Patients' vulnerability does not necessarily mean that they have impaired competence or that they lack competence to consent. An acute crisis, for example with the development of psychosis and involuntary hospitalisation, affects patients' cognitive and emotional capacity, while patients who have been undergoing treatment for a while will probably be more stable in the preferences they express. In cases where the basis for informed consent is uncertain, patients will need protection and the option of reconsidering.
At the same time, the degree of pathology is not always a good indicator of ability to consider participation. Patients are not devoid of preferences and values even if they are vulnerable or regarded as belonging to a vulnerable group (Hem et al. 2007). Patients may have a sound understanding of the risks, benefits and side effects associated with participation. Nor does not having research conducted on them make patients less vulnerable (Storosum et al. 2002). There is therefore reason to reflect over the risk of underestimating the competence of vulnerable patients by depriving them of the option to consent to participation. Subjective factors such as attitudes, values, motives and context have a strong bearing on the decision as to whether or not to take part. This is true irrespective of whether a person has a mental or somatic illness, or is healthy (Roberts and Roberts, 1999). There are also aids for assessing competence to give consent (ACE) or decision-making capacity, which is the expression that is used increasingly. According to section 17 of the Norwegian Health Research Act, the next of kin must give consent on behalf of persons who lack competence to consent. This may be a good solution in some cases, but not always. Relations between persons with mental health problems and their next of kin may be fraught with conflict and antagonism. This is especially true of persons with long-term, serious mental health problems that entail the use of coercion. Researchers should therefore tread carefully and base their assessments on the individual participant and their next of kin (Baronet 1999, Berlung et al. 2003).
The ability to take decisions is not a stable capacity, but changes depending on the patient, the situation and social relations. For example, trust in the researcher is important for the patient's decision as to whether to participate or not. The use of standardised tools for determining decision-making competence must therefore not replace the use of time to create a relationship of trust. In many cases, consent is a process (process consent); i.e. consent is not necessarily given once and for all. Some may maintain that it is unethical to obtain only an advance consent, as though it were possible to predict the research process (Hem et al. 2007, Parker 2007). In some research approaches, a consent may be a relational and sequential process: the development of the study is closely linked to the development of trust in the relationship between participant and researcher (Terkelsen 2010). This is different from establishing agreement at the outset in the form of a contract (informed consent form). At the same time, it is important to reflect over the power relationship between researcher and participant (Hem 2008, Juritzen et al., 2011). In this perspective, it is important to include assessments of what is being negotiated, and what reaching agreement means. Trust is important, but it also means that the requirements made of the researcher's moral judgement are more stringent.
Confidentiality is first and foremost about safeguarding and protecting the privacy of the participants. This means anonymising characteristic features so that their identity cannot be recognised. (See also Privacy and personal data.) But how is anonymisation to be carried out, and what challenges may be encountered? Anonymisation in which fundamental characteristics such as gender are changed may affect the quality of the data (Skarbæk 2007). A dilemma of a different kind may arise when participants do not wish to be anonymous. There may be cases where participants are eager to form an alliance with the researcher/research to promote an issue that they think is important; for example by wanting to be co-authors, or to take part in presentations (Mattingly 2005). In such cases, anonymisation may be an expression of exercise of power, in that participants are silenced and the researcher retains the power of interpretation (Brinkmann 2007). Such situations may present the researcher with a dilemma because the requirement of protecting participants may be in conflict with the participant's own wishes in cases like these (Hem 2008).
There is another, related challenge: the fact that the researcher usually anonymises in such a way that patients under no circumstances are recognised by outsiders. But what if they are identified by insiders, such as the personnel providing treatment? Such situations may present a challenge, for example when case studies or life-history interviews are used (Etherington 2007).
A situation that touches on the issue of confidentiality in a different way may arise when the researcher, in her capacity of also being a clinician, discovers matters of concern regarding the treatment, for example that the patient is not getting the most competent treatment available. Should the researcher make the patient aware of this, and recommend a practitioner who the researcher knows will provide more competent treatment?
One topic that is increasingly relevant is research on Internet chat sites for vulnerable groups. Discussion fora, which it is not unusual for patients and others to take part in, may be considered private by the participants, but they are available to the public (Tsao et al. 2008). Although they are open for all, does that give researchers the unrestricted right to use the material from the participants in their research without informing them? Should the researcher inform the participants, and perhaps obtain informed consent?
The question of how one should relate to the media in research projects that attract attention also touches on the confidentiality issue, in that one is responsible, as a researcher, for presenting research in a sober and objective manner.
