House of Commons - Science and Technology

Select Committee on Science and Technology Fifth Report

Relevance of test results

30. Insurers are understandably concerned about the impact that genetic testing could have on their industry. Insurers operate on a system of "mutuality - namely, the pooling of similar risks and the achievement of broad equity between persons of similar circumstances who are in an insured risk pool".[75] Essentially, people with similar risk pay a similar premium, because there is the same probability that they will claim on their policy. If you are deemed to be at a higher risk, for example you live in an area prone to flooding, your premium (in this case, for your building insurance) will be higher because there is a greater chance of your claiming. In addition, as the ABI told us, there will "always be a section of the population whose level of risk, whether or not genetically determined, will be so great as to render them commercially uninsurable."[76] As an example of the proportions involved, of all applicants for life assurance at Norwich Union, 90% were currently accepted on standard terms, 8% on increased premiums and 2% were declined insurance.[77]

31. Such a system depends on the disclosure of all relevant information by both sides, on a principle of "utmost good faith... If there is not equivalence of information between the applicant and the insurer, the latter will be unable to set the correct premium level".[78] Many insurers are afraid that increased knowledge of future ill-health, which genetic testing brings, "could lead to behavioural changes in insurance choices"[79]: those who know they are at greater risk will take out larger insurance policies than they would have normally, while those who feel themselves to be less at risk may choose not to take out any insurance at all. This is usually referred to as 'adverse selection'. Insurers argue that if they do not know the risk involved, they cannot correctly calculate the right premiums or put aside the necessary reserves to pay future claims. We acknowledge insurers' concerns about the risk of adverse selection and accept, as a principle, that commercial insurance companies should have access to the same information as applicants, where it is relevant, and reliable - but only if there are no adverse consequences for society as a whole (for example, by discouraging people from taking tests).

PREDICTIVE SIGNIFICANCE OF POSITIVE TEST RESULTS

32. It does not appear to be certain, at present, that the information obtained from positive genetic tests is relevant to the insurance industry. Although the understanding of genetics has made enormous leaps forward in recent years the tests do not, at present, have the predictive power which they have been accorded. Professor Bobrow told us that "Genetic tests are very good at distinguishing those who carry a particular gene from those who do not. They are somewhat less accurate at identifying those who will and will not eventually get the disease."[80] This distinction between identifying the existence of a defective gene and being able accurately to predict what will happen to the gene was supported by the British Society for Human Genetics. A particular genotype may raise the risk of the development of a disease but there remains a wide spectrum of age of onset or severity and course that cannot be predicted with certainty.".[81] Professor Bobrow informed us that, even for Huntington's Chorea, which is generally considered to be one of the most straightforward of the purely genetic disorders, he was unaware of any scientists who believed that the age of onset of the condition could be predicted to any really accurate degree. He believed that long-term studies were needed to show the importance and interaction of genes in causing diseases and that any "validated, verified test" with the sort of predictive powers that insurers need "was certainly decades away".[82] In effect, we were told, genetic tests can tell you whether, or not, a person has a particular gene defect you are looking for but cannot say, with any degree of certainty, whether they will actually get the disease, when they will get it, how severe it will be, and how responsive to treatment they will be.

33. These concerns over the relevance of the tests are shared by at least one insurance company, the CIS, who told us that there were "too few genetic tests of significant predictive value"[83] currently available. The ABI's decision that four of the tests it had previously strongly recommended are now no longer considered reliable would also seem to support such doubts. According to Mr Beddell-Pearce of the Prudential, the tests for familial adenomatous polyposis and hereditary motor and sensory neuropathy are irrelevant to insurers owing to their, typically, early onset; that for myotonic dystrophy is apparently not sufficiently predictive, and multiple endocrine neoplasia has too wide an age of onset.[84] How the ABI felt it had enough evidence to pronounce these to be relevant and useful in 1998 and now suddenly finds it does not, is beyond the Committee's understanding. The ABI's decision that four of the tests it recommended insurers use for the past three years ago are now no longer relevant or reliable casts the gravest possible doubts on the validity of all the tests not explicitly approved by the GAIC, being currently used by insurers. Insurers have given the test results a predictive significance that cannot, at present, be justified.

