Français | Government of Canada BioPortal | |||||||||||
Home | Site Map | News Room | FAQ | Search | ||||||||
|
Home Publications Project Reports 2004 Protecting Privacy in the Age of Genetic InformationOf Volume, Depth and Speed: The Challenges of Genetic InformationAuthors: Trudo Lemmens and Lisa Austin Table of Contents1. Introduction 2. What is Genetic Information? 3. Characteristics of Genetic Information: The Claim of Genetic Exceptionalism 4. How Genetics Highlights Existing Problems 5. Conclusion 1. IntroductionThe futuristic movie GATTACA pictures the struggle of a 'genetic proletarian,'1 born out of a stubborn mother who refused to abort her genetically inferior fetus, towards the fulfilment of his life-dream of becoming a cosmonaut. In order to do so, he has to cheat the genetically monitored GATTACA society that performs regular DNA testing to systematically screen out people of his kind from any reasonable job and from insurance coverage. One of the scenes of the movie portrays the activities of a local "gene shop." Clients of this shop have a mouth swab to recuperate DNA traces of the person they just dated and kissed. The gene shop conducts a computerized DNA analysis on the spot, providing a summary genetic portrait of the potential partner they just 'caught.' On a one-page summary, clients get basic information about the person's behavioral traits, life-expectancy and potential progeny. Obviously, a simple mouth swab and DNA analysis based on saliva currently cannot give us such detailed information and certainly not in a time span of three minutes, as suggested in the movie. Also, several scientific hurdles have to be overcome before we will be able to conduct simultaneous, and affordable, tests for a variety of conditions and traits. But even though the gene-shop scene is very much a caricature of genetic testing, it provides a strong metaphor both for what type of issues are likely to be raised by genetic testing and for why we have to be concerned about the social consequences of unbridled use by third parties. Research into the development of DNA chip and microarray technology is already taking place and in the near future will likely allow us to scan entire genes for the detection of a variety of mutations. Genetic tools will become faster, more efficient, and cheaper.2 In this paper, we will argue that the combination of the following three elements constitute the primary reason why we have to develop appropriate regulatory measures or adapt existing ones:
These are the main reasons, we argue, why genetic testing, if inappropriately used, can have detrimental social consequences. Other characteristics have been identified as making genetic information 'unique', but we will argue that many other types of health information share these characteristics. However, when combined with the three factors we identify, many of our traditional concerns regarding health information are augmented. In other words, the concerns raised by the advent of genetic testing are related more to what one can call an amplification of existing concerns about the use of health information than to the specificity of genetics. It is a matter of degree, or depth, more than a matter of newness. But even if these concerns are not in themselves new, the new contexts in which they are raised may require different types of responses, or additional responses, than those pertaining to more traditional health information. This paper aims at identifying the relative specificity of genetic information and analyzing the arguments invoked to support specific regulation and legislation that singles out genetics. 2. What is genetic information?It is difficult to analyze the use of genetic information in a comprehensive manner without treating genetic information as a one-dimensional concept. However, one always has to keep in mind that genetic information can be obtained in a variety of ways and can refer to very different forms of health information. The genetic information referred to in this paper is generally the information resulting from genetic research undertaken with new genetic technology developed in the last decades and that has led to the identification of specific associations between genes and genetic diseases and traits. It is this relatively new form of genetic information that much of the debate about the potential negative social impact of genetic testing focuses on. But the term genetic information also includes family history of disease, information from chromosomal testing and data gathered from twin studies, for example, all which have been used in research and in health care for the most part of the last century without receiving the same attention. In the debate over what constitutes genetic information, some even point out that all health information is to some extent genetic.3 Any type of regulation or legislation developed in the context of genetics will have to be attentive to the problem of defining what constitutes genetic information and how one can distinguish in fact, and as a matter of principle, the different types of genetic data. In order to highlight this problem, we will enumerate here some of the major differences between types of genetic information arising from the type of testing, the purpose and manner of collection of the information or sample, and the method of storage. 2.1 DNA TestingIn most current discussions, genetic information is understood as information resulting from the analysis of an individual's DNA. Startling developments in molecular genetics and DNA technology (closely linked also to developments in computer technology) over the last decades are directly responsible for what has been termed 'the genetic revolution'. When people talk about 'genetic information,' they are most likely thinking of information derived from the use of this new technology. Through the use of a variety of techniques such as electrophoresis, somatic cell hybridization, cytogenetic mapping, multiplexing, and radiation-induced breakage of chromosomes, scientists have been able to make physical maps of the human genome. The physical maps portray the position, size, order and numbering of base pairs in the different genes. Comparison of the maps of different people allows researchers to find specific mutations associated with genetic conditions or traits. Even when the mutation directly related to a genetic condition has not been identified, DNA techniques can be used to find markers for the disease. Markers are characteristic DNA sequences that enable scientists to determine whether a mutation present in that DNA region has been inherited or not. A variety of tests have been developed on the basis of these techniques. Currently available DNA-based genetic tests include tests for: Amyotrophic lateral sclerosis (Lou Gehrig's disease), Alzheimer disease, ataxia telangiectasia, inherited breast and ovarian cancer, Cystic Fibrosis, Duchenne muscular dystrophy, fragile X syndrome, Huntington's disease, myotonic dystrophy, sickle cell disease, thalassemia, Tay-Sachs disease and many other conditions.4 2.2 Indirect Genetic TestingAlthough some of the most spectacular advances in medicine have been obtained by DNA analysis, other forms of testing can clearly be identified as 'genetic tests.' The identification of phenotypic characteristics associated with genetic conditions such as cleft palate or Spina Bifida, for example, is a form of genetic testing. Testing can also occur at the chromosomal level. Chromosomal abnormalities can be detected, for example, through amniocentesis. Other forms of 'genetic tests' involve the testing of urine, blood or other body fluids to discover abnormal metabolite levels that are indicators of genetic disorders such as phenylketonuria (by measurement of phenylalanine in blood) or Lesch-Nyhan disease (by identification of high urinary uric acid levels). Finally, genetic disorders can be detected through measuring proteins, which are the products of genes. Defective genes often lead to identifiable deficiencies in protein production. The observation of mutant proteins can be used as a measurement to determine the presence of a genetic condition such as Tay-Sachs. 