Comparative Architecture of Genetic Privacy


COMPARATIVE ARCHITECTURE OF GENETIC
PRIVACY

Khadija Robin Pierce

INTRODUCTION: PRIVACY: A CONCEPT

"Perhaps the most striking thing about the right to privacy is that nobody
seems to have any very clear idea what it is."' Despite that hurdle, or perhaps
because of it, privacy has received enormous attention in the literature of
numerous disciplines, including law, sociology, anthropology, philosophy, and
medicine. Both the definition and the paradigm of privacy can vary depending
on the disciplinary lens through which it is viewed.2 Furthermore, the general
absence of a comprehensive legal framework regarding the protection of
privacy can make it difficult to develop a common understanding of what
interests are protected by privacy rights. This makes it virtually impossible to
devise a common concept that covers the full range of considerations across
disciplines. Thus, the scope of this Article is necessarily limited. This Article
addresses the issue of biomedical privacy and, more specifically, genetic
privacy. The information that follows focuses on two related aspects of genetic
privacy-informational autonomy and decisional autonomy.

1. Privacy Architecture

Privacy interests and rights are constructed differently across societies.
The interplay between the structure and the substantive right can be significant.
These variations in the construction of rights have been the subject of some
debate regarding the degree of protection afforded those rights by virtue of their
architecture. Frederick Schauer has examined the comparative architecture of
freedom of speech rights with regard to constitutions, noting that certain
constructions render rights "seemingly absolute" while other constructions are
qualified, allowing for overrides; some are universal while others are situational
and temporal; and some are worded broadly while others are more precise and
narrow in their articulation.3 Schauer's focus on the freedom of expression

* Robin Pierce received her J.D. from Boalt Hall School of Law and her Ph.D. from

Harvard University. The author wishes to thank Professors Francesca Bignanii, Frederick
Schauer, Allan Brandt, and Jonathan Beckwith of Harvard University for helpful suggestions
and comments. Dr. Pierce has been an invited speaker in Europe and North America on
biomedical law and ethics regarding innovative technologies.

1. Judith Jarvis Thomson, The Right to Privacy, in PHILISOPHICAL DIMENSIONS OF
PRIVACY: AN ANTHOLOGY 272, 272 (F. D. Schoeman ed., 1988).

2. See generally DAVID M. O'Brian, PRIVACY, LAW, AND PUBLIC POLICY (1979).
3. Frederick Schauer, Freedom of Expression Adjudication in Europe and America: A



IND. INT'L & COMP. L. REv.

bears some parallels to the topic of this Article. The Fourth Amendment allows
a warrant to be issued solely upon the showing of probable cause, and thus it is
similar to Schauer's view of free speech privacy rights in that it has also been
construed in "absolute terms." Yet we understand privacy to be a qualified
right, the contours of which are largely crafted by case law (often tort) and, in
the biomedical context, regulatory law and agency guidance that set margins
based on countervailing interests. Other legal scholars have considered the
architecture of rights pertaining, for example, to enforcement, arguing that the
structure of enforcement can affect the meaning of the substantive right.4 There
are several ways counties have structured privacy laws. One of which is the less
categorical model.

The less categorical model employed in many European countries
recognizes privacy as a fundamental right. However, the less categorical model
also accommodates other important competing interests. Resolution of
competing interests typically calls for a balancing of interests and rights in
which the privacy rights of the individual may be weighed against other
interests that fall within certain designated categories, for example, "the public
interest.",5 The proportionality rule, a multifaceted test, poses three primary
questions: 1) Can the intended action achieve its stated goal?; 2) Is the action
necessary in order to achieve the goal, and are there less burdensome means of
achieving it?; and 3) If a non-economic right is involved, is the burden on the
right an acceptable one? 6 Furthermore, it requires a determination of whether
the burden on the right to individual privacy is proportionate to the public
interest being privileged. However, it has been demonstrated in several types of
analysis7 that this seemingly more flexible model does not necessarily expose
fundamental rights to the subordination of societal interest any more than the
"absolute" model. Rather, as this Article explores, certain safeguards
protecting the underlying principle of the protections seem to remain fairly
intact. What this proportionality model does is make the process and rationale
that may result in the override of a fundamental right transparent, ensuring that
such override occurs only in certain circumstances and in the least restrictive
way possible. Indeed, Schauer notes that critics of the American approach say
that the European architecture is more transparent in its open declaration that
rights are subject to a weighing process as against other interests, whereas the
stringent categorical approach of the U.S. system merely obscures the weighing
process that has already occurred in the drafting of the right.8

Case Study in Comparative Constitutional Architecture, in FACULTY RESEARCH WORKING
PAPERS SERIES RWP05-019 1 (2005).

4. See, e.g., Colin J. Bennett & Charles D. Raab, The Governance of Privacy
(2003).

5. Council Directive 95/46/EC, Protection of Individuals with Regard to the Processing
of Personal Data and on the Free Movement of Such Data, 1995 O.J. (L 281) 31 (30).

6. See, e.g., Francesca Bignami,. European Versus American Liberty: A Comparative
Privacy Analysis of Antiterrorism Data Mining, 48 B.C. L. Rev. 609, 642 (2007).

7. See e.g., id.
8. Schauer, supra note 3, at 3.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

In the biomedical context, innovative uses of genetic information, and the
technologies that use it, implicate a wide range of privacy issues. Because of
the nature of genetic information, protection of privacy interests and rights in
the biomedical context must be viewed anew; traditional protections do not fit
the contours of the new privacy vulnerabilities and interests. Indeed, it has
been suggested that, for a variety of reasons (including significant technological
innovation), a new taxonomy of privacy is needed in order to more accurately
understand modem privacy violations.9 Consequently, lawmakers and judges
have great difficulty articulating the privacy harm in contrast to opposing
interests like free speech, market interests, and national security, which are
more easily articulated.'0 This skewed articulation of rights and interests may
well serve to unseat highly valued privacy rights merely because they are poorly
articulated.

This Article examines the comparative architecture of privacy in the
biomedical context and analyzes specific aspects of genetic privacy, assessing
comparative approaches to three pivotal issues in the biomedical context: tissue
use, disclosure of genetic test results, and reduced capacity to consent. I select
these issues because they illustrate different levels and aspects of privacy and,
as such, suggest a composite picture of both the degree and nature of
protections and to what extent architecture affects the meaning of the
substantive right. The third issue, reduced capacity to consent, while not
exclusively a genetic issue, is addressed because it reveals something of the
non-negotiated boundaries of privacy not modifiable by individual consent.

Often, one thinks of medical privacy as being a matter of protection of
medical information and confidentiality. These concepts are indeed central to
the scope and force of medical privacy, and, like autonomy, are often
considered among the fundamental principles of biomedical ethics. Yet, one of
the most important aspects of patient rights in this context is a notion that
combines both concepts to create a hybrid interest in what I shall refer to as
"informational autonomy"-the right to control information about oneself."
This encompasses what Brandeis referred to as the right to be let alone. 12 The
European Union, in its Data Protection Directive, 13 also adopts a view of
privacy that renders this concept a core concern in privacy protections.

9. Daniel Solove, A Taxonomy of Privacy, 154 U. PA. L. REV. 477, 483 (2006).
10. Id. at 480.
11. See, e.g., GRAEME LAURIE, GENETIC PRIVACY: A CHALLENGE TO MEDICO-

LEGAL NORMS 4 (2002); DAVID H. FLAHERTY, PROTECTING PRIVACY IN SURVEILLANCE
SOCIETIES: THE FEDERAL REPUBLIC OF GERMANY, SWEDEN, FRANCE, CANADA, AND THE
UNITED STATES 7 (1989); ALAN WESTIN, PRIVACY AND FREEDOM (1970) (for usage of
terms describing privacy interests).

12. Samuel D. Warren & Louis Brandeis, The Right to Privacy, 4 HARV. L. REV. 193
(1890).

13. See Council Directive 95/46/EC, Protection of Individuals with Regard to the
Processing of Personal Data and on the Free Movement of Such Data, 1995 O.J. (L 281) 31
[hereinafter Directive].

2009]



IND. INT'L & COMP. L. REv.

Concepts of both informational and decisional privacy are implicated in this
discussion and serve as the focus for this analysis.

Informational autonomy in the biomedical context arises in many forms.
In the age of genetic medicine, the genetic information of one patient may be
significant for another. Thus, when deciding whether to share an individual
patient's medical information, one should attempt to strike a balance between
the rights of a patient to keep his medical information confidential and the
rights of a patient who may benefit from the release of another's personal
medical records. Often the question comes down to whether the patient has a
superseding right to make decisions about who has access to his private medical
information when others may benefit from knowing it. As biomedical
technology proceeds, the question of individual control of medical information
arises with increasing frequency in clinical practice, medical research, and even
public health and law enforcement. As the cases of both tissue use and storage
and disclosure of genetic test results show, the individual privacy right can
come into direct conflict with the pursuit of societal benefit. 14 In the former
case, the societal benefit derived from the compromise of personal data
protections comes in the form of more informed medical research and potential
individual and collective benefits to society. In the later case, the interests of a
smaller but less speculative circle of potential beneficiaries can come into
conflict with those of individual right to privacy.

With the growing case for encroachment on informational autonomy in
the name of public interest, it becomes increasingly important to ascertain the
force of privacy protections. By examining this question through a comparative
analysis of the structure of rights that protect privacy in the biomedical context,
some light can be shed on the nature of privacy protections and where the true
source of their comparative force lies. As biomedical innovations increase the
value of sharing private medical information, identifying and crafting the
protective mechanism of privacy rights becomes increasingly important.

2. Overview of the Sources of Privacy Protections

This comparative analysis begins with an identification of the sources of
privacy protections and the hierarchy of legal norms regarding privacy rights.
The sources of privacy protections, like many laws and regulating mechanisms,
do not operate in a vacuum, but rather rely on various institutional, social, and
political factors for their implementation and ultimate effect. While not
discounting these extra-legal factors, they generally lie outside the scope of this
Article.

Starting with European treaties and the Data Protection Directive of the
European Union, this analysis examines the European Union Directive 95/46

14. See, e.g., Bartha M. Knoppers, Confidentiality of Health Information: International
Comparative Approaches, in Protecting Data Privacy in Health Services Research, App. D, 180 (Inst.
of Med. 2000), available at http://newton.nap.edu/htmlldata-privacy/appD.html#FOOT19.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

(Data Protection Directive) and related national legislation to assess the nature
and effect of the architecture of privacy protections. Then crossing the
Atlantic, the Article examines the U.S. structure of privacy protections in the
biomedical context.

2.1 European Union Membership and Its Effect on Member
National Law

The European Union was formed as an outgrowth of a number of treaties
and agreements with the primary goal of facilitating trade. Beginning with a
core membership of eight member nations, it expanded to a membership of 27
by 200715. Membership in the EU held out a number of benefits to member
states, particularly economic benefits, and became a foreign policy goal for
several post-communist eastern European states. 16 Of course, membership in
the EU carries with it numerous obligations, particularly in the observance of
EU law and directives. For some relatively late membership candidates, some
East European countries, for example, the obligations operated as both a
passive and active leverage, resulting in what has been referred to as
"asymmetric interdependence."' 17 Some countries that joined the EU after the
completion of the internal market but were already members of the European
Economic Area (EEA) experienced less EU leverage. This is most likely
because, as EEA members, they had already complied in principle or practice
with general EU norms, particularly those stemnming from status as a market
economy.18 This relatively lower degree of leverage may help to explain some
aspects of implementation and application of privacy and freedom of speech
law in Norway, an EEA member.