34. It is unclear why the ABI chose the ten tests for seven conditions in particular, out of the large number extant at present. The Royal Society told us that "this selection appears arbitrary"[85] and Professor Bobrow agreed, stating that, "these seven [tests] are just exemplars; there are already tests of greater and lesser precision.".[86] Mr Beddell-Pearce of the Prudential implied during his oral evidence that the ABI was less interested in the merits of these particular tests than in establishing the right to use genetic tests in principle.[87] Insurers appear to have been far more interested in establishing their future right to use genetic test results in assessing premiums, than in whether or not they are reliable or relevant.

ACTUARIAL RELEVANCE

35. Doubts have also been raised about how genetic test results are converted into actuarially useful information by insurers. The Institute of Actuaries told us that, "In one sense all information pertaining to the risk propensity of policy-holders and potential policy-holders is relevant from an actuarial point of view."[88] This information has traditionally included age, gender, family history, occupation, place of residence, lifestyle (smoker or non-smoker, for example) and several other factors. Many insurers see the results of genetic tests as a natural extension of this information, to be used in actuarial calculations in the same manner. Extensive data showing how the results of genetic tests would affect the likelihood of someone claiming on their insurance policy, however, appears to be difficult to obtain. The Wellcome Trust told us that "specific research correlating gene changes with mortality experience is still necessary before insurance implications can be safely assessed."[89] Professor Bobrow similarly stated that there was a dearth of actuarial evidence. We were informed by the Alzheimer's Society that they had experienced great difficulty in establishing exactly why these tests were considered 'actuarially significant'. Indeed, when they tried to obtain a definition of 'actuarial significance' from the Institute of Actuaries, they were informed that "if an actuary says something is significant, then it is."[90]

36. We were assured by the insurance companies attending the oral evidence session that the actuarial information they based their decisions upon, which largely came from re-insurance manuals, was reliable and was in the public domain.[91] The work currently being undertaken at the Genetics and Insurance Research Centre at Heriot-Watt University should help to clarify further the link between genetic test results and insurance premiums. This research is funded, though not directly controlled, by a consortium of insurers and its remit is to develop mathematical and actuarial models to estimate the costs of genetic knowledge to individuals, to insurers and to service providers.[92] We also appreciate the distinction made by Professor Raeburn, in his memorandum, between the prediction of the onset of a disease for a particular individual, and the mean age of onset for sufferers, which can be used in actuarial calculations.[93] (By comparison, it may not be possible for an actuary to predict exactly when one particular house may catch fire, but they could calculate the mean probability for such a fire and calculate the insurance premium on that basis.) As Professor Raeburn points out "underwriting is based on group data, not the exact timings for each individual."[94] However, if the insurance industry wishes to reassure the public that they are acting in a reasonable manner, they must publish more data which unequivocally supports the changes made to insurance premiums based on positive genetic test results.

37. It is also unclear as to what weighting and consideration is given in actuarial calculations to such factors as early treatment and diagnosis. For example, if a woman is shown to have the BRCA1 breast cancer gene, which is associated with familial breast cancer, but then had a double mastectomy, it is unclear whether the insurers would take into account the test result, the treatment, both or neither. We recommend that insurers should publish clear explanations as to exactly how such factors as early diagnosis and treatment are factored into their actuarial calculations.

NUMERICAL SIGNIFICANCE

38. The seven conditions chosen by the ABI's Genetic Committees are all extremely rare diseases. For example, there are estimated to be a total of approximately 6,000 people with Huntington's Disease in the UK.[95] In insurance terms the number is even fewer: Prudential dealt with seven applications involving genetic test data in 2000; the Norwich Union, 50 out of 150,000 applicants for insurance in 2000; and the CIS only 14 out of 460,000 in three years.[96] In the latter case, that is a minuscule 0.003% of all business transacted. We agree with the Standard Life insurance company, who stated that "in the view of the numbers involved we believe the effects to be insignificant."[97] We suggest that at present the very small number of cases involving genetic test results could allow insurers to ignore all genetic test results with relative impunity, allowing time to establish firmly their scientific and actuarial relevance.