2.3 Family HistoryFor a very long time, people have been aware of the fact that diseases are 'running in families' and have been involved in studying the familial character of diseases. Not only family physicians, but also interested third parties such as insurance companies have been aware of this and have been collecting information on people's family history of disease. Indeed, from time immemorial people have talked about people having a disease 'running in the family.' The history of behavioral genetics contains a remarkable example of a lay person's contribution to genetic research. The first association of a particular gene with a tendency to violence was established with crucial help from detailed records of a Dutch family's history of crime and violence, kept by one member of that family. Clearly, family history of disease is genetic information that can lead to the identification of 'at risk families', in which all members are identified at increased risk for developing certain conditions. For example, breast cancer, Huntington's disease, Tay Sachs, and some mental illnesses, have all been 'running in families' and people identified as members of these families have both benefited from knowing this (e.g. for making life choices, taking preventive action if possible, improved monitoring) as well as been harmed by it (e.g. being excluded from insurance, discriminated in employment, being stigmatized as members of diseased families, suffering emotionally). 2.4 Differentiation of Genetic Testing According to its Health Care Purpose and its TimingGenetic testing can be differentiated according to why it is used in health care and at what stage.
2.5 Identification Purposes: Forensic DNA and Military DNA banksDNA is now widely used to identify tissue samples such as hair, skin particles, blood and so on, left at the scene of a crime or found attached to clothing, vehicles or other instruments used by potential suspects. The technique used to match the DNA of these samples with the DNA of identified suspects or victims is different from DNA sequencing undertaken for health care purposes. It is also unlikely that tissue or samples from a crime scene could be used to identify genetic traits or conditions. The way these samples are collected, and the often minimal amount of usable DNA that is discovered in this way makes these samples unfit for uses other than identification. Issues raised by the use of forensic samples are therefore often very specifically connected to criminal law and evidence. Nevertheless, while crime scene samples may not be fit for uses other than mere identification, law enforcement agencies have started to establish DNA banks of convicted offenders, missing persons and unsolved cases as well as population frequency databases for comparison.12 Samples in these databases, collected in more clinically reliable circumstances, could be used for further testing. For the same purposes of potential identification, the United States military is now one of the largest collectors of DNA samples. 2.6 Existing Samples that were Collected for Other PurposesAlthough genetic testing is often conducted on samples that are collected for the purpose of a specified test, genetic information can also be extracted from sources that were not provided with that aim. A couple of different scenarios are worth pointing out:
2.7 Genetic Information Distinguished According to the Way it is StoredGenetic information can also be distinguished on the basis of the way it is kept and expressed. As already suggested, genetic information can be contained in a blood sample, which then has to be further analyzed in order to release any of its secrets. Forensic DNA can be retraceable from objects, tissue or hair samples collected at crime scenes and stored by law enforcement agencies. Genetic information such as gene sequences can be available on paper or on computer files. It can be kept as a printout of a strip of DNA. Or it can be written out in the format of the sequences containing the four letters representing the chemical compounds that make up DNA. A family linkage study can be expressed in the form of a family diagram. Finally, results of genetic tests can be written down in medical files. 2.8 Which Definition?We cannot expand here on the relevance of these distinctions to the various debates on ethical, legal and social issues raised by genetic testing. But as has been argued elsewhere in more detail, regulation that defines genetic information narrowly - as information resulting from DNA analysis - may lack coherence and become unenforceable as a result of the problem of defining what falls under the regulation.15 Privacy legislation, legislation aiming at curbing genetic discrimination, or other regulatory interventions in the context of genetic testing will have to be carefully drafted so that a narrow scope does not undermine the efficacy of such efforts. For example, if regulation of genetic testing is deemed to be desirable, it would seem odd to regulate only DNA testing, and leave protein testing unregulated. Similarly, if limits on the use of genetic testing for insurance purposes are to be introduced, these should not single out pre-symptomatic screening while ignoring susceptibility testing and the specific issues raised by the advent of pharmaco-genomics. Several earlier initiatives in other countries can be cited in this context. Belgium introduced a new insurance statute in 1992 in which the use of "genetic data" for insurance purposes is prohibited.16 The statute specifies that a medical examination in the context of insurance can only be based on medical history and not "on techniques of genetic investigation that aim at determining a person's future health situation."17 An Austrian federal law of 1994 also contains a sweeping prohibition on the use of the results of genetic testing by insurers as well as employers.18 There is nothing in these statutes that clarifies how and why genetic information other than genetic test results, which is often revealed when establishing a medical history, will be excluded from further consideration by insurance companies. The Council of Europe prohibits very generally "discrimination against a person on grounds of his or her genetic heritage" and as has been argued elsewhere, this sweeping prohibition might also be untenably wide.19 A Norwegian Act on biotechnology is somewhat more precise. While it also contains a general prohibition on the use of genetic information resulting from genetic tests, it makes an exception for diagnostic information resulting from genetic tests.20 But as our discussion of the definitions above indicate, even this precision can hardly be called satisfactory. 