Abstract
Introduction
This article addresses the scenario, more commonly encountered by doctors in primary care, where a third party such as a relative, friend or other acquaintance (the terms will be used interchangeably) approaches a doctor (will be referred to using the male gender pronoun) expressing concern about the health or wellbeing of a patient (will be referred to using the female gender pronoun) registered with the practice or known to the doctor. The exact circumstances and content of the disclosure will inevitably vary, but can be linked to the friend's concern for the mental or physical health of the patient or to a perception that she is reluctant to seek help or that she is not receiving sufficient or the right type of care. Any of these issues may be voiced incidentally during a visit, to address his personal health, to the doctor's surgery where both the friend and the patient are registered. But visits or phone calls are sometimes arranged by friends to doctors with whom they are not registered and for the specific purpose of expressing concerns about the patient's health. Especially problematic are instances where the friend couples the disclosure with a request that the approach to the doctor not be disclosed, possibly because of the friend's concerns that disclosure may harm his relationship with the patient or that it may further hinder her willingness to access care in the future. One of the obstacles to such contact is absent in health-care systems such as the UK National Health Service where services are free at the point of delivery. But it does not follow that approaching a doctor in this manner is without significant barriers: these may include the friend's uncertainty about the rightness of his action or of the doctor's reaction. Different considerations may apply to patients who lack mental capacity or those who may be subject to mental health legislation; these will not be considered in this article.
A patient's husband may, for example, advise the doctor that his wife who has a psychiatric illness is skipping her medication or that she is not disclosing some aspects of her health. Concerned friends may approach a doctor expressing their fear that the patient may commit suicide or with specific requests such as that the doctor pays a visit to the patient or convinces her of the need for surgery. They may add that because of their concern about its impact on the patient, they wish their approach to the doctor to be kept confidential.
This article explores how a doctor may handle such scenarios in light of available guidance, particularly those of the General Medical Council (GMC). I argue that current guidance tends to construe such encounters as a challenge to confidentiality and to outline the doctor's duties within that framework, but this approach is insufficient to address the complex issues that need to be considered such as the challenge to autonomy, and the importance and interdependency of family and friends.
Available guidance
Guidance for doctors on how to handle such encounters is difficult to find. There is no reference to family or friends in the American Medical Association's (AMA) Code of Medical Ethics. 1 Doctors who refer to UK GMC guidance will note that it does not address the issue in depth, and that the recommendations that have some applicability are to be found in its section on confidentiality. 2 The GMC states that a doctor ‘should make it clear to a third party that whilst it is not a breach of confidentiality to listen to their concerns, the doctor cannot guarantee not to tell the patient about the conversation’. One situation where the patient may learn that family members have been in contact with a doctor would be if the patient requests access to her medical records, but here the GMC suggests that doctors may use an exception in the UK Data Protection Act 1998 which allows for non-disclosure if that identifies another person as the source of the information. 3,4 The GMC also advises that doctors should not refuse to listen to a concerned third party on the basis of confidentiality and justifies this by pointing out that third-party views and input might be helpful. Thus doctors are not only allowed, but appear to be encouraged to listen to a third party so long as doctors do not disclose confidential information. But whether non-disclosure is at all achievable is questionable. Conversations between doctors and family carry a real risk of disclosure. Reassuring anxious relatives can in itself confirm their suspicions or provide an indication of disease severity or prognosis. Family may interpret verbal as well as non-verbal clues, and it is not difficult to see how friends can deduce a considerable amount from circumstantial evidence and fragmentary conversations.
Role of the family
It should be acknowledged that family and friends have a significant presence and important roles in relation to a patient's health. Health matters are often subject to free exchanges between patient and family, and keeping some issue in total secrecy may be an exception driven by a specific motive. Indeed, input of family and friends can be very beneficial as they may act to safeguard a patient's interests, for example in facilitating her access to care, providing first aid or raising the alarm in cases of medical error. There are thus situations where excluding the family from near the bedside can be to the patient's disadvantage. The complexity of health care, the importance of social factors and the role of the family all favour that they be well informed. While it is arguable, within the framework of autonomy, that informing the family is a matter for the patient herself, there are situations where the patient may not be able to do so to a sufficient degree. This may be due to lack of understanding of the health problem, inadequate appreciation of the value of involving the family or embarrassment. On the other hand, family input can (and in many instances does) influence decision-making; some of this may operate by altering the patient's behaviour or access to care. 5–7 Whether the family should be allowed to influence doctor's decisions is more controversial. But family interests are real and can even extend beyond those of the patient, such as when adverse events lead to serious incapacity or death. Furthermore, families have their own needs: they require the resource of space and time and that their own questions or anxieties be addressed, and it is perhaps in this context that the GMC advises doctors to ‘be considerate to relatives, carers, partners and others close to the patient, and be sensitive and responsive in providing information and support’. Thus, it could be judged that actions (or inactions) that may alienate the family can, in themselves, signify poor practice. But it could be difficult to be considerate of relatives without factoring their needs into the equation, or how to provide information that is relevant to the patient's condition without disclosing something about her health.