39. Furthermore, while most of the companies who submitted memoranda to the inquiry would consider genetic test results in assessing risk, three do not: the CIS; Standard Life; and Virgin Direct, which is not an ABI member.[98] Their reasons are various but the CIS and Virgin do not feel the tests are sufficiently reliable as yet to be relevant and all three feel that the cost of adverse selection to insurers are at present minimal. The view of the ABI and some insurers, that ignoring genetic test results would be too costly, is contradicted by the actions of at least three insurance companies who have decided to ignore tests for the short-term and do not seem to regard it as an act of commercial suicide.

NEGATIVE GENETIC TEST RESULTS

40. The evidence presented to us seems to indicate that the only tests that are currently of any real relevance and reliability in an insurance context are those which show a negative result, i.e. an absence of the defective gene. Traditionally, for highly heritable diseases, close relatives of a known sufferer will have their insurance premiums substantially loaded owing to the perceived increased risk. In the case of Huntington's Chorea, for example, this weighting is usually around 300%, according to the Huntington's Association.[99] If such a person takes the appropriate genetic test and the result is negative, it is almost certain that he or she is not at risk from that particular condition and therefore his or her premium can be set at normal. We were very pleased to hear from the Norwich Union that the majority of applications involving a genetic test made to its company involved negative results and resulted in a cheaper premium for the applicant than if they had been judged on family history alone.[100] We are therefore concerned by the ABI's statement (not included in the Code of Practice) that should the GAIC decide that a particular test is not relevant or reliable, "The decision to ignore genetic test results will apply equally to positive and negative results."[101] This seems to ignore the clear scientific evidence that a test can prove that a person does not have a specific genetic defect, even if it cannot accurately predict onset, severity, etc. for those with a positive result. Legal & General clearly disagree with the ABI's viewpoint, as it already takes account of negative results for tests not currently included in the ten submitted by the ABI to the GAIC.[102] We welcome the policy of companies who take account only of negative test results in the calculation of premiums. The scientific and actuarial evidence currently available seems to indicate that this is the only justifiable use that can currently be made of genetic test results. We strongly recommend that the ABI and other insurers consider this when reviewing their policies.

FAMILY HISTORY

41. One issue that has been raised several times during this inquiry, is the use of family history by insurers when assessing premiums. The Cancer Research Campaign (CRC) told us that there "is no regulation of the use of family history in assessing applications for policies"[103] and the Genetics Interest Group (GIG) stated that, "those with a family history and those with existing disabilities or illnesses are likely to constitute the largest group who experience problems in the insurance context for the foreseeable future".[104] This is outside the scope of the current inquiry, but the issues are clearly closely related and cannot be ignored. We recommend that insurers explain publicly how they use family history information in the assessment of insurance premiums, and publish the supporting data.

Impact on health

42. The developments in genetic science have had enormous implications for clinical medicine. Through the network of regional genetic centres, many people whose diseases would have otherwise escaped detection can now be diagnosed quickly and accurately, and receive treatment and advice that will significantly improve their quality of life. For example, individuals shown to have the gene associated with familial polyposis are at high risk of developing colorectal cancer. Once the gene is detected the individual can be kept under regular surveillance and preventative measures put into effect should polyps appear in the large bowel. Family members shown not to have the gene on the other hand, do not need such surveillance. The take up for many of the tests is still quite small, however, especially for those diseases for which there is no known cure as yet. The Huntington's Society, for example, believe that only 15% of those at risk actually take a genetic test.[105] As the identification and understanding of more of the genes underlying human disease increases, so too does the capacity of genetic testing to bring real benefits in terms of early diagnosis, treatment and prevention.