'Diagnostic information' cannot be separated easily from other types of genetic information. And it is also not clear whether such a distinction is always appropriate in the context of insurance. George Annas, Leonard Glantz, and Patricia Roche try to give a more precise definition of genetic data in their widely cited Draft Genetic Privacy Act, but are thereby limiting the scope of their draft act to a degree that would make any legislation based on this model ineffective and unfair.21 The Draft Privacy Act defines genetic information narrowly as information derived from DNA analysis. The authors defend their exclusion of protein tests and family history-based genetic information by referring to the need for a tight focus of their draft act.22 Interestingly, in an article discussing the drafting of this act, they argue that including other genetic information in their proposal would have undermined the distinction between genetic information and other health information.23 This raises the next question that we want to address. Leaving aside for a moment the difficulty of defining what exactly constitutes genetic information, and accepting the fact that the line between 'ordinary' health information and genetic information is often blurred, are there still some distinct issues raised in the context of the increasing use of genetic information? 3. Characteristics of genetic information: the claim of genetic exceptionalismOur discussion of the different forms of genetic information already makes clear that we are not necessarily dealing here with something totally new. Genetics has been around for some time in one form or another. But that does not necessarily address the arguments invoked to defend the exclusive nature of genetic information. Indeed, it could very well be that the exclusivity of genetic information was less of an issue in the past because it was less widely used. In other words, it still could be true that genetics is truly distinct and merits a distinct regulatory approach. We therefore have to address the arguments invoked to support 'genetic exceptionalism.' Lawrence O. Gostin and James G. Hodge define this as "the societal practice of treating genetic data as different from other types of health data for the purposes of assessing privacy and security protections."24 We will indicate first why genetic information shares most, if not all, of its characteristics with other types of health information. We will then concentrate on those aspects of genetics that have caused many to be more concerned about its use now than, say, thirty years ago. As already suggested, we will argue that the fear for abuse is connected not so much to the inherently new characteristics of genetics, but rather to the link between some of its shared characteristics and the advent of new computer technology. In a fascinating chapter in which he also discusses the Draft Genetic Privacy Act, philosopher Thomas Murray analyzes four different arguments invoked to support "genetic exceptionalism": 1. Genetic prophecy 2. Concern for kin; 3. Concern about discrimination; 4. Generalizability of data to families, communities, racial and ethnic populations.25 Discussing his analysis in an article in the McGill Law Journal, one of us also addressed a fifth characteristic often invoked in discussions about genetics: the argument about lack of control over one's genome.26 3.1 Genetic Prophecy'The concern for genetic prophecy' refers to the predictive character of genetics. Most genetic tests reveal a risk factor without necessarily saying anything about actual health. They give us a snapshot of what can or, in some cases even is likely to happen in the future to our health. Genetics would differ substantially from other health information obtained in 'ordinary' medical examinations, according to this argument, because the latter only give us information about current health. Third parties such as insurers, employers, immigration authorities and perhaps even judicial authorities have, for various reasons, an innate desire to know more about the future health status, behavior or performance levels of employees, immigrants or convicts. They could be interested to use this genetic technology in a way that may harm these individuals. However, as Murray and others indicate, information on future health can and has been obtained from various other sources: family histories of disease, various medical tests indicating high blood pressure, cholesterol levels or iron levels in blood have been widely used in medicine for decades.27 Eating habits, life-style and often even physical appearance will give others some clues about our future health. Tests for HIV/AIDS and hepatitis have been around for some time. With the advent of new forms of treatment for HIV/AIDS, a positive test result for HIV/AIDS is no longer an imminent death sentence. People are surviving longer with the infection. They are in a situation that is to some extent similar to some people who test positive for cystic fibrosis. They may suffer greatly from their condition and may be seriously limited in their daily life as a result of the disease, but they are often capable of controlling the progress of the disease. Moreover, some people seem to be immune to HIV-infection. Others who are infected with HIV do not develop full-blown AIDS, or only much later. Specific genetic mutations are likely the reason for the variable expression of the disease.28 This brings us to a point that we already mentioned in the discussion of the various forms of genetic information: genetic research is increasingly showing us how variable the impact of genetic factors on disease really is. For example, studies now indicate that some people who test positive for Huntington's disease do not develop the disease.29 With respect to cystic fibrosis, the disease has shown to be much more complex than originally conceived and involves not only one common mutation, which accounts for about 70% of the disease, but also a staggering number of more than 850 different mutations.30 Many differences in phenotypic expression of the disease are associated with these different mutations. Environmental and secondary genetic factors (interaction with other genes) are now considered to have an impact on cystic fibrosis. This affects the predictive power of the genetic tests for cystic fibrosis. In a way, this pre-symptomatic testing for cystic fibrosis is in these cases not fundamentally different from susceptibility testing. For other diseases, it is even clearer that environmental factors and gene-gene interactions play a major role in their development. It is fair to say, then, that the risk factors associated with specific genetic mutations and the predictive power of the existing genetic tests vary greatly. The severity of the disorders associated with one or more known genetic mutations is very diverse. The traditional severe single-gene disorders are rare in comparison with the variety of conditions that have now been identified as being associated with specific genes. Moreover, a severe disease such as cystic fibrosis, which was previously untreatable, has become more controllable. Cystic fibrosis also highlights how genetic disorders can be expressed variably, thus making prediction of future health more uncertain. Genetic disorders, even the very determinant ones, are often characterized by "incomplete or reduced penetrance" meaning that the time of onset of the disease varies.31 Other diseases can be avoided or their impact reduced by changes in diet or lifestyles. In other words: it seems difficult to defend the uniqueness of genetics on the basis