2.2 European Convention on Human Rights

The European Convention for the Protection of Human Rights and
Fundamental Freedoms (European Convention) was adopted by the Council of
Europe in 1950.19 It articulates the right to privacy in two clear and succinct
provisions, arguably leaving no room for suggestion that privacy rights are
without meaningful legal basis. Article 8 of the European Convention
articulates the right to privacy in two provisions:

15. The History of the European Union. Europa, http://europa.eu/abc/history/
indexen.htm (last visited Feb. 13, 2009).

16. MILADA ANNA VACHUDOVA, EUROPE UNDIVIDED: DEMOCRACY, LEVERAGE,
AND INTEGRATION AFTER COMMUNISM 108 (2005).

17. Id. at 107-10.
18. Id. at 111.
19. Convention for the Protection of Human Rights and Fundamental Freedoms as

amended by Protocol No.11. Rome, 4.XI.1950. available at http://conventions.coe.int/
Treaty/en/Treaties/Html/005.htm.

2009]



IND. INT'L & COMP. L. REv.

1. Everyone has the right to respect for his private and family
life, his home and his correspondence.

2. There shall be no interference by a public authority with the
exercise of this right except such as is in accordance with the
law and is necessary in a democratic society in the interests of
national security, public safety or the economic well-being of
the country, for the prevention of disorder or crime, for the
protection of health or morals, or for the protection of the
rights and freedoms of others.2°

This Convention and its implementing and institutions of enforcement are
critical to the protection of privacy interests.2 '

3. European Directive on Data Protection

In 1995, the European Commission enacted the Data Protection Directive
(Directive) addressing the member states of the European Union.22 This
document was designed to facilitate the free flow of information among the
member states without compromising the privacy of the citizens in each of the
member states.23 The Directive sets forth with greater particularity the
provisions that ensure the processing of personal data in a manner that is
consistent with Article 8 of the Convention while aiming to assist in the
protection of privacy interests in the domestic and international transfer of
information. Article 1 ("Object of the Directive") states: "In accordance with
this Directive, Member States shall protect the fundamental rights and freedoms
of natural persons, and in particular their right to privacy with respect to the
processing of personal data., 24

20. European Convention for the Protection of Human Rights and Fundamental
Freedoms, Art. 8, Nov. 4, 1950, Europ. T.S. No. 5, (entered into force on Sept. 3, 1953), as
amended by Protocol No. 3, Europ. T.S. 45 (entered into force Sept. 21, 1970); Protocol No. 5,
Europ. T.S. 55 (entered into force on Dec. 20, 1971); Protocol No. 8, Europ. T.S. 118 (entered
intoforce on 1 Jan. 1990); and Protocol No. 1l, Europ. T.S. 155 (entered into force on Jan. 11,
1998) [hereinafter European Convention].

21. Taking the 1948 Universal Declaration of Human Rights as its starting point, the
European Convention was designed to further the goal of collective enforcement of the
Universal Declaration of Human Rights. To that end, the European Commission of Human
Rights, the European Court of Human Rights, and the Committee of Ministers of the Council of
Europe were set up in subsequent years to enforce the terms agreed to by the contracting states.
See Universal Declaration of Human Rights, G.A. Res. 217A U.S. GAOR, 3d Sess., 1st Plen.
Mtg., U.N. Doc A/810 (Dec. 12, 1948).

22. Council Directive 95/46/EC, Protection of Individuals with Regard to the Processing
of Personal Data and on the Free Movement of Such Data, 1995 O.J. (L 281) 31.

23. Id. at 10.
24. Directive, supra note 13, at Art. 1.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

3.1 Sensitive Information

Although the Directive does not specifically address medical information
or research, much can be derived from its provisions with regard to
informational privacy protection issues as they arise in the medical arena.25

Article 8(1) of the Directive on the processing of special categories of data
states in relevant part: "Member States shall prohibit the processing of personal
data revealing racial or ethnic origin, political opinions, religious or
philosophical beliefs, trade-union membership, and the processing of data
concerning health or sex life.",26

Thus, although medical research is not specifically addressed by the
Directive, Article 8 indicates that personal health information falls into a
special category of data deserving of special protections in its processing. This
is frequently referred to, both in EU contexts as well as in many national laws,
as "sensitive" data.

3.2 Derogations, Exceptions, and Overrides

The protections against the processing of this special category of personal
data are not absolute. Indeed, Article 8 subsections (2)-(5) serve to derogate the
privacy protections afforded in Article 8(1), stating that Article 8(1) shall not
apply to the processing of data relating to offences, criminal convictions or
security measures, to exemptions laid down by member states in the
"substantial public interest," and notably, to instances in which the processing
of data is required for "purposes of preventive medicine, medical diagnosis, the
provision of care, or treatment or the management of health-care services. 27

Recital 34 elaborates on the derogation contained in the articles of the Directive
stating:

Whereas Member States must also be authorized, when
justified by grounds of important public interest, to derogate
from the prohibition on processing sensitive categories of data
where important reasons of public interest so justify in areas
such as public health and social protection - especially in
order to ensure the quality and cost-effectiveness of the
procedures used for settling claims for benefits and
services in the health insurance system - scientific
research and government statistics; whereas it is incumbent
on them, however, to provide specific and suitable safeguards

25. See, e.g., IMPLEMENTATION OF THE DATA PROTECTION DIRECTIVE IN RELATION
TO MEDICAL RESEARCH IN EUROPE (Deryck Beyleveld et al., eds., 2004) [hereinafter
Research].

26. Directive, supra note 13, at art. 8.
27. See id.

2009]



IND. INT'L & COMP. L. REv.

so as to protect the fundamental rights and the privacy of
individuals. 28

While the Directive does not specifically address medical research and
the use of personal medical information in that context, its inclusion of health
data as "sensitive information" and its designated derogations in the public
interest (as stated in Recital 34, in the interests of public health, social
protection, health delivery and administration, and scientific research), indicate
that the architecture of privacy protections as designed in the Directive is such
that the protection of substantive rights is not stringent. The list of broad
derogations appears to yield no predictable precise boundary for the privacy of
health data. Nor does the Directive interpret or direct the precise scope or
application of these categories of derogation. Thus, social protection in the
public interest may actualize more broadly in some countries as an override of
privacy interests than in others. Where this is contested, it may be brought to
the European Court of Human Rights or Court of Justice.29

In accordance with the proportionality rule, which allows for a balancing
of interests and rights, implementing member states may derogate individual
privacy rights if the countervailing interests constitute an important public
interest, such as public health or social protection, and otherwise meet the
criteria of the proportionality principle. However, the protection of
fundamental rights figures prominently into the balancing of interests, even in
the face of broadly articulated exceptions. In order to prevail over a
fundamental right, the proportionality rule requires that the countervailing
interest may not be frivolous or dismissive of individual privacy. Thus, the
proportionality rule does not open the floodgates; rather it operates within
certain confines and assurances that the essential character of fundamental
rights is not undermined.3° Still, health information, while accorded special
protected status, in many ways stands to lose some important aspects of that
status by virtue of how it is used in the field of health and medical research and
how the broadly constructed derogations are interpreted and applied nationally.

As this Article explores, the relevant parameters of privacy are likely to
be created and enforced by national legislation, the structure and conception of
privacy rights, and by the priority placed on privacy by society.

3.3 Implementation of the Directive

One of the Directive's primary objectives is to harmonize data protection
across member states in order to facilitate the use of such data across borders
within the EU. Additionally, the Directive also addresses standards and
practices regarding data protection among non-member states in order to ensure

28. See id. at Recital 34 (emphasis added).
29. See, e.g., Z v. Finland, App. No. 22009/93, 25 Eur. Ct. H.R. (1997).
30. See Schauer, supra note 3.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

an adequate level of protection. This is manifest in the requirement that the
transfer of personal data from an EU member state can only be done if the
protections in the receiving country are equal to those in the EU, or if not, a
certificate of adequacy must be acquired from the appropriate authorities. 3'

Each member and EEA state was required to effect implementation of the
Directive by 24 October 1998.32 This deadline for compliance was
subsequently extended by amendment until 2001.

The implications of the Directive for medical research are substantial.
With increasing transnational collaborative research projects, and the
desirability of using existing tissue collections, regulations regarding
transferability are pivotal to the medical research enterprise. The commitment
to fundamental rights and the protection of individual privacy is really called
into question when weighed against the range of societal interests that may fall
within the list of derogations.

The process of implementing the Directive, however, was not merely a
matter of rubber-stamping of Directive provisions. Indeed, it sometimes
required changes that involved numerous procedural, institutional, and political
hurdles in the member states. Consequently, when the first report on
implementation by member states was to be made in 2001, the slow process of
transposition among many member states resulted in delay of the first report33.

At the end of 1999 the European Commission decided to file against France,
Germany, Ireland, Luxembourg and the Netherlands in the European Court of
Justice for failure to fully implement Directive 95/46.34 In 2001 the
Netherlands and Germany documented their compliance, and eventually the
cases against the remaining nations were closed.

National implementation of the Directive also had to be sensitive to
national notions, norms, and practices regarding privacy. Particularly in the
context of biomedical research and genetic research, the standards set by the
EU provide for a fair margin of discretion for Member States in the substantive
implementation of certain areas. This broad discretion is very significant to the
protection of sensitive information, particularly regarding health information.
For example, one of the consistently mentioned justifications for derogation
pertains to scientific purposes and research that presumably benefits society.

Despite the fact that the Directive contains no specific reference to
medical applications, the transposition of the Directive by Member States has
an effect on the policies and regulations governing significant aspects of

31. Research, supra note 25.
32. Commission Report on the Transposition of the Data Protection Directive, Analysis

and Impact Study on the Implementation of Directive EC 95/46 in Member States, available at
http://ec.europa.eu/justice-home/fsj/privacy/docstlawreport/consultation/technical-annex-en.pdf
(last visited Nov. 11, 2008) [hereinafter Analysis].

33. See actions against France, Germany, Luxemburg, Ireland, and the Netherlands,
discussed below.

34. See id.

2009]



IND. INT'L & COMP. L. REv.

biomedical research that use genetic technologies and information. 35 It is
generally recognized, although not without controversy, that genetic
information has a special character and, at least in some instances, may require
separate legislation and oversight.36 While the Directive does not require this
kind of particularized regulation in implementation, several Member States
have taken up this matter. In addition, the Commission's Article 29 Data
Protection Working Party (Working Party 29) has established a Working
Document on Genetic Data in recognition of the concerns arising from the
unique nature of genetic information.37

Variation with regard to protection of genetic data is considerable.
Working Party 29 has noted that regulation of the processing of genetic data is
uneven across the EU: "Indeed, while some Member States have explicitly
listed genetic data as sensitive data in their Data Protection law with all the
safeguards and restrictions associated, in most Member States the issue of the
processing of genetic data is not as such regulated by specific legislation. 38

Thus, while some Member States provide for complementary rules in their laws
on patients' rights and have enacted legal regulations for the processing of
genetic data, the Working Party 29 anticipates a trend toward increased national
regulation of the processing of genetic data specifically. For example, in 2005,
Portugal, a non-member state, enacted genetic legislation applying many of the
principles and safeguards that regulate general medical information, including
the nature of permissible overrides.39

Where the unique aspects of genetic information have consequence for
privacy protections, Member States face the challenge of finding ways to
provide the necessary protections, whether in sectoral legislation or broader
protective legislation, while leaving sufficient room in which to derive the
benefits of developing technologies that challenge the margins of privacy. This
Article looks at two largely unsettled issues involving the processing of genetic
information - tissue use and storage and disclosure of genetic test results.
Regulation of these issues is uneven. The following section provides a very
brief view of three national systems of privacy protections pertaining to health
information.

35. See Research, supra note 25.
36. Sonia Suter, The Allure and Peril of Genetic Exceptionalism: Do We Need Special

Genetics Legislation? 79 WASH. U. L.Q. 669, 747 (2001).
37. Article 29 Data Protection Working Party, Working Document on Genetic Data,

12178/03/EN WP 91 (Mar. 17, 2004), available at
http://ec.europa.eu/justice-home/fsj/privacy/docs/wpdocs/2004/wp9l-en.pdf [hereinafter
Article 29 Data Protection Working Party].

38. Id. at 3.
39. Law No 12/2005 of Jan. 26,2005, Personal Genetic Information and Information on

Health [hereinafter Portugal].