43. Many involved in the field of clinical genetics are worried that the use of test results by insurers, and possible discrimination against them, may actively deter people from seeking the help and treatment available. As the Breakthrough Breast Cancer charity told us, "those deciding whether or not to take a test already have to face a number of emotional and social issues".[106] Adding the fear of losing your ability to obtain insurance at a reasonable price to these issues will, arguably, only make the situation worse. As Professor Bobrow observed, clinical testing "requires willing participation... and things that may compromise that are genuine causes for concern".[107] At least one insurance company, the CIS, has recognised that there may be a "possible conflict for individuals between undergoing a test as recommended by their medical advisers and not doing so because of the potential adverse effect for insurance purposes."[108] and this concern has apparently informed its policy on the use of test results.

44. It is difficult for either clinicians or insurers to prove what effect, if any, the use of results is having.[109] There are examples of anecdotal evidence from clinicians and patients that people are being deterred, but, as yet, no quantifiable research has been done, if indeed any is possible.[110] Mr Beddell-Pearce of the Prudential commented that, though this was one of the most frequently expressed arguments against the use of results, there was in fact little proof and it was probably "just a perception" that people were being deterred from seeking the help they need.[111] The GIG agrees that there is little evidence and states that a greater concern, as regards the BRCA1 breast cancer test, is probably "the assumed inability of the NHS to afford to purchase the test for all those demanding it."[112]

45. We accept that there is as yet little hard evidence that insurance issues are deterring patients from taking genetic tests. However, there is concern over this issue and if the use of tests by insurers is to be allowed, adequate research must be carried out to discern the impact of the use of genetic test results by insurance companies on the likelihood of patients seeking clinical advice. This research must be independent and have the confidence of all those involved but, as it is the insurers who wish to use the tests to protect their own interests, they should provide the funding.

Impact on research

46. For progress to be made in understanding the underlying genetic basis of human diseases, and particularly those involving more than one gene, it is essential that disease-association studies are carried out in populations of both affected and non-affected individuals. People will need to be encouraged to volunteer for genetic testing to acquire this essential data. One of the most ambitious such research project currently being considered by geneticists is the planned MRC / Wellcome database of tissue samples from a cohort of up to half a million people aged between 45 and 65.[113] By comparing which diseases, if any, volunteers contract and their genetic make-up, it is hoped to be able to obtain a much greater understanding of the genetic component in many common diseases. In such circumstances where there is no immediate or direct benefit to the individual, fears of insurance penalties may act as a deterrent. As the Association of the British Pharmaceutical Industry (ABPI) informed us "clinical trials involving patients are essential. Any issue that would inhibit patients' willingness to participate in such studies would be a matter for concern."[114]

47. Genetic testing for research purposes is quite different from that done in a clinical context, where tests are carried out for a specific purpose with clear potential benefits for the patient, for example testing for the BRCA1 gene for breast cancer in women. In genetic research tests, volunteers donate DNA samples, usually for no diagnostic purpose, to aid researchers correlating genetic defects with particular diseases. Volunteers, unlike patients, very rarely receive back the results of any tests they have undertaken. The tests being carried out at this research stage may also be very unreliable, which is why they are not being used in a clinical setting. The ABPI tells us that the distinction between diagnostic and research test results is often misunderstood by many, including some insurers.[115] This confusion may create difficulties as volunteers are not sure whether they will have to disclose the results of any tests in which they take part. The distinction between genetic tests carried out for purely research purposes and for diagnostic purposes must be clearly understood by all those who seek to use the results.

48. Concern over the impact on research is shared by several other contributors, including the Royal Society, who argued that, "A statement from the ABI that research results will not be used, and need not be declared, would be very helpful."[116] We asked the insurers present at the oral evidence session to make such a statement, and were pleased when they agreed to. We were also very pleased to receive a confirmation by the ABI that, as a result of our concerns, it would advise members not to seek research results when assessing policies.[117] The statement by insurers and the ABI that they will never use results from genetic tests carried out for research purposes is extremely welcome. We recommend that this principle be incorporated into the ABI Code of Practice without delay.

