CU D66 GL PvdA PvdD PVV 63

64 SGP SP VVD other party [ EM political parties currently in the Parliament, ordered alphabetical / NKO st wave ] {party} Personal background 14. Year of birth a number between 1900 and 2009 [ NKO st wave 1 ] {birthyear} 15. Gender male female [ NKO st wave 2 ] {gender} 16. Philosophy of life / religion not religious believer, but not a member of a church or religious association Roman Catholic Reformed Presbyterian / Calvinist Protestant Islam other church or religious association [ NKO st wave 2 ] {religion} 17. Occupation employed student caretaker retired unemployed and searching other [ EM ] {occupation} 18. For which education you followed for at least 2 years have you achieved your latest certificate or diploma? Primary school Lower vocational education Lower secondary education Higher secondary education Intermediate vocational education Bachelor, higher vocational education SGP SP VVD andere partij [ EM political parties currently in the Parliament, ordered alphabetical / NKO st wave ] {party} Persoonlijke achtergrond 14. Geboortejaar een getal tussen 1900 en 2009 [ NKO st wave 1 ] {birthyear} 15. Geslacht man vrouw [ NKO st wave 2 ] {gender} 16. Levensovertuiging niet gelovig gelovig, maar geen kerkelijke gezindte of levensbeschouwelijke groepering Rooms Katholiek Nederlands Hervormd Gereformeerde kerken Protestantse Kerk Nederland Islam andere kerkelijke gezindte of levensbeschouwelijke groepering [ NKO st wave ] {religion} 17. Dagelijkse bezigheid werkend studerend verzorgend vut, wao of pensioen werkzoekend anders [ EM ] {occupation} 18. Voor welke opleiding waarmee u 2 jaar of langer bezig bent geweest heeft u als laatste een akte, getuigschrift of diploma behaald? Lagere school (basisschool) Lager Beroepsonderwijs, VMBO basisberoepsgerichte of kaderberoepsgerichte leerweg 64

65 Master, doctoral, PhD, university [ NKO st wave 46 with ] {education} Comments 19. Space for comments. [ EM ] {comments} Mavo, VMBO theoretische of gemengde leerweg, ULO, MULO Havo, VWO, Gymnasium, HBS, MMS Middelbaar beroepsonderwijs (MBO, BOL, BBL) Bachelor, Kandidaats, Hoger Beroepsonderwijs Master, Doctoraal, (semi-) Wetenschappelijk Onderwijs, Universiteit [ NKO st wave 46 met ] {education} Opmerkingen 19. Ruimte voor opmerkingen. [ EM ] {comments} A summary of quantities, variables, types, and ranges is provided in the attachment hereafter. 65

66 15.2. Att: datafile addendum 66

67 15.3. Att: party encoding in the data file Table 12: Encoding of the party variables in the SPSS database party party_leftright party_conservative party_christian not n/a other cda cu d gl pvda pvdd pvv sgp sp vvd Values used to create the ordinals taken from Kieskompas / Krouvoet (2010) party random nominal ordering party_left-right ordinal ordering on the left-right scale (left = low, right = high) party_conservative ordinal ordering on the progressive-conservative scale (progressive = low, conservative = high) party_christian boolean ordering of christian (1) and secular (0) political parties (-1 = missing) 67

68 15.4. Att: SQL queries Some SQL queries used in the statistical analysis. Using mysql syntax. 1. select gender, avg(2009-birthyear) from respondents group by gender; 2. select religion, count(*), count(*)/218*100, avg(principle) from respondents group by religion; 3. select principle, avg(2009-birthyear) from respondents group by principle; 4. select party, (sum(females)/94)/(sum(males)/124) < 1, sum(total) from ( ( select party, count(*) as females, '' as males, '' as total from respondents where gender='f' group by party) union ( select party, '', count(*), '' from respondents where gender='m' group by party) union ( select party, '', '', count(*) from respondents group by party) ) as A group by party; 68

69 15.5. Att: Dutch elections of 2006 Table 13: Dutch political parties in parliament after the elections of party seats full name (Dutch) full name (translated) CDA 41 Christen-Democratisch Appèl Christian Democratic Appeal PvdA 33 Partij van de Arbeid Labour Party SP 25 Socialistische Partij Socialist Party VVD 22 Volkspartij voor Vrijheid en Democratie People's Party for Freedom and Democracy PVV 9 Partij voor de Vrijheid Party for Freedom GL 7 GroenLinks GreenLeft CU 6 ChristenUnie ChristianUnion D66 3 Democraten 66 Democrats 66 PvdD 2 Partij voor de Dieren Party for the Animals SGP 2 Staatkundig Gereformeerde Partij Reformed Political Party 69

