Dementia is a progressive condition, that leads to impairments in memory, behaviour and cognition. It affects a person’s ability to maintain function and activities of daily living. The World Health Organization considers dementia a global epidemic as it is the leading chronic disease contributing to disability and dependence among older people [1]. The Global Action Plan on the Public Health Response to dementia 2017-2025 highlighted the importance of quality research as a crucial step in ensuring an evidence-based approach to manage this global crisis [2].

In response to the call for increased research on dementia, Alzheimer’s Disease International organised a ‘Let’s talk about dementia research’ global webinar series between from November 2018 to May 2019 [3]. These webinars provided an opportunity for members of the public, dementia associations, health and social care professionals, and companies involved in dementia research to interact with each other. The topics covered ranged from: why research is needed, the lived experience of research participants, barriers and opportunities to access dementia research, as well as how to maintain hope when trials ended, particularly if trials ‘failed’ to show the benefits of disease-modifying drugs. This forum served as an initial introduction to open discussions and brainstorm ideas for collaborative research involving people with dementia.

Dementia research tends to emphasise the medical model, reducing the person with dementia to neurobiology and neuropsychology. The experience and manifestation of dementia usually focuses on the disease process, ignoring the psychosocial context surrounding the individual. There needs to be renewed focus on the person with dementia, particularly the individual’s sense of self; the person’s rights; and the value gained from perspectives of people with dementia [4]. However, a common problem researchers face is determining the capacity for informed consent in people with dementia [5]. In this paper, consent issues are discussed with several approaches to obtain consent from people with dementia to enable their participation in research.

Informed consent is required before participation in research. Consent must be voluntary, informed, provided by someone who is competent to provide consent and understand what they are consenting to. It is also continuous, in that a person who initially agrees may withdraw participation during the study. The person must understand the proposed research, make an informed decision regarding their willingness to participate, and communicate this decision. People with dementia are vulnerable because each of these steps may be affected by the disease. However, these abilities exist on a continuum as well, where a person’s true understanding of the research or implications are frequently incomplete.

There are several principles when consenting a person to participate in dementia research. A dementia diagnosis does not preclude competence to provide informed consent to participate. Competency is the ability to make an informed decision about participation in the particular context of the specific study. Competency may also change over time, which leads to a shift from obtaining an ongoing consent to getting assent. As consent should occur without risk of coercion, there should be distinctions between the clinician’s role and while conducting research. If conflicts arise in a research context, the opinion of another clinician not involved in the research may be necessary to ensure that decisions regarding treatment and research participation are made in the best interests of the person [6].

Consent rates are affected by multiple factors; such as the risk versus benefits of the study, relationships with caregivers, the researcher’s ability to identify participants who can consent, characteristics and attitudes of the person providing consent, and the method of obtaining consent, including timing, location, method of presentation and type of consent requested. There are also different aspects of consent refusal. Complete refusal means not participating at all, while a person may have partial refusal or consent to certain parts of the research protocol only. Consent may also be contingent on other factors, such as the agreement of another person, caregiver, spouse or children. When a person gives time-contingent refusals, the investigator should clarify whether a different time will suit the participant, or whether it is a non-assertive refusal to consent [7].

An approach to taking consent from people with dementia was proposed by Warner et al. [8]. Participants should be provided with sufficiently detailed, salient written and verbal information in a form they can understand. This should include the objectives of the trial, potential risks and inconveniences of participation, product being tested, possibility of receiving a placebo, concept of randomisation, length of the trial and outcomes measured, and the opportunity to withdraw at any time. This information may need to be repeated or if appropriate, involve caregivers while providing information. Sufficient time should be given for the person to understand and retain the information. If the researcher is uncertain regarding capacity, they should ask the person to repeat relevant information and assess understanding before a decision is made. It is also useful to provide introductory written information before approa-ching them, using a range of background materials and anticipated questions with pre-printed answers [7].

Many people participating in dementia research lack the capacity to give informed consent. In a randomised controlled trial for people with mild to moderate dementia, most of the participants (76%) were unable to give informed consent according to the legal criteria for capacity [8]. Cognitive testing with the Mini-Mental State Examination did not predict the presence of capacity. For those lacking this capacity, process consent by obtaining assent and respecting dissent is an ethical approach to safeguard participants. Assent is an affirmative agreement to participate in research, while dissent is an objection to participation. However, it may be difficult to identify if behaviour is an indication of dissent or a symptom that needs to be identified and treated. For people with severe cognitive impairment, the ability to clearly state willingness to participate may be affected, limiting this approach. There is also no consensus on how assent and dissent should be defined or what procedures should be used regarding them.

