Introduction

The ethos of clinical genetics is to consider the risks of a genetic condition not only for the consultand attending the genetics clinic but also to relatives of that person.1 The extent to which the genetic specialist should shoulder the responsibility for ensuring that family members are aware of their risk is a perennial debate in genetics. In contrast to infectious diseases, communication of genetic risk to those considered at risk is ‘ethically dubious’ in nature, with a lack of clarity about what individuals should reasonably be expected to do, and how professionals should respond when they are aware that communication within a family has failed or is blocked. This will depend in part on the nature of the information available (risk information only or genetic test results) and implications of the condition. For instance, a greater imperative may be felt when preventive treatment options or reproductive choices are being restricted. In general, there seems to be an uneasy consensus that genetic services should rely predominantly on the consultand to convey information, except in exceptional circumstances.2

The consultand, however, does not always convey risk information to their at-risk family members. First-degree relatives are consistently given genetic risk information directly by the proband more frequently than the second- and the third-degree relatives. Wilson et al3 have comprehensively reviewed the complex factors influencing whether information about genetic risk is shared with family members, describing individual and family characteristics, disease, and sociocultural factors that can affect this process. It has been suggested that disclosure of genetic information is best described as a process rather than an act,4 implying a collection of actions rather than a single event. If genetics professionals are to continue to rely on family members to convey information – or, conversely, if they seek greater participation in this process – it is important to extend beyond identifying the factors that influence who in the family will be informed and gain an understanding of the process of communication of genetic information within families and its consequences. We conducted a systematic review of papers to illuminate these aspects of communication, with the intention of informing future practice in the facilitation of communication within families by health professionals.

Methods

Search strategy

Six electronic databases (Medline, CINAHL, Cochrane Library, EMBASE, PsycInfo, and National Research Register) were searched from 1985 to January 2006. The search strategy was broad (high recall and low precision), as this review was part of a wider review of risk communication in genetics.5 Searches focussed on risk and communication; genetics, predisposition, screening, or counselling; and service delivery, health services, and organisation. The strategy used for Medline is shown in the Supplementary Material. Follow-up searches included both electronic (Citation Indices) and manual searches (key authors encountered most frequently in the field and reference lists of included papers).

Assessment for inclusion of studies

The inclusion criteria were peer-reviewed publications on the process and outcome of communication of genetic risk to family members. Communication related only to actual discourse, dissemination, or behaviours with case reports, hypothetical, or anticipated communication excluded. Genetic risk included information about the condition, numeric risk, or genetic test results. All papers meeting these criteria and utilising qualitative methods were included; quantitative data were only included if it specifically provided information on outcomes of communication. Two reviewers assessed all titles and abstracts and examined full-copy papers of potentially relevant references. Decisions were made independently about inclusion or exclusion, with reference to a third reviewer if there was disagreement between the first two.

Data extraction and analysis

Papers identified as containing information about communication of risk were read in detail by two investigators (CL Gaff and AJ Clarke). An aggregative synthesis approach, focussing on summarising data, was adopted.6 Relevant findings were coded on data extraction templates in an Access database and are presented in narrative form. Multiple papers from a single study have been extracted and presented as a single study.

Results

Twenty-nine papers from 26 studies were subjected to data extraction. Figure 1 illustrates the search results and a summary of relevant findings is presented in the Supplementary Material. While this review focusses on the communication of risk information, none of the papers included describe the content communication in these terms. Nineteen studies discussed transmission of test results, while the remainder were about the diagnosis or genetic condition or were not specific about what was communicated. The majority of papers (n=21) investigate families with hereditary cancer syndromes, with very few on more ‘traditional’ Mendelian genetic conditions such as Huntington disease (HD) (n=3) or cystic fibrosis (CF) (n=3) or chromosome anomalies (n=1).

Figure 1
figure 1

Search results.

Three themes relating to the review topic emerged from analysis of the data: deliberation before communication, communication strategies, and outcomes of communication (Table 1).

Table 1 Summary of results

Deliberation before communication

If an individual does not ‘immediately’ communicate information about genetic risk to their relatives, it appears that they undergo a period of deliberation in which decisions about disclosure are made. Hamilton et al7 describe these deliberations as encompassing (1) consideration of the effects of disclosure, (2) selection of what information to disclose, and (3) planning the timing of disclosure. Forrest et al4 describe an additional phase of ‘making sense of personal risk’ before being able to deal with deciding who to tell and what to say, with those who continue to perceive their risk as uncertain or ambiguous most likely to have problems knowing what message to give their relatives. However, the ways in which individuals construct and make sense of their risk is beyond the scope of this paper and are presented elsewhere.5 The remaining three aspects are discussed here in more detail.