70 15.6. Att: EPR mentioning per Platform Only a few parties who made it to the Parliament made a reference to Electronic Patient Records in their election platforms: VVD NL: Als iedere patiënt beschikt over eigen gegevens in een elektronisch patiëntendossier wordt de afhankelijkheid van één arts of instelling kleiner. EN: If each patient has an electronic patient record at his disposal, then the dependency on one single doctor of institute will decrease. D66 NL: D66 wil daarom een snelle invoering van het burgerservicenummer en het elektronisch patiënten-dossier, met voldoende waarborgen voor de privacy. EN: D66 want for that reason a fast implementation of the citizen ID number and the electronic patient record, with sufficient privacy guarantees. SGP NL: In de ziekenhuizen moet er meer aandacht komen voor digitalisering van de patiëntendossiers. EN: In hospitals, more attention must be given for the digitizing of patient records. CDA The CDA did not mention health records, but did mention a related electronic record for young persons who often cause problems. NL: Voor probleemjongeren wordt een elektronisch volgdossier aangelegd. Hulpverleners, politie en Openbaar Ministerie krijgen de verplichting om informatie aan te leveren, hebben hierin inzage en voeren overleg over de aanpak. Deze gegevens blijven beschikbaar nadat de jongere meerderjarig is geworden. EN: An electronic record will be kept for problem youths. Social workers, the police and the Public Prosecutor will be obliged to add information, and have access to it and can discuss the approach to the problems. These data will be kept accessible after the young person has become adult. 70

71 15.7. Att: all comments All comments are anonymised. The respondents are from the Netherlands, so all comments are in Dutch Comments received from the survey free text input electronische patientendossiers: kon helaas niet kiezen voor de optie: op dit moment tegen Vraag 1 vind ik niet zo handig. Ik ben vóór EPD maar uitsluitend regionaal of lokaal. Voor gebruik door anderen moet je toestemming geven, of iemand anders machtigen toestemming te geven, voor het geval dat je bijv. ergens een ongeluk krijgt. Een dergelijke optie vind ik niet terug in vraag 1, noch in vraag 2. Ik wil je wel helpen een paar anderen te vinden je enquete ook in te vullen. Groet van ****. ik ben uitgesproken tegen het EPD. Medische dossiers kunnen verouders zijn of 'gekleurd' door de behandeld arts destijds. Hiermee loopt men een risico niet de juiste behandeling te krijgen. Ik spreek uit eigen ervaring: de medische informatie in het dossier van mijn eerste huisarts klopt niet. Deze informatie is echter wel meegegaan toen ik van huisarts overstapte. de vragen over privacy: ik denk dat het waarschijnlijk is dat sommigen graag genetische informatie zoudne willen gebruiken, maar dat dat op ene gegeven moment verboden wordt. Het vakje dagelijkse bezigheid moet je anders doen. Nu ben ik werkend maar ik ben ook zorgend. Ik ben voor het EPD mits de privacy beschermd wordt en alleen toegankelijk voor mensen met een medisch beroepsgeheim. Dossier mag niet gebruikt worden door opsporinginstanties vind ik. mijn stemgedrag uit 2006 was niet zo zeer omdat ik een aanhanger ben voor die partij, maar ik vond alle andere partijen niet ok. te slap te wispelturig of te strikt. EPD: beter zou het zijn dat iedereen alle gegevens opgeslagen heeft op een eigen pasje (ziektekostenpas o.i.d.) en dit altijd bij zich draagt. Dat maakt het EPD overbodig, terwijl bij b.v. een ongeval gegevens direkt beschikbaar zijn. Uitvoering is ook minder kostbaar dan EPD. Deze enquête geeft volgens mij niet echt een goed beeld; ik ben bijvoorbeeld tegen het EPD, maar vooral omdat de beveiliging er niet goed genoeg van is en het voor zover ik weet niet mogelijk is om er niet in te staan. Ik kan wel zien wat de voordelen ervan zouden kunnen zijn (mensen vergeten immers makkelijk welke medicijnen ze allemaal gebruiken enzo), maar vind de huidige manier van invoeren niet correct. Ook zou de Islam moeten worden opgedeeld in de verschillende hoofdstromingen (soennieten, sjiieten, etc). Op zijn minst iets wat deze verschillen erkend. Ook zou het denk ik helpen als je niet alleen maar moet invullen wat je in 2006 stemde, maar ook wat iemand nu zou stemmen. De vorige verkiezingen zijn tenslotte alweer zo'n drie jaar geleden. Ik heb met plezier meegewerkt aan het onderzoek, maar ik ha dgraag ook de achternaam van Evert gehad, en geweten hoe hij aan mijn adres was gekomen. Dat deze informatie niet 71

