Defining Supported Decision-Making

There are several models of supported decision-making, which differ conceptually, especially in how they understand the role of supporters. Silvers & FrancisReference Silvers and Francis ⁹ envision supported decision-making as a “mental prosthesis,” whereby supporters help the supported person reason and communicate about their “personalized notion of the good” without substituting their own values. McCarthy & HowardReference McCarthy and Howard ¹⁰ challenge the mental prosthesis model with their “non-domination” model, which focuses on reaching a decision endorsed by the supported person with “genuine collaboration” with the supporter that is responsive to their feelings and values.

Alongside Ishida,Reference Ishida ¹¹ we see value in both models. Supported decision-making is not one-size-fits-all, and varying types and levels of support can benefit PWIDs depending on their specific needs. For example, a PWID who is comfortable making decisions generally, but has trouble understanding medical or research jargon, might find their needs best addressed by a supporter who acts more as a prosthesis, clarifying information without offering their own suggestions. Alternatively, a PWID who values their supporter’s perspective might benefit from hearing the supporter’s views, so long as they are offered without pressure. Some PWIDs may also prefer other types of decision-making accommodations, such as more time discussing the protocol with a researcher in non-technical terms.

Others have raised concerns that supported decision-making may be inappropriate for some PWIDs, especially those with complex support needs. Arstein-Kerslake et al., ¹² for example, argue that supported decision-making for individuals with “more severe” cognitive disability risks becoming “merely positive rhetoric as opposed to transformative practice.” BarclayReference Barclay ¹³ argues that such “positive rhetoric” can be disrespectful if decision-makers rely on their own preferences but claim to be engaged in supported decision-making. Contrary to calls from the United Nations Convention on the Rights of Persons with Disabilities (UNCRPD) to fully replace surrogate decision-making with supported decision-making, ¹⁴ Barclay argues that surrogate decision-making is more respectful when the preferences or values of PWIDs are unknowable; otherwise, supported decision-making can become surrogate decision-making under a different name. In these situations, the UNCRPD proposes implementing the best interpretation of a person’s will and preferences rather than a “best interest” standard; Barclay argues that calling this process “supported decision-making” is still disingenuous because it denigrates the privacy and opacity of the PWID’s inner life by assuming or replacing their preferences while suggesting the decision is the PWID’s own.

Enrollment in research typically requires the participant’s consent, or the permission of a legally authorized representative along with the participant’s assent (i.e. affirmative agreement). Ideally, both surrogate and supported decision-making involve the PWID as much as possible and center their preferences, values, and aspirations. However, surrogate decision-making with assent typically requires a simple yes or no from the PWID, while supported decision-making actively involves the PWID in the decision-making process and prioritizes their decisional authority. Wendler & Kim ¹⁵ thus argue that reliance on supported decision-making, rather than surrogate decision-making with assent, might increase the chances that decisions whether to participate in research are consistent with the participant’s own values and aspirations.

Given that supported and surrogate decision-making have similar goals, and that there is substantial variation between supported decision-making models, we believe the exact boundary between supported and surrogate decision-making with assent is not always clear. We focus here on supported decision-making in order to anticipate and attempt to address challenges that might arise as it becomes more common in research. However, the ethical challenges we address overlap with those arising for surrogate decision-making. As we will discuss, how we categorize and carry out decision-making processes matters, but our concern in this article is less the label that one uses and more the extent to which an interaction realizes the goals of engaging and centering PWIDs in the decision-making process.

Challenge 1: Communication

Communication between PWIDs, researchers, and supporters is an essential part of decision-making. In the present section, we focus on PWIDs whose communication can be at least partially understood by supporters and/or researchers. In Speculation and Displacement, we address PWIDs whose communication is not understood, even with significant attention, patience, and caution.

In a large cross-sectional study by Smith et al.,Reference Smith ¹⁶ about 58% of PWIDs reported communication difficulties such as understanding and expressing spoken, signed, and written language, and 64% reported that caregiving staff only partially understood their communication. To ease these communication difficulties in healthcare settings, advocates suggest PWIDs schedule appointments with the same doctors to establish familiarity with their history and communication styles. Some doctors suggest PWIDs bring advocates to their appointments to assist with communication, and both advocates and doctors emphasize that all parties should work together throughout the consultation process.Reference Ziviani ¹⁷

These suggestions emphasize involving individuals the PWID knows and trusts. However, the high proportion of PWIDs who report being only partially understood by their own caregivers raises questions about how best to interpret PWIDs’ communication. To address this challenge, McCarthy & Howard endorse a public process, with third parties (such as social workers or members of the medical team) serving as “external checks.” ¹⁸