Considering the effects of disclosure

Underlying accounts of the decision whether or not to disclose risk information to relatives is a dilemma: the desire to protect relatives from potential harm is weighed against the wish to provide them with information that may have important health consequences.4, 8, 9 In weighing these conflicting responsibilities, the individual assesses their family members' vulnerability to the information as well as their receptivity.7 Vulnerability is assessed by considering life circumstances and mental and/or physical condition, as well as the potential harm that may be caused by news that may be perceived as bad or harmful and able to cause upset or worry.4, 8, 10, 11, 12, 13, 14, 15, 16 Receptivity is assessed by predicting the response of the family member to the information (see Outcomes of communication).

Selecting what to disclose

Individuals can be selective about what information they disclose. In one study, individuals disclosing their HD test results only provided more information if asked to do so, while those from families with hereditary breast ovarian cancer (HBOC) were more forthcoming.7 In contrast, some men undergoing testing for HBOC mutations disclosed only limited information about testing and its implications to their children.16 Sometimes, selectively conveying information may delineate what constitutes ‘family information’ (ie news of a mutation in the family) and what is ‘personal’ (individual predictive test results).17

Timing of disclosure

Disclosing at the ‘right time’ is of concern to informants.4, 7, 14, 18, 19 Generally, communication with close relatives happens within a week and late disclosure is less common.12, 17, 19, 20, 21, 22 Forrest et al4 distinguish pragmatism, where an active approach (such as arranging an appointment or mailing letters) is adopted and prevarication, where the right opportunity is sought. The right ‘time’ appears to refer to both life stage, for example, availability of surveillance for cancer, forthcoming marriage, or children, and to the right opportunity, for example during normal social contact.4 It may also depend on the perceived emotional readiness of the recipient.7 It should be noted that communication of results in families with hereditary bowel cancer occurred as part of normal social interaction, but this appeared to be due to a more matter-of-fact approach to hereditary nonpolyposis colorectal cancer (HNPCC) rather than prevarication.17, 20

Communication strategies

In some cases, the outcome of the deliberative process will be to decide that the information is not disclosed. Information may be deliberately withheld or kept totally secret.23 Alternatively, there may be a ‘passive’ failure to inform due to poor or nonexistent communication pathways and a low sense of responsibility to that individual.20, 24 Even when communication does occur, it may not be completely open. Hallowell et al18 describes the different strategies observed in men undergoing testing for HBOC: complete openness, where children were informed about testing at the outset and were one of the first to hear of their father's results; limited disclosure, a cautious approach in which involvement in testing was deliberately downplayed; and total secrecy. Kenen et al23 focus on the interaction between informant and recipient. As before, open and supportive communication is observed. Blocked communication was reported with recipients refusing to receive the information directly (eg hanging up telephone) or indirectly (apparent discomfort, lack of responsiveness, or failure to disclose information themselves). Aspects of selective disclosure were described in this study as ‘self-censoring’, with the informant not pursuing difficult conversations or seeking to protect other family members by only giving them reassuring information. Finally, some women used intermediaries if they felt uncomfortable contacting a particular person. These different patterns were observed within families as well as between families.

The use of intermediaries is observed particularly in relation to parents seen as having the responsibility to inform their children (generational responsibility or vertical transmission).4, 17, 20, 25 Thus, an individual may directly inform their siblings but would (perhaps implicitly) expect those siblings to inform their own offspring. A cascading of responsibility is apparent, with responsibility for informing others transmitted along with the information conveyed. The assumption that at-risk family members have been informed by other relatives11, 15 is not necessarily borne out, parents do not always inform adult children21, 25, 26 and children under 18 years of age are less likely to be told.10, 18, 21, 27 Intermediaries are recruited to inform other relatives,9, 14, 24, 28 particularly by men.14, 20 This appears to be a different concept to that of a ‘pivotal’ person who perceives him/herself to take responsibility for passing on the information and encourages relatives to consider testing.17, 29, 30

Outcomes of communication

An obvious outcome of communication is the reported number of family members informed, but others are also apparent in the literature, namely: uptake of genetic testing, knowledge of the recipient, and family responses to the informant.

Uptake of genetic testing

Seven studies investigated uptake of genetic testing by informed relatives.11, 15, 17, 25, 28, 31, 32 Most of these studies determined uptake in relatives informed about testing: the percentage of those informed who underwent testing ranged from 13%32 to 57–64%.17 Peterson et al17 found somewhat lower uptake of predictive testing for HNPCC (46%) in a family where disclosure had been delayed compared to families were there had not been a delay (57–64% uptake). Blandy et al28 found that second- and third-degree relatives, when aware of the mutation, are less likely to proceed with testing than first-degree relatives and also suggested associations between uptake of testing in first-degree relatives and greater family support of the index case and knowledge of risk of transmission. Landsbergen et al15 compared the characteristics of index cases of families with a low uptake of predictive mutation tests to those of other families, finding significant relationships between low uptake and, not disclosing in person, less involvement with psychosocial workers during testing, and other variables relating to family dynamics. The authors suggest that less attention to psychosocial issues in counselling sessions may have had an inhibiting effect on the process of dissemination.