of a commonly shared predictive character of the information provided by genetic tests. With respect to predicting future health, there is too much variability in genetic tests to support that claim. It is also important to keep in mind how later research often corrects earlier overstatements about the predictive character of specific genetic tests. We mentioned the case of Huntington's disease and cystic fibrosis. As discussed elsewhere, BRCA1 and BRCA2 tests for breast cancer are also good examples of how later findings may tone down the predictive power of genetic tests.32 3.2 Lack of Control Over One's GenomeThe lack of control over one's own genome is sometimes invoked to differentiate genetics from some other types of health information. In discussions about the need to protect people against third party use of genetic information, some state that it would be unfair to use the presence of certain genes to the detriment of those who carry them, because genetics escapes individual control.33 Indeed, people do not choose their genetic make-up; they are born with it. Or, as Gostin and Hodge frame it, "[p]eople feel stuck with their genes."34 Several points should be raised about this argument to show that it cannot be invoked to differentiate genetics from other health information. First, if GATACCA-like predictions come true, individuals will increasingly be able to select perhaps not their own genes, but certainly the genes of their offspring through carrier screening and pre-natal diagnosis. The GATTACA "pre-mating ritual" of having a genetic screen of one's partner clearly is a caricature of what is going on now, but some selection of traits does take place. Various methods are used to obtain certain traits in offspring or to avoid other traits. Results of genetic tests are used to inform couples about their risk for giving birth to children with genetic disorders. People who select sperm for artificial insemination do select to some extent half of the genes of that baby through selecting the profile of the sperm donor. If control is the morally relevant criterion to assign blame, then parents may increasingly be held accountable for the genetic make-up of their children since they partly selected it. Second, as Gostin and Hodge point out, "[g]enetic flaws, like environmental diseases, can increasingly be altered or corrected through clinical interventions."35 Developments in gene therapy will bring genetic traits more under control, even for the living who are 'stuck with them.' Moreover, as we pointed out in the introductory part, genetic testing will often only refer to an increased susceptibility, and individuals who test positive for such susceptibilities may be able to control the onset of disease through changes in life-style and other preventive measures. Third, the idea that genetics is different and merits special protection because it is 'beyond our control' may contribute to some extent to claims of genetic determinism. Genetic determinism refers to the belief that everything, including human behavior, is ultimately determined by people's genetic structure.36 If one were to say that genetics has to be protected because it is not controlled, it could contribute to the perception that genes really do fully determine who we are and what kind of life we may expect. In a way, even if well intended, regulatory initiatives in this area, if justified by reference to lack of control, could contribute to an erroneous public perception of genetics. Moreover, there is also considerable debate about the appropriateness of using the notion of 'control' to assign moral blameworthiness. What does it mean to have control over one's health? People who have control over their health and still 'choose' to become sick or disabled by unhealthy life-styles or risky behavior, according to this mode of thinking, may deserve less protection and less access to health care.37 But what constitutes choice? How do we decide on what constitutes a morally relevant contribution to disease and disability? One would not want to argue that, for example, battered women are more responsible for their health risks than someone who has a susceptibility to cancer and does not manage to follow a risk-reducing diet.38 As one of us argued elsewhere, "[t]o hold people accountable for "lifestyle-related" increased health risks without taking into consideration the social, cultural and environmental context is, in many cases, to further discriminate against populations already vulnerable to the negative effects of discrimination."39 Furthermore, research in behavioral genetics suggests that genetic factors may contribute to some of these 'lifestyle choices' such as alcoholism and nicotine addiction, or even engaging in risky sports.40 Using some form of genetic determinism as the basis for distinguishing genetically 'caused' from 'self-inflicted' diseases thus creates a stark contradiction. If genetic determinism holds true, then those self-inflicted behaviors are equally beyond our control. Finally, our discussion of this issue should not be seen as an endorsement of the moral relevance of this distinction in the first place. Indeed, a just society may very well have to disregard the fact that one is partly responsible for one's illness and expand a protective gaze to embrace all those who suffer from illness and poverty, irrespective of its cause. 3.3 Family and Ethnic CommunityThe arguments about family relevance and ethnic relevance of genetics are more interesting and more complex to discuss. The concern for kin, as Murray phrases the former, refers to the way genetics links us to our families. Genetic tests necessarily reveal information about family members. Children of a parent who tests positive for Huntington's disease, a dominant genetic disorder, have a 50% chance of inheriting a copy of the mutant gene and thus of developing the disease. Clearly, a positive test result of a parent also provides information about the risk status of his or her child. Children from parents who both carry the thalassaemia gene have a 25% chance of inheriting the two mutant copies from their parents (one from each parent). If a woman diagnosed with breast cancer subsequently tests positive for a known disease causing mutation of the BRCA1 or BRCA2 gene, then her first-degree relatives (children, siblings) have a 50% likelihood of carrying the same mutation and therefore of having an increased cancer risk.41 In particular, when it comes to potentially preventable or treatable conditions, it can be very important for people to be informed of a family member's genetic test results and thus of their own risk. Some types of genetic research (genetic linkeage studies) are impossible without the participation of family members. These factors raise questions with respect to the duty of physicians and patients to inform or disclose genetic risk information to family members, family members' right not to know this information, and family members duty to collaborate in research.42 Other family issues are also raised in the context of genetics. DNA techniques allow for an accurate determination of parentage. This can be particularly important in paternity disputes. It has been used in courts, for example in child support and immigration cases.43 The wide-spread use of non-paternity testing clearly undermines the value of the legal 'presumption of paternity' as social stabiliser within marital relations - provided it still has this value. It changes the way paternity issues can