Introduction

Patients attending genetic clinics are often the main gatekeepers of information for other family members.1 However, patients do not always pass on this information to relatives.2, 3, 4, 5 This can be important because these relatives may then be denied the opportunity to make informed health choices, for example, in relation to reproduction or the management of the risks of disease.6, 7, 8, 9, 10, 11 Furthermore, there may be other family members, not related by blood, who could benefit from the knowledge.6, 9, 10, 12 Nondirectiveness has long been held to be a central principle of genetic counselling,13 despite doubts about whether this is always possible in practice14 or appropriate to all genetic counselling situations.15 While disclosure without consent would violate the client's privacy,16 any steps taken to persuade or encourage a counsellee to disclose information to another family member would clearly amount to directive counselling.

Family communication

Empirical research suggests that family communication about genetic risk is not always straightforward.17, 18, 19, 20, 21 Although clients may acknowledge their obligation to pass on relevant genetic information to family members, this may be balanced against a desire not to cause anxiety or alarm.22, 23, 24 Difficult issues may arise about the timing and content of disclosure and the judgement as to who ‘needs to know’, which, of course, may depend upon how the disorder in question is (thought to be) inherited.4, 9, 20, 25 While communication of genetic information may follow the usual family channels,25 practical barriers such as geographical distance, family rifts, divorce, separation, adoption and large age gaps between siblings may impede communication.17, 22, 23 Complex family dynamics may make it more difficult to convey information to siblings than adult children.26 Communication may be influenced by patterns of mutual surveillance for signs of disease, and moral scrutiny and judgement as well as beliefs about inheritance and disease.19, 27, 28, 29 In addition, information provided in genetic counselling may be imperfectly understood, leaving genetic professionals and family members with different views because of their different understandings of the risks and benefits of disclosure.

Ethical and practitioner issues

The failure of a client to disclose important information to relatives will raise ethical issues for the professionals involved,23, 30 but professional disclosure without consent could undermine trust in the counselling process.16 To address this issue, profession-led guidelines in the UK, Australia, and the USA recommend an approach which emphasises the primary importance of maintaining confidentiality but which permits disclosure by health professionals without the patient's consent in ‘exceptional circumstances’.8, 31, 32 These guidelines provide a framework within which clinicians can work but there has been no specific legislation regulating the flow of genetic information in families.33, 34, 35 An inquiry by the Australian Law Reform Commission and the Australian Health Ethics Committee6 concluded that privacy legislation inappropriately constrains health professionals' decisions about the disclosure of clinically relevant information to genetic relatives and that this situation should be remedied by legislation and the modification of professional guidelines, but this has been criticised in a response by the Human Genetics Society of Australasia as promoting, or perhaps making obligatory, forced disclosure by professionals. While English law underlines a strict interpretation of a doctor's duty of confidentiality based on an individualistic perception of patient autonomy and an overriding utilitarian principle of prevention of harm to others, it has been argued that family members deserve more legal recognition and that the current individualistic legal approach to confidentiality is too narrow.36

The attitudes of patients and clinicians towards the breaking of confidentiality may vary according to the cultural context1, 37 and the nature of the health care system.10, 38 Hypothetical studies suggest that many professionals would disclose information to family members without consent under some circumstances.39, 40 Others have argued that the obligation to disclose to relatives is not a question of ‘rights’ but rather of family responsibilities41, 42 and some caution that pressure to disclose genetic information may undermine family values and disrupt relationships.43

Little is known about the frequency with which these ethical problems arise within everyday clinical genetics practice.44 Retrospective surveys of members of the American Society of Human Genetics and/or American College of Medical Genetics and National Society of Genetic Counsellors found that a quarter of the clinical geneticists and half the counsellors reported having had patients who refused to inform family members at risk. While a significant minority seriously considered informing relatives without consent only a single geneticist and a single counsellor reported having done so. Genetic counsellors reported that emotional issues were the major reason not to disclose without consent, while clinical geneticists more often cited patient confidentiality and their own legal liability.45, 46

An interview-based study of US patients affected with a variety of genetic or nongenetic conditions found that their views and experiences of disclosure within families did not differ between genetic and nongenetic conditions. The authors suggested that a focus specifically on genetic information, treating this as a special case, may therefore be unwarrented.47

Aims of Study

In order to inform discussion about confidentiality and family communication in the context of genetics, we need to know more about what actually happens in clinical practice. The aim of this prospective collaborative study was to collect information on the frequency with which clinical geneticists and genetic counsellors become concerned about the refusal of clients to disclose important genetic information to their relatives, the circumstances in which these situations arise and the actions then taken by the genetic health professionals.