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

4. Europe: National Hierarchies of Privacy Norms

4.1 France

The Constitution of France does not mention privacy or set forth rights of
this type.4° However, by virtue of its declaration of membership in the
European Union,4 1 France is obligated to abide by the principles of the
Directive. There is no single comprehensive privacy provision in France.
Provisions for the protection of privacy interests are found in both the civil and
criminal codes of this civil law country.

Article 9 of the Civil Code recognizes privacy as a fundamental right
stating that "everyone has the right to respect for his or her private life. 42

What constitutes private life is not defined, but rather has been a matter for the
courts, who have over time constructed a meaning that has been said to include
a person's "love life, friendships, family circumstances, leisure activities,
political opinions, trade union or religious affiliation and state of health. 43

This protection applies to both public and private spaces. Furthermore, Article
9 authorizes the court to take measures to prevent or halt invasions of personal
privacy and, in general, at least regarding publication, privacy protections in
France may be among the most protective. 44

Privacy in the health sector is generally provided for at the national level
by the Penal Code in Articles 226-13,14). In the section on "Professional
Secrecy," Article 226-13 states that disclosure of secret information by one
entrusted with such information by virtue of his profession or position is
punishable by a year of imprisonment and a fine of 15,000 Euros.45 It has been
noted that, unlike most obligations of physicians, which are "obligations of
means," the obligation of secrecy is an "obligation of result"; thus, it is not only
what is explicitly or implicitly communicated, it is also that which is
understood. 46 It has been noted that, unlike most obligations of physicians,
which are "obligations of means," the obligation of secrecy is an "obligation of
result"; thus, it is not only what is explicitly or implicitly communicated, it also
refers to that which is understood.47 Therefore, whatever is "communicated,"
verbally or non-verbally, to a physician by any means within the context of the

40. 1958 La Constitution (Fr.).
41. Title 15 on the European Communities and the European Union, and its agreement, as

a result of the Treaty on European Union signed on 7 February 1992.
42. Code civil [C. CIV.] art. 9 (Fr.).
43. Legal and Technical Office of Information and Communication for the Embassy of

France, Embassy of France in Washington, French Legislation on Privacy, http://ambafrance-
us.org/spip.php?article640 (last visited Nov. 12, 2008).

44. Id.
45. Code penal [C. pen.] art. 226-13 (Fr.).
46. See Knoppers, supra note 14, at 180.
47. Analysis, supra note 32, at 24.

2009]



IND. INT'L & COMP. L. REV.

doctor-patient relationship is privileged. Thus, protections in France are
constructed broadly, resting on both the Data Processing Act and relevant
provisions in the Codes.

4.2 Implementation of the Data Protection Directive: Act on Data
Processing, Files and Individual Liberties

The French Data Protection legislation 48 came into compliance with Data
Protection Directive 95/46/CE in 2004, after having been taken to the European
Court of Justice for failure to notify all the provisions of the Directive.49 The
French Act on Data Processing, Files and Individual Liberties does contain a
chapter specifically regulating the processing of medical data.5°

5. Norway

The Norwegian Constitution, adopted in 1814, specifically stating that the
"search of private homes shall not be made except in criminal cases," 5'
somewhat resembles privacy protections as expressed in the Fourth
Amendment of the U.S. Constitution. Additionally, Article 110(c) of the
Norwegian Constitution, which broadly sets forth protections of human rights,
sets forth further provisions that pertain to privacy rights.52

Giving effect to international agreements requires both ratification and
implementation of national legislation. The European Convention on Human
Rights was incorporated into Norwegian law by the Human Rights Act of 21
May 1999 no. 30. According to Article 3 of the Human Rights Act, the
Convention takes precedence over conflicting legislative provisions. The
ECHR does not, however, enjoy constitutional status although it clearly has
strong force by virtue of its precedence over conflicting legislative provisions.53

This provision, not surprisingly, is the subject of some debate.54 Additionally,
as in much of Europe, under "the principle of legality," unwritten norms rooted
in customary law may also acquire constitutional status.55

Since the European Convention on Human Rights (ECHR) specifically
provides for protection of individual privacy and Norway is a signatory to the

48. Decree No. 2005-1309 of Oct. 20,2005, Journal Officiel de la Republique Francaise
[J.O.] [Official Gazette of France], Oct. 22, 2005, p. 16769.

49. Analysis, supra note 32.
50. Decree No. 2005-1309 of Oct. 20, 2005.
51. Grunnloven [GrlI [Constitution] § 102 (Nor.).
52. Id. at art. 110(c).
53. Lee A. Bygrave & Ann Helen Aaro, Privacy, Personality and Publicity--An Overview

of Norwegian Law, in INTERNATIONAL PRIVACY, PUBLICITY, AND PERSONALITY LAWS,
333, 333 (M. Henry, ed., 2001).

54. See Synne S. Maehle, Limits of Rettsanvendelsesskjonn: About the Legal Legitimacy
of a Tension Between Flertallsmakt and Rettighetsvern, in GYLDENDAL ACAD. 285, 285-99
(2005).

55. Id. at Ch. 22.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

Convention and has passed implementing legislation incorporating the ECHR,
those privacy protections set forth in the Convention are protected under
Norwegian law. Therefore, under Article 8 of ECHR, provisions relating to
privacy protections are incorporated into Norwegian law by way of the
Norwegian Constitution.

5.1 Data Protection Directive Implementation: Norwegian
Personal Data Act of 2001

Norway, a member of the EEA and not the EU (but similarly required to
bring its laws into compliance with the Directive), locates primary protections
for personal information in the Norwegian Personal Data Act of 200156

(replacing the Data Registers Act of 1978). 57 In the context of medical privacy,
other laws, regulations, and provisions interact with the Norwegian Personal
Data Act to provide the national standard of privacy protection of medical
information. Of particular interest in the realm of genetic privacy is the
Biobank Law of 2004, which sets forth regulations pertaining to the use of
genetic technologies.

Initially, certain provisions of the draft Norwegian Act proved
problematic both with regard to the Directive as well as for certain
constituencies in Norway, particularly those concerned with research. It has
been argued that the final Norwegian Data Protection Act that was passed in
2001 was modified to satisfy the research community as well as comport with
Directive provisions as a result of political idiosyncrasies. 58

The new Norwegian Personal Data Act occasioned a shift in privacy
protections, both in content and, significantly, in underlying orientation.59 The
former Norwegian legislative tradition regarding privacy took its rise from a
"model of control" in which privacy, as a responsibility of society and the
government, was a matter of external supervision, monitoring, and licensing.60

The new Act, following the lead of the Directive, is oriented toward a "model
of consent" and, according to Bygrave and Aaro, introduces more substantive
rules and regulation of specific applications rather than the previously favored
"framework legislation." 61 Now the Act includes numerous and more detailed
substantive rules and clearly delineates specific principles such as purpose

56. Act of 14 April 2000 No. 31 relating to the processing of personal data (Personal Data
Act) available at http://www.datatilsynet.no.htest.osl.basefarm.net/upload/Dokumenter/
regelverk/lovjforskrift/lov-20000414-031-eng.pdf (last visited Feb. 16, 2009).

57. Vigdas Kvalheim, Implementation of the Data Protection Directive in Relation to
Medical Research in Norway, in RESEARCH, supra note 25, at 291.

58. Id. at 290 (describing idiosyncrasies such as who participated and controlled the
process that formed the final law).

59. Id, at 290.
60. Id. at 291.
61. Bygrave & Aaro, supra note 46.

2009]



IND. INT'L & COMP. L. REv.

62specification (as may be applied to tissue use and storage).
The Norwegian Personal Data Act has been perceived as not only being

consistent with the Directive, but exceeding it in the standards for data
protection.63 It has been observed that the Norwegian Act may actually serve to
value privacy over furthering scientific knowledge. 64 To the extent that this is
true, it is reasonable to expect that this elevated status of individual privacy is
consistent across sectors.

The Norwegian government, through its oversight agency, Datatilsynet,
states that the express purpose of the Act is to protect persons from violations
of their right to privacy when personal information is processed and to ensure
that any processing of personal information is done in a way that accommodates
a fundamental respect for privacy rights; it aims to protect personal integrity
and private life and ensure that any personal data that is processed meets
acceptable standards of quality.65

In virtually every aspect of the Norwegian approach to privacy
protections, the shift toward consent seems to predominate. Consistent with
this is the government website for the Data Protection Act: its implementing
government agency, Datatilsynet (Data Supervision/Oversight), has as its
headline slogan: "Protection of Persons: Your Right to Choose." 66 Consent is
the predominating principle and a decisive factor in much of the processing of
personal information in the biomedical context in Norway.

Three pieces of legislation constitute the Norwegian Data Protection Act:
1) The Personal Data Act; 2) Personal Data Regulations; and 3) the Personal

Health Data Filing System.67 However, since this Article looks at specific
applications involving biotechnology, such as tissue storage and genetic test
results, the Biobank Law, a relatively new piece of legislation implemented in
2004, is also implicated. Together, these regulations provide personal data
protection in the biomedical context.

6. United States

It is frequently asserted that there is no fundamental right to privacy in the
United States, but rather that it is a social construction,68 undergirded by no

62. Id.
63. See Kvalheim, supra note 57, at 290.
64. Kvalheim, supra note 57, at 292.
65. Datatilsynet, About the Data Inspectorate, http:llwww.datatilsynet.no/templates/

Page 194.aspx (last visited Feb. 17, 2009).
66. Id.
67. Act of 18 May 2001 No. 24 on Personal Health Data Filing Systems and the

Processing of Personal Health Data (Personal Health Data Filing System Act) available at
http://www.regjeringen.no/en/dep/hod/Subjects/The-Department-of-Public-Health/Act-of-18-
May-200l-No-24-on-Personal-Health-Data-Filing-Systems-and-the-Processing-of-Personal-
Health-Data-Personal-Health-Data-Filing-System-Act-.html?id=224129 (last visited Jan. 26,
2009).

68. See Frederick Schauer, Free Speech and the Social Construction of Privacy, in THE

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

directly relevant constitutional provision. 69 Yet, the Fourth Amendment would
seem to protect at least some aspects of privacy interest. It states in part, "[tihe
right of the people to be secure in their persons, houses, papers, and effects,
against unreasonable searches and seizures, shall not be violated, and no
[w]arrants shall issue, but upon probable cause . . ,,70 The notion of
protection against unreasonable search and seizure has been interpreted to apply
to a wide variety of activities, communications, and entities that are not
specifically enumerated. Wiretapping is a frequently used example.7 1

One might note that, like First Amendment protection of freedom of
speech, there is no list of derogations to the Fourth Amendment; it appears as
an "absolute" non-derogable right. Arguably, the warrants that may be issued
according to specific criteria may be said to provide the opportunity for the
balancing that is transparent in European legislation.

But the reality is that there is no overarching privacy legislation in the
U.S. comparable to the European Directive on Data Protection or most of the
national implementing legislation following a model of comprehensive data
protection. Rather, the U.S. has adopted an approach characterized by sectoral
privacy legislation. Thus, in the U.S. there is special privacy legislation
regarding, for example, credit, banking, communication, and health.

6.1 Sectoral Health Privacy Legislation: HIPAA

Privacy in the biomedical context is governed by the Health Insurance
Portability and Accountability Act (HIPAA).72 Through this 1996 Act and its
subsequent amendments, the U.S. system of privacy protections in the
biomedical context would appear to offer a strong statement of protections.
Yet, in its application in the context of biomedical research, much of the
protection is in fact left to the discretion of implementing agencies. Following
guidelines and laws set forth in HIPAA, administrative agencies are charged
with oversight of the conduct of medical research within the confines of the law
largely as applied by institutional review boards (IRBs).73 This is especially
true in the case of new technologies and the novel ethical and legal issues they
raise. Because the process of law-making is so slow, the use and application of
new technologies often goes forward without regulatory oversight of the
vulnerabilities created by such technology. The lacunae created by the

JOAN SHORENSTEIN CENTER ON THE PRESS, POLITICS AND PUBLIC POLICY: THE FIRST
AMENDMENT SERIES (2000), for an exemplary analysis of the nature of this claim of social
construction.