49. We note that there are a number of complex issues of confidentiality and consent involved in genetic research tests. These issues are being addressed by the House of Lords Select Committee as part of its inquiry into Human Genetic Databases. We await their conclusions with interest.[118]

Impact on under-insurance

50. The problem of under-insurance in the UK is one recognised by insurers. The CIS told us that "the British public is under-insured and under-protected."[119] It is a genuine concern, therefore, that fear of disclosing genetic test results may increase the number of people without insurance. The CIS, for example, goes on to state in its memorandum that "It is our opinion that the widespread use of genetic test results for insurance purposes without general acceptance or understanding of these tests by the general public would increase this problem of under-insurance". This was affirmed by Mr Clarke, of the CIS, at the oral evidence session.[120]

51. There is also clear public concern about the risk of creating a 'genetic underclass': a group of people for whom, owing to their genetic make-up, insurance will either be too expensive or actually unobtainable. Insurance is a necessity, in many respects, for participation in modern life, for example in obtaining a mortgage or some credit cards. There are examples from the USA where people with genetic defects face severe problems obtaining healthcare insurance, although the situation in the UK is not fully comparable owing to the provision of universal healthcare by the NHS.

52. We welcome the positive statements made by several insurance companies with regard to the provision of insurance for those with genetic diseases.[121] It may be that the mechanism set up in the 1930s, whereby, if three insurance companies refused a disabled driver motor insurance, the first company applied to had to accept them, is a useful comparison in this context.[122] The concept of the Government as 'insurer of last resort' could also be usefully explored. The Government and the insurance industry must collaborate to provide an alternative form of insurance for those who would be denied it because of their genetic make-up. Alternative insurance arrangements may help to calm public fears about the risk of a genetic underclass.

53. It has been suggested to the Committee that, in the future, some insurers may wish to 'cherry-pick' customers: offer those with a perceived good genetic make-up lower premiums than others. This would be completely unacceptable discrimination. We note that the ABI's Code of Practice states that "insurers must not offer lower than standard premiums on the basis of genetic test results".[123] The Government should ensure that this position continues to be upheld.

Regulation

GENETICS AND INSURANCE COMMITTEE

54. As discussed above (paragraphs 15 and 16) the GAIC is the principal means through which Government influences the activity of insurers in this area. We have a number of concerns about the way it operates.

Membership

55. Our first concern is about the Committee's membership. Currently, there are seven members, all of whom are nominated by different organisations and individuals:

a geneticist nominated by the ABI;

a geneticist nominated by the Chief Medical Officer (England);

an actuary nominated by the Institute and Faculty of Actuaries;

an insurance practitioner nominated by the ABI;

two members of patient support organisations nominated by the Genetic Interest Group (GIG); and

an academic with a background in epidemiology and genetics nominated by the Director of Research, Department of Health.[124]

56. Currently, Professor Raeburn sits on the Committee as a "geneticist nominated by the ABI". He is also retained by the ABI as its genetic adviser and was therefore responsible for drawing up the applications to the GAIC on behalf of the ABI. As a result, there is the rather curious and worrying situation of a geneticist being asked to assess his own application. In the Prudential's words, Professor Raeburn cannot "continue to run with the hare and hunt with the hounds" and his dual role may make any decision made by the GAIC be "seen as somewhat tainted".[125] Whilst we wish in no way to cast doubt upon the integrity or ability of Professor Raeburn, who assures us that he takes all possible steps to declare his relevant interests,[126] we believe that it is an unacceptable conflict of interests for a geneticist nominated by the ABI to judge his own test application, as a member of the GAIC.

57. The membership of the Committee more generally could also be usefully reviewed. At present there are only two geneticists (including one nominated by the ABI) on the Committee. This number should be increased to add to the expert nature of the group.[127] The 'nominated' elements of the Committee should be removed, and replaced with independent experts who are members because of their experience and expertise, not because they represent a particular organisation. Lay members have a key role to play in advisory committees. As we state in our Report on the Scientific Advisory System, however, it should be clear that their role is to bring an alternative perspective to the Committee, not to be representatives of a particular body.[128] We recommend that the membership of the GAIC should be throughly reconsidered if it is to inspire public confidence in its decisions.