Methods

A total of 12 Regional Genetic Services in the UK and two in Australia agreed to collaborate in this study. Clinical geneticists and genetic counsellors in the UK centres provide a comprehensive genetic counselling service for adult (including cancer), paediatric and reproductive referrals for their population; the two Australian centres jointly provide a similar service, with one centre dealing with cancer genetics only. Ethics Committee approval was obtained for the collection of nonidentifiable information to be collected prospectively, without patient consent. Episodes of nondisclosure were collected over a 12-month period (commencing on a different date at each centre between October 2000 and July 2001). Nondisclosure was defined as: ‘a situation in which a clinician believes (i) that a family member should disclose information to another family member because failure to do so could lead to potential harm for that relative and to her family members, but (ii) that this disclosure seems unlikely to take place.’

For each episode, details were recorded on the diagnosis, sex, age, ethnicity and genetic status of the proband, how the issue of nondisclosure arose in the clinic, the reasons for the nondisclosure as given to the professional by the proband, and the professional's response (any subsequent actions and/or ongoing concerns). Approximate annual consultation figures for the collaborating centres were also obtained to estimate the frequency of episodes of nondisclosure.

The cases were discussed between the UK collaborators and lead researchers at two meetings to ensure a common approach to the reporting of cases and to establish agreement on inclusion criteria. Separate discussions were conducted in person and by mail with the Australian contributors. After all cases were reported, the lead researchers identified common themes among cases, and these categories were refined after clarification with the collaborators. Follow-up data on cases were obtained in January 2003 (6–31 months after cases first reported).

Results

Cases

A total of 65 cases of nondisclosure were reported, which amounts to a very small proportion of the nearly 40 000 genetic clinic consultations conducted annually in the collaborating centres. There was a wide range of Mendelian and chromosomal conditions represented but the conditions most commonly reported were Huntington's disease (HD), chromosome translocations and hereditary cancer syndromes (Table 1). These included 39 cases of the failure of parents to transmit full information to their adult offspring, 22 cases where siblings or other relatives were not given information and four cases where information was withheld from partners – including former and prospective partners.

Table 1 Approximate frequency of nondisclosure in consultations for different disorders
Full size table

Nondisclosure to adult offspring

There were 39 cases where adult children were not given full information by their parents about the family's hereditary disorder. In half of these cases involving nondisclosure from parents to adult children, the genetic diagnosis had recently been confirmed or revised through molecular testing. There were three women affected with breast cancer, who did not intend to inform their adult daughters that a BRCA1 or BRCA2 mutation had been identified, and one male, who had predictive testing for a BRCA mutation while intending not to inform his adult daughter. There were 11 cases where a parent with a confirmed HD mutation (whether or not currently affected) and/or their spouse withheld knowledge of the diagnosis from their adult children. In one case, an affected man and his wife chose not to disclose nonpaternity to their adult children, who believed themselves to be at 50% risk of HD. In two cases with an unfavourable result of an HD predictive test, the consultant indicated that they had told their children their test result but the children still believed they were at 25% risk (rather than 50%) when they subsequently and independently attended the genetic clinic. These cases were included in the study but, in such circumstances, it may be difficult to know whether it was the ‘parents’ who had failed to provide the information or the ‘children’ who had failed to appreciate it. This group of nondisclosers has also shown that significant life events can diminish the immediate importance of the genetic information: there was one case each of heroin addiction, imprisonment and major mental illness affecting the at-risk ‘adult’ children to whom full disclosure had not been made.