69. Id.
70. U.S. CONST. amend. IV.
71. See, e.g., Videotape: In Search of the Constitution: Mr. Justice Brennan (Films for

the Humanities & Sciences 1987).
72. Health Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191,

110 Stat. 1936 (codified as amended in scattered sections of 18 U.S.C., 26 U.S.C., 29 U.S.C., 42
U.S.C.) [hereinafter HIPPA].

73. 45 C.F.R. § 46.107 (2005).

2009]



IND. INT'L & COMP. L. REv.

outpacing of ethical analysis and the regulation spurred by the scientific
developments are considerable. One area where this is very apparent is that of
tissue use and storage, and the derivation, compilation, and storage of personal
data taken from such tissue. The significance of the fact that HIPAA makes no
specific reference to genetic data is unclear. HIPAA does not distinguish
among different types of personal medical information 74 . However, because of
the structure of rights, permissible processing of this data will not necessarily
be the same. A use that falls through the cracks in HIPAA legislation and,
therefore, might go unregulated in Europe will still receive rigorous analysis if
it burdens a fundamental right.

Under HIPAA, protected health information (PHI) refers to individually
identifiable health information. This includes information such as demographic
information relating to a person's past, present or future health state; the
provision of care, payment for health care, or anything that otherwise makes it
possible to identify the individual. Common identifiers include birth date,
social security number, name, and address.

6.1.1 Explicit Exceptions

Along with the protections, HIPAA contains an extensive list of
exceptions to non-disclosure of PHI.76 This list includes disclosures: 1) for
public health activities; 2) about victims of abuse, neglect, or domestic
violence; 3) for health oversight activities; 4) for judicial and administrative
proceedings; 5) to avert a serious threat to health or safety; 6) for specialized
government functions; 7) for research purposes and; 8) for law enforcement
purposes.77 These categories of exceptions are carved out in the legislation,
reflecting, as Schauer points Out, 78 that the balancing has taken place in the
drafting of the legislation. Importantly, these are also distinguishable from the
broad derogations characteristic of European legislation precisely because of
their specificity and the largely absent opportunity for further balancing.

This absence of opportunity for further balancing is not absolute
regarding the use of private health data. The carved-out exceptions allow for
some flexibility for the arbiter of disclosure. Thus, "research purposes," for
example, absent a specific provision to the contrary, could allow a wide range
of practices that would normally be considered in violation of privacy rights.
However, typically, an IRB will have the opportunity to conduct a form of
balancing 79. The important point that distinguishes this kind of balancing from

74. See HIPPA, supra note 72.
75. 45 C.F.R. § 160.103 (2005).
76. 45 C.F.R. § 164.512(2005) (providing: "Uses and disclosures for which consent, an

authorization, or opportunity to agree or object is not required").
77. 45 C.F.R. Section 164.512 (a)-(f).
78. Schauer, supra note 3.
79. See, e.g., Nat'l. Inst. of Health, Nat'l. Comm'n. for the Protection of Hum. Subjects of

Biomedical and Behav. Res., The Belmont Report: Ethical Principles and Guidelines for the

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

that performed in accordance with the proportionality principle is that the
nature of the balancing at this level is rarely transparent. As a result, it is often
not clear what interests are being balanced and what weight is being accorded
them. Thus, a decision by an RB to permit a waiver of consent for the
indefinite storage and use of existing genetic samples in an ongoing study may
be justified by an oversight committee based on the inconvenience to the
research staff, time constraints, or even on the reputation of the researcher, not
to mention personal idiosyncrasies of a given IRB in which members
expressing reservations about a study may be subjected to subtle (and not so
subtle) expressions of disapproval.8° Still, HIPAA provides a baseline level of
protection, which is viewed by some, particularly researchers, as "overly
protective. ' 81

7. Genetic Information in Research

It is widely recognized, although not undisputed, that genetic information
has characteristics that set it apart from other medical information.82 For
example, if one parent carries the genetic mutation for Huntington's disease,
the children have a fifty percent chance of developing the same fatal and
untreatable disease. Information about this genetic information can be of great
interest to the offspring as well as siblings.83 The power of information on a
single gene disorder like Huntington's can be overwhelming, but when
multifactorial diseases also carry genetic markers which can be tested for, the
value of the genetic test result becomes more speculative. For example,
questions arise regarding whether a forty-five-year-old who tests positive for the
genetic mutation associated with colon cancer should be compelled to disclose
this to family members in the interest of providing health benefits.

The ability to access and interpret genetic information as a biomedical
tool can offer a range of benefits by virtue of the kinds of information it can
provide and interventions to which it might lead. However, genetic technology
also presents enormous complexities regarding the use of this information.
Indeed, genetic information can trigger the involvement of interests,
obligations, and rights of persons that extend beyond those of the individual
whose genetic information is at issue. Furthermore, as the Article 29 Data
Protection Working Party points out, genetic information can provide personal

Protection of Human Subjects of Research (1979) available at http://ohsr.od.nih.govl
guidelines/belmont.html#goc2.

80. Sohini Sengupta & Bernard Lo, The Roles and Experiences of Nonaffiliated and Non-
Scientist Members of Institutional Review Boards, 78(2) ACAD. MED. 212, 212-218 (2003).

81. See, e.g,. Robert E. Erard, Release of Test Data Under the 2002 Ethics Code and the
HIPAA Privacy Rule: A Raw Deal or Just a Half-Baked Idea? 82 J. PERSONALITY
ASSESSMENT 23 (2004).

82. See, e.g., Sonia M. Suter, The Allure and Peril of Genetic Exceptionalism: Do We
Need Special Genetics Legislation? 79(3) Wash. U. L. Q. 669 (2001).

83. See, e.g., ALICE WEXLER, MAPPING FATE: A MEMOIR OF FAMILY, RISK, AND
GENETIC RESEARCH (1995).

20091



IND. INT'L & COMP. L. REv.

information relevant throughout an individual's remaining life. 84

As personal information that, in most circumstances, falls in the category
of "sensitive information," such as health data, it is also subject to derogation.
The assurance of the protection of personal data that is paramount in the Data
Protection Directive is tested in new ways in the context of the collection, use,
and dissemination of genetic information. Inasmuch as the Directive lists "the
public interest" among the categories of overrides that may be used to justify
exemption from certain privacy provisions,8 5 individual genetic information
may routinely be caught in the paradoxical position of being classified both as
"sensitive information" and as information whose benefit to society outweighs
the potential harm to the individual's privacy interests and, thus, is less
deserving of protection than even non-sensitive personal data.

The rationale for the selection of these two issues is manifold. First, there
is no consensus on the resolution of either issue and there is considerable

86variability in the current approaches to these issues . Second, these issues
present profound questions regarding the future of privacy protections. Third,
policies regarding tissue use and storage will have long-term effects on privacy
interests and the weight of the countervailing potential benefit derived from
incursions on those privacy interests87 . Finally, these two questions gauge
different levels of privacy protections as well as suggest key aspects of the
nature of the application of the proportionality rule in the balancing between
individual rights and third party benefit.

This Article also references the issue of research involving persons with
reduced or no capacity to consent. I shall refer to this category of persons as
"incapacitated persons." The reasons for the selection of this issue are related
to its place as a largely unsettled and controversial aspect of biomedical
research and to what it tells us about attitudes toward privacy. In a sense, the
issue of research on incapacitated persons forces clarity on the limits of privacy
since consent, as a mobile parameter, is not available. As this Article explains,
the current trend is to limit research on such persons to that which will provide
a direct benefit, and in some instances, only when a surrogate decision-maker
has provided consent. This has been a very unpopular policy with researchers,
and many countries are re-examining their position on this issue. This issue
calls forth a declaration of the balance between incursion into the private sphere
without consent and the weight of the public interest. So, while this particular
issue is governed primarily by Directive 2001/20/EC (Clinical Trials Directive),
the privacy implications are significant, and, in the case of incapacitated
persons, the policy resolution of the balance between individual privacy and

84. Article 29 Data Protection Working Party, supra note 37, at 4.
85. Directive, supra note 13, at recital 34.
86. See, e.g., Privireal.org, http://www.privireal.org/index.php (A European Commission

Framework 5 Project on the Implementation of the Data Protection Act among Member States).
87. See, e.g., E.W. Clayton, et al., Informed Consent for Genetic Research on Stored

Tissue Samples, 274 J. Am. Med. Ass'n 1786 (1995).

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

public benefit can be very telling.

7.1 Tissue Use and Storage

Medical research has taken on new ethical and legal challenges with the
advent and development of genetic technologies and various computerized
database and advanced sequencing capabilities. Some of the most promising
developments are tissue banking and biobanking. The former is generally
understood to refer to the storage and processing of human biological material
("HBM"), e.g. blood, tissue, saliva, etc.; whereas the latter is primarily
concerned with the collection of data derived from the tissue. A biobank allows
for the processing (for research purposes) of HBM and known phenotypic
characteristics and thus may include both. This kind of banking allows for
longitudinal studies as well as studies on related or unrelated diseases or other
epidemiologic phenomena. Tissue and biobanking potentially implicate
property and privacy interests, while data-banking more strongly implicates
privacy interests, although not exclusively. In most instances, it is the
information derived from the tissue that is the basis for the privacy concerns.8

It is this coupling of information and tissue into a "biobank," an optimal
research tool, that is the primary focus of this inquiry. Interestingly, as far back
as the 1970's, there was growing concern about the increase of data banks of
various sorts, which permit "computerized pools of information" about virtually
every aspect of people's lives.89

Several issues have proven controversial with regard to various aspects of
biobanking. First is the issue of informed consent as a mechanism of both
decisional and informational privacy.90 One of the great benefits of biobanking
- the ability to analyze the same tissue over time for various characteristics as
scientific developments would merit - also constitutes one of the great
dilemmas of the enterprise.

The Helsinki Declaration 91, embraced by the European Convention on
Human Rights, declares that informed consent is a foundation for biomedical
research on human subjects.92 A typical scenario in biobanking involves a
subject (human participant) agreeing to participate in research that involves the
taking of a tissue sample for analysis for a specific research project. However,
it is increasingly common for researchers to acquire a sample with the idea of

88. The most notable exception to this would be the use of tissue that results in patent.
See, e.g., Moore v. Regents Univ. of Cal., 793 P.2d 479 (Cal. 1990).

89. See Hyman Gross, Privacy-Its Legal Protection 100 (1976).
90. See, e.g., E.W. Clayton, et al., Informed Consent for Genetic Research on Stored

Tissue Samples, 274 J. Am. Med. Ass'n 1786 (1995).
91. World Medical Association, Declaration of Helsinki: Ethical Principles for Medical

Research Involving Human Subjects, (2008) available at http://www.wma.netlelpolicy/b3.htm.
92. Research on identifiable human tissue is generally considered human subject research

(See, e.g., 45 CFR 46.102(f) and 45 C.F. R. 46.116). Variations on this construction do exist
and I examine them here as appropriate.

20091



IND. INT'L & COMP. L. REv.

storing that sample so that it may be analyzed at a later date either for a related
study or for a completely new purpose, which may be unforeseen at the time the
sample was taken.93 In order to secure a sample that can be used both for the
instant research and for unspecified future uses, the researcher must obtain
informed consent from the prospective participant. Herein lies the dilemma: it
is much debated whether or not an individual can actually give "informed"
consent to unspecified future uses since a participant would be consenting to
something about which he is not actually informed at all. There is little
resolution about the ethical nature of "blanket consent." However, there is
considerable discussion about the degree to which restrictions on blanket
consent hinder research efforts. 94

A second related issue involving informational privacy is that of the
identification and anonymization of data derived from HBM, as well as links to
other information about the donor, e.g. phenotypic information (observable
properties) or family history. Identification and anonymization in this context
unfold into a fairly complex set of configurations that is designed to enable the
researchers to link analyzed data to individual phenotypic characteristics
(possibly the optimal research resource scenario), while simultaneously
protecting the privacy of individuals. Simply put, samples can be identified
(labeled with a person's personal identifying data), coded (linkable to personal
identifying information not readily available to researchers), de-identified
(collected with identifiers, but subsequently stripped of all identifying
information and links) and anonymized (collected and stored with no
identifying information). Among the most central issues on which there is
variation in regulation are: 1) which form of identification constitutes human
subject research; 2) whether blanket consent can be given to use of identifiable
samples; and 3) whether new use of an identifiable (coded) sample requires re-
consent, and under what circumstances it can be exempted.