Knowledge of the recipient

Little is known about the knowledge of the family member who has been informed by the proband. Denayer et al22 assessed knowledge of CF among siblings of parents of a child with CF, finding that most had only a superficial knowledge about the inheritance of this condition. While 80% reported that the family was a source of information about CF, mentioning CF parents as information source was not significantly associated with the genetic knowledge score. Binedell et al33 found that those presenting for HD predictive testing were able to give a risk figure of 50% but had misconceptions about inheritance. Hallowell et al16 found uncertainty about risk in women whose father had tested negative for a BRCA1/2 mutation. This is consistent with data showing that the person communicating the information may themselves have poor knowledge.11

Impact on individuals, relatives, and family relationships

Study design has precluded direct assessment of the responses of family members to the risk information given by consultands, with almost all accounts provided by the consultand.10, 18, 20, 21, 23, 24, 27, 28, 34 In only one study did some of the recipients of the information directly give their reactions to researchers, indicating their dissatisfaction with the limited disclosure or secrecy of their parents and demonstrating misconceptions about their risk.18

Anticipating or actually communicating with relatives can be experienced as burdensome or difficult by the informant,4, 8, 10, 13, 14, 26, 28 with consultands reporting adverse reactions by some family members, often typical of responses to bad news, or a high emotional temperature.18, 20, 23, 24, 28 A list of these is provided in Table 2. Even where there were no difficulties in communication, respondents expressed concern about the quality of their relationship with their relative or spouse in the future.10 There are reports that relationships had been strengthened after disclosure.9, 36

Table 2 Summary of communication difficulties

The impact on children and young adults of family communication might be expected to be different from that of adults. Carriers of BRCA1/2 mutations have reported concerns expressed by their children about their mother's future health.21 Although disclosure of BRCA1/2 results to children aged under 18 years did not relieve parental distress,34 neither did it impact on parent–child functioning.27 Indeed, some carrier mothers in another study perceived their relationship as closer after testing.21

Discussion

This review has summarised literature on process and outcome of family communication. As might be expected, communication is not uniform between or within families and the nature of interaction between informants and recipients differs. For example, a person may convey information openly and easily to some family members, but in others conversations may need to negotiate the direct refusal of the information by the family member or subtle cues that the information is unwelcome. It is therefore unsurprising that communication is, in some situations, immediate, while in others there is a delay and deliberation. This process may not be completely conscious on the part of the individual who undertakes, or is given, the task of communicating, but is evident in accounts of disclosure and nondisclosure.

It appears that communication occurs when a sense of responsibility to provide the family member with potentially important information outweighs concern about harming the individual by imparting ‘bad’ or potentially unwelcome news. This could be termed a ‘calculus of responsibility’ and demonstrates one side of the exchange of mutual obligations of trust among kin. Taking responsibility for transmitting information implies acting responsibly towards fellow family members. To act responsibly is to assess the potential consequence of sharing or withholding information and is embedded in a moral economy consisting of (mutual) assessments of competence and maturity, as exemplified by the deliberative communication strategy. Communication of genetic risk information is the result of intricate, existing ties of differing strengths.

Arguably, to genetic professionals, the potential psychosocial harm to family members is a relatively abstract concept compared to the health implications of nondisclosure. Consequently, there may be an inclination to prioritise the sharing of information over a desire to protect; reports of the experiences of genetics professionals tend to focus on overcoming problems that arise in family communication,38, 39, 40 despite recognition that these are the exception rather than the rule.38 An active approach to communication of genetic information by health professionals has been actively debated, particularly for familial hypercholesterolaemia.41 However, to date it has been rare that genetic professionals intervene directly and other means to foster the client's capacity or willingness to communicate are preferred.39, 40 There are hints that this may be possible: the provision of psychosocial support is associated with greater family communication and, where emotional problems blocked the flow of information, the desire for more support was expressed retrospectively.15 While families believe that the responsibility lies with them to inform family members, they also want health professionals available in a supporting role.4, 21 This seems to go beyond the simple provision of written materials. Suggestions about the way professionals may enhance family communication have been made,2, 15, 16, 42 but this systematic review did not identify any papers evaluating in practice the ways to foster the capacity or willingness to communicate in families. Caution should be exercised when moving beyond the descriptive studies identified in this review to those testing an intervention. Rather than considering all failure to communicate as the same and testing a ‘one-size-fits-all’ solution, it may be more fruitful and ethical to recognise the diverse nature and causes of poor communication and address each with a distinct approach. The categorisation of Forrest et al4 could be particularly useful here. They suggest that declining to inform may be perceived as (1) positive – acting in order to produce benefit (such as preventing harm), (2) negative – failing to act or overcome barriers, or (3) neutral –perceiving that nothing is needed. Each of these categories is likely to require a different approach but distinct strategies for these different types of communication blocks were not evident in the literature.