72 beschikbaar is bij deze enquete, heeft ertoe bijgedragen dat ik mijn adres niet heb achtergelaten. Commentaar/Anders veld (cq. vrije invoer) zou handig zijn, voor motivatie beschrijvingen Als je officieel tot een kerkelijke groepering behoort hoeft dat niet te betekenen dat je die ook navolgt, of dat je naar de kerk gaat. Ieder heeft zijn eigen opvattingen, uit dat soms door naar een kerk naar keuze te gaan, maar anderen doen nauwelijks iets geloof, of geloven heel sterk, maar gaan niet naar de kerk. Officiële groepering/gezindte zegt dus weinig succes, overigens: de pc liep vast na het invullen en verzenden. De kilometerkastjes, de poortjes van het OV, het EPD, de vingerafdrukken in de paspoorten.. Ik maak me zorgen over mijn vrijheid. Jammer dat je op de vragen over wat in nederland belangrijk is en wat mijn dagelijkse bezigheid is niet meerdere antwoorden kunt geven (vrijheid van meningsuiting kun je alleen binnen een zekere orde garanderen - beide dus belangrijk- en ik werk en studeer bijv. tegelijk) succes Evert Ik mis enkele belngrijke zaken in de enquete. Bijv. de omgnag met vertrouwelijk gegevens, en de rechten van de consument. Daarnaast is bij sommige vragen een optie nodig als men niet wil vertellen op wie men stemt, welk geloof e.d.. Aan het eind van de enquete is het verstandig aan te geven wie de info gebruikt en welke grenzen er zijn. Dus wie wel, en vooral wie niet. Agv de verzendfout komen soms adressen in de adresbalk te staan die niet van de respondent zijn, zo heb ik een adres in mijn eigen adres veranderd. ik werk met het epd doordat ik werk bij een GGZ-instelling. Het is dat ik jouw naam herkende als Explolid, anders had ik niet zomaar meegewerkt. Ik wens je succes met je studie. Explogroet Sommige vragen vond ik vrij lastig om in te vullen. Zo weet ik eigenlijk te weinig over het EPD. Daarnaast vond ik de volgende vraag lastig: Nederland moet...orde handhaven... mensen een stem geven over overheidsbesluiten...stijgende prijzen tegengaan... de vrijheid van meningsuiting beschermen... geen mening. Ik heb hier voor A gekozen, maar vind dat C en D ook moeten gebeuren. Verder vind ik dat de overheid wel andere taken heeft die heel belangrijk zijn, waarvan de belangrijkste misschien wel is om onderwijs te stimuleren. Ten slotte wil ik ook nog aangeven dat ik nu anders zou stemmen dan in Toen heb ik SP gestemd, nu zou ik D66 stemmen. Ik weet niet of dit belangrijke informatie is voor je, maar misschien kan je er iets mee. Verder: Leuk onderzoek en succes ermee!! Groetjes **** Bedenk goed:dossiers zullen altijd misbruikt kunnen worden, ook het EPD.Wat weegt zwaarder? Het evt. nut/voordeel of het evt. misbruik/nadeel? Persoonlijke noot: ik liep rond 1946/47 met schaapherder Willem Mouw over de Vierhouter Heide en woonde rond 1954/58 op de Nunspeterweg 80, vlakbij de Maatweg, waar de boerderij van Mouw was. Overigens: succes bij de verdere studie! 72