Another set of concerns relates to unforeseen subsequent use. The issue
of secondary uses is somewhat less controversial than the issues of consent.
Some policies limit secondary use to related research, yet a determination must
be made about what constitutes "related" research. This determination must
generally be made on a case-by-case basis. Additionally, uses by third parties is
a very important issue implicating both informational and decisional privacy,
particularly given the Directive's concern with the free flow of information
across borders. 95 If consent is given to a particular researcher, a determination
must be made about whether that consent extends to a third party who also

93. Philip Reilly, et al., Ethical Issues in Genetic Research: Disclosure and
Informed Consent, 15 Nat. Genetic 16-20 (1997).

94. See, e.g., Sandra Chandros Hull, et al., Genetic Research Involving Human Biological
Materials: A Need to Tailor Current Consent Forms., 26 IRB: Ethics & Hum. Res. 1, 1, 6
(2004).

95. See Council Directive 95/46, On the Protection of Individuals with Regard to the
Processing of Personal Data and On the Free Movement of Such Data, 1995 O.J. (L 281) 3 1.

[Vol. 19:1



COMPARATIVE ARCHrrECTURE OF GENETIC PRIVACY

wants to use the sample and data. Clearly, the reasons that may motivate a
person to permit one researcher to use her tissue and data may not apply at all to
an unknown third party. Various approaches have been taken in this regard,
with a common policy being to state this possibility of third-party transfer in the
consent form. There is an argument that when a new use of a sample is
proposed that the researcher should obtain the participant's re-consent. Matters
of practicality have resulted in different approaches to this issue, with perhaps
the greatest variation being in who makes the determination of whether an
exemption for impossibility or impracticability should be permitted. 96

Participants' post-consent control over samples is complex. Consistent
with basic principles of research ethics, a participant must be able to withdraw
from research at any time.97 Accordingly, a participant is generally allowed to
withdraw her sample at any time, up until it has become anonymized or until
the data derived from an individual's sample has been compiled with others.

Another related issue is that, given the nature of genetic research, analysis
of genetic data could reveal information that would be of beneficial interest to
the donor-participant. It is far from resolved whether a researcher should re-
contact a research participant to inform them of an incidental finding of a
genetic mutation that indicates the likelihood of serious disease with the
possibility for early intervention.98 This issue implicates an interesting aspect
of privacy in two ways: 1) the right not to know; and 2) the right to access
information being processed about oneself. There are both practical and ethical
reasons for not adopting a contact-and-inform99 approach. Difficulty of follow-
up and the need to respect persons' right not to know are just two of the critical
considerations.

Finally, another unsettled issue involves the handling and processing of
samples taken from minors. In many countries, a parent can consent to the
involvement of a minor child in research if certain conditions are met (usually
requiring the child's assent). 100 However, when a sample is taken from a minor,
some policies permit the indefinite retention of that sample without obtaining
consent from the donor when she reaches the age of majority. °10 While there
are legitimate issues of practicability, the implications for individual autonomy
and privacy are significant. Indeed, if such recruitment occurs on any
significant scale, a substantial collection of data could be obtained and stored
indefinitely without the consent of the participants.

96. See Research, supra note 25.
97. See, e.g., 45 CFR 46.116 et seq.
98. See E. W. Clayton, Incidental Findings in Genetics Research Using Archived DNA,

36 J.L. MED. & ETHics 286 (2008).
99. Id.

100. See Marina Cuttini, Proxy Informed Consent in Pediatric Research: A Review, 60
Early Hum. Dev. 89 (2000).

101. See, e.g., Partners Healthcare System Research Consent Form (2005), available at
http://healthcare.partners.org/phsirb/consfrm.htm (follow "Tissue Repository Consent Form"
hyperlink). [Hereinafter Healthcare].

2009]



IND. INT'L & COMP. L. REv.

The challenge presented by the ability to access genetic information that
provides information about probable future health states of individuals
complicates considerably the analysis of privacy implications and optimal
resolutions. Although there is general agreement on the governing principles,
the regulation of tissue use and storage is far from harmonized. Principles of
self-determination, decisional autonomy, and "control over information about
oneself' tend to drive many policy approaches. Yet, even within application of
these principles, there is a perhaps surprisingly wide berth for variation in
implementation. This Article isolates just a few of the issues involved in tissue
use and storage. I intend for these issues to serve as indicators of privacy
orientation, sensibilities, and how the architecture serves to create, enforce, or
fail to establish substantive boundaries.

7.2 National Approaches to the Regulation of Biobanking: Tissue
Use and Storage Issues

Biobanking regulation is still very much in flux. A major debate has
been underway in Norway between researchers and those charged with
regulating and overseeing human subject research. Countries like the U.K.,
Estonia, and Iceland have undertaken the establishment of national biobanks for
research purposes, sometimes facing considerable resistance.' 0 2 While other
countries have not (yet) initiated efforts to establish national biobanks, the
establishment of smaller-scale biobanks is undertaken more frequently and
involves some of the same issues. °3

Resolution of issues involving national databases impacts many of those
smaller-scale projects. One of the most controversial issues regarding informed
consent in national biobanks arose with the DeCode project. The Icelandic
Health Sector Database initially instituted a model of "presumed consent."
Presumed consent requires that the potential participant opt out of participation.
The default is inclusion and use of the sample, so that if the potential

participant does nothing, his sample and information are included in the
databank. Perhaps not surprisingly, this issue eventually became one of the
most contentious of the DeCode project and the model was ultimately modified.

Another of the more problematic issues was the blending of data/samples
collected in the health care or therapeutic context and data/samples collected
specifically for research.' 4 This distinction between residual tissue and tissue

102. S.B. Haga & L.M. Beskow, Ethical, Legal, and Social Implications of Biobanks for
Genetics Research, 60 Advances in Genetics 505 (2008). The DeCode project in Iceland moved
swiftly in the beginning, but as concerns mounted, the project was substantially slowed down
and forced to address many of the concerns. See Skuli Sigurdsson, Yin-Yang Genetics, Or the
HSD deCODE Controversy, 20 New Genetics and Soc'y, 103 (2001).

103. E.g. Jocelyn Kaiser, NIH Ponders Massive Biobank of Americans, 304 Sci. 1425,
1425 (2004); College of Medicine and First Genetic Trust Form Biobank, How. U. Capstone,
June 2, 2003, available at http://www.howard.edu/newsroom/capstone/2003/June/news2.htm.

104. Jacquelyn Ann K. Kegley, Challenges to Informed Consent, 5 EMBO Rep. 832,833

[Vol. 19:1



COMPARATIVE ARCHrIECTURE OF GENETIC PRIVACY

collected specifically for research is an important one in European regulatory
and ethical analysis. 105 Consent regulations frequently differ for the two types
of tissue collection and storage, typically with fewer restrictions on the
collection and storage of clinical samples. (There is some suggestion that the
norm in many European countries is to treat clinically-derived samples as
"abandoned."' 1 6)

The establishment of smaller-scale biobanks creates a number of
challenges for researchers, depending on the governing regulations. As
mentioned, one of the highest hurdles is that of protecting individual privacy
while facilitating legitimate and potentially beneficial research efforts. The
issues most heavily implicated in this attempted balance are: 1) the limitation
of use to the original research purpose; 2) blanket consent; 3) permissible
duration of storage; 4) right to withdraw one's sample after initial consent; and
5) re-consent for new uses. Below is a snapshot of policy positions on these
issues among two European Union member states, Norway, and the U.S.

Table I Tissue Use and Storage'0 7

Limited to Blanket Permissible (Re)Consent
Original Consent Duration of Storage Exception
Purpose To Un- approved by
w/o Specified

Explicit Future Use
Consent Permitted

France Yes Yes Necessary CNIL
Germany Yes No Only as long as Supervisory

necessary Auth 10 8

Norway Yes No Only as long as Dept, REC
necessary to achieve
purpose

United No Yes Indefinite REC
States I 1 771

7.3 European Union: Data Protection Directive

The provisions of the Directive suggest that their direct application to the
above-mentioned issues of tissue use and storage would result in the regulations

(2004).
105. Ben-Evert van Veen, Letter to the Editor, Human Tissue Bank Regulations, 24 Nature

Biotech. 496, 496 (2006).
106. See, e.g., Bartha Maria Knoppers, et al., Ethical Issues in International Collaborative

Research on the Human Genome: The HGP and the HGDP, 34 Genomics 272, 274-75 (1996).
107. Admittedly, this and other tables in this paper grossly oversimplify the nature of the

policies and otherwise fail to reflect the nuance of both the underlying reasoning and the
application. Nevertheless, as a summary, it serves as a useful point of departure.

108. See Bundesdatenschutzgesetz [Federal Data Protection Act], Jan. 1, 2002, § 38
(F.R.G).

109. Biobankloven [Biobank Law], 20 Juni 2008 nr. 44 § 13 (Nor.) [Hereinafter Biobank
Law].

2009]



IND. INT'L & COMP. L. REv.

contained in Table 1. Notably, with the exception of Norway (which amended
its policy effective in 2007), the policies suggested by the Directive are the
most restrictive. Article 6(1)(c) and Recital 28 require that member states
ensure that the processing of personal data is "adequate, relevant and not
excessive in relation to the purposes for which they are collected/further
processed." This would suggest that it would be impermissible to keep tissue
and data beyond the time necessary to accomplish the original research
purposes. Yet, France does not adopt this approach °. The architecture of
privacy rights in Europe can, in large part, answer why the policies of the
member states can be less restrictive. The derogations in Article 13 and Recital
29 affect the Directive provision prohibiting the processing of personal data
without the consent of the individual. The basis for different policy outcomes
are within the interpretation and applications of these derogations, particularly
the "public interest" and "historical, statistical, and scientific" purposes. If
application of a provision permitting blanket consent to unspecified future use
were permitted and consequently challenged in ECHR, the court would employ
the proportionality test to determine the legality of the act.

Consider the following scenario: Claude, a French citizen, voluntarily
participates in a diabetes study in which DNA samples are taken in addition to
other information, although genetic associations are not the primary focus of
this study. Claude, who may have perceived a direct benefit in the form of
close monitoring of his health, gave blanket consent to use of his health
information. Several years later the French research project sells the samples
with personal information to an Estonian research enterprise whose study
intends to explore a genetic link between diabetes and alcoholism. Claude
reads about the transfer in the newspaper and objects. His success in blocking
the transfer in France is not guaranteed. In a similar situation in Norway,
Claude would likely prevail. In the U.S., Claude would almost surely lose. I
examine resolution of this scenario below.

7.3.1 Specific Provisions Regarding Medical Research (Tissue Use)

7.3.1.1 France

The national legislation of France regarding data protection requires a
description of the purpose, population, and nature of data to be involved in the
study and processing.'11 This legislation explains some of the reasoning for the
determination that blanket consent may be permissible.

The French National Consultative Ethics Committee (CCNE) has taken
up the question of tissue storage and use in Opinion No. 77, "Ethical issues

110. Decree No. 2005-1309 of Oct. 20, 2005, supra note 48.
111. Decree No. 2005-1309 of Oct. 20, 2005, Journal Officiel de la R6publique Frangaise

[J.O.] [Official Gazette of France], Mar. 25, 2007, p. 16 (Fr.).