Even with improved communication, it seems likely that there will remain relatives with whom the individual is reluctant or unable to communicate. Unwritten rules of family communication dictate who is the most appropriate person to inform certain family members.4, 17, 29, 30 Further, family rifts are a common cause of failing to communicate. To attempt to alter existing family dynamics in an endeavour to meet the ends of the genetic services seems to be a dubious undertaking and, in the context of short-term genetic counselling rather than family therapy, possibly unachievable. Rather than valuing the outcomes of communication identified here (eg number of family members informed, knowledge of the recipient), attention may be better focussed on the professional's exploration of existing family dynamics and communication patterns: that is, assessing the process of counselling about family communication rather than these more quantitative or focused results.

Strengths and weaknesses

The strengths and weaknesses of this review arise in two areas: the protocols used to conduct the systematic review and the nature of the literature identified. The review meets most of the AMSTAR criteria,43 although meta-analytic combination of study findings was not possible due to the heterogeneous nature of results, but study characteristics and quality are reported (summarised here and fully available in Edwards et al5). The search criteria for the overall review focussed on risk communication and also included designated-specific conditions, consequently some relevant papers may have not been identified. However, we sought to overcome this by including papers cited by those studies that were identified.

With respect to the literature reviewed, studies were predominantly on communication about hereditary cancer syndromes and, more specifically, hereditary breast cancer. Differences in communication between families with HD and those with HBOC have been described,4, 7 and it may be hypothesised that family communication is influenced by the implications of the condition: the perceived treatability, perceived preventability, and reproductive implications. Similarly, communication about conditions with autosomal-dominant inheritance may not reflect communication about recessive or X-linked conditions. Therefore, conclusions drawn from data on families with familial cancer should only be extrapolated to other genetic conditions with caution.

The studies reviewed may be considered small studies and criticised for lack of representativeness. It is possible, perhaps even likely, that the results represent the views of those at the extremes of communication (very difficult–very easy). Finally, although the review intended to address ‘risk communication’, none of the studies framed the content of communication as solely ‘risk’. It seems reasonable to cautiously assume this relationship, as one of the strongest motivations to communicate test results to family members is the desire to inform them of their risk or options,12, 25 but this is certainly not the only one and communication may be used as a strategy to cope with bad news or obtain family support.25, 32, 44

Further research

There is a noticeable lack of examination of cultural, genetic, and gender diversity in these studies. As the implicit rules governing family communication of genetic information are likely to be dependent to an extent on cultural context,45 there is a need for studies within other cultural groups. A similar comment may be made about gender, which has been recognised as a factor affecting disclosure and nondisclosure.3 While there are hints in the literature reviewed that gender may affect the process and outcome of communication about genetic information, wider conclusions are hampered by the gender-associated nature of the conditions usually studied (ie HBOC). It would be useful to explore in more depth the experiences of men communicating about conditions such as HNPCC, HD, and CF, as different strategies may be more effective in assisting men to discuss genetic risk. Similarities and differences in the process of communication for genetic conditions with differing implications and inheritance patterns also need to be explored. There would also be value in considering the family communication processes that lead to an understanding of risk, which may occur over a long period of time rather than in the context of conveying a piece of information. The findings of such investigations may be relevant to communication about common complex conditions. We also propose that greater attention needs to be paid to the use of theoretical models of communication to complement empirical studies. The study of communication about genetic information needs to move beyond the simple sender–receiver model of information transfer assumed by most of the studies reviewed, which usually focus on the sender of the information. Rather, understanding needs to be developed about the meaning of such information to both the individuals concerned, the dynamics between them and also the influences on their relationship. Theories from the family communication literature specifically addressing information sharing within families may prove useful to inform such studies.

Conclusion

The objective of genetic services, implied by the content of research studies on family communication and guidelines in this area,46 is to maximise the flow of important information to relevant family members. But the myriad of factors that affect communication means that obligations to family members flow along channels that are not always identical with those through which genes are passed between generations: a sense of responsibility does not always map onto current biomedical models of inheritance.

Family communication is a complex process and the discussions between professional and client about dissemination of information need to recognise and respect this complexity. However, the nature of interactions about genetic information remains poorly understood. Greater attention needs to be paid to the process of communication with family members. This would inform professional discussion and possibly also the development of strategies that assist – but do not compel – clients to communicate with at-risk family members. Concurrently, desirable outcomes (beyond a one-dimensional assessment of the number of relatives informed) should be defined and include those valued by both the person communicating the information and the recipient.