Procedures

58. There have also been significant doubts raised about the procedures by which tests are examined by the GAIC, and in particular its decision to approve the test for Huntington's Chorea. Professor Bobrow, in his oral evidence, questioned the criteria set by the GAIC for the approval of tests and the quality of data it had based its decision upon, particularly the actuarial data: "if I had had that on a grant proposal I would have gone back and looked at it a bit more carefully."[129] The Wellcome Trust agreed, stating that "the justification for their decision was not subject to the rigours of peer review required for publishing in a respected journal or recommended by the HGAC."[130] The Alzheimer's Society also cast doubts upon the degree to which the actuarial evidence submitted by the ABI to the GAIC would stand up to thorough, independent peer-review.[131] In the light of these concerns over the membership and procedures of the GAIC, we recommend that the reformed GAIC re-examine the decision to approve the use of the Huntington's Chorea genetic test for use by insurers, with extensive peer review both for the data supplied by insurers and its own decisions, prior to publication.

59. At present, it appears that only the ABI submits genetic tests for approval by the GAIC. The evidence we have received on the reliability of negative results leads us to recommend that the GAIC should be given the ability to approve the use of negative test results alone for use by insurers, without allowing the use of positive results. If this happens, many patient groups and charities may wish negative tests for specific conditions to be approved, to allow their members with positive family histories but negative gene tests to benefit from reduced insurance premiums. It is doubtful, however, that these bodies would have the necessary expertise and resources to complete the GAIC application procedure. We recommend that some mechanism be established to facilitate applications for the approval of tests by the GAIC, from bodies other than insurers.

Resources

60. The quality of the GAIC's work will largely depend on the resources it has available. We were, therefore, extremely concerned by the supplementary memorandum we received from the GAIC in which we asked them to describe and comment upon the resources they currently receive (as mentioned in paragraph 14 above). We were shocked to read the break-down of the secretariat in fractions, with the total number of staff amounting to 1.6 people.[132] The GAIC has also apparently been dependent upon the goodwill of actuaries nominated by the Faculty and Institute of Actuaries to provide independent actuarial reviews free of charge, instead of paying the normal rate.[133] The situation can only worsen as the GAIC is expected to undertake more decisions, as well as fulfilling its monitoring role effectively. It is imperative that the GAIC be properly resourced for the work it is doing.

SELF-REGULATION BY INSURERS

61. The ABI's Genetic Testing Code of Practice is seen by most insurers as the principal form of regulation in this area. We have highlighted many of the good aspects of the Code above (paragraph 20), including the voluntary moratorium on the use of test results for life assurance linked to mortgages under £100,000. The establishment of the UK Forum for Genetics and Insurance (UKFGI) and the actuarial research being carried out at Heriot-Watt University are also indications that some insurers are aware of the importance of this issue.

62. In addition, we should like to note that the ABI, and insurers, were generally acting in something of a regulatory vacuum. The Government chose not to take account of recommendations made both by our predecessor Committee and the HGAC advising they adopt a more interventionist role in this controversial area. Both committees predicted that the issue would not go away and could create serious problems in the future. The ABI were left to formulate their own policies, and, arguably, did the best job that could have been expected of them. It is interesting to note that when specifically asked by this Committee to comment on the regulatory system in their written memoranda, of the eleven companies none mentioned any direct Government oversight. The previous unwillingness of Government to become involved in this area has contributed to the atmosphere of confusion and ignorance that pervades the use of genetic test results.

63. The Committee appreciates the arguments put forward by the insurers in favour of self-regulation over legislation. Professor Bobrow argued that "science moves very quickly and legislation moves very slowly" and that the exact contingencies which would need to be covered would be very difficult to predict.[134] The insurers, similarly, stated that, in this rapidly changing area of science and insurance practice, a "code of practice could respond more rapidly to developments".[135]

64. On the other hand, many feel that self-regulation in this controversial area is simply not enough. The Royal Society stated emphatically that "Self-regulation is not an adequate regulatory system."[136] There must be doubts whether the ABI, a trade organisation funded by insurers to represent their own interests, is the right body to regulate the use of genetic test results.