[Vol. 19:1



COMPARATIVE ARCHITECrURE OF GENETIC PRIVACY

raised by collections of biological materials and associated information data:
'biobanks' and 'biolibraries."' The CCNE has determined that blanket consent
to unspecified future use is acceptable if the sample is "scrupulously
anonymized." This blanket consent is allowable partly because of the
impracticality of re-contacting a participant, but primarily for the research value
of such tissue and data information.' 12 Basing its reasoning on a principle of
"solidarity," the CCNE states that "[tihe principle of solidarity would in this
case be ajustification for concessions regarding the rules observed to safeguard
individuals, but it would be true solidarity dependent on voluntary
decisions."

'1 13

The answer to this seeming contradiction can be found in the architecture
of French privacy rights. Claude's case against the transfer of his data via
blanket consent could conceivably go either way depending on a number of
factors. Article 39 of the French Data Processing Act requires consent. The
CCNE, while condoning blanket consent, does emphasize that the participation
must be voluntary. Claude's case becomes considerably stronger, however, if
the sample has been transferred with any identifying information. As the
CCNE has stated, blanket consent is acceptable only if the sample is
"scrupulously anonymized." Therefore, even with consent, Claude's sample
could not be transferred with any personally identifying information. The
Directive also requires that the receiving country have the equivalent
protections as the transferring country, or at least provide proof of adequate
protections. Here, Estonia, as a member of the EU that has ratified the
provisions of the Directive, meets the requisite criteria for a receiving country.
Thus, if Claude's sample is anonymized, its sale and transfer would be
considered legal by a French Court, even if the sample was linked to
(unidentifying) phenotypic information.

However, this case might be handled differently in the European Court of
Human Rights. The Directive states that storage of personal information is
limited to that period and purpose necessary to effect the original goal. The
court's application of the proportionality rule would determine: 1) if the
intended action can achieve its goal; 2) if the action is necessary to achieve the
goal; and 3) if the burden on the right is proportionate to the public interest
being privileged at the expense of the right. Here, the issue is whether the
burdening of Claude's right is proportional to the public interest being served at
the expense of his right. If the sample is anonymized, the right involved is only
that of decisional privacy. Although Claude has given blanket consent, he can
now claim that this use of the sample is not permissible, since Claude would
not have given permission to the use of his sample in a study connected to

112. Nat'l Consultative Ethics Comm. for Health and Life Scis., Opinion n° 77: Ethical
Issues Raised by Collections of Biological Material and Associated Information Data:
"Biobanks," "Biolibraries" 19 (2003) [hereinafter Ethical Issues Raised by Collections of
Biological Material], available at http://www.ccne-ethique.fr/docs/en/avis077.pdf.

113. Id. at 16.

2009]



IND. INT'L & COMP. L. REv.

genetics and alcoholism, especially if the study took place in a foreign country.
In this case the burden on decisional privacy may be viewed as either minimal
or considerable - minimal in that a completely anonymized sample cannot
cause Claude direct harm by virtue of disclosure of sensitive information;
considerable since Claude knows that he is now participating in research that he
does not support. His right to withdraw his sample has been lost since the
sample has been anonymized. Thus, while the harm is minimal, one could
argue that a dignitary violation has occurred from which the psychological
burden could be considerable. Therefore, although Claude's claim is likely to
lose in a French court, it is quite possible that he could prevail in the ECHR.

7.3.1.2 Norway

The Biobank Law largely governs tissue use and storage. 14  The
Norwegian law has been regarded as very restrictive as compared to its
European counterparts. 115 The problematic restrictions invalidate blanket
consent and serve to limit the involvement of incapacitated persons. In April
2006, the Norwegian government announced some important changes to the
Biobank and Patients Rights Laws 1 6 in response to a very public controversy.
The controversy concerned researchers decrying the degree of restrictiveness
regarding consent and the accompanying short and long term negative effects
on the conduct of biomedical research in Norway. 117 The changes to DeCode
will allow the collection of HBM from persons who do not have the capacity to
give consent, including in circumstances of emergency medicine, or in cases of
persons with physical or mental disturbances, dementia, or developmental
limitations.

Therefore, if Stein, a Norwegian citizen, also wishes to claim that his
sample may not be used for a purpose not related to that of the original study,
Stein need only show that the researchers acted in violation of the law.
Violations will be relatively easy to prove, as Norwegian law prohibits blanket
consent, limiting the storage and use of samples to the original purpose.

7.3.1.3 United States

Tissue use and storage in the United States has been something of a
maelstrom for the past five years. As recently as 2004, the annual convention
of research ethics committees had as its focus the regulation of tissue use and
storage, with part of the program designed to solicit input on what the policies

114. Biobank Law, supra note 109.
115. Kvalheim, supra note 57, at 291.
116. Otprp. nr. 64 (2005-2006) Om lov om endringar i pasientrettslova og biobankova

(helsehjelp og forsking - personar utansamtykkekompetanse), http://odin.dep.no/filarkiv/277845/
Otp0640506-TS.pdf (last visited Dec. 1, 2006).

117. Johan Votvik, Apner For Forskning Uten Samtykke, Helserevyen Online, Apr. 11,
2006, available at http://www.helserevyen.no/index2.asp?newsid=3797.

[Vol. 19:l



COMPARATIVE ARCHITECrURE OF GENETIC PRIVACY

should be. 18 Guidelines were issued in August 2004' 19 but were non-binding
and have been interpreted in inconsistent ways across IRBs. For example,
some IRBs permit both blanket consent and indefinite storage of identifiable
tissue of minors without attempts to consent upon the age of majority, 20 while
others restrict the use of blanket consent and limit storage of minor HBM until
the donor reaches the age of majority. While the European Directive permits
variation in the application of blanket consent, indefinite storage of the tissue of
minors seems well outside the scope of the permissible boundaries of European
law. Indefinite storage would occur after the original research purpose was
achieved, violating the law. Additionally, other aspects of the proportionality
test would likely fail. For example, the necessity component would seem to
present insurmountable challenges to this practice. Balancing the burden on the
right versus the goal to be achieved would likely also cut against indefinite
storage.

Consider the following scenario: Debbie, a U.S. citizen, has a family
history of alcoholism and objects strongly to the use of her sample in the study.
In the United States, although the National Bioethics Advisory Committee has
criticized blanket consent, there is no enforceable provision prohibiting it. The
terms of the permissible future use of Debbie's sample may be different
depending on the identifiability of the sample. However, even if the sample is
identified or identifiable, the transfer may still be permissible if the researcher
can show that re-consent is impossible or impracticable and the harm is
minimal. Thus, Debbie may only have a case if she is likely to suffer actual
harm from the further use of her sample and the approval of that use and
processing was negligent.

7.4 Unconsented Disclosure of Genetic Tests Results to Relatives

It is widely recognized that genetic test results may provide important
information for the person undergoing the test regarding his possible future
health state. However, the same test results may be of significant interest to
relatives of the tested individual. A test result indicating the presence of a
mutation associated with a hereditary disease can have enormous implications
for certain relatives, both in terms of future health risks and the possibility of
initiating interventions. Indeed, this issue highlights the tension between
individual privacy rights and the interests of others in a most dramatic and
profound way. To honor the proband's right to control the dissemination of
information about him is to disregard an opportunity to confer a potentially life-

118. Public Responsibility in Medicine and Research (PRIM&R), an organization whose
mission is to create, implement, and advance "the highest ethical standards in the conduct of
research." See http://www.primr.orgl.

119. Off. for Hum. Res. Protection, Guidance on Research Involving Coded Private Information
or Biological Specimens (2004), available at http'//www.hhs.gov/ohrp/humansubjects/guidancel
cdebiol.pdf.

120. See Healthcare, supra note 101.

20091



IND. INT'L & COMP. L. REv.

saving benefit on the relative. Strong arguments can be made on either side.
One pro-disclosure argument makes the point that disclosure to relatives does
not violate an individual's privacy since family members share genetic
information. The conflict between the individual's right to control information
about herself and a relative's interest in having access to information that could
potentially afford the opportunity for life-saving measures is one that has not
met an easy solution.

Consider the case of Grete, a Norwegian citizen, who has received a
positive indication of the presence of the BRCA mutation associated with
breast cancer. Knowing that at least one grandmother died from breast cancer,
she opted to take the test. Her doctors urge her to tell her three sisters and also
two daughters who are considering starting a family. Grete refuses, knowing
that her sisters will immediately have prophylactic surgery if they are informed.
Grete also does not want to have her daughters live their lives in fear of

developing the disease. She has been unable to return to her normal life and
deeply regrets having taken the test. The doctor feels strongly that Grete should
inform family members since interventions are available, and decides to inform
Grete's family members himself.

In Norway, France, and the United States, her doctor would be required
to honor Grete's wishes; and in France particularly, the doctor could be subject
to criminal penalties for breaching confidentiality. In Portugal 21 and Italy 122,
for example, the doctor might be allowed to disclose based on permissible
breach of confidentiality to save the lives of third parties. The physician is
prohibited from disclosure, 123 but may not be penalized for doing so if he can
show that it was necessary to save a life.

7.4.1 Background

In 1992 the Council of Europe took a position that straddled both sides.
While recognizing the need for confidentiality, its recommendation called for
consideration of disclosure to family members in the case of serious disease
risk. Principle 9 of Recommendation on Professional secrecy No. R (92) 3 on
Genetic Testing and Screening for Health Care Purposes stated that:

. . . in the case of a severe genetic risk for other family
members, consideration should be given, in accordance with
national legislation and professional rules of conduct, to
informing family members about matters relevant to their

121. See Helena Moniz, Privacy and Intra-Family Communication of Genetic Information,
21 L. Hum. Genome Rev. 103, 103-24 (2004).

122. See Article 29 Data Protection Working Party, supra note 37 at 4.
123. See Portugal, supra note 39.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

health or that of their future children. 124

This recommendation seems to imply that in certain situations, e.g. in the
case of severe genetic risk, that professional secrecy rnay be, and even ought to
be, breached. It is important to note, however, that this 1992 recommendation
predates the Directive on Data Protection.

Relevant provisions in the Directive could be interpreted to draw a clear
line against any unconsented disclosure of genetic test results to anyone.
Genetic test results clearly fall within the category of health data and, as such,
constitute sensitive information deserving of special protections. The Article
29 Data Protection Working Party also notes that genetic data can also be used
to contribute to the identification of a person's ethnic identity. 125 In the United
States, recent studies have claimed not only to be able to identify a person's
ethnicity or race, but also identify the degree of admixture. 126 The Working
Party makes the point that even though this information may not be health data,
as an indicator of race and ethnicity, this type of genetic data nevertheless falls
in the category of sensitive information 127 deserving of special level of
protection. 128 Thus, novel issues deriving from the implications of test results
revealing racial or ethnic information could also present problematic
complexities. For example, a test result that reveals admixed membership in an
ethnic group known to have a higher risk of a particular disease, e.g. breast
cancer in Ashkenazi Jews or sickle cell anemia in African-Americans, could
also have implications for the future health states of relatives.

In such an instance, the architecture of the protections may actually have
significant impact on both the substance and procedure of handling of such a
matter. By contrast, a European proportionality model could accommodate
such a situation. The Working Party currently advocates a case-by-case
approach that could allow disclosure in some instances but not in others, albeit
at some sacrifice of predictability and universal application. Because of the
practice of defensive medicine in litigious societies, discretionary disclosure is
likely to lose much of its discretionary quality. 129

One of the exceptions to the Directive's prohibition of processing
sensitive data is when such processing is required "for the purposes of

124. Comm. of Ministers, Council of Eur., Recommendation No. R (92) 3 on Genetic Testing
and Screening for Health Care Purposes (1992), http://wwwl.umn.edu/humanrtsfmstree/coerecr92-
3.html (last visited Jan. 9, 2009).

125. Article 29 Data Protection Working Party, supra note 37 at 4-5.
126. Mark D. Shriver, et al., Skin Pigmentation, Biographical Ancestry and Admixture

Mapping, 112 Hum. Genetics 387, 387 (2003).
127. Directive, supra note 13, at Art. 8(1).
128. This issue is further complicated by predictions in the scientific literature claiming

that it may soon be possible to draw links between degree of ancestral admixture (suggesting
racial and ethnic identity) and disease predisposition.