be dealt with. Family secrets may now be brought into the broad daylight without the consent of the mother involved.44 Genetic testing may also open the door for many orphans and for children created by artificial insemination to retrace their biological parents, even when clear adoption or donor records are missing. Genetic testing can also open the books on generation-old myths of parentage. For example, DNA testing is being used to assess heritage claims related to various historical figures such as former U.S. president Jefferson.45 This raises interesting questions about whether and to what extent we ought to respect people's desire to keep biological secrets after death. Non-paternity will more often be discovered by accident during routine clinical testing or genetic research. When searching for specific genes, family linkage studies are often conducted to determine how specific markers for genetic diseases are inherited across families. Non-paternity may also be discovered during routine genetic testing when researchers or clinicians simultaneously test samples of different family members. In these cases, researchers or clinicians may face a dilemma between their duty of confidentiality towards the parent involved, and their duty of disclosure towards the children. In practice, it appears that clinicians and researchers do not always warn family members of the risk of discovering non-paternity and do not discuss with them the option of non-disclosure. When non-paternity is discovered, they may withhold that information. This creates the odd situation whereby physicians or researchers may not be willing to tell individuals the real results of their tests (e.g. the real reason why they are not at risk for developing a disease running in the family). While there are certainly important familial issues surrounding genetics, we would argue, as Murray does, that family relevance of genetic information per se does not make genetic information unique. As we pointed out, genetic testing has created some interesting issues with respect to how families can be affected by this information, and how they have to deal with it, but family relevance itself is not a distinguishing feature. The difference, so we shall argue, lies not so much in the familial nature of the information, but in the level of detail and in the way that the familial aspects of genetics may play out in some areas of the law. Genetic information is not unique in disclosing information about other people. Other forms of health information also do so. For example, our medical files often include information about our family members as physicians routinely ask questions about family background. Some of this may be genetic information in that it discloses familial pre-dispositions to particular conditions. But some of this may be non-genetic information regarding issues such as past and present lifestyle, mental and sexual health. Indeed, there are many other types of information that affect families. The physical and social environments of families tell us something about all their members. When one person develops cancer because of exposure to toxins in the family home, others know that they are at increased risk. Tuberculosis in one family member tells us something about the risk of people living with them. Alcoholism of one family member may be an indicator of risks for other family members' health and wellbeing. Gambling problems affect the family's financial security. And if the "breadwinner" has a serious life-threatening illness, Murray argues, surely other family members have an interest in knowing about it. One comparison that is often made when it comes to discussing the duty to disclose genetic risks, is the comparison with disclosure of sexually transmitted diseases to partners. Even the case of not divulging the 'true' results of genetic testing is not necessarily new. Family secrets have been divulged in the past to physicians, psychiatrists, and other professional confidants. In some cases, this may have been relevant for family members' health care choices but physicians have traditionally respected their duty of confidentiality. Another characteristic of genetics is that it may affect or tell us something particular about ethnic, racial or local groups. Genetic information is to some extent shared between larger communities. The most traditional examples of community relevance of genetic information relate to the higher presence of specific diseases within some ethnic groups, or in specific local communities. Tay Sachs, for example, is common among Ashkenazi Jews and some French Canadians.46 Ashkenazi women who have a family history of breast cancer are at a higher risk of developing cancer than women of other ethnic groups.47 Sickle cell trait, which is also associated with a higher resistance against malaria, has a very high incidence among Africans and African-Americans. From 8% to 10% of African-Americans are carriers of the sickle-cell trait, and 1 in 400 to 600 has sickle cell anaemia.48 Cystic Fibrosis is more common among Caucasians.49 When research leads to the development of genetic tests for conditions that are affecting a particular ethnic community, members of this community may be stigmatized. The development of better knowledge about this particular genetic disease and the publicity that often comes with new discoveries may attract attention and may lead to increased discrimination. Often, these communities are already stigmatized and affected by discrimination. Genetic testing could then become yet another, more sophisticated tool of discrimination. In the United States, sickle cell screening programs and erroneous scientific interpretations about the risk of being a carrier of the trait, lead to discrimination against African-Americans.50 Behavioral genetics research linking intelligence, criminality, attention deficit hyperactivity disorders, and other behavioral traits to specific genes that may be more prevalent in some ethnic communities than in others creates even greater risks for stigmatization and discrimination. It is also interesting to note that some communities have participated very intensively with genetic research. This genetic research has, in turn, often been of benefit to the general population. It seems unfair that those who contributed in this way would also become particularly vulnerable as a result of their participation. New DNA technologies have also contributed to various types of research with ethnic or population relevance. DNA research may impact on the cultural identity of these communities. For example, DNA research on the South-African tribe of the Lemba confirmed their claims based on oral tradition and respect for Jewish religious rules that they are direct descendents of Jews led out of Judea by their religious leader Buba. Genetic researchers found that many Lemba men carry a set of DNA sequences that is distinctive of the Cohanim, the Jewish priests believed to be the descendants of Aaron. This DNA sequence is particularly common among those Lemba men who, according to their tradition, are also members of a priest-like clan.51 DNA research is increasingly used to determine the migratory history of people. This research has significant archeological and historical value. Some stress it could also be a social tool, since it shows how much different ethnic groups have in common. At the same time, it could also raise sensitive issues about our ancestors and culture. What if DNA research is used to rebut