Consequence assessment is first and foremost about risk and benefit considerations. Risk/benefit assessments can be related directly to the Hippocratic Oath, where the focus is primarily on protection of the individual. (See also the "do no harm" principle). Protecting the integrity of vulnerable patients is important in this perspective. Risk/benefit assessments do not only involve no-harm considerations, however. They are also concerned with how the researcher can promote the welfare (beneficence principle) of the individual participant and, not least, of the group in question. It might be said that the assessments in question are based on the assumption that patients do not know what is good for them (Oeye et al. 2007, p. 13), i.e. a paternalistic point of view. This can be seen as in conflict with the focus in research ethics on respect for patients' autonomy and self-determination. There is a divergence between protection and voluntariness (Oeye et al. 2007) which, when different research methods are applied, can create paradoxes that must be thought through in all projects.
However, it is important that researchers are aware of the requirement that the research must entail minimal risk (Declaration of Helsinki 2013), i.e. that taking part in research must not entail greater risk than the normal risk of everyday life. This requirement may be difficult to interpret and assess, so the researcher's ethical sensitivity, awareness and judgement will be crucial.
A dilemma that researchers may be presented with is that participants want to take part because they believe it will benefit a cause that they regard as important. The assumption that the researcher and the research can create the necessary attention for an important cause implies viewing the researcher as a spokesperson for problematic conditions in mental health care. This may, for example, be expressed in connection with patients' experience of violation through the use of both formal and informal coercion. Such expectations may not be consistent with what the researcher herself believes to be realistic, important or ethically defensible (Hem et al. 2007).
Other problematic consequences may be that patients with a mental illness regret their participation in research because they feel they have revealed too much of themselves. The good relationship of trust in the researcher leads to untold stories being told (Dickson-Swift et al. 2007, p. 338). Participants may experience their confessions – or the revealing of very personal matters – as emotionally painful. In such situations, the researcher has a particular responsibility to act wisely. Participant observation is a powerful research approach that – through the researcher's constant presence – it tends to influence the lives of those being studied, and this can cause discomfort. What is at stake and for whom, in such cases, should be a part of the contextual assessment rather than a preconceived notion of aspects and characteristics of the situation. Another type of dilemma arises when the role of researcher entails being both insider and outsider at the same time (Connolly and Reilly 2007, Norvoll 2007). In order to obtain data, it is important to be an insider, but in the interests of the analytical scrutiny, it is important not to go native (Oeye et al. 2007, p. 18). This is a relevant problem when a researcher comes close to the participants in psychiatric health care both in long-term field work and in interviews on private and vulnerable aspects of their lives.
The role of the researcher
Researchers are always placed within a knowledge situation, in a culture. Researchers have values, preferences, preconceptions and perspectives (Etherington 2007) associated with mental disorders. Factors such as this are of significance for research in mental health care (with respect not only to what has been found, but also to the procedure employed) whether the factors are reflected over or not. Continuous critical reflection (Roberts et al. 2003, Ellis 2007, Malterud 2011) involves establishing reflexivity as an ethical practice, whether the role of researcher is characterised by distance or by involvement with the participants. A researcher role characterised by willingness to be open for discussion – which employs ethical discretion (Wifstad 2007) – has both ethical and methodological/scientific significance in that the research is as transparent as possible. Carrying out research on personal matters and experiences in people's lives, in vulnerable situations and at pivotal points of time (Clark and Scharf 2007), as research on mental health problems does in various ways, involves complex ethical dilemmas, for example in connection with power. Researchers have been questioning for a while whether duty ethics provides an adequate guideline for preserving human dignity and worth in connection with research, particularly where vulnerable participants with mental health problems are involved. Guillemin and Gillam (2004, 2006, 2008) have developed the concept of "ethical mindfulness", which they describe as ethics in practice, in contrast to procedural ethics (see also Guillemin and Heggen 2009). The concept includes five aspects (the list is not intended to be exhaustive):
- acknowledge the significance of ethically important moments
- acknowledge the perception that something can be uncomfortable even if this perception is vague and unclear
- be able to express what is ethically important, or what is at stake, in the research
- be capable of reflexivity, "to research the researcher"
- have the courage to pursue one's own ideas, listen to one's own doubts and get to grips with what is challenging
Another contribution, "relational ethics" (Ellis 2007) has roots in the ethics of care (see for example Gilligan 1982) and places emphasis on the fact that it is our common humanity that creates a basis for relations and that makes mutual understanding possible. Relational ethics is another contribution stressing the importance of researchers assuming personal responsibility for acting in an ethical manner in research in which persons with mental health problems are involved. A reflexive relational ethics can help to achieve a balance between our needs as researchers and our obligations in relation to research participants when persons with mental health problems are involved.
This article has been translated from Norwegian by Jennifer Follestad, Akasie språktjenester AS.