Response to HGC consultation on personal genetic information from Human Genetics Alert

March 2001

1. What is personal genetic information/Is genetic information special?

1.1 The distinction between genetic information and family history

Section 2 of the paper whilst disavowing genetic reductionism and determinism actually encourages them. The section gives the impression that family history and phenotype are genetic information, and suggests that they may be equivalent to molecular genetic data. Family history and phenotype information may be used to make inferences about a person's genes, but they are not genetic information; they are phenotypic information. To suggest that they are 'forms of' genetic information is the crudest type of reductionism.

This distinction underlies the key practical difference between molecular genetic data and family history/phenotypic (medical) information. Whereas the latter allow deductions of varying degrees of reliability, molecular genetic data are precise (about genes). This is why predictions of future health derived from molecular genetic data are likely to be more accurate than those derived from family history information, even though both will be often subject to environmental influences. It is only because of this increased predictiveness of molecular genetic data that we are seeing massive investment in genetics research and the creation of an industry based upon it. Family history information is not leading to similar developments because of its limited usefulness in prediction, as well as its complete uselessness in understanding molecular and cellular causes of disease.

1.2 The distinction between genetic information and medical information

It is true that personal genetic information is a form of medical information. However, because it is difficult to find a single criteria to draw a clear line between genetic and other types of medical information, this does not mean that therefore genetic information is no different from the other types. There are a number of different factors, which collectively add up to making the meaning of genetic information distinctively different, as well as its impact both on individuals and society.

1. Notwithstanding the incorrectness of popular overblown accounts of genetics, it is impossible to deny that the molecular genetics revolution of the last 20 years has given us access to a whole new level of biology. This opens major new possibilities, both good and bad, which are the raison d'etre of the HGC. Although genes are not the 'blueprint' of an individual, there is an important sense in which genetic information is more intimate and fundamental information about a person than what their parents died of, or their cholesterol level.

2. Genetic information defines something more enduring and constitutive about a person than whether they have high blood pressure or have contracted cancer. Whereas the latter may fluctuate during a person's life, and are seen as semi-random events that may or may not occur to a person, a person's genes are fixed at conception and remain constant. To whatever extent genes determine the person's characteristics, they are enduring and constitutive facts about that person. Amongst the characteristics where the influence of genes is strong are many of the external facets that a person presents to the world, and through which we recognise them throughout their life: sex, facial features, body shape, skin and hair colouring. It is not simply ignorant genetic determinism that leads people to view genetic facts about a person in a different light to whether they have had the bad luck to have had a heart attack.

3. Genetic information is hereditary information, unlike almost any other medical information. Thus information about a person's genes will often immediately raise very sensitive issues about reproductive choices. Issues of equal sensitivity arise about the person's own parentage.

4. Genetic information is about families much more strongly than most other medical information, and thereby raises issues, such as that addressed in section 6, which are not raised by most other medical information.

5. Most importantly, the social meaning of genetic information is very different to that of most medical information. Primarily because genes define crucial and enduring aspects of a person, and because they are passed on to offspring, perceived and real genetic differences continue to attract discrimination and oppression. A fully-fledged theory of social development (eugenics) is based on the importance of genetic differences. It is ridiculous to imagine anything comparable based on people's medical histories, even though these can certainly lead to discrimination.

We cannot consider the meaning of genetic information in the absence of this social context. It is important that the HGC does not fall into the trap, unfortunately common amongst expert committees, of approaching this subject with a narrow technical focus, and dismissing the meanings that lay people attach to it as 'uninformed' and 'emotional'. The HGC should beware of labelling public concerns as due to uninformed genetic determinism.

In summary, these inherent features of genetic information, when taken together, mean that it is distinctively different from other forms of medical information. These features are at the root of well-justified public concern about the uses of genetic information, especially in a climate of rampant commercial exploitation of genetics with few legal safeguards. It should be clear that ethical and legal approaches for dealing with medical information are not adequate as a framework for dealing with genetic information.