A study sought to get agreement from key informants, namely experts on dementia and research ethics, dementia researchers, and dementia caregivers and advocates regarding this [9]. The overall opinion is that assent and dissent should be based on assessing how people without capacity express or indicate preferences verbally, behaviourally, or emotionally. Assent was viewed as necessary if a person has the ability to assent, while dissent should be binding if unequivocal or sustained after efforts to relieve concerns or distress.

The following approach is recommended for process consent [10]. Firstly, permission for access to the person should be gained from staff, relatives or family. Researchers should be transparent about intentions to achieve process consent; this first step seeks permission for access, rather than proxy consent. As health practitioners have a legal and professional duty of care towards people with dementia to act in their best interests, initial social engagement is usually only permitted when the person is in a state of well-being. The researcher should consider how the person usually ‘consents’ to activities within their daily life as well as existing assessments or opinions on capacity. If initial consent is obtained, ongoing consent monitoring is still required. Feedback and support should be encouraged. The researcher may give feedback to carers if requested by the participant, taking into account confidentiality issues; while carers may feedback to the researcher changes in well-being from participation.

Process consent offers an ethical and practical framework to ensure that consent is continually assessed in people with dementia during research [11]. This was illustrated in a qualitative research scenario, where research data were obtained from a participant’s diary containing their thoughts and feelings [12]. Although a proxy decision maker gave written consent and the participant gave initial assent; she mentioned that nobody should read her diary later on in the course of the study. Consequently, the researchers decided not to use the information obtained from that diary. Overall, respect for the research participants should be emphasised over the potential benefits of the research.

1. Surrogate or proxy decision makers

If a person lacks capacity, another alternative approach is to ask surrogate decision-makers to consent on behalf of a person with dementia. Theoretically, these surrogates should decide based on what they think the individual would prefer, should they have capacity. Research shows that consent for dementia research by family surrogates is broadly supported by older people. Older participants from the Health and Retirement Study were offered four randomly assigned surrogate-based research (SBR) scenarios: lumbar puncture study, drug randomised control study, vaccine study, and gene transfer study. Participants were asked whether society should allow family surrogate consent, whether they would want to participate in the research, and whether they would allow their surrogate some or complete leeway to override stated personal preferences. At least two-thirds felt that family surrogates should be allowed to consent for research, while almost 80% stated they would participate. Up to two-thirds would grant some or complete leeway to surrogates. However, there was a trend towards lower willingness to participate in surrogate-based research among ethnic or racial minority groups [13].

When individuals with cognitive impairment and their study partners or surrogate decision makers were interviewed, subjects and surrogates often differed in perspectives on how decisions are made regarding participation in research. Most agreed that the participant ultimately made the decision and engaged to some extent in decision making. For future proxy research, there was a range of preferences; mostly to use of the ethical standard of best interests, to substitute judgement and to consider the interests of others for complex discussions. There was also support for obtaining proxy consent and subject assent if the person lacks consent capacity [14].

There were also differences in the focus of discussions between the participants and surrogates. When audiotapes of informed‐consent encounters in a study of genetic markers for sporadic Alzheimer's disease were analysed, patient interactions mainly consisted of 49% agreement and approval and, 16% psychosocial information; while companions’ interactions were mainly positive statements (37%) and biomedical information (19%) [15]. Overall, the patients were more silent than their companions and the physician, and when they spoke, they primarily agreed on what was said. While this may signal assent, such an interpretation should be made with caution for people with dementia. Although informed consent discussions emphasise cognitive aspects, emotional and social dimensions also warrant attention.