Nondisclosure to siblings and other relatives

In all, 22 cases fell into this category. Eight involved a balanced chromosome translocation, including two where the consultant had been adopted and did not want to contact their biological family, although this was feasible. One patient with a confirmed colorectal cancer (HNPCC gene) mutation, felt unable to discuss this with their adult mentally handicapped sibling, and accepted the advice of the sibling's GP and carers not to pass on the genetic information. The patient therefore acquiesced in a decision made by health and social care professionals about the sibling, raising important issues about the welfare of those with intellectual disability cared for in our communities. One consanguineous couple would not inform their possible carrier siblings of their child's diagnosis, which entailed gender ambiguity. One HD mutation carrier declared an intention to emigrate without disclosing any information to their siblings, unaware of the family history or to their doctor. Another HD mutation carrier had for many years been the victim of extreme violence from their sibling and would not make contact for this reason. Two cases involved carriers of cystic fibrosis (CF), one male and one female, who were providing gametes for assisted conception and would not disclose their CF carrier status to relatives, who were unaware that a family member was undergoing fertility treatment.

Nondisclosure to partners

In four cases, information was being withheld from partners or expartners. In one case, a pregnant woman was unaware of her partner's family history of HD. In two cases, parents (one mother and one father) would not disclose information about a genetic diagnosis in their child to an estranged expartner who could be at risk of having affected children. In one family, a young woman and her parents withheld information from her new husband about the risk of an autosomal recessive condition in the family.

Professionals' reports of reasons for nondisclosure given by consultants

When reporting episodes to this survey, our colleagues described the reasons given by their clients for not disclosing information to relatives. Most clients gave complex reasons for withholding information from relatives, more often citing concern and the desire to protect their relatives rather than poor family relationships (Table 2). The desire to avoid causing anxiety was the most frequently given reason for the decision not to disclose information. Many clients made their own judgements as to whether their relatives personally needed the information or could cope with it. Other frequently cited reasons were problematic family dynamics, including loss of contact, poor family relationships and fear that disclosure might disrupt a family member's marriage plans – there were three families with arranged marriages pending. In three further cases, it was another relative – not the client – who was ‘banning’ disclosure. Five consultants feared being blamed if the information was disclosed, and eight felt unable or unwilling to take responsibility for informing relatives. Eight consultants did not want to inform their relatives because they felt it was better for them not to know about the genetic risk; five of these eight were still expressing strong emotion over learning about the genetic diagnosis themselves. For just six consultants was the issue of privacy given as the principal reason for failure to disclose.

Table 2 Professionals understandings of clients' reasons for nondisclosure
Full size table

Professionals' actions

In most cases, clinical geneticists and genetic counsellors reported that they ‘took further steps to persuade the consultant to make a disclosure’. These steps included further discussion with the consultant, both to reinforce the professional's view that disclosure was important and to clarify the consultant's reasons for nondisclosure. Discussion took place at the first consultation in which the nondisclosure was identified, and again at follow-up contacts in clinic, by telephone or at home visits. Other actions included involving experienced colleagues in the discussions with the consultant and using written reminders in an effort to prompt disclosure.

While there were no reports in this series of genetics professionals disclosing information to relatives without the consultant's permission, there were several instances of active offers made to facilitate disclosure. These included enclosing copies of the summary letter with the suggestion that these be passed to relatives, and sending clients ‘open letters’ – including the offer of a genetic clinic appointment – which could be given to relatives.

Discussion

In this prospective study, clinically important episodes of nondisclosure within families were identified in less than 1% of all consultations. It may be that patients usually do disseminate important genetic information to their relatives, or genetics health professionals may be unaware of the extent to which communication within families does not happen. A collaborative study of this type, relying on a large number of professionals to report their subjective concerns, can only provide a rough estimate of the frequency of professional concern about nondisclosure. Professionals will undoubtedly vary in the attention they give to the issue of the dissemination of information to other family members; they will often not know whether family members actually do pass on information and there may or may not be follow-up to enquire as to whether this happens. Finally, some obstacles to the transmission of information may be less likely than others to be noted in such a survey. In particular, patterns of family communication that result in blocks to the passing of genetic information from parents to their (adult) children have been recognised for decades48 but such longstanding behavioural patterns may not be noted by professionals recording discrete episodes. Where parents know the results of genetic carrier tests on their children, for example, they may omit to pass on this information as the child matures,49, 50 but it would be difficult to pinpoint an episode of nondisclosure. Despite these notes of caution, however, our figure is the first population-based estimate of such clinical episodes giving rise to professional concern.