2. Genetic information in medical practice: consent

2.1 Question 5.1

This section does not deal with the issue that will be most crucial if genetic testing is to become widely used in medicine: the standard of counselling given. It is widely expected that most genetic tests will be performed in primary care, yet general practitioners are generally not well informed on genetics and not trained in genetic counselling. It will be expensive to provide adequate genetic counselling, and there will undoubtedly be pressure from government and technologically enthusiastic NHS bureaucrats to introduce genetic testing without adequate provision of counselling. If this happens, it will invalidate the consents obtained, because they will not be adequately informed. While this may be cheaper for government, the costs for individuals of taking genetic tests which they do not properly understand may be very high. Furthermore, this will rather quickly lead to a backlash against genetics amongst patients.

2.2 Question 5.4

The section and this question seem to assume that people's DNA will be subject to more tests than is necessary, simply because the cheapest testing technology, 'gene chips' demands this. In other words, technological rigidity and economic pressure will override common sense medical practice and basic medical ethics. We reject this suggestion. Unless patients consent to a broad range of genetic tests, and receive genetic counselling that explains the impact of all of them, consent will not be informed and will therefore not be valid. Where a patient is undergoing genetic testing for a particular purpose the technology chosen should be appropriate to that purpose.

3. Confidentiality

Section 6 does not provide an adequate description of medical privacy law. However, if the situation is as described the law is not adequate. There is little clear legislation and therefore no reliable protection for individuals. If individuals can only gain redress by sueing, rather than relying on public prosecution services, many people will be excluded from any form of justice. There should be clear legislation that guarantees confidentiality of personal genetic information.

4. Genetic information in research

This section assumes that conventional models of doing research are basically sound. In the conventional model scientists do research on people rather than with them, and tend to treat them as objects rather than equal participants in the research. This basic structure is at the root of ongoing scandals in medical research and over the use of human tissue without consent. These scandals have eroded public confidence in the medical profession to the point at which research is seriously threatened. In the present climate, where commercial companies are increasingly involved in research with human subjects and tissue samples, the need for strict standards is even more apparent.

A new model of research is needed in which people who donate samples are seen as partners in research and are involved in the planning and governance of the research. This approach is very suited to the creation of gene banks, since they require thousands of people to altruistically participate for the good of the community, as they do in blood donation. Such an approach would make the term 'community genetics' a reality. These kind of approaches are increasingly common, in social science research but have not been used in medical research.

To make such an approach work in practice would require a change in the institutional organisation and control of research. Before any gene bank was created there would be an extensive public consultation process, including a social impact assessment conducted by independent academics or consultants. The community must be asked whether it wants such a project, not merely how the project should be implemented. Since they are community projects, gene banks would have to be managed and controlled by the community, in partnership with scientists. Typically, a gene bank would have a management committee comprised of a majority of community representatives, which would control the use of funds. Decisions would need to be consensual. The committee would set ethical guidelines for the type of research that could be done, and set conditions on the transfer of samples to third parties. Wherever possible, the gene bank would be located in the geographical area in which the particular community lived. The operation of gene banks would be as transparent as possible, and there would be the maximum amount of consultation with the community about important decisions. Of course, the management body must also be subject to external audit.

Questions 7.2, 7.3 and 7.4

We cannot provide a comprehensive list of all the information that should be given to someone donating tissue for use in a research project. However, we are most concerned that people are given adequate information about the potential uses of their sample (including commercial use, see below). Treating people as genuine participants in research requires a higher standard of consent than is common in conventional medical research, and means that their wishes must be respected to a high degree. Research projects involving samples taken from individuals should set clear ethical guidelines on the types of research permitted. These guidelines should exclude research in behavioural genetics. There are many possible harms, or other outcomes that people might object to that might arise from research on gene banks, such as stigmatisation of people with particular behaviours or ethnic or other social groups, development of genetic 'enhancement' technologies, patenting of genes (see below), use for biological warfare and the genetic engineering of animals (this is not as comprehensive list). For particular ethnic and social groups there may be additional concerns. Where research will take place that participants might object to, they must be informed of this. Participants must be able to set conditions on the use of their individual samples, as part of the consent ('opt-out system'). Since the aim should be to respect participants wishes, not merely to prevent direct harm to individuals (as assumed in conventional medical ethics guidelines such as the MRC's), if the original consent is inadequate, it is not possible to do further research with samples even by anonymising them: the scientists must re-contact the person for consent.