land claims of aboriginal groups based on their right of first arrival? Or if DNA testing is used to determine whether someone truly is a member of an aboriginal group? In the United States, compensation offered by the United States' government created tension between different members of the Seminole Nation. Descendants of escaped slaves who for generations considered themselves members of this Nation and seem to have been accepted as such were excluded from those who would receive compensation on the basis of blood lineage.52 Similar debates take place elsewhere, including in Canada, and it does not seem impossible that some would call for the use of genetic in this type of debate. The impact of genetic tests on communities raises important questions with respect to obtaining informed consent, respecting community values and practices, the need to involve communities in the design, conduct and analysis of research, and so on.53 Genetics has certainly created particular dilemmas for communities and for researchers conducting research with these communities. However, concerns about the impact of research on communities are again not unique to genetics. Statistics indicate differences in the incidence of cancers among local communities,54 the lower incidence of high cholesterol levels among certain ethnic groups, the fact that HIV/AIDS is more prevalent among gays, intravenous drug users and specific ethnic communities, and so on. Postal codes are used as indicators of a higher chance of being infected by HIV/AIDS, or of bad housing and living conditions which may affect one's life expectancy. Poverty rates affect particularly certain ethnic communities, such as aboriginal peoples in Canada and African-American communities in the big American cities. Poverty is the most direct indicator of low life expectancy and ill health. Research on, for example, alcoholism in communities and life style, can affect and stigmatize the community that is the target of research. 4. How Genetics Highlights Existing ProblemsSo far, we have argued that genetic information shares various characteristics with other types of health information and that genetics does not raise inherently new ethical and legal questions. However, that does not mean that there is no need for increased scrutiny of existing regimes or the development of new regulatory responses. Three reasons justify careful analysis, public debate and regulatory intervention where needed. First, even though genetics does share many of its characteristics with other health information, that does not mean that we have found satisfactory ethical, legal and regulatory answers to address issues raised by these characteristics. More systematic use of genetics deserves special attention because it could exacerbate existing problems. Murray claims that the increasing number of genetic tests "broadens the pool of possible factors that might be used to discriminate against an individual, and it likewise increases the number of individuals who might become the subjects of discrimination."55 It is worth pointing out that significant ethical and social problems could also be created by the fact that these 'exacerbated problems' increasingly play out in a much more commercial health care context. Second, genetics brings many of these issues to a different level. As we argued in the context of the familial impact of genetics, genetics gives a new spin to some of our concerns, which may require a different regulatory approach. In our analysis, we have given examples of the way genetics brings these issues up to a different level: genetics gives us more detailed risk information than most other tests; genetics increases dilemmas surrounding the duty to inform third parties such as family members of particular health risks; and genetics does impact in particular ways on communities. Third, a combination of the following factors, which are not necessarily "unique characteristics" of genetics, makes it necessary to look at how adequate current regulatory approaches are: 1) the volume of information that can be extracted from one sample which can be kept indefinitely; 2) the speed of testing; and 3) the link between genetics and computer technology. This third issue is clearly related to the first two issues and is contributing to the ease by which genetic information can be obtained. For the remainder of our paper, we want to analyse some of these elements further. First, we will analyse in more detail how genetics may shed new light on an existing issue. We will use as an example the issue of duty to inform others of health risks. We will then discuss how the nature of a genetic sample, in particular the volume of information contained in it and its potential longevity raises particular issues. We will conclude by discussing how computer technology impacts on this debate. 4.1 Family and Disclosure of Risk InformationWe have already indicated that the fact that an individual's genetic information also reveals information about biological family members, raising concerns regarding the disclosure of such information to relatives, does not render genetic information unique. Other information may also be of concern to an individual's relatives. However, characteristics of genetic information and the way in which it is relevant to third parties calls into question traditional justifications for disclosure or non-disclosure of health information to third parties and suggests that there may be a need for unique guidelines for the disclosure of genetic information. There is considerable controversy over whether a physician has a privilege (or, more strongly, a duty) to breach duties of confidentiality in order to warn a patient's relatives as to their risk of a genetic condition.56 The breach of confidentiality involved in disclosing a patient's health information against his or her wishes is generally only justified where there is a threat of serious and imminent harm to others or there is a serious public health risk involved.57 An example that would fall into this traditional rationale would be infectious disease. When an individual has an infectious disease this is of concern to that individual's family (and those who come into close contact with him or her) because it alerts them to a potential risk to them arising from their physical and social environment. Genetic risks are to some extent different. For example, as we pointed out, if a woman diagnosed with breast cancer subsequently tests positive for a known disease causing mutation of the BRCA1 or BRCA2 gene, then her first-degree relatives (children, siblings) have a 50% likelihood of carrying the same mutation. They therefore have a significantly increased cancer risk.58 But she has not created the risk for the relative. It is a preexisting risk that is identified because of her genetic information.59 The risk is not so much related to something external to the other person. It is a risk associated with genes that are part of that person. This distinguishes genetic information from the type of harm involved in the recognized exceptions to the duty of confidentiality. Furthermore, the harm disclosed by a pre-existing genetic condition is not necessarily preventable. If it is not preventable then disclosure likely will not help the third party and may even harm the third party by causing distress. To the extent that the manifestation of the disease associated with a particular genetic condition is preventable, because of complicating biological and