There are many possible scenarios where friends contribute to the patient–doctor interaction. These include situations where the friend's role is providing information or advocacy or where he may act to support decision-making in instances such as when dealing with coma patients or those who temporarily or permanently lack capacity. Such input may even be invited to help ascertain the patient's preferences or best interest – much of this has been discussed in the literature. 8,9 There is also literature debating the legitimate limits of family interests in an individual's health, for example instances where there is an arguable case for disclosure of health matters that have relevance to family members. 10 A friend may accompany the patient during a consultation or is asked to communicate with health professionals on her behalf, e.g. collecting a prescription or calling an ambulance. Here the friend's role is supportive and possibly (but not always) welcomed by the patient. Situations like these demonstrate the interdependency and relevance of family in health matters, but an important distinction is that in the scenario discussed here the concerned friend makes the approach without the patient's authority and possibly without her knowledge.
Arguments from confidentiality
Codes of medical ethics have traditionally guarded confidentiality with much zeal. The AMA, for example, refers to legally required disclosure as a major threat to trust in the patient–doctor relationship, and in its guidance it refers to a historical standard when information obtained from a patient was – using their expression – ‘sacrosanct’, and when any disclosure was ‘ethically impossible’. It is not difficult to see how applying this absolute sense of confidentiality can alienate family and even result in a patient's disadvantage. Another significant challenge to excluding the family based on confidentiality is that they often have considerable if not complete knowledge of the patient's condition, which makes the doctor's endeavour appear contrived if not superfluous. Confidentiality does not derive its importance from a specific property of the information under consideration; equally the concept does not derive moral worth with reference to who is included or excluded from an exchange. 11 But confidentiality may best be seen as having instrumental value towards gains or losses, as a framework that facilitates exchange between parties, or as an assurance of patient-centred fidelity. 11 It is, nevertheless, perhaps true that current views do not regard confidentiality with the same strictness as historical standards, at least insofar as there can be allowance for limited disclosure based on third-party interests. Practically, there is evidence that doctors find difficulties interpreting or applying their duty of confidentiality. 12–15
In 1982, Siegler 16 concluded that medical confidentiality in the traditional sense of a relationship between a doctor and a patient had become old, worn out and useless. The challenge stems from the argument that, because of the increased complexity and involvement of multiple providers, a patient's interest in maintaining confidentiality is now coming into conflict not with the interest of others – and hence necessitating vows of fidelity on the part of the doctor – but with her own interest in receiving best care. Current practice is to allow what is often a large number of ‘team members’ access to personal information. GMC guidance advocates presumed consent in this area, and includes as team members administrative staff and others with whom patients might not expect information to be shared. 2 The only proviso is that information is made generally available to patients informing them that this takes place, but it remains unclear how a patient can reach any informed view on the matter given that the extent of disclosure and the implications of non-disclosure are largely unquantified. 16 The point here is not to undervalue the role of confidentiality, but to argue for a more realistic appreciation of the way it could be put into practice.
Siegler 16 recounted that 75 health professionals had access to one of his patient's personal information following a hospital admission for a non-complex clinical need. Others have made similar observations. 17 But the role of many of these individuals is far more marginal and removed from a focus on a patient's best interest compared with the role of most family and friends. All of this must raise the question as to why provisions similar to those seen fit for the wider team cannot be applicable to family.
It is notable that some clinical services such as sexual health, human immunodeficiency virus testing and psychiatry have developed their own standards of security for personal information, including separate case-notes. That this was judged to be necessary reflects clinicians' scepticism about the prevailing standards or a perception that the professional pledge of confidentiality is insufficient to assure patients. Nevertheless, it does not appear that the prevailing weaker application of confidentiality has hindered doctor–patient exchanges. This may be advanced to cast doubt on the instrumental value of the pledge.
