BSHG STATEMENT ON GENETICS AND LIFE INSURANCE - May 1998

Note: More recent developments (April 2001) are covered in Human Genetics Commission announcement of 3 year moratorium on use of genetic tests results for insurance (mentioned in Alan Milburn's recent speech on Clinical Genetics) , and the joint statement on use of research results.

1. The special nature of genetic disadvantage

The goal of medical genetics, and the guiding principle of the British Society for Human Genetics (BSHG) in the matter of insurance, is to help families at a genetic disadvantage live and reproduce as normally as possible. We recognise, however, that this must be reconciled with the legitimate requirement of insurers to avoid an unacceptable degree of anti-selection.

Clinical practice has confirmed that the disadvantage is not just the disease and the suffering it causes, but also the threat that the disease may develop in later life or be transmitted to offspring. The impact of a diagnosis and any 'lost opportunities' it brings spreads wider than for non-inherited disease. It sends ripples through the family. This ripple effect is one reason why inherited disease might justify special measures.

A second reason why inherited disease might justify special measures lies in the nature of the decisions that individuals and families have to make in order to cope - decisions that are made all the more taxing by advances in genetics. People threatened by genetic disease use various, often complex, strategies to 'carry on as normal'. They juggle their well-being and peace of mind with discharging what they see as responsibilities to themselves (in terms of healthcare) and to their family members, born and unborn. The decisions that are coloured by genetic risk, or the threat of it, tend to fall into that sphere of personal responsibility and choice where undue outside influence or discrimination, commercial or otherwise, are usually seen by our society as inappropriate - choosing a partner, decisions to have children or continue a pregnancy, choosing a home or job, and, not least, consenting to genetic or other testing and to medical treatment. Here again, it is the perceived threat of undue discrimination that tends to modify behaviour rather than what might actually be the case. There is a perceived threat that, once tested, others might use that personal information against you in relationships, in employment, in social opportunities, or in health and welfare services. That such a worry is a real one is shown by the emergence of genetic privacy laws in the USA.

A distinction is made between people who are already affected and those who are healthy. This Statement concerns only presymptomatic risk assessment by genetic testing (usually by DNA analysis).

Monogenic (mendelian) diseases raise different issues from multifactorial disorders, so each is considered in turn.

Monogenic disorders
The relevant mendelian disorders are highly penetrant, but of adult onset. These are relatively rare, and the effect of any anti-selection by a potential customer exploiting knowledge of his (undisclosed) genotype comes largely from the size of the sum for which his life is insured, not whether he decides to insure when he might not have done so (Macdonald, 1996).

A compromise that would be beneficial to families at genetic disadvantage (and permit confidence that they can obtain the necessary life insurance cover for a house mortgage, for example) would be the offer of life insurance, without genotype disclosure, up to an agreed maximum sum. The agreed sum should be no less than the limit for each age bracket that the company sets for a Medical Examination. Macdonald (1996) gives as typical �300,000 up to age 40 years, falling to �25,000 at 61-65 years, and zero above 65 years. This is in keeping with the 'as normal as possible' goal, and may allow people to shop around for the best deal.

Cover, up to the agreed ceiling, should be available for all life insurance purposes without genotype disclosure. Whilst mortgage-associated life insurance was highlighted in the Science and Technology Committee report (1995), and has been subject of a concession by the ABI, there is no reason to confine these arrangements to one type of life insurance. Where there is no disclosure of genotype (as above), it follows logically that the insurance companies should not request or use family history information either.

Multifactorial disease
These diseases are most likely triggered by specific combinations of functional DNA polymorphisms interacting with the environment in ways that are subject to behavioural changes. The complexity of their causation makes theoretical prediction unlikely. For two reasons it is going to be very difficult to establish genotype-specific predictive empirical risk figures for multifactorial diseases, and therefore we see little or no role for genetic data in underwriting decisions.

First, data showing that a certain genotype is associated with susceptibility to a disease do not prove that the genotype is associated with an increased overall risk of early death or morbidity. Current genetic research is showing increasing success at identifying genetic susceptibility factors for particular common diseases. Many of these diseases carry an increased risk of early death. But given the complexity of genetic susceptibility, it is entirely plausible that there may be a compensating protective effect against some other disease. We are not aware of good data that address this problem. Studies of the concordance of monozygotic twins for insurable events might be initiated to explore the limits of prediction, given total knowledge of the genotype.

Second, and more crucially, the past is a poor guide to the future. The main aim of research into genetic susceptibility factors is to help understand the pathology and devise lifestyle changes or prophylaxis to avoid the risk of disease, or treatment to minimise its impact. Over the timescales covered by insurance policies, the likelihood of substantial progress in prevention cannot be ignored. The diseases that we are most likely to be able to prevent are precisely those where we are best able to predict susceptibility.

 

We firmly endorse Paragraph 2 of the ABI Code of Practice, stating that applicants must not be asked to undergo a genetic test in order to obtain any type of insurance.

In so far as any genetic test data are sought, it is inappropriate that the client be asked to disclose all existing genetic test results. Must an applicant disclose that a prenatal test, carried out on him as a fetus, showed that he had a normal karyotype? A blunderbuss approach would result in the insurance company holding personal information for which it had no use. There is a requirement for evidence-based underwriting. Any requirement for disclosure should be restricted to results that the company can show, on the basis of published and actuarially validated evidence, is reasonably likely to affect the risk insured, over the timescale insured. For life insurance, this is the risk of premature death in the 21^st century, not the risk of developing a specific disease in the 1990s.

In addition to the provision of life insurance up to an agreed sum without any genotype disclosure, as described above, any disclosure of common 'normal' genotypes would also be inappropriate. Current life insurance practice in the UK results in premiums for 95% of applicants being based only on age, sex and smoker status, without reference to occupation, health or lifestyle. By analogy, no adjustment to life insurance premiums should be made on account of a genotype that is present in 5% or more of the population. If no use is to be made of the genotype by insurers, then it need not be disclosed. The reinsurance scheme suggested in the Science and Technology Committee report (paragraph 246) allows non-disclosure of genotypes at the time the contract was made. The handling of the contract would be simplified by confining it to the few genotypes of intermediate prevalence and substantial predictive value.

It would be inappropriate to offer 'preferred life' insurance at below standard premium on the basis of genetic test results, and we welcome the prohibition of this in the ABI Code of Practice.

One great concern is that the perceived threat of undue discrimination will tip the balance against somebody seeking testing to obtain improved medical management and/or reassurance. A second concern is that people may be encouraged to seek out over the counter genetic tests and not share the result with their general practitioner (GP), for fear that the GP will be obliged to disclose it to an insurance company. We welcome the recommendation of the Advisory Committee on Genetic Testing that would limit over the counter genetic testing to tests for carrier status for recessive conditions, because by-passing the GP will lessen the chance of optimal health care.

There is an opportunity for the insurance industry to counter these fears by acknowledging that those facing genetic risks need special consideration (and probably small concessions) that would prove of great emotional and practical help to the families and yet pose little if any threat to the life insurance industry. The principle that firms may have to incur modest extra costs to enable certain minorities to be treated as normally as possible is not novel, being enshrined in the Disability Discrimination Act of 1995.

House of Commons Select Committee on Science and Technology (1995). Human Genetics: the science and its consequences. 3^rd report. Vol 1. HMSO, London.

Macdonald AS. How will improved forecasts of individual lifetimes affect underwriting? Discussion meeting of the Royal Society and Actuarial Profession, Sept 25th 1996.