73 Nederlands Hervormd zijn die gemeenten die een aparte verklaring ondertekend hebben in PKN verband. Dus niet de Herstel Hervormde kerk, de benaming is in de enquette niet correct. Evert, bedankt voor het vullen van mijn mail box!!!! En misschien volgende keer de adressen in de bcc zetten ipv de cc Hoi, de enquete leest wat onduidelijk vrij veel tekst. miss tip voor volgende x? M.b.t. EPD: een EPD op zich kan heel nuttig zijn, mits goed beveiligd, dus uitsluitend toegankelijk voor zorgverleners die op dit bepaalde moment deze bepaalde patient zorg moet verlenen en daarvoor over zijn medische gegevens moet kunnen beschikken, en mits met toestemming van de patient gebruikt. Deze zin kwam mij een beetje raar over: Anderen vinden dat euthanasie mogelijk moet zijn als de patiënt daarom vraagt. Anders lijkt het me gewoon moord... Over overige privacy - gebruik genetische info - vragen zijn te algemeen - het zal naar mijn verwachting niet gebruikt gaan worden voor alle verzekeringen, banen en leningen - wel indien er een bijzondere situatie is of lijkt te zijn. Iets wat ik overigens niet toejuich. prioriteit overheid: onderwijs en milieu Bij levensovertuiging zou je ook de moglijkheid moeten geven geen antwoord te geven. Veel succes met het verdere onderzoek! En natuurlijk met het afstuderen!! Groetjes, **** Hoi Evert, Ik ben zowel werkend als studerend. Daarnaast werk ik in de zorg en zie ik alleen maar voordelen van een goedwerkend en goed beschermd EPD ten opzichte van hoe de patientendossiers nu zijn opgebouwd. Heeft bij mij dus misschien minder te maken met mijn politieke voorkeur maar meer met mijn ervaringen in de zorg. Succes en ik hoor graag de uitkomsten (de korte versie dan) Groetjes **** ik weet echt niet meer op wie ik in 2006 heb gestemd. Ik ben zeer tegen het EPD, omdat er veel te veel fouten mee gemaakt kunnen worden en het risico op betutteling groot is. Een mondige patient kan zelf vertellen wat relevant is. Alles is niet zo zwart/wit. Vreemd; in de ene vraag wordt men met u aangesproken en bij andere vragen weer met jij. Ik vraag me af of de vragenlijst mijn mening wel goed pakt. De discussie over het EPD speelt zich op twee niveaus af. Zorginhoudelijk is er veel voor te zeggen om één EPD te hebben. Daar staat tegenover het "Big brother is watching you" gevaar. Aangezien dat laatste voor mij een schrikbeeld is wat ik vrij realistisch acht, ben ik niet zo gecharmeerd van een EPD. Voor mij is een belangrijke voorwaarde tot welke digitale database dan ook die persoonlijke gegevens bevat een zeer strenge en frauderesistente toegangscontrole. De overheid heeft in dit opzicht tot nu toe GEEN goede reputatie. Zie ook stemcomputer, OV-kaarten, waarschijnlijk ook rekeningrijden, et cetera. Nadeel van het electronisch dossier is, wat ingevoerd staat niet gemakkelijk eruit te krijgen is c.q. verwijderd kan worden. Uit eigen ervaring opgedaan. Bij een verkeerde patiënt 73