How might this approach be implemented in the research setting? Familiarity and exposure over time can help people to understand the verbal and nonverbal communication of PWIDs.Reference Kruithof ¹⁹ However, with the possible exception of longitudinal studies, researchers are typically not familiar with potential participants. Take, for example, a PWID who uses echolalic phrases to convey discomfort or enthusiasm.Reference Cohn ²⁰ On the surface, the words in an echolalic phrase may not reflect their meaning; instead, phrases may be connected to other emotionally significant events or related moments that a close supporter might understand after hearing it many times. ²¹ A third party, who lacks familiarity with the PWID, might doubt the supporter’s assertions regarding what the phrasing means. This problem, which is present in the clinical setting, can be especially pronounced in research given that clinical researchers tend to be highly specialized and often lack training and experience with ID. As a result, they may provide poor external checks on communication between PWIDs and their supporters.

We see several ways to address this concern. At a minimum, researchers should speak directly to the PWID as much as possible, honor the PWID’s communication preferences, and be flexible about trying new communication strategies.Reference Badcock and Sakellariou ²² For researchers who are unsure about their interpretations of the PWID’s communication, recruiting an additional researcher or member of a participant protection or ethics team could help to confirm, challenge, or add nuance to the discussion; their inclusion could also serve as an opportunity for researchers who lack confidence to learn new communication strategies. At the same time, inclusion of additional team members may make the PWID uncomfortable and potentially undermine the focus on the PWID’s perspective. Researchers should ask the PWID before including others and, regardless of the group’s size, make sure to give PWIDs ample time to think about their options, discuss them with their supporter(s), and ask questions.

Clinical research relies on specialized procedures and methods, such as blinding, randomization, and crossovers, which can be unfamiliar and intimidating to research participants. For example, in one ethics consultation, a PWID who made his own medical decisions and required very little support day-to-day became unnerved by the research environment, including the unfamiliar setting and complex concepts and jargon. In response, he chose to have his sibling act as his surrogate decision-maker for research. This individual, and other PWIDs struggling with navigating communication challenges specific to research, might benefit from supported decision-making. Instead of placing a surrogate in charge of communication, the supporter(s) can assist with communication and hopefully help PWIDs feel comfortable making their own decisions.

Importantly, though, the research community should not treat supported decision-making as the solution to consent process failures; instead, researchers should work to improve the consent process based on existing recommendations. ²³ For example, despite considerable efforts to improve research consent forms, they tend to be written at higher reading levels than recommended, are typically very long, and are often not understood by participants.Reference Foe and Larson ²⁴ Consistent with revised US regulations, consent forms should begin with a concise and clear summary of the key information individuals need to decide whether to participate. ²⁵ To minimize miscommunication, and provide supporters and PWIDs the best chance of understanding the decisions at hand, researchers working with PWIDs should pay careful attention to the length and complexity of consent forms. In addition, researchers should consider other consent form adaptations that have been endorsed by PWIDs, including easy-to-read information, verbal description, visual aids, checking understanding, and allowing more time for the consent process.Reference Cameron and Murphy ²⁶

Challenge 2: Influence

PWIDs are at heightened risk of social victimization, including manipulation and persuasion.Reference Fisher, Moskowitz and Holdapp ²⁷ Persuasion can occur with malicious or benevolent intent. For some caregivers, it is a “constant and unremarkable feature” of everyday decisions with PWIDs who demonstrate conflicting preferences or resist activities the caregiver considers essential.Reference McKearney ²⁸ Expectations that PWIDs will comply begin at a young age, and decision-making is often framed around protection and familial norms.Reference Saaltink ²⁹ In a study by Cameron & Murphy, ³⁰ researchers noted that some family members and caregivers did not understand their role in terms of promoting participant choice and instead pressured PWIDs to comply, with one caregiver noting that “gentle persuasion is often all that is needed” to convince PWIDs to participate in research.

Training, including explanation on the role of supporters, may help them to move away from more directive approaches and actively work to center the PWID’s preferences. However, supported decision-making, and decision-making generally, inherently involve influence from others. ³¹ This is often a good thing, as new information and ideas can encourage people to evaluate their choices and move forward with nuanced perspectives. However, researchers and supporters need to ensure that, in the process, they do not devalue the PWID’s expressed preferences and exert excessive influence over their decisions.