We disagree with the MRC's view that where samples have been coded or anonymised in historical collections research can be done without going back to the donor for consent. Where research is envisaged that has not been consented to, researchers should make the maximum possible effort to contact the donor and obtain consent. If it is assumed that it is acceptable to routinely conduct non-consensual research on historical samples, this undermines the basis for requiring consent for present and future uses.

Question 7.8

We are very concerned by the suggestions in the Medical Research Council's recent guidelines suggesting that feedback of genetic research results to donors would be common. This is an extremely dangerous policy. In research, the clinical value of genetic tests is usually unknown and there is therefore no case for feeding back potentially misleading information, even when it is wanted. If the test is clinically validated, feedback should only occur under exceptional circumstances. If it is done on a regular basis then the project becomes genetic screening, not research, and will have to budget for the necessary genetic counselling for all donors. The only case where feedback would be justified was if one found a genetic predisposition in someone to a disease that would cause them to collapse suddenly with serious consequences, or where prophylactic action is available that would not be effective after the person developed symptoms.

5. Commercial use of personal genetic information

Question 8.1

This section is not a fair representation of the issues in this area, since it does not give adequate discussion of the question of patenting of genes and other biological material. While we understand that the HGC intends to address these issues separately, and has therefore not gone into them in any depth here, it is not reasonable, in the midst of a major policy debate on the patentability of human genes, to neglect to mention the arguments against patenting of genes. Paragraph 8.14 is an uncritical regurgitation of industry arguments for patenting, which must be balanced by the counter arguments. Paragraph 8.1 is positively misleading, since it would suggest to an uninformed reader that human genes were simply not patentable. It is not true that the patenting issues are 'of limited application', as the experience of the patenting of indigenous people's genes, the John Moore case and other more recent examples demonstrate.

This section also fails to mention the question of compensation for sample donors or communities, even though there has been considerable debate in this area. The question has been especially inflammatory in connection with the Human Genome Diversity Project, and patients organisations and even individual donors in the USA are beginning to demand compensation or joint ownership of patents. If commercial companies develop products based on research using gene banks, this is only possible because of community participation. Therefore, companies should share the financial benefits arising by donating a fair percentage of their profits to the healthcare systems of the gene bank community.

Questions 8.2 and 8.3

Firstly, sample donors must be informed of commercial access and give explicit consent. Human Genetics Alert is opposed to the patenting of human genes and other body parts. Sample donors must be explicitly informed of the fact that companies may patent genetic information, even if this does not include the patenting of an individual sample of genes derived from it. Since many donors would object to patenting per se, it is vital that they are aware of this possibility: if they are not then consent is not valid.

Secondly, publicly or charitably funded gene banks should not permit the patenting of genes by companies that use the banks. We agree with the majority of respondents to the People's Panel survey, who said that ownership of the results of research should remain in the public sector.

Thirdly, samples given to commercial companies must be securely coded so that individuals cannot be identified. Information about the donor should be restricted solely to that necessary for the research project and large amounts of other personal information, such as job, home location, etc that might allow identification of the person should not be transferred.

Fourthly, companies should not be allowed exclusive access to a gene bank, and it is vital that gene banks be financially and managerially independent of commercial companies. In particular, the scientists and doctors must not have any financial interests in companies that use the gene bank.

6. Insurance and employment

Question 9.1

This section is a poor summary of the issues and is biased in favour of the insurance industry. Paragraph 9.2 repeats the reductionist mistake of section 2, in describing family history as a form of genetic information. The clear aim of this paragraph is to use this 'fact' in order to allay public concern. Paragraph 9.3 suggests there is no evidence of genetic discrimination in the UK. In fact a large survey by Wellcome Trust researchers found considerable evidence of discrimination by insurers. Paragraph 9.5 states that test results will only be taken into account if the test has been approved by the GAIC, whereas in fact insurers are already using test results before GAIC approval. Most importantly, it fails adequately present the case against the use of genetic tests by insurers. It does not even mention the main cause for public concern: that large numbers of people may be made uninsurable as insurance companies compete to reduce their risk, and that genetics may catalyse a shift from insurance based on pooling of risk to a system of 'cherry picking' and large variations in premiums for different individuals.

