Whole‐genome approaches, which are replacing targeted tests in research and clinical practice, increase the chances of ‘incidental findings’ (IFs) – that is, those unrelated to the reason for the test. IFs raise several challenging questions, such as are researchers obliged to disclose IFs, and does this change if the researcher is also a clinician? How can the clinical significance of IFs be determined, and what significance level should determine disclosure? Could family members be tested to help to clarify significance, and if so, how? What should happen if adult‐onset risks are found in children or prenatally? No consensus currently exists about disclosing IFs from research, or about how participants can be helped to make decisions about and give consent (not) to receive them. We recommend that as more research studies that use genome‐wide tests are launched, longitudinal empirical work be conducted to explore participants' experiences and inform best practice for consent and, where relevant, feedback.
Key Concepts
Using genome‐wide tests increases the likelihood that findings outside the target area will be made. These are often called ‘incidental findings’ (IFs). Early empirical data show that the chances of finding IFs range between 1% and 7%, depending on the test used.
IFs in genetic and genomic medicine differ from those found in imaging or biochemistry because they can predict future risks and risks to family members.
Researchers generally are not thought to have a duty of care to research participants, but some experts argue that they ought to return IFs that are clinically valid, medically important and actionable, or even ‘hunt’ for additional findings that fit these criteria.
However, research funding for such individualised approaches is often insufficient: protocols do not always include quality assurance procedures, and research teams lack health professionals who can communicate findings to participants.
The research‐clinical practice boundary is often blurry: participation to some research studies is offered to patients in the clinic and some promise clinical feedback (e.g. the United Kingdom's 100 000 Genomes Project).
Clinical validity is difficult to determine in both research and clinical practice, and this will not necessarily become easier with time, meaning upon discovery, IFs are better thought of as ‘potential incidental findings’.
Paediatric IFs introduce more complexity: guidelines and legislation suggest that testing children for adult‐onset conditions should be deferred, but with IFs, the question is not about whether to test the child, but whether to reveal information already found.
Prenatal IFs raise the additional issue that women might terminate pregnancies based on information that has uncertain significance.
Research about participants' preferences shows that people want to know about all IFs, but this research has been hypothetical and cross‐sectional; thus, satisfaction with real‐life decisions has not been explored.
The problem of hypothetical decision making is also relevant in real‐life research settings: at the time of consent, participants are unlikely to have thought about their choices or their implications. Empirical research is thus required to determine best practice in consent and disclosure.