In another consult, researchers noted that a PWID enrolling in a study with a surrogate seemed to have a particularly agreeable demeanor, and worried that the PWID might not feel comfortable expressing dissent. Ethics team members met with the PWID and surrogate and asked the PWID about their understanding of their rights as a research participant, as well as their general level of comfort expressing dissent. For this PWID and others who tend toward agreeableness or discomfort with dissent, supported decision-making seems well-suited to addressing these concerns. In surrogate decision-making, a PWID’s quick agreement with the surrogate’s decision might simply be accepted without comment or discussion. In contrast, by centering the PWID in the decision-making process, supported decision-making is less likely to prompt and accept too-ready agreement on the part of PWIDs.

While supporters can buffer PWIDs from researcher influence, they might exert their own undue influence on the PWID. Take the parent of an adult PWID who finds an ongoing study on the PWID’s disability. The parent is excited about the possibility of contributing to science and helping families like theirs. The PWID has a fear of needles and is not interested in the study’s goals, but their parent repeatedly emphasizes the study’s importance and helps the PWID to enroll in the study. Permutations of this scenario could involve more or less influence, for better or worse reasons: perhaps the parent downplays or ridicules the PWID’s fear of needles, perhaps the PWID is morally opposed to the study because its aim is to identify genes that could be used select against their disability, or perhaps the parent wants to enroll the PWID in a clinical trial as the last resort for a terminal illness and pressures the PWID to enroll despite their understandable concerns about severe side effects. Regardless of the parent’s motivations, this process could result in undermining the PWID’s opportunity to decline.

These permutations raise the question of how much influence is acceptable in supported decision-making. PWIDs may seek out supported decision-making because they hope to receive guidance but, depending on their relationship with the supporter, their life experience, and the nature of their disability, they may benefit from more or less directive support. For example, a PWID who struggles to conceptualize long-term outcomes might put more weight on their discomfort traveling to a research facility than the possible benefit of an experimental trial for their degenerative disease. If the supporter has traveled with the PWID before, they may recognize that the PWID has a pattern of anticipatory anxiety about travel that dissipates during the travel itself, and they could try to redirect the PWID to consider the decision without thinking about the travel. We would see this as acceptable influence, with the supporter’s knowledge of the PWID helping to center their values; however, this redirection could be unacceptable with a PWID for whom the experience of travel is extremely distressing. For this PWID, travel may be very relevant to their decision. If their distress has been downplayed or reprimanded in the past, asking them to discount it might make them feel as though their input and strongly held preferences do not matter.

To address this complexity, we suggest PWIDs and supporters explicitly discuss the supporter’s role prior to the decision-making process, including how much input the PWID prefers, what kinds of input are acceptable, and what the supporter(s) can specifically do to help make decisions. This way, the PWID and supporter(s) will have an opportunity to address their existing relationship in the context of the decision, giving them a better chance of recognizing and avoiding undue influence as defined by the PWID themself.

Of course, it can be difficult to assess the amount of influence a supporter is exerting on a PWID and whether it is appropriate. If members of the research team have concerns that the supporter is exerting excessive influence, they should bring their concerns to the principal investigator, participant protection team, and/or ethics team; however, overt suspicion from the researcher runs the risk of alienating the PWID and supporter, working against the goal of ID-inclusive research.

One approach to this dilemma might be to remind the PWID of their rights and provide opportunities to interact with researchers without the supporter present. This additional space for the PWID to dissent could serve as an extra check on their willingness to participate. Unlike an expectation to correctly analyze complex social interactions, we do not think this is too much to ask of researchers. When possible, researchers could recruit participant protection or ethics teams to speak with the PWID instead. Their independence from the research team and the supporter, as well as experience speaking with participants about consent, may reduce undue influence and increase the PWID’s comfort and confidence expressing their preferences.

In addition, when PWIDs dissent, researchers should try to determine the source of their objection and whether it can be addressed short of study withdrawal. For example, some PWIDs find certain sensory experiences very upsetting, and researchers may be able to accommodate them by altering or skipping the relevant procedures or tests. Of course, researchers should be very careful here, emphasizing that they take the PWID’s dissent seriously before offering potential accommodations, and explicitly stating that these accommodations are only options if the PWID does decide to participate.

Challenge 3: Speculation and Displacement

At the intersection of communication and influence, speculation about PWIDs’ preferences and displacement of these preferences emerge as distinct ethical challenges in supported decision-making. The less supporters reliably understand and engage with the PWID, the more difficult it will be to discern the PWID’s preferences — leading, in some cases, to displacement of the PWID’s preferences with the supporter’s preferences. McCarthy & Howard ³² capture this tension with testimony from the mother of a minimally verbal autistic child, who explains that perceptions of nonverbal and minimally verbal communication as “untranslatable and unknowable” can lead supporters to “displace [these] voices with [their] own.”Reference Chew, Davidson and Orsini ³³ This practice risks decentering PWIDs from the supported decision-making process and raises questions about the line between supported and surrogate decision-making.