65. While all ABI members subscribe to the Code of Practice, there is some concern that it is not always complied with. The Alzheimer's Society informed us that in a small survey they undertook, eight of the nation's leading insurers were asking customers applying for life assurance linked to mortgages under £100,000 to disclose genetic test results, in clear breach of the ABI Code. The Society observed that, "The ABI's failure to ensure compliance of its own members on this issue cast doubts on the industry's ability to regulate itself.".[137] The ABI must act as a matter of urgency to convince the Government and the public that the Code of Practice is being complied with.

66. It is also unclear whether the ABI has the means to compel companies to comply with the Code. Mr Beddell-Pearce of the Prudential, spoke of "stiff penalties and meaningful sanctions",[138] but it has not been made clear what these are. No member of the organisation ever seems to have been expelled from the Association. It is also uncertain how, the admittedly small number of, companies outside the ABI are regulated. Insurers must prove that they are capable of regulating themselves effectively and thoroughly, with sanctions in place to ensure compliance. The ABI's Code of Practice is a welcome step in the right direction by insurers but it is inadequate in its present form. The reformed GAIC should make recommendations to the ABI for its Code of Practice. The GAIC should also closely monitor insurers' compliance with the Code.

67. One of the strongest and, in our view, most justified criticisms of the way in which the ABI has regulated the industry has been in its attitude toward the GAIC's approval of genetic tests. Nearly all of the ABI member companies who submitted memoranda to this inquiry state their faith in the GAIC's decision-making process and their willingness to respect its decisions. Yet, the ABI has acted in what Professor Bobrow described as "an irrational and unfriendly"[139] manner and insisted that its members have the right to use tests prior to their approval, or otherwise, by the GAIC. This seems especially irrational in the light of the grave doubts that exist over the relevance of these tests, and the relative infancy of the whole science upon which they are based. The ABI's decision that four tests, which their own Genetics Adviser had approved for use, are no longer supported by sufficient data to be used, seems to support those who argued the need to 'look before you leap'. Insurance companies were wrong to use the results of genetic tests in assessing risk before they had been approved by the GAIC. We recommend that all insurance companies should immediately cease to use the positive results from any genetic test that has not been explicitly approved by the GAIC.

68. The arguments for self-regulation have also certainly not been helped by what the Wellcome Trust described as a "significant mistrust among customers of insurance companies".[140] Deserved or not, there is a general opinion that insurers cannot be trusted, and would immediately seize any opportunity to enrich themselves at the customers' expense. For example, only 7% of those questioned in a MORI poll (commissioned by the HGC) thought that insurers could be trusted to use genetic data responsibly.[141] The evidence has not borne out this conclusion but we note the observation of the Virgin Direct insurance company, which is not a member of the ABI, that "After ripping many of them [customers] off over personal pensions and endowment policies, we believe the industry might perhaps better focus its energies on rebuilding consumer trust.". The insurance industry would certainly benefit from concentrating efforts on building public confidence in its actions and motives, rather than giving itself the right to extend its ability to load premiums.

The way forward

69. In considering how to take the issue of regulation forward, there are strong arguments on both sides. We welcome the fact that the insurance industry clearly has made some attempts to regulate its own actions and we do not wish to undermine the ability of insurers to operate in a commercially competitive manner. On the other hand insurers have acted in a precipitate manner, ignoring the doubts over scientific and actuarial relevance.

70. We do not believe that legislation banning the use of all genetic information by insurers, as has been enacted in some countries, for example Austria or Denmark,[142] is the correct way forward on this issue. This would not only harm the insurers, but more importantly it would deny those with a 'bad' family history and a 'good' genetic test result the opportunity of obtaining insurance on a standard rating. As GIG told us, "policy makers should take care to ensure they do not make things worse by simply banning the use of data altogether".[143] We do not believe that legislation denying insurers access to all genetic test results would be appropriate.