129. Khadija Robin Pierce, Setting Margins for Genetic Privacy (June 2007) (unpublished
Ph.D. dissertation, Harvard University) (on file with the Harvard Kennedy School Library).

20091



IND. INT'L & COMP. L. REV.

preventive medicine, medical diagnosis, the provision of care or treatment or
the management of health care services." This exception is subject to national
codes of professional confidentiality and other relevant provisions and
safeguards. 130 Therefore, as the Working Party points out, in many European
countries, an individual's genetic data could conceivably be processed under
one of the exceptions of Article 8(3).

7.4.2 The Finality Principle

The Article 29 Working Party takes an interesting approach to the
applicability of the Data Protection Directive in the reference to Article 6(1 )(b)
and (c). The Working Party states that implementing national legislation must
provide that personal data may only be "collected for specified, explicit and
legitimate purposes and not further processed in a way incompatible with
purposes."' 13 1 The Working Party adds that personal data must be "adequate,
relevant and not excessive in relation to the purposes for which they are
collected and/or further processed."'' 32 The Working Party subtitles the latter
segment of the provision as the "proportionality principle" and the former, the
"finality principle." 

33

Application of the "finality" rule to the question of unconsented
disclosure suggests that test results collected for the specific purpose of
diagnostics of the individual submitting to the test should not be further
processed for some other purpose. That the Directive specifically prohibits
processing that is incompatible with the original purpose begs the question of
whether disclosure to the proband's family constitutes incompatible processing.
Moreover, even the permitted processing for purposes of various health care

services as set forth in Article 8(3) renders such processing subject to norms of
professional secrecy. Yet, a plausible rationale for disclosure may be that
members of a proband' s biological family share this information and, as such,
may be considered "data subjects" with the attendant rights. 34 Though
plausible, this argument is unpersuasive given that the relatives have neither
consented to the test nor voluntarily assumed any risks associated with taking
the test. Such a convenient but flawed construction of their status in order to
confer rights rapidly unravels. The alternative rationale put forth by the
Working Party 135 that family members could assert a right based on the
potential effect on their personal interests is more convincing. However, this
reasoning currently has no legal basis for superseding the rights of the tested
individual. 136 The Working Party concludes that the complexity of the balance

130. Directive, supra note 13, at art. 8(3).
131. Id. at art. 6(1)(b).
132. Id. at art. 6(l)(c).
133. Article 29 Data Protection Working Party, supra note 37, at 5-6.
134. Id. at 8.
135. Article 29 Data Protection Working Parry, supra note 37 at 7.
136. The Working Document on Genetic Data does reference an Italian case in which a

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

of interests and rights is such that no clear resolution has yet emerged and that
"at this stage" a case-by-case approach should be considered.

It is not entirely clear what the implications are for resolving the conflicts
inherent in the unconsented disclosure of genetic test results to relatives.
Arguments asserting that a new legally relevant social group has emerged in the
biologic family 137 have not found widespread acceptance. Many European
countries have adopted a policy that disclosure is desirable in certain specified
circumstances, but that rules of confidentiality and privacy prohibit such
disclosure where the patient has not consented. However, the option to
consider the weight of countervailing interests would permit an individual act
of discretionary disclosure. See Table 2 below.

Table 2 Unconsented Disclosure of Genetic Test Results to
Relatives

Absolute Confidentiality
European Union (WP) Unsettled-case-by-case
France Generally No-case by case
Norway Yes
United States Yes
Portugal1 38  Generally No--case by case

7.5 Approaches and Architecture

The French National Consultative Ethics Committee for Health and Life
Sciences (CCNE) has issued an opinion on the question of disclosure of test
results to relatives. The opinion takes a very careful approach in which it
weighs the nature of the patient's interests and those of the family, finding
merit in both. However, as with disclosure of HIV status,' 39 CCNE holds fast
to the principle of absolute confidentiality. Opinion 76 carefully outlines
procedures to prepare a proband for disclosing to family members, but it stops
short of taking the decision out of the hands of the tested individual, stating that
"a strict application of the principle of medical confidentiality, as well as
utilitarian attitudes, argue against systematic breaching of confidentiality."''14

Despite the unsettled nature of the balance between individual rights and

woman sought disclosure of her father's genetic data so that she could make "a fully informed
reproductive decision." The relevant authorities granted the woman's request despite her
father's refusal to consent, stating that the "father's right to privacy was to be overridden by the
daughter's right to health - the latter meaning her 'psychological and physical well-being."' Id.
at 9.

137. Id.
138. Portugal is included here for purposes of illustration as a country that does not

distinguish between genetic and other medical information in its privacy protections.
139. Nat'l Consultative Ethics Comm. for Health and Life Scis., Opinion n°76: Regarding

the Obligation to Disclose Genetic Information of Concern to Family Members in the Event of
Medical Necessity 6 (2003), http:/Iwww.ccne-ethique.fr/docs/en/avisO76.pdf (last visited Jan. 9,
2009).

140. Id.

2009]



IND. INT'L & COMP. L. REv.

the interests of biologic family members, the legal order seems to favor an
individual rights model that allows persons to control the flow of information
about themselves. Application of a proportionality approach may be seen in
some instances, e.g. Sweden, where disclosure is discretionary under specified
circumstances if the benefit outweighs the potential harm to the patient and that
harm is minimal.14 1 In Portugal, genetic data, like other medical information,
can be disclosed without consent if the benefit will outweigh the harm and the
harm is minimal. 42 Sweden and Portugal implemented the Data Protection
Directive in 2003 and 1998, respectively.143  These widely divergent
implementations of the Directive speak to several aspects of the dilemma. First,
the categories justifying the overriding of the privacy protections are flexible in
their interpretation. Secondly, the issue is very complex with strong competing
interests, both of which can find support in the wording of the Directive.
Thirdly, the categories of derogation are constructed such that national cultural,
legal, and professional norms may resolve the dilemma in a manner most
suitable to the national context.

7.5.1 Analogy: HIV Notification

The European Court of Human Rights has not taken up the issue of
unconsented disclosure to genetic test results, but one of the few cases
involving medical secrecy, Z versus Finland,144 decided by the European Court
of Human Rights, may provide some insight on the court's likely position.
Given the paucity of case law in this area in Europe, this analysis is well-served
by examination of analogous issues. In Finland, a husband stood accused of
rape. The HIV status of the accused's wife was disclosed, without the wife's
consent, to the alleged rape victim. This disclosure was undertaken for the
benefit of criminal proceedings against the husband. The wife filed suit,
alleging that this unauthorized disclosure of her HIV status to the alleged rape
victim constituted unauthorized disclosure of confidential personal data in
violation of Article 8 of the Convention. While the court acknowledged the
highly sensitive nature of HIV status, it reasoned that an overriding public
interest could justify unauthorized disclosure, and thus the burdening of a right
to special protection of sensitive information. 145 Here, striking the balance
between protecting sensitive information and facilitating the public interest was
a complex one. This complexity was due in part to the nature of the
information and the weight of that privacy interest, but also to the context in

141. Lag (1998:544) om virdregister [Health Data Law] (Swed.).
142. Eur. Comm'n, Genetic Testing: Patient's Rights: Insurance and Employment: ASurveyof

Regulations in the European Union 100 (2002), ftp://ftp.cordis.europa.eu/pub/life/docs/
genetic_testingeur20446.pdf (last visited Jan. 9, 2009).

143. See Report on the Implementation of the Data Protection Directive 95/46/EC
http://ec.europa.eu/justice-home/fsj/privacy/lawreport/index_en.htm.

144. Z v. Finland, App. No. 22009/93, 25 Eur. Ct. H.R. (1997).
145. Id.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

which the information was sought and the possibility that additional judgments
(such as how important the information was to the investigation) would have to
be made. 146

The court in Finland applied a version of the proportionality test in which
the burdened right of privacy was outweighed by the state's interest in resolving
the criminal investigation.1 47 It was not just that the public interest in resolving
a criminal investigation was sufficient to override protection of this sensitive
information, but that the probative value of this information was significant
enough to justify the override. Therefore, it is unlikely that sensitive
information that would merely assist in the resolution of a criminal proceeding
would be subject to disclosure. Rather, that information, albeit by local
standards, must have high probative value, and presumably, must not be
obtainable by less burdensome means.

This case may have significant implications for unconsented disclosure of
genetic test results to relatives. Like HIV status, information about a genetic
disease may be probative regarding someone other than the individual who has
been tested. Finland suggests that highly sensitive information may lose its
protection in the face of its value to a criminal proceeding going to the question
of manslaughter.1 48 However, because of the mode of transmission of a genetic
disease, it is unlikely that an action of manslaughter would ever lie based on the
knowing transmission of a genetic disease. However, if a scenario could be
constructed in which the genetic information of a relative were probative in a
criminal trial, it is possible that this might create the outer bounds for
permissible disclosure of highly sensitive information.

Still, one can imagine a situation in which a breadwinner knows that he
carries the genetic mutation associated with Huntington's disease, a
neurodegenerative disorder. He takes out several major unsecured loans and
gives the money to his girlfriend. These gifts result in a criminal proceeding
against the girlfriend in which the breadwinner's genetic test results (indicating
the presence of the mutation associated with Huntington's disease, which has
roughly 99% penetrance) are relevant and highly probative. Finland would
suggest that the privacy interests of these test results would be outweighed by
the public interest.149  Furthermore, here, unlike in Finland, the tested
individual is also culpable in some way even though he may not at the time of
trial be alive or competent.

By contrast, the process of adjudication of the case of Finland in the
United States would be very different, largely because of the architecture of the
privacy protections. As noted earlier, HIPAA provides for the unauthorized
disclosure of personal health information in certain circumstances that are

146. See e.g., id.
147. Id.
148. Id.
149. Id.

2009]



IND. INT'L & COMP. L. REv.

outlined in section 164.512 of the Code of Federal Regulations. 50 This list
includes disclosures for judicial and administrative proceedings as well as for
law enforcement purposes.' 5' Thus, even though these exceptions have their
parameters, the exceptions have been carved out in advance, presumably as a
result of a balancing between the weight of the privacy right and the
countervailing public interest. 52 The result in this case is probably the same,
but one could easily imagine a less clear case in which the PHI of an innocent
third party could be disclosed simply by virtue of one of the exceptions;
whereas in Europe, a balancing of the rights and interests could conceivably
result in a different outcome regarding the permissibility of disclosure. In
recent years, a few courts have authorized the subpoena of medical records and
other personal data (believed to be of a genetic nature) of the parents of child
plaintiffs in lead paint cases, thus allowing a defendant to pursue a defense
alleging that the low I.Q. of the plaintiffs was inherited and not the result of
lead poisoning.1 53 The architecture of the rights, the transparent balancing of
the proportionality rule versus the "absolute" articulation with carved out
exceptions (e.g. relevance) can, indeed, affect the degree of privacy protection.

The more likely scenario is that involving a civil action. At least one
such case has been litigated at the national level in Europe. As noted earlier,
the Data Protection Article 29 Working Party mentions an Italian case in which
a daughter seeks disclosure of her father's genetic test results to assist her in
reproductive decision-making. 54 The court determined that the father's test
results should be disclosed to the daughter even against the wishes of the
father. 55 How ECHR would resolve this case would probably turn on such
factors as the gravity of the heritable disease and the extent to which disclosure
burdens the father's privacy interest. However, based on the Working Party
recommendations, it seems unlikely at this time that the court would find an
overriding interest in preventing a speculative genetic disease. Furthermore,
since the override would require a determination that a less burdensome
alternative was not available, i.e. that infringement on the right was necessary,
the fact that individuals can seek out a genetic test themselves may preclude the
authorization of disclosure. This preclusion should at least be the case in
instances where gatekeeping arrangements do not rely on the known presence
of the disease or mutation in a relevant biological relative. However, in the
case of autosomal recessive diseases, genetic testing of the parents may be
inadequate to inform the degree of risk.