Sometimes supporters and surrogates recognize that, despite their best efforts, they must speculate about the PWID’s preferences. In one ethics consultation, the father of a PWID, who had been her primary caregiver for many years, stated that, although he had learned many small signs over time that made him better than others at knowing when his daughter was uncomfortable or upset, he did not always know what was “in her head.” The small signs described by the father are suggestive of “tacit knowledge,” or implicit interpretations based on cues learned through very close and longstanding relationships. ³⁴ We see tacit knowledge as a valuable resource for supporters and surrogates, especially when they recognize the limits of their knowledge and combine it with more explicit forms of communication for validation. For example, a PWID who is apprehensive or unsure how to articulate their discomfort to a researcher could benefit from a supporter who knows them well, recognizes their discomfort, and takes steps to pause the interaction or help the PWID voice their concerns. Familiarity with the PWID’s cues is valuable in a supporter if, importantly, the supporter also checks with the PWID to confirm their interpretations of these cues.

Subtle or enigmatic signals can be primary or even exclusive forms of communication for some PWIDs, making it very difficult for supporters to confirm their interpretations. In these cases, as the father in the ethics consultation above observes, speculation is unavoidable. Here, the boundaries between supported and surrogate decision-making become less clear. When a decision is based entirely in speculation about a PWID’s preferences, with no explicit communication, calling it supported decision-making seems erroneous. But how much speculation is acceptable, and how much explicit communication is required for supported decision-making? To address these questions, we will start with a clinical case from the supported decision-making literature and then discuss how it translates to a research context.

Watson et al.Reference Watson, Wilson and Hagillassis ³⁵ describe a man named Neil with profound ID facing life-or-death circumstances, which his family and caregivers addressed through what they described as supported decision-making. Neil “did not use or appear to understand formal communication,” and relied on those who knew him well to interpret his facial expressions, sounds, and body language. When Neil developed aspiration pneumonia, he and his advocates had to decide whether to proceed with a tracheostomy recommended by Neil’s medical team. Neil had a temporary tracheostomy in the past, which his advocates described as “horrible” and very distressing to Neil. In the authors’ account, the advocates first talked to the doctors about Neil’s options, then discussed their knowledge of Neil and his past experiences, shared documentation such as videos and a communication diary to analyze Neil’s behavior change after his temporary tracheostomy, and concluded (in Neil’s father’s words) that “if he had a choice he wouldn’t want that bloody thing again.” The advocates decided against the tracheostomy, and Neil died shortly thereafter.

In discussing this case, Watson et al. highlight the importance of historical knowledge and relational closeness in making “a collective decision that they [Neil’s advocates] thought reflected his will and preference.” Neil’s involvement in the decision-making process was limited to his past expressions of discomfort, and his team speculated that based on these expressions, Neil would probably prefer not to undergo the procedure to continue living.

This speculation highlights a central debate about the boundaries between supported and surrogate decision-making. Neil’s advocates collaborated using their close knowledge of Neil to draw conclusions, which aligns with the UNCRPD’s suggestion to use a “best interpretation” of will and preferences rather than a “best interests” standard in supported decision-making. ³⁶ However, Neil himself never directly engaged in the decision. His distress from having a tracheostomy was clear based on past experiences, but whether Neil would have chosen death over a tracheostomy if he understood the options is impossible to know. As Watson et al. mention, it is unclear whether Neil understood others’ communication. Hence, even with direct and sustained assistance, it seems unlikely that his advocates could identify Neil’s enduring preferences and values as defined by him.

On one influential philosophical account, a decision can be said to be the individual’s own only if it “meshes” with their “psychology” or “identity.” But:

Under this definition of supported decision-making (as well as Barclay’s ³⁸ ), Neil’s lack of direct input in the decision-making process would make this a case of surrogate decision-making. To clarify, although we see supported decision-making as a promising avenue, we do not think that surrogate decision-making is always objectionable. There are many good reasons Neil’s team may have been unable to engage him in the decision. Perhaps Neil was in too much pain or distress to communicate. Perhaps his care team judged that his implicit communication required too much speculation for such a complex decision. With both approaches — supported or surrogate — we believe the goal should be to engage the person affected as much and as directly as possible, taking into account the context and their past preferences.