In another study, 25 proxies were interviewed with a hypothetical scenario related to enrolling relatives with dementia into a trial for investigational drugs for Alzheimer’s disease [16]. Proxies described looking for consistent behavioural or verbal indications of assent or objections when trying to determine patients' preferences. However, proxies sometimes expressed willingness to override patients’ desires in favour of their presumed best interests. The amnestic nature of dementia led some proxies to justify overriding temporary dissent or discomfort in the interest of promoting patients’ values. The dependence on caregivers for decision-making, and caregivers’ awareness of their ability to persuade relatives also emerged in descriptions of the decision-making process. Examining the actual decision-making process is further warranted to determine how best to operationalise the concepts of assent and dissent in the context of research involving decisionally impaired adults.

In reality, the absence of functional surrogates is a major barrier to research and clinical decision‐making for hospitalised patients with advanced dementia. In a randomised controlled trial of palliative versus usual care for those with advanced dementia, surrogates or proxies were approached for consent. Among the 146 eligible subjects, 49% could not be enrolled as surrogates and could not be engaged in the informed consent process and 22 did not have a functional surrogate to consent for research [17].

Another approach for research consent is advance research directives (ARDs), particularly for those with the capacity to give but with anticipated cognitive impairment, such as in prodromal or early dementia. If a person does not wish to relinquish this decision to another person, ARDs allow competent individuals to prospectively consent to future or ongoing participation in research, in the event they lose the capacity to give consent. This approach emphasises the principles of autonomy and respect, rather than rely exclusively on decisions by surrogates or substitute decision makers.

There are several approaches to specifying ARDs. A person may name a surrogate and provide a clear indication of their wishes, to serve as guidance for the surrogate. A person may instead decide to consent to participate in specific research projects or categories of research. Alternatively, they may consent to future research participation and name a surrogate, with a clear delineation of the relationship between the surrogate and the consent-by-directive.

There are several ethical and policy considerations for the use of advance directives in clinical dementia research. Advance directives were originally used to influence medical treatment and care in the case of future incompetence, but only recently extended to the research context. In an advance care directive, a person describes situations in which they do not want treatment started or would prefer treatment to be discontinued. This is usually motivated by fear of medical overtreatment if they end up in a state of incompetence. However, there is no parallel motivating factor for ARDs, which tend to be altruistic, rather than self-interested. Refusal of treatment in specified situations also has stronger moral and legal weight than research because the negative effects of privacy, body integrity and self-determination through laws against assault have stronger claims than positive willingness to be a research participant. Research is also innovative, which makes it difficult to give advance consent for participation in a future experiment, without knowing in advance the nature of the study.

The likelihood of successfully carrying out ARDs is also quite low. A very small proportion of people complete advance medical directives, even among those who are informed about this option to explicitly state their wishes. There is less public awareness and interest in advance research directives or participation in non-therapeutic research. For a person to issue ARDs, early detection and diagnosis of dementia is required, as well as awareness of ARDs early in the phase of their illness. For some studies, this may raise ethical issues of screening appropriateness and disclosure of diagnoses. Competence or decision-making capacity should also be determined. This is less obvious in research contexts when there is no promise of benefit to the person and if more than minimal risks or burdens are involved. This usually requires a high level of decision-making capacity, which people in the early phases of dementia may not meet. Advance directives also require a person to anticipate the future and express wishes taking into account current knowledge, circumstances and predictions about the course of what the future will take. Usually, there is a difference between what is anticipated and what actually eventuates. These wishes should also be expressed in a way that is neither too vague nor restrictive. Overall, while ARDs are an option, they are unlikely to compensate for a person’s inability to consent in non-therapeutic dementia research [18].

There are several recommendations for those who consider ARDs. It may be useful to consider integrating preferences for research and treatment in advance directives. As progression from amnestic mild cognitive impairment to Alzheimer’s disease is often studied over a three-year period, advance research directives should be executed within 3 to 4 years or reviewed and updated within this time frame. Recent advance directives are also more likely to be valued, as it is seen to reflect a person’s contemporaneous wishes and views. If possible, it would also be useful to identify circumstances where it is appropriate and ethical for a person to override their ARD and consider external oversight so that the absence or variation of surrogate commitment or capability does not compromise participant protection [19].

Dementia has widespread implications on a person’s life and those of caregivers and family and research to provide an evidence-based approach is required. Obtaining consent from people with dementia presents challenges to researchers, which may limit the participation of people with dementia in research. For those unable to give informed consent, several approaches such as process consent, surrogate or proxy decision makers and advance research directives may be considered.

The author declares no conflict of interest.