It is important to note that case law, as a source of rights, does not play

150. 45 C.F.R. § 164.512 (2008).
151. Id.
152. Id.
153. See, e.g., State v. Idellfonso-Diaz, Tenn. Crim. App. 2006, 2006 WL 3093207

(unpublished).
154. Article 29 Data Protection Working Party, supra note 37, at 9.
155. Id.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

the same role in Europe as it does in the United States. This is particularly true
in the area of health privacy law. Health privacy law tends to be statutory and
regulatory. To a lesser degree this is also true of U.S. medical privacy. Doctor-
patient confidentiality has a line of case law beginning as early as 1920.156

The issue of disclosure of genetic test results has appeared in American
courts. Interestingly, U.S. case law has demonstrated a leaning toward
embracing the importance of disclosure to relatives, but has stopped shy of
imposing a duty. The closest the court has come is in Safer v. Estate of Peck. 157

In Safer, the court found that the physician did owe a duty to the daughter of
his patient who had a hereditary disease; however, because the daughter was
put on notice by other means, the physician was not held liable.158 The
important thing about these cases, of which there are very few, is that the
closest the courts have been willing to come is in the case of vertical disclosure
- disclosure to an offspring. No case has embraced horizontal unconsented
disclosure to siblings or other related non-offspring. Consequently, if the courts
move toward unconsented disclosure of genetic test results to relatives, it is
likely to be in the case of offspring wanting to avert serious illness or death or,
more likely, to assist in reproductive decisions. 159

Nevertheless, there are strong reasons why ECHR would hesitate to take
this step even in the cases set forth above involving offspring. One of the most
important aspects of privacy law is its relationship to cultural and social norms.
Unconsented disclosure of a hereditary disease to other family members could

be disastrous in some cultures and situations. A decision by ECHR that such
disclosure is permissible, and thereby implicitly creates a duty, would mean that
all national laws would have to be consistent. Thus, a Norwegian or German
court could not deny disclosure to a son or daughter wishing to know the
genetic test results of a parent in certain circumstances. Such a decision, while
possibly saving lives, may serve to unravel fundamental aspects of social
relations and cultural norms. It is not clear that the EU, which was originally
created to facilitate trade and pursuit of economic interests, is ready to take this
step. Furthermore, permitting unconsented disclosure to relatives further erodes
protection of sensitive data.

Finally, it can be argued that this private sphere may be more
appropriately regulated by cultural and social norms. The pressure imposed by
the norms may be sufficient to achieve the described result in most instances.
Additionally, it may be excessive and unnecessary to impose a legal burden on
the privacy right in order to force disclosure in the rare cases where it would not

156. Simonsen v. Swenson, 177 N.W. 831, 832 (Neb. 1920) (statutory doctor-patient
privilege must be subject to a public health exception).

157. Safer v. Estate of Pack, 677 A.2d 1188, 1192 (N.J. Super Ct. App. Div. 1996).
158. Id.
159. This would also be consistent with principles of the duty to rescue between a parent

and a minor child where such statutes are in force (e.g. Vermont).

20091



IND. INT'L & COMP. L. REv.

otherwise occur.'60

8. Conclusion: Measuring the Force of Privacy Protections:
Architecture or Substantive Law?

8.1 France

The story told about France by the analysis of these two medical
indicators is one of interesting paradoxes, but it is consistent with a host of
social and political contextual aspects. We can deduce three key points from
the foregoing analysis.

First, the permissibility of blanket consent for future unspecified uses of
tissue may suggest a more permissive research climate regarding protections.
The architecture of French privacy rights probably contributes to this result by
virtue of the heavy reliance on CNIL as an oversight body charged with
considerable active and practical regulatory authority and with the
administration of privacy protections. Furthermore, the French Data Processing
Act looks to the cultural climate of "solidarity" to embrace a risk-benefit
calculation when determining whether to allow the processing of sensitive
data. 161 This embrace suggests that the processing of sensitive data is allowable
because of the use of strong oversight.

Second, the use of a risk-benefit analysis in the case of incapacitated
persons rather than a requirement of direct benefit may be inconsistent with a
policy of absolute confidentiality honoring the proband's right to control the
flow of information about himself. However, as openly stated by the French
national ethics committee, a strong value of "solidarity" and the "public
interest" may be behind this policy. The policy of absolute confidentiality of
genetic test results helps to define the limits of "group consciousness."

Third, consent is exempted when it is impracticable or impossible to
obtain; however, saving the life of a relative does not justify unconsented
disclosure of genetic test results. Interestingly, this could suggest that the right
of privacy receives the greatest protection when it is exercised.

8.2 Norway

The picture we get about privacy protections in Norway shows a strong
allegiance to individual rights and autonomy. This may seem surprising, given
that Norway has been a highly functioning social democratic welfare state for
several decades. However, strong protections of individual rights permeate
Norwegian law.

First, Norway's previous prohibition of the involvement of incapacitated

160. Pierce, supra note 129.
161. See Article 29 Data Protection Working Party, supra note 37.

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

persons in medical research is that of a tiny minority in the western world. This
approach departs from two of the other leading views: 1) research that provides
a direct benefit to the incapacitated person is permissible; and 2) the European
Union's risk-benefit calculation. However, as mentioned, this law was changed
in 2006. Second, the disallowance of blanket consent and the limiting of tissue
use to those uses necessary for the original purpose for which it was collected
both depart from the European approach. Norway's departure represents a firm
commitment to individual autonomy and informed consent. The policy against
unconsented disclosure of genetic test results is also consistent with this
commitment, even in its convoluted articulation. Third, the approval of any
consent exemptions for impracticability or impossibility must come directly
from a government entity, consistent with the former Norwegian approach to
privacy that relied heavily on government oversight and protection rather than
the model of individual consent.1 62

Health privacy practices, the scant case law, and analogous rulings in
Norway reveal a strong national commitment to the protection of individual
rights even in the face of strong countervailing societal or third party benefit.
Recent cases involving freedom of speech have upheld the right of individuals
to engage in hate speech, receiving considerable criticism from the rest of
Europe. 163 Synne Sather Mahle, a legal scholar on judicial review, has
explained that the rigidity of the Norwegian Constitution and the privileging of
that document over European texts help to explain this result. 164 There are
indications that Norwegians may be unwilling to open the door to derogations
of the freedom of speech. 165

Indeed, the rationale for changing the law regarding participation in
research by persons with reduced capacity to consent was ostensibly based on
the fact that the former degree of restrictiveness made it impossible to include
such persons in research. Not including such persons in research hindered the
ability to investigate methods and interventions that may improve the services
that could be offered to them. 166 Thus, arguably it was not societal benefit that
served to outweigh individual privacy rights, but a move towards better serving
individual rights by providing a direct benefit.

Thus, it seems reasonable to conclude that the approach to privacy in the
biomedical context, even with the recent changes, remains on the restrictive end
of the European continuum. The Norwegian human subject research
regulations, which incorporate by reference the Patients Rights Law, were
recently legalized. 167 (They had "guidance" status up until December 2005.)

162. Kvalheim, supra note 57, at 289-290.
163. See Hvit Valgallianse-kjennelsen, Rt. 1997 s. 1821, Hoysterett. 2004; and Boot Boys-

dommen. Rt. 2002 s. 1618.
164. Interview with Synne Sether Mxhle in Bergen, Norway (Dec. 2005).
165. See Hvit Valgallianse-kjennelsen and Boot Boys-dommen, supra note 163.
166. Johan Votvik, supra note 117.
167. See De nasjonale forskningsetiske komiteer [Norwegian Regional Ethics Committee],

http://www.etikkom.no/English/NEM/REK (last visited Jan. 9, 2009).

2009]



IND. INT'L & COMP. L. REv.

These regulations provide that "historical, statistical, and scientific" reasons, as
well as "important societal benefit" can exempt data from the data protection
provisions. Thus, the balancing does occur in Norwegian decisions involving a
burden on individual privacy, but the scale seems to tip in favor of individual
rights. However, the scale does not tip in favor of individual rights unless two
elements are present: 1) a clear and convincing need to burden the right in
order to achieve an important benefit, as in the case of permitting research on
persons who do not have the capacity to consent; and 2) the state, through some
responsible agency, can provide effective oversight, as was the case before its
adoption of the individual consent model upon the passing of the Data
Protection Directive.

That a change in this regard has been legislated may represent a move
toward a more permissive approach in support of the biomedical research
enterprise. In a climate in which public hearings are being held to discuss
whether the ban on stem cell research should be lifted, it may well be that
Norway is taking steps toward the research imperative. Arising perhaps from
the earlier "oversight model"'' 68 of protection, it is possible that the model of
consent with relaxed application may be sufficient to assure society that the
research enterprise is not necessarily at odds with individual rights.

8.3 United States

The reality of privacy in the United States, as constructed by the analysis
of the three medical indicators, puts forward a sectoral model containing
absolute privacy protections with carved out exceptions, along with a heavy
reliance on individual consent. This may actually result in fewer privacy
protections than the seemingly more elastic proportional approach taken in the
European Union.

First, the permissibility of blanket consent for unspecified future uses
suggests a "market model" of research participation. There is no overarching
oversight body with administrative and enforcement powers, thus leaving the
injured plaintiff to file a private tort action as his primary recourse. Unlike in
France, the permissibility of blanket consent is not accompanied by a strong
oversight body, but rather is left to application and monitoring by individual
research ethics committees that employ different standards and policies to
review in-house research proposals.

Similarly, that samples/data may be stored indefinitely if the participant
has given consent suggests a very strong reliance on individual consent as the
mechanism for privacy protections. Official guidance suggests that blanket
consent is ethically problematic;' 69 yet, to date, this issue remains largely
without enforceable regulation.

168. See Kvalheim, supra note 57.
169. Nat'l Bioethics Advisory Comm'n, Research Involving Human Biological Materials:

Ethical Issues and Policy Guidance, 1, at 33 (1999).

[Vol. 19:1



COMPARATIVE ARCHITECTURE OF GENETIC PRIVACY

Third, the requirement that incapacitated persons may participate in
medical research only if they will directly benefit from the research suggests a
boundary that may be more a matter of historical sensitivity to vulnerable
populations than a clear commitment to privacy boundaries. Finally, where the
sectoral legislation and exceptions do not specifically address issues arising
from new technology, there is no reliable protection of privacy rights.

9. Concluding Observations

There are reasons to believe that architecture does matter in the force of
privacy protections. In the United States, a stringent prohibition of physician
disclosure of genetic test results to relatives does not allow for breach in even
the most compelling of circumstances. To permit discretionary disclosure by
the physician would require a legislative act carving out an exception (even in
specified circumstances) or a high court ruling rendering such disclosure
permissible. By contrast, in Europe, even legislative provisions prohibiting
discretionary disclosure can be overridden on the basis of the weight of the
countervailing interests. However, this same flexibility allows a country like
Norway to maintain a policy of strict confidentiality and countries like Sweden
and France to decide on a case-by-case basis. Thus, the architecture does
matter.

The case of human biological material exemplifies this well. Policies
allowing permissive use of HBM, from blanket consent to indefinite storage of
a minor's tissue have determined that the balance weighs in favor of competing
research interests against individual privacy interests. In contrast, in Europe,
the fundamental right of privacy can only be overridden when countervailing
interests are found to be greater, thus maintaining the essential privacy right
while allowing for circumscribed exceptions with adequate oversight, as in the
case of France. That Norway tends to maintain the privacy line speaks to the
ability of the European model to accommodate national norms. This leaves us
with the conclusion that national law and norms determine the strength of
privacy protections in Europe, but that this is enabled by the architecture.
Furthermore, the regard for privacy as a fundamental right may be what
ultimately sets the margins for privacy.170  Thus, while the architecture
accommodates a range of policy positions, from the very restrictive to the
permissive, it is the regard for privacy as a fundamental right that ensures that
limits remain regarding how much this essential right may be burdened by
countervailing interests.

170. The phrase "sets the margins for privacy" comes from the title of Khadija Robin
Pierce's Ph.D. dissertation, Setting Margins for Genetic Privacy. Pierce, supra note 129.

2009]