74 gerapporteerd, pas maanden later bleek dat. Zelf kun je het niet meer corrigeren. Het dossier is te beperkt. Aanullende, vaak uiterst belangrijke, papieren informatie kun je er niet in kwijt. Vertrouwelijke informatie -cliënt specifiek voor de directe hulpverlener(s) - kun je er niet in noteren. Euthanasie? Wel het recht op, maar men spreekt te gemakkelijk over recht op Euthanasie. Vergeten wordt wat men vraagt aan degenen die het uit moet voeren. Ofwel men vraagt iemand zijn leven te beëindigen, te doden, zonder eraam te denken dat deze daar zodanig last van gaat krijgen dat deze flinke psychische problemen oploopt of kan lopen. Het enorm hoge aantal zelfmoorden, qua beroep, is niet voor niets zo hoog. Onder verpleegkundigen is bekend dat dit ook meespeelt ingeval van zelfdoding. Recht zeker, maar er wordt zodanig over gesproken dat het een uiterste van egoïsme is zonder aan anderen te denken. gebruik concequent je of u, mixen is falen :) En een cijfer tussen 1 en 7 kiezen is ook lomp. De bedoeling van de opdracht is erg onduidelijk wat zorgt dat mensen afhaken Comments received by Dag Evert, Ik vul de enquête graag voor je in: niet alleen ben ik politiek geïnteresseerd, maar ook beroepsmatig heb ik met het EPD te maken! Echter: de link werkt bij mij niet. Zou je mij deze opnieuw kunnen toesturen? En is het voor mij ook mogelijk om de link naar anderen door te sturen of werkt dat niet en zou ik jou dan hun adres moeten geven? Hartelijke groet en sowieso succes met de laatste studieloodjes, **** Ik denk dat er iets is mis gegaan met je . Bij mij is het spam geworden. Ik kreeg 76 het zelfde mailtje binnen. Ik hoop niet dat dat bij iedereen is gebeurd. Maar ik zal kijken of ik tijd kan vinden. Al is het voor mij niet relevant waarschijnlijk. (Als het goed is zit ik vanaf 26 december 2009 in Canada, permanent.) Net de enquette in gevuld jammer dat de keuze voor het zelf beheren van de gegevens op een b.v memorystick of pasje er niet bij staat. Dan kun je zelf bepalen aan wie je de gegevens overhandigd. Is er niemand op dat idee gekomen? Dag Evert, Wees er nog voorzichtiger in want het toeval wil dat Scouting Nederland nogal allergisch reageert op dergelijke acties omdat Scouting Nederland heel veel waarde hecht aan het juiste gebruik (scouting doeleinden) van de bij haar bekende adressen. We krijgen in Leusden namenlijk behoorlijk wat klachten van mensen die berichten van andere leden als vervelend ervaren. Ik ga de enquete invullen, maak je geen zorgen maar bij een volgende keer, kun je veel beter in je inleiding iets sterker verwijzen naar het zogenaamde spam-artikel. Succes met je onderzoek en wellicht tot ziens op een van de evenementen van Scouting Nederland. -- Met vriendelijke groet, **** Best Evert, Omdat ik de enquete niet op mijn scherm krijg, doe ik het maar zo. Ik geef geen toestemming voor een EPD. Mijn gegevens liggen binnen de kortste keren op straat. Zelf werk ik in een van de takken van sport waarvoor het Electronisch Kind Dossier op stapel staat en uit ervaring weet ik dat gegevens in handen komen van personen die gegevens gebruiken voor oneigenlijke doeleinden. Verzekeraars vragen naar zaken die minderjarige kinderen betreffen en waarvoor een geheimhoudingsplicht moet worden nagekomen. Ouders worden onder druk gezet om die informatie te geven op straffen van het weigeren van rechtmatige vergoedingen. De huisarts van mijn moedere heeft haar hele medische hebben en houden, en detail, in een brief aan mij toegestuurd omdat ik voor haar, hoog bejaard en zeer ernstig ziek, een vakantie 74

75 had geannuleerd. Zijn assistente had de gegevens uit mijn moeders dossier naar mij toegestuurd en was hoogst verbaasd dat ik de dokter liet weten dat zo niet kon en ook niet hoefde. Ze was zich van de prins geen kwaad en de dokter had de brief wel ondertekend, vandaar. Ik weiger iedere medewerking aan de plannen van de overheid ten aanzien van het EPD. Succes, ****, 66 jr. Evert, Mooi onderwerp! In mijn afdeling zit de functionaris gegevensbescherming van LNV. Ik heb het ook naar hem g d. groet, **** Hallo Evert, Heb je enquette blanco ingevuld. Toch maar even m n adres. Heb aan de enquette toegevoegd wat mijn negatieve ervaringen zijn met het electr. dossier. Bij een verkeerde patient rapporter bijv. Ook hoe ik over Euthanie denk. Werk zelf met beademingspatiënten. Te gemakkelijk wordt over recht op gesproken. Je hoort niet; wat leg je meer bij degenen die het moeten doen. Heb hier collega s, maar ook een arts bijna aan onderdoor zien gaan. Vanzelf dat er recht op zijn, maar het "gemak" hoe men er over spreekt is wat mij betreft een uiterste vorm van egoïsme. In een heftige discussie verwoordde een collega ooit het nogal erg duidelijk! Als ze verdomme zo graag willen dat ze aan hun eindje moeten, laat die families het dan zelf doen, maar het niet op onze schouders leggen. Gr. **** 75

76 15.8. Att: The Medisch Contact survey This is the first page of the results (and questions) of the survey held by Medisch Contact (Katzenbauer, 2009). In Dutch. 76