While it would appear that confidentiality alone is unable to provide a sufficient framework for the exchange between the family and doctors, a professional vow to guard confidentiality may be understood within the overall pledge to promote the patient's welfare, and a breach of confidentiality could be seen as morally wrong insofar as it reflects breaking a promise. 11 But defending confidentiality within this construct must rest on its assumed value in relation to welfare, not on it being a promise. The GMC advises doctors ‘to consider whether listening to concerns of others to be a breach of trust’, and quotes the example where a patient ‘may have asked the doctor not to listen to particular people’. Absent prior undertaking, it is not clear why listening to a third party would in itself constitute a breach of trust. Third parties are invited as witnesses and are a necessary part of legal procedures without that jeopardizing trust. An action undertaken against patients' express wishes may conflict with their autonomy, but any challenge to trust must stem from violations of the essence on which such trust is based. Trust has a number of facets, including trust that the doctor has the necessary skill, knowledge and judgement, and in all these areas trust needs to be earned and maintained. An additional promise of secrecy is not in itself conducive to trust. The promise to keep something confidential provides some reassurance about the status of communication, and perhaps reduces the stress of the encounter, but this is neither a necessary nor a sufficient condition for trust.
The legal implications of breaches of confidentiality are also important, for although the basis for an obligation of confidence in jurisprudence is a matter of debate, breach of confidentiality can result in legal action. In English law, information is not characterized as property over which an individual may exercise property rights, but in relation to patient–doctor interaction there may be action in negligence for a breach of duty of care. 18 The obligation of confidence is also recognized by the common law, as well as in specific statute such as the National Health Service (Venereal Disease) Regulation 1974, the Abortion Regulation 1991 and the Human Fertilization and Embryology Act 1990. 18 In addition, the European Convention on Human Rights now provides part of the legal basis for confidentiality. Article 8(1) states that everyone has the right to respect for his private and family life, his home and his correspondence. But the law may still place different weight on disclosures based on the particulars of the case. Baroness Hale, for example, opines that ‘Not every statement about a person's health will carry the badge of confidentiality or risk doing harm to that person's physical or moral integrity’. 19
Conflict with autonomy
Framing the exchange between family and doctor within confidentiality seems to miss the essence of the contact. The scenario described here is not the one where friends attempt to gain confidential information, but the one where they seek to influence care, often towards a course or an outcome that differs from that formulated through the patient–doctor contact. Thus what is primarily at stake is the question of patient autonomy, and the challenges that arise are perhaps best seen by considering the situation where attempts are made to keep family–doctor contact secret. Autonomy will be significantly challenged if a doctor were to factor information received from friends into patient management without full disclosure. It is arguable that the source of information on which doctors formulate their advice is pertinent to the exercise of autonomy. The risk to patient–doctor interaction will be further compounded if the doctor were to attempt, at the same time, to keep the source of information secret. This necessarily challenges the GMC guidance that doctors should advise family that ‘they cannot guarantee not to tell the patient’. Instead, it is argued that doctors have an obligation to inform patients of any conversation with family and friends. Withholding such information or the use of data protection legislation as suggested by the GMC brings into question doctors' veracity and fidelity as may be explicitly or implicitly promised. A patient may come to suspect that her doctor has had a conversation with a friend, or the discussion may be disclosed later by the friend himself, resulting in damage to the patient–doctor relationship. Communication between family and doctors establishes a parallel relationship. Family may harbour an expectation that the doctor will act on the information they disclose and therefore become disappointed, resentful or persistent if their approach has not generated action (or the action they envisaged or demanded). It is also possible that attempts to protect the information source could influence doctors' choices. All of this could result in escalation that jeopardizes care.
Beneficence
A physician's obligation to live up to a patient's reasonable expectations of privacy and confidentiality is one way to specify the general obligation of fidelity. 19 Patient-centred fidelity may be a core facet of patient–doctor interaction, 11,20 but its requirements may differ significantly in health care compared with adversarial legal conflicts, for example. Indeed it could be interpreted to require the doctor to seek to maximize a patient's welfare, including in circumstances where this invites the input of others – be they health-care workers or family. Here it would appear that disclosure of some information may be justified. The question is whether this requires implicit or explicit consent and whether provisions similar to those seen appropriate for the wider health-care team could be applicable to close family members. This is again primarily a question of autonomy, not confidentiality.