Question 9.2

In our view it would be better if the insurance industry abandoned the use of family history in underwriting. Family history information is notoriously unreliable and of questionable predictive value: the result is bound to be a considerable amount of unfair discrimination.

However, it should be clear from the points made in response to sections 2 and 3, that the use of family history and medical information does not constitute a justification for the use of genetic information.

The key point is that although describing genetic information as a form of medical information captures part of the meaning of genetic information this is not an adequate way to deal with the issue, because it ignores crucial individual and social meanings of genetic information. To treat the issues as a purely technical one of comparing the predictiveness of molecular genetic data, family history and medical information falls into a trap that the HGC was created to avoid. Treating insurance as a purely technical matter also ignores the fact that the provision of insurance is very much an issue of social policy.

Viewed from the perspective of the public interest, it is apparent that the use of genetic tests by insurers is very undesirable. Firstly, it is very clear that it discourages people from taking genetic tests that may benefit their health. Secondly, it may make large numbers of people uninsurable. Thirdly, it is part of a larger tendency, partly driven by genetics' own tendency to locate risk in individuals rather than society, and partly driven by governments' efforts to reduce welfare spending and place the responsibility for financing healthcare provision on individuals. These synergistic tendencies are at the root of public concern about the use of genetic tests by insurers, which is rightly seen as one element within a larger eugenic tendency in society. Fourthly, the sensitive and personal nature of genetic information means that its deserves special protection, especially from organisations, such as the insurance industry, which do not have an exemplary track record on the issue of discrimination. Clearly, from a social perspective which is conscious of the history of genetics and eugenics, there is nothing 'arbitrary' about demanding that genetic information be excluded from insurance.

Question 9.3

The insurance industry has not demonstrated convincingly that it has any need for genetic information or that the lack of it would do any noticeable damage to its interests. Human Genetics Alert supports a ban on the use of genetic tests by insurance companies.

Question 9.5

The section on employment, like that on insurance, is biased towards industry perspectives, without acknowledging that fact. Paragraph 9.14 fails to acknowledge the major political issues involved in testing people for susceptibility to occupational disease. By failing to even mention these, and presenting the issue as if it were a straightforwardly good idea to test people for susceptibility to occupational disease, the Commission takes the side of employers. The section also fails to mention a survey by the Institute of Directors which indicated substantial interest in the use of genetic tests by employers.

Questions 9.6 and 9.7

The HGAC principles are not acceptable. Points 2 and 3 assume that it is automatically desirable for people to be tested for susceptibility to occupational disease, and even go so far as to put the onus on employers to offer such tests. By failing to acknowledge the power imbalance between employers and their employees/job applicants, and the tendency of employers to fail to maintain worker health and safety, in order to maintain profit margins, the HGC implicitly takes the side of employers. We would argue that in particular working environments or practices meet health and safety standards but still pose a risk to some individuals, the first thing to be done is that the standards should be tightened. Standards should be set at levels which will protect even those who are genetically susceptible. It is only in very rare instances that a small group of individuals has a greatly increased susceptibility - in most cases susceptibility will be on a continuum. In these situations, a tightening of standards will benefit all employees, and it will be arbitrary discrimination to exclude those with a particular gene variant from employment. Employers should use the best available technology to clean up their workplaces. In a situation in which employers were routinely 'offering' genetic tests, and the only thing protecting employees and applicants was the individual's 'right not to know', there would be intense pressure, particularly on job applicants, to accept the 'offer'.

Whilst it may be acceptable under certain circumstances to require testing for people who may endanger the public, these circumstances should be defined as narrowly as possible.

The HGAC principles should be revised in order to prevent employers using genetic testing as an easy way to reduce costs.

March 2001