Direct-to-consumer genomic testing is a thriving market – for example by June 2020, 23andMe had sold over a quarter of a million testing kits in the UK. The House of Commons Science and Technology Committee has been looking into the opportunities and challenges raised by direct-to-consumer genomic tests, and recently published a report making recommendations as to how these tests are regulated.

Direct-to-consumer genomic tests present an interesting regulatory challenge – the tests available are hugely varied in scope and quality, their aims vary from commenting on ancestry to informing about health risks, and the providers are based in many different countries. Used well, some direct-to-consumer genomic tests have potential to improve customers' health, but without scrupulous communication around their risks and limitations, there is a serious risk that even tests of a very high technical standard could do more harm than good.

It was very encouraging to see the report's recommendation that 'the Government should extend the scope of the performance requirements on direct-to-consumer genomic tests to explicitly cover clinical performance as well as analytical performance'. Many tests advertise impressive-looking technical accuracy, and at present, this risks disguising issues like very poor sensitivity from consumers. For example, some direct-to-consumer tests designed to tell a customer whether they have a cancer-predisposing BRCA variant (that would put them at increased risk of developing certain cancers, particularly breast, ovarian and prostate) only check for three specific variants in BRCA1 and BRCA2 despite there being thousands of potential variants. Often these have a high technical accuracy of over 99 percent, meaning that if they say that you have a particular variation you are likely to have it. However, their clinical sensitivity may be less than 20 percent as they are not testing for all of the variants of concern. Clinical sensitivity of 20 percent means the test will miss around 80 percent of people with a cancer-predisposing variant in one of these genes. If tests had to make their clinical performance figures as prominent as their analytical performance figures, this would reduce the risk of consumers being seriously misled by their results, and might also incentivise development of better tests.

The report also suggested that the government consider 'requiring medical supervision or the provision of genetic counselling for at least some types of genomic testing offered directly to consumers'. They highlighted that when genetic tests are requested via clinical pathways, a person's medical history and family history influence both which test is performed, and how its findings are interpreted. In contrast, direct-to-consumer genomic testing usually operates via a 'one-size-fits-all' model and so cannot adapt either the test or its interpretation in response to the clinical context. A requirement for companies offering direct-to-consumer genomic testing to offer genetic counselling for customers requesting certain tests, would go some way to mitigate this. While it could not remove potential inadequacies with the test itself, at least these could be highlighted such that a consumer would be less likely to be falsely reassured, or unduly concerned, by their result.

A further welcome recommendation was that 'the Government should consider which, if any, genomic tests for asymptomatic children should be able to be provided directly to consumers, including whether there should be a ban on the provision of genomic tests for use on children that do not meet the criteria of the UK National Screening Committee'. At present, there is no legislation preventing parents buying tests for their children, and some companies specifically target this market, for example claiming to diagnose child talent. The ethical issues here are huge, from the potential for children to find they have a high chance of developing conditions like Alzheimer's disease in late adult life, to being expected to become a musical prodigy on the basis of a test that can identify no such propensity. Legislation introduced here could be used to ensure that tests could only be performed on children if this were likely to further the children's interests, or at the very least not harm them (more guidance is forthcoming as the British Society for Human Genetics 2010 guidelines on genetic testing in children are currently being revised). The current situation means there is not much to prevent a direct-to-consumer genomics company exploiting parental anxiety in whatever way they think will make them the most money.

The report from the Science and Technology Committee also made various other recommendations, including that 'the Government should require manufacturers of direct-to-consumer genomic tests to have the performance of their tests assessed by an external body prior to placing their products on the UK market', and that this performance should be regularly reassessed. Here again, it is useful to think of a test's clinical performance and not just the technical accuracy. The committee also highlighted the need to assess the impact that direct-to-consumer genomic testing is having on the NHS. This is timely, especially given the contrast between suggestions from companies that their tests will ease pressure on the NHS by preventing illness, and concerns from professional societies such as the British Society for Genetic Medicine and the Royal College of General Practitioners that they might put pressure on the NHS when recipients turn to it to help make sense of their test results.

Direct-to-consumer genomic testing is here to stay, and new approaches to regulation are much needed. Currently, most providers are able to self-certify that their tests meet requirements, and whilst some companies may be very scrupulous in ensuring that they do, the potential for misleading customers is high. The Science and Technology Committee report recommendations, if adopted, would guide further provision of direct-to-consumer genomic tests in a more responsible and accountable direction. This involves clearer communication about both what such tests are actually capable of, and what their results might mean in the context of a person's life, as well as clamping down on unethical practice such as providing tests with major health implications without appropriate support, and marketing testing of children without evidence that it will cause more benefit than harm.