Our exploration of clients' reasons for failure to disclose information was second hand, but it was of interest that the motivation appeared more commonly to be a reluctance to cause anxiety than poor family relationships. The nondisclosing clients were often making judgements as to whether a particular relative could cope with, or would want to be given, the information in question – parents in particular making these decisions on behalf of their (adult) children. The extent to which such explanations are post hoc justifications for the messenger's reluctance to be blamed for the ‘bad news’ is unclear, but considerations about the best interests of the potential recipient of the information are a consistent feature of accounts from other sources too.9, 27 In only eight cases did the client take the general stance that it is better ‘not to know’ one's genetic predicament than to know it.

Nondisclosure rates vary between conditions with nondisclosure issues arising relatively frequently in HD families, often from a desire to protect family members from distressing knowledge in the absence of effective therapy. Studies of communication and decision-making in HD families have described how family members may reach decisions using rather different considerations than those adopted by professionals.51, 52, 53 Another factor contributing to the higher number of nondisclosure episodes reported in families with HD may relate to the greater number of contact hours between the professional and the consultant in this disease. The consequence may be a deeper understanding of the inter-relationships between the consultant and family, alerting professionals to the potential issue of nondisclosure. There may be less motivation to disclose risks in families carrying translocations where the risks of a liveborn child with a chromosomal aneuploidy may be small and the miscarriage of affected pregnancies may be interpreted as ‘nature taking its course’. There may also be a greater wish for privacy in relation to lost pregnancies and the use of assisted conception or prenatal diagnosis. The particular problems in cancer families may arise because healthy, but at-risk individuals may be reluctant to ask affected individuals to provide a sample for analysis54 while the affected individuals may be reluctant to raise anxieties in their healthy relatives in the absence of a clear practical benefit.55

The duration of follow-up for the cases reported here has been variable but has certainly not been sufficient to let us draw any conclusions about long-term outcomes of reported episodes of nondisclosure. The diagnosis had recently been confirmed or revised in the light of molecular testing in more than half of the 28 cases involving nondisclosure by parents to their adult children, so disclosure may progress with time with the new test results acting as a catalyst.

There is no evidence from our study that geneticists or genetic counsellors are ‘breaking confidentiality’ when nondisclosure becomes apparent or is declared by the genetic counselling client, but they do regularly try to persuade clients to disclose relevant information within the family, sometimes quite actively. This policy of active persuasion, recommended by professional bodies, may be one of those contexts within genetic counselling where an adherence to the ethos of nondirectiveness would be inappropriate.56 A recent analysis of the current legal situation in USA has also concluded that professionals should encourage but not coerce the sharing of important health risk information by genetic counselling clients with their at-risk relatives.57

It has been argued that genetic information cannot by its very nature be private and should therefore not be bound by the usual professional codes of respect for confidentiality15 – with genetic information generated about one individual being treated as essentially familial and therefore to be shared with other family members on a ‘joint account’ model58 – although, of course, a case can also be made for genetic information being regarded as the most private information of all.59 While we all have moral obligations to our kin, not all such obligations are enforced by health professionals. It would be difficult to define the circumstances in which a professional should disclose genetic information without consent, however, and perhaps it is best if these circumstances are left ill-defined so that they remain truly exceptional instead of becoming applied in a routine and formulaic manner that could impact upon the trust with which patients and clients approach clinical services.60

Our findings suggest that many cases of nondisclosure arise because of the practical difficulties encountered by clients in managing the disclosure rather than from their wish not to inform their relatives. Other studies3 have highlighted that patients are often worried about how to disclose rather than whether or not it would be best to do so. The general issue of how to support families to promote the appropriate sharing of genetic information is something for the genetic counselling community to consider further.17 There may be a role for professionals to be actively involved in family disclosures in support of anxious clients. Indeed, in an earlier study of women attending a clinic with a family history of breast or ovarian cancer, one of the dissatisfactions reported by several women was the lack of support or help in informing relatives who might be at risk.61, 62 Families offered follow-up genetic counselling services, including those provided through genetic family registers, report better preparation for discussing genetic issues with their relatives.23 We should be working on ways of helping families to communicate effectively and sensitively across the generations and should include support for this in the training of genetics professionals.63 This is likely to be much more fruitful than developing regulatory approaches to deal with the most unusual confrontations between counsellors and clients.