environmental factors, then it looks a lot less like the serious and imminent harm at issue in the traditional justification for a breach of confidentiality (even though the information is in that case so much more relevant for the family member). There are further important questions that need to be answered regarding what counts as a serious or imminent harm in this context. The first is the degree of likelihood of having a genetic condition. That is, is a 50% likelihood of carrying a mutation of the BRCA1 gene sufficiently serious? The second element is the degree of likeliness that the genetic predisposition will lead to the actual physical manifestation of the disease. In other words, if having a particular mutation is associated with a 63% increase in the likelihood of a particular condition then is this sufficiently serious?60 Does it matter that this represents the known lifetime risk, rather than some imminent health risk? The Science Council of Canada has adopted the same guidelines for physician disclosure to third parties as the President's Commission for the Study of the Ethical Problems in Medicine and Biomedical and Behavioral Research:
While these guidelines help to clarify some of the issues outlined above, questions remain. One question is whether harm is the appropriate way to think about disclosure of genetic information to third parties. In general, an individual has a right to know his or her own genetic constitution, based on the value of individual autonomy.62 This rationale can also be used to justify the right not to know one's own genetic constitution.63 But this rationale would also provide a reason for biological relatives to gain access to the genetic information of a family member: it is also information about their own genetic constitutions. Therefore if they seek this information, in situations where they can only learn about their own genetic information through gaining access to information about relatives (e.g. linkage studies), then there is a reason for disclosing this information that is based on autonomy rather than harm. An autonomy rationale would provide more reasons to allow access to another's genetic information than a harm rationale. It therefore at least raises the following questions:
Of course, there are competing autonomy interests at stake here, which are difficult to balance. It may be that the claims of relatives are not strong enough to outweigh health care professionals' duties of confidentiality. At a minimum, what these questions highlight is the need for genetic counseling services to be widely available to aid families as they deal with these issues. 4.2 Information, Longevity and IdentificationAn analogy can be made between genetic information and information technology. For more than two decades, concerns regarding the practices made possible by burgeoning information technology have spurred the adoption of data protection regimes. These concerns include:
One commentator has distilled these concerns into two features of information technology: its ability to shift information from the context in which it was gathered to a different context, and its ability to aggregate data.67 If we deal with genetic information in the form of DNA samples, or substantial sequencing information, then similar issues are raised as in the example of information technology. The amount of information about an individual that can be gleaned from a DNA sample is potentially staggering. If the sample is stored, then it has the potential to reveal information about the individual far into the future, beyond the purposes for which the individual first provided the sample, and even long after the individual's death, limited only by available genetic knowledge and technique. Furthermore, DNA can uniquely identify an individual, which can provide the basis for linking together information from disparate sources. A clear example of this is in the forensic context, where biological material gathered from a crime site can be compared to existing "DNA fingerprints" of suspects. As the analogy with information technology indicates, the fact that genetic information allows for the aggregation of information and its shift between contexts is shared by information technology. And, as other commentators have pointed out, the science of biometrics allows for other ways to uniquely identify an individual than by using DNA fingerprinting.68 However, the particular contexts in which these concerns arise with respect to genetic information may be unique and warrant a specific focus, and response, when regulating health information. For example, it is quite common, when developing data protection regimes, to exempt non-personally identifying data from protection.69 Similarly, existing human tissue samples used in research are often exempt from consent requirements when used in non-identifiable form.70 There are serious questions, however, regarding the extent to which genetic information can be collected or used in a non-identifiable form. Genetic information, unlike other health information, is inherently linked to a particular individual.71 This fact, in combination with computer technology, makes the linkage of genetic information to an identifiable individual always a possibility. As the Tri-Council Policy Statement on Ethical Conduct for Research Involving Humans states, "[g]enetic testing has greatly narrowed the concept of anonymous tissue... [T]he concept of traceable tissue is now wider, since it is now possible to identify biological relatives by using genetic markers."72 Interestingly, the new Canadian Tri-Council Policy Statement does impose a more stringent consent requirement for collection of tissue and for the use of previously collected tissue, but it leaves the door open for the use of previously stored anonymous tissue without consent.73 It is worth noting also that the Tri-Council Policy Statement requires researchers to specify in a research protocol future use of the genetic material. In other words: DNA banking of material should be justified at the time of collection. Moreover, the Policy Statement also suggests that there should be a time limit on the storing of samples, which should be congruent with the purpose of storage.74 Very often, samples collected for research purposes are not made anonymous. They can be directly identifiable through the use of an identifying tag or through patient or research subject numbers. Or they can be traced back through a specific identifying code, which may be kept at a different location and/or by a different person than the person who conducts research on the sample. Linkage between the samples and the clinical file can be of crucial importance either for clinical or for further research purposes. Researchers are worried that rules that are too stringent with respect to the requirement of anonymizing data will hamper research. Given the concerns, it is not advisable to completely exempt "anonymous" genetic information from data protection regimes. However, this does not mean that full informed consent for all uses of "anonymous" samples must always be obtained.75 Other options include security provisions, or practices similar to Statistics Canada for handling aggregate information.76 Research Ethics Boards will have to pay particular attention to the issues raised by using stored genetic material in research. The amount of information potentially available from a DNA sample as well as the longevity of the sample raises other concerns regarding the uses of samples and informed consent. For example:
The fact that DNA can be used to uniquely identify an individual also raises concerns regarding the uses that could potentially be made of stored samples. For example, the federal DNA Identification Act (not currently in force) puts in place a framework, including safeguards, for the establishment of a national databank containing a crime scene index and a convicted offenders index in order to facilitate the identification of persons alleged to have committed certain designated offences.78 However, as we have already discussed, DNA databanks are being set up outside of the forensic context as well, and contain genetic information on individuals who have never been suspected of an offence, let alone convicted of one. These may not be research databanks, either. We mentioned already the "Guthrie cards" containing blood spots obtained from newborns for the purpose of screening for PKU and other disorders. DNA is relatively stable in dried blood and so such samples provide a de facto DNA bank.79 Nonetheless, the police could gain access to this information through seeking a warrant. This highlights a need to examine the warrant procedures to see if specific guidelines regarding genetic information are required, including regulations regarding the uses of this information once obtained. For example, should the police be allowed access to genetic information that they would not be allowed to obtain under the DNA Identification Act? Under what circumstances will the police be allowed to retain the information, for how long, and with what safeguards? These issues are currently dealt with by ss. 487.05 to 487.09 of the Criminal Code.80 However, these provisions address the conditions under which a warrant may be issued to obtain a bodily sample from a particular individual and do not seem to address the situation where what is sought is an already existing sample.81 4.3 Uses of Genetic InformationThe increasing prevalence of genetic databanks raises questions not only regarding genetic information but also about the combination of genetic information and information technology. This issue has particularly come to light with respect to the Human Genome Project. Gene sequencing on such a large scale requires information technology. Not surprisingly, for computer giant Bill Gates, DNA research and computers seem a natural fit.82 As Dr. Karp, a leader in computer science who works on sequencing projects at the University of Washington, has stated:
The combination of information technology and genetic research makes possible types of research that may raise unique issues, particularly with respect to informed consent. One such issue arises from the possibility of combinatorial rather than hypothesis-driven research. In hypothesis-driven research, researchers note that a particular family, or community, experiences a high incidence of a particular disease. This leads to the hypothesis that the disease has a genetic basis and leads to the search for the particular gene mutation responsible for that disease. Combinatorial research, by contrast, makes use of information technology to compare medical records, genealogical records, and gene sequences in order to discover correlations that may indicate the genetic basis for particular diseases. Such research does not require particular hypotheses to test, but rather relies upon brute computational power to discover genetic links to disease. For example, this is said to be the revolutionary potential of the Icelandic database.84 But this type of research raises the question of whether individuals should be allowed to give broad consent for the use of their genetic information, as well as their other health information. Broad consent would mean consent to future unidentified uses of their information, and the question is whether such consent can be meaningful. As we have already discussed, genetic information also relates to one's biological relatives. This reason makes families particularly important for genetic research, but increasingly whole populations such as in Iceland, or Newfoundland, or various aboriginal communities, are of interest to researchers. As the kind of whole population study such as in the Icelandic database becomes more prevalent, made possible by the combination of genetic research and information technology, then community concerns need to be taken into account.85 The following are some issues that need to be dealt with:
Finally, the drive to identify the genetic basis of disease often has a commercial aspect, particularly when researchers seek patents on the genetic sequences they identify, and the diagnostic tests they develop on the basis of these sequences. While gene patenting is controversial for many reasons beyond the scope of this discussion, one issue that is of relevance here is informed consent. To what extent and how do patients and research subjects have to be informed of the potential commercial applications of research? The Tri-Council Policy Statement explicitly imposes a duty on researchers to discuss this with research subjects.86 A recent controversy around the patenting of the Canavan disease highlights how sensitive these issues are. In this case, parents of children who suffered and died from Canavan disease sued researchers involved in the research and subsequent patenting of the gene. Family members of children with Canavan-disease were particularly upset because those who contributed to the discovery of the gene now had to pay for testing of other family members.87 5. ConclusionIn this paper, we have indicated that different characteristics attributed to genetics do not necessarily make genetic information something entirely new and unique. However, the combination of some of these characteristics, and the way some of them play out in the context of genetics, do warrant a careful assessment of existing regulatory approaches towards health information. These characteristics, combined with the particular volume of data, the potential speed of testing and the link with computer technology, make it necessary to investigate how we can better protect private health information. More comprehensive protective measures are warranted. One of us has argued elsewhere that it may be inappropriate to single out genetics in specific genetic anti-discrimination legislation, particularly in the context of insurance.88 However, it was also stressed that we need to develop flexible regulatory structures to analyse how developments in genetics impact on society and to intervene when protection is needed. In many cases, existing regulatory regimes could be adapted, for example to make sure that regulatory definitions capture new genetic information. As we stressed in our discussion, when developing new regulations, attention will have to be paid to the various ways in which genetic information is gathered, circulated and used. In the coming years, additional efforts will have to be made to analyse carefully whether existing regulations and laws capture all the issues raised by genetic technologies, how we can adapt them, and what new initiatives are needed. Computer technology has accelerated genetic research. Social concerns about some consequences of genetic research will hopefully accelerate the pace of regulatory change aimed at improving the protection of individuals and communities against informational and other harms.
|
http://cbac-cccb.ca | ||||
Created: 2005-07-13 Updated: 2006-07-17 |
Top of Page |
Important Notices |