Whether to act on the information received from friends is one of the first dilemmas the doctor faces. Clearly, the case for action and the degree of urgency will depend on the illness and the content and credibility of the disclosure. The requirement of beneficence appears to constitute the basis for action, but this may not necessarily be valid. A doctor's duty of beneficence can only exist within a framework that requires an initial (and renewable) approach by the patient to define her care needs. Absent such an approach, patient–doctor interaction and related obligations cannot be said to exist. A critical consideration is the degree of concordance between the patient and the third party. A case could be made for the doctor to act to prevent harm based on a general obligation to rescue if the health issue is more pressing, but this still requires the patient's concordance and is a matter of autonomy.
One important factor that could be advanced in favour of doctors' listening to concerned family is that they may have information that enables better care. The general point is acknowledged and could be supported by examples, but this should not be accepted without qualification, and a distinction needs to be made between communication to and communication from relatives. Beneficence is based on assumptions and valuations about what is best for the patient, but this ought to reflect the patient's wishes which, in turn, requires her input. 21,22 A patient's wishes may not be in full, or in any, agreement with the views of the family. Some patients ask a particular friend or relative to accompany them to a consultation, others prefer to rule them out, and it seems odd for someone left in the waiting room to seek to impart their views at the margin of a consultation. The principle of beneficence affirms the doctor's obligation to minimize harm and maximize benefits. But in relation to competent adults, that obligation is necessarily constrained within a framework defined by the patient, and it should be accepted as such even when faced with gaps or omissions that could conceivably be filled by uninvited (and perhaps unwelcome) family. Full disclosure of information by patients is desirable and potentially useful, but this may not always be forthcoming for reasons that the doctor may explore but not over-ride. Suggestions such as that made by the AMA that patients have a ‘responsibility to be truthful and to express their concerns clearly to their physicians’ 23 is not necessarily helpful in the context considered here, and may be more applicable to instances where patients may seek unwarranted advantages. Patients can refuse to undergo examinations or investigations advised by doctors even if these are potentially useful. Such refusal ought to be seen as an exercise in autonomy, not as a violation of an obligation, ethical or otherwise. Similarly, patients may not disclose aspects of their clinical history, and even if this were to impact on the care they receive, it does not follow that doctors have a duty to verify or complete the information from other sources unless authorized by the patient. Thus, the GMC advice that ‘doctors might need to share with a patient information they received from others in some situations such as when it has influenced doctors’ assessment and treatment' is problematic, as it allows for non-disclosure. Indeed it could be argued that doctors should not allow information obtained from others to factor into their assessment and management unless they have discussed it and verified it with the patient. But dismissing information once received may not be practically possible, a point which could be made to strengthen the case for avoiding communication unless it could be reasonably assumed that this is consistent with a patient's wishes.
There could perhaps be agreement that keeping doctor–family communication secret while acting on the information provided is not possible without creating more elaborate barriers to truthful information exchange, and it is arguable that openness is more conducive of a better relationship between all concerned including the family. Yet the GMC advises that doctors ‘should make it clear to a third party that … the doctor cannot guarantee not to tell the patient about the conversation’. This creates unhelpful ambiguity. Favouring frank disclosure can only deter those who are most concerned to keep the exchange secret, and will arguably reflect favourably on trust in the profession. Ambiguity can leave relatives wondering how doctors, who have been receptive to their approach about others, will handle their health matters if or when they themselves fall ill.
Conclusion
The complexity of health care and interdependency of family and friends mean that communication often does take place between a third party and clinicians. Practical challenges and the difficulties inherent in the concept itself suggest that confidentiality is not a sufficient framework within which such communication can be addressed. It is suggested that the occurrence and role of such communication be acknowledged more explicitly, and that they are positioned within a framework where all relevant factors including fidelity, veracity, beneficence, trust and – more prominently – autonomy are taken into consideration. The legitimate extent of any contact with family and friends may be dependent on the clinical scenario, but if communication with family were to take place, it ought not to be kept secret. On the other hand, there is a need for guidance to define the limits of exchanges with a third party more convincingly.