Speculation, and its implications for the boundaries between supported and surrogate decision-making, can take on even more complexity in clinical research. Imagine “Nicole,” like Neil, communicates primarily via facial expressions, had a distressing temporary tracheostomy in the past, and has recently developed aspiration pneumonia. Her doctors offer a tracheostomy as standard-of-care but also tell Nicole and her family about an experimental procedure under evaluation at a nearby research hospital that could address Nicole’s acute illness. As in Neil’s case, Nicole’s family discusses the options and Nicole’s past experiences with different medical procedures. However, unlike Neil’s case, Nicole’s family must consider two additional research-specific factors: the potential unknown risks associated with clinical research, and the risks that research participants take on for the benefit of future patients.

The experimental procedure was designed to be less invasive and distressing than a tracheostomy, but experimental procedures are unproven, and the extent to which this procedure poses risks is largely unknown. If Nicole undergoes the experimental procedure, her data and outcomes will be used to determine its safety and efficacy for future patients, but it is unknown whether she will benefit personally. As a result, Nicole’s family has little basis to judge how the experimental procedure might play out for her.

This introduces another element of speculation in the decision: would Nicole accept the risks inherent in clinical research for a potentially less burdensome procedure, and would she want to contribute to a study designed to benefit others? These are abstract concepts that can be very difficult for research participants to understand. So how can Nicole’s family, unable to communicate with her verbally, make an educated guess as to what Nicole herself wants? In Neil’s case, his supporters substituted their best interpretation of his will and preferences based on data about the tracheostomy’s success, as well as their observations about Neil’s personal experience with his past procedure. When it comes to experimental procedures, though, Nicole’s family does not have access to this kind of information — they are dealing with a novel procedure, and a research landscape that has goals beyond Nicole’s well-being. While the speculation involved in Neil’s case brought the process to the boundary between supported and surrogate decision-making, we think the additional speculation introduced with clinical research pushes Nicole’s family’s deliberation clearly into surrogate decision-making.

Why does this boundary between supported and surrogate decision-making matter? In clinical research, the type of decision-making can provide justification for a participant’s exposure to different levels of risk. For example, when a child’s parents consent to enroll them in a clinical trial, there are strict limits on the extent to which the child may be exposed to research risks for the benefit of others.Reference Spriggs ³⁹ We think surrogate and supported decision-making with PWIDs should be treated similarly. When a surrogate consents for someone else, the research risks cannot be justified on the grounds that the individual who is facing them has consented. Instead, the research risks are justified by the extent to which facing them is consistent with the participant’s preferences and values, a determination that is made by the surrogate. Supported decision-making, on the other hand, offers the potential to justify research risks based on the fact that the PWID themself is involved in the decision to enroll.

The level of research risks that supported decision-making justifies depends on the extent of the PWID’s understanding and involvement in the decision-making process. The more the PWID makes the decision, with assistance from others, the greater the research risks that can be justified. The more that others speculate and make the decision to enroll on behalf of the PWID, the lower the research risks that can be justified. (For further discussion of the extent to which supported decision-making can justify research risks, see Segal et al. in this volume.)

For a decision to be considered a supported decision and thereby justify greater research risks, supporters should, at minimum, elicit direct involvement of the PWID, encouraging as much explicit and validated communication as possible, and the PWID should demonstrate basic understanding of the options at hand. The PWID should then be able to communicate a decision with final authority. We see two reasons for this. First, as we have discussed in Influence, treating PWIDs as central decision-makers in supported decision-making can help them feel more connected to the decision and encourage them to have closer involvement in the process. This is especially important in research because, unlike clinical contexts, the participant’s interests are not the only ones at stake. Second, to ensure the research risks and burdens are justified, it is important to be clear about who is the source of the decision to participate. When a PWID is not involved, or their decision is not treated as authoritative, we think calling the decision theirs is illusory and potentially harmful: the PWID in question did not actually consent to the research risks. In the long run, this process might result in supported decision-making becoming indistinguishable from surrogate decision-making, diluting the benefits of centering PWIDs in the decision-making process.

Earlier in this section, we mentioned a father who acknowledged that he did not have access to his adult child’s inner life. We think this kind of honesty and reflection aids in good decision-making. Speculation will be involved in many decisions, supported or surrogate, and to reach a decision that is respectful to the person affected — with as little displacement as possible — those involved should do their best to recognize when they are speculating, reflect on why, seek out explicit communication when available, and make judgments that take as much information about the PWID into account as possible. In clinical research, it is especially important to be clear about the PWID’s role in the decision-making process in order to account for additional complexity introduced by novel procedures, unknown risks, specialized methods, and goals that go beyond the individual participant’s care.