71. We suggest that, the best way forward for the Government and industry would be a voluntary moratorium on the use of all positive genetic test results by insurers for at least the next two years. During this time more research should be done to establish the actuarial and scientific relevance of genetic test results to the assessment of premiums, and the possible consequences for research and healthcare. If the insurers are unable, or unwilling, to regulate themselves and enforce this moratorium, we recommend that Government enforce its will by legislation. We further recommend that insurers should still consider negative test results in assessing insurance applications throughout any moratorium.

72. One point that has been made clearly to the Committee is that the use of test results by insurers is extremely unpopular with the general public. In a MORI poll commissioned by the HGC, only 8% of people thought that results should be used for setting insurance premiums, and 78% disagreed (57% strongly disagreed) with insurance companies being able to see genetic test results.[144] Were a 'scare' to emerge over genetic testing, the Government might feel compelled to ban the use of genetic data completely. This would benefit none of those involved in the area. It would be better for the insurance industry to act responsibly now, rather then be forced into a commercially compromised position in the future.

73. Our inquiry has been largely concerned with the situation currently in operation in the UK, and, in particular, with the ten tests for seven conditions that the ABI's Genetic Adviser advised companies to take into consideration. We are aware, however, that the science of genetics is developing at a rapid pace, and in the recent past has progressed much faster than had been originally anticipated.[145] Government and the insurance industry need to give serious consideration to the full ramifications of the effects of significant developments in genetic testing and screening, and in the ability to predict diseases more accurately. The HGC will have a key role in advising Government on the wider effects of these continuing developments. We strongly recommend that the HGC should continue to monitor the use of genetic test results by insurers and the consequences of their actions, in their widest possible context, and advise Government on further developments as they arise and in a timely manner.

74. The HGC's effectiveness will depend, however, on its resources in the same way as the GAIC's does. In its supplemental memorandum to us, the HGC makes it clear that the increased workload that they are being asked to cope with and the added financial burdens involved in holding meetings in public are straining their budget to the limit.[146] The HGC's programme and, in particular its efforts to involve as many of the public as possible in its consultations, should not be undermined by a lack of resources. The Government should, as a matter of urgency, review the funding and resources the HGC is allocated. If the HGC is to receive extra work as a result of our recommendations, this should also be reflected in its budget.

Conclusion

75. As is often the case with Select Committee inquiries, it is clear that our deliberations have helped to 'focus the minds' of the key players in this area, especially the insurance companies. We have been pleased to receive supplementary memoranda from two insurers altering their policies in the light of our questioning and to have encouraged the ABI to issue a statement on research and a clarification of its policy. We regret that the insurance industry insisted on using genetic tests before their reliability had been fully established. In hindsight it would have been better if the insurance industry had proceeded far more cautiously in this difficult area, which at present can bring them little financial return but a great deal of adverse publicity.

76. The remark that struck us most strongly was one made by Professor Bobrow:

"the simple inherited diseases we are looking at today are an area in which people who are pretty seriously disadvantaged do not need more trouble in their lives; and trying to load insurance issues on top of their medical issues could only be justified if the evidence was really clear and if the financial implications were significant.".[147]

From the evidence we have received, we believe that neither of these conditions has been satisfied. If the choice has to be made between exposing insurance companies to a small degree of short term risk, and increasing the stigma and discrimination many of these sufferers feel, then the choice for Government, and our society, is clear.

75 Evidence, p 53, paragraph 11. Back

76 Ibid., paragraph 13. Back

77 Evidence, p 19, paragraph 4. Back

78 Evidence, p 53, paragraph 16. Back

79 Evidence, p 49, paragraph 2.1. Back

80 Evidence, p 2, paragraph 8. Back

81 Evidence, p 55. See also: evidence, p 80 and p 83, paragraphs 9-12. Back

82 Q 8. Back

83 Evidence, p 17, paragraph 5.6. Back

84 Q 99. Back

85 Evidence, p 56, paragraph 2. Back

86 Q 6. Back