Bearing Others’ Burdens: Helping Surrogate Decision-makers
Notes from a Family Meeting, Vol. 3, No. 7
Things have not gone well for Paul Smith. Hours after the flock of white-coated physicians left his room he started to yell about intruders. No one else was there. Shortly after that, he lost the capacity to meaningfully communicate and only writhed in bed, picking at his sheets and batting at unseen assailants. None of the circumstances that prompted the physicians’ meeting with Paul had changed. Decisions still needed to be made about a course of action, but now Paul couldn’t make any of them.
A medical student sat next to her attending as they called Mr. Smith’s wife. “Mrs. Smith, this is your husband’s doctor at the hospital. He’s taken a turn for the worse.”
“What do you mean? I just stepped away to get a shower and handle some things at home!”
“He’s more confused now since our meeting. We’ve checked some blood work. His lactic acid level is very high.”
“Lactic acid? But I thought there was an issue with his kidneys? You know, I was worried something was going on even before our meeting. I wanted to say something but y’all left so quickly I didn’t want to interrupt.”
“Yes, ma’am. Lactic acid is…”
The medical student nodded along as her attending released an eloquent explanation of lactic acidosis, sepsis, and acute renal failure.
Ten minutes passed as Mrs. Smith and the physician exchanged questions and answers about Mr. Smith’s perilous pathological predicament. And then, making an abrupt turn back to the realm of the practical, the physician asked, “Would he want to go to the ICU? I don’t think he needs to go there right this moment, but if he gets sicker he might.”
“Go to the ICU? Good heavens, I don’t know. Who wants to be in an ICU? Do what you can for him. I’m on my way back.”
Paul Smith - at one time unable to decide because of toxic information, and now unable to decide because of delirium. In such cases where patients lose the capacity to make their own healthcare decisions, clinicians will turn to someone to speak for them.1 The standard by which this person is expected to make decisions is substituted judgment. That is, they should choose what the patient themselves would choose, if they could. The alternate standard is the best interest, where, because we don’t know what the patient would have chosen, the surrogate makes a decision based on what might best serve the patient. This latter standard can be unwieldy because what’s in someone’s best interest is dependent on a number of factors, not least of which would be the goal of care, a highly value-laden choice.
So, the physician has turned to Mr. Smith’s wife and asked, “What would he want to do?” It’s a straight-forward question. Does it allow for a straight-forward answer? Is it a question that helps Mrs. Smith make this decision?
Well, let’s step back and take a look at some of the data around surrogate decision-making.
Surrogates and Substituted Judgment
It seems surrogates are able to predict patients’ treatment preferences with about 68% accuracy - better than chance, though not perfect. Things vary by disease state, though. For example, in another study of patients on maintenance dialysis, only 27% of surrogates had spoken with patients about stopping dialysis. They had discordant views on prognosis (38% agreed) and general values about longevity versus comfort (45% agreed). Even still, in this population, surrogates and patients agreed on things like CPR (83% agreed).2 That discordance regarding prognosis extends to other disease states like cancer. It also varies based on the type of surrogate: for example, in dementia, spouses were less likely to opt for life-sustaining treatments than children when acting in the role of surrogate. You wouldn’t expect the type of surrogate to matter if they’re all using the substituted judgment standard.
In another study, patients who participated in their own decisions about end of life care were less likely to receive mechanical ventilation, artificial nutrition, chemotherapy, and treatment in an ICU than those who had decisions made for them by a surrogate. While this doesn’t prove that those surrogates didn’t know the wishes of their loved one, it is a strange association and suggests that there may be a discordance between what patients would choose for themselves and what surrogates would choose for them.
What does “substituted judgment” look like for surrogates? Well, they do rely on conversations they’ve had with the patient as well as documents that may been created by the patient. Conversations are general, if not downright vague, and advance directives have their own problems. Surrogates also rely on shared experiences to inform their decision making (“We’ve lived together for so many years, so I know…”), as well as their own values as a surrogate and the help of those upon whom they rely.
Indeed, surrogates are influenced by all sorts of things. For example, when considering prognosis, less than 2% of surrogates based their beliefs entirely on information provided to them by the patient’s physician. Instead, most factored in their “perceptions of the patient’s individual strength of character and will to live; the patient’s unique history of illness and survival; the surrogate’s own observations of the patient’s physical appearance; the surrogate’s belief that their presence at the bedside may improve prognosis; and the surrogate’s optimism, intuition ,and faith.” For some, these other factors were more important than the physician’s prognosis. Another study supported this, noting that surrogates were routinely more optimistic when physicians felt a patient had a low chance of survival: “Two main themes from the interviews explained this trend: surrogates’ need to register optimism in the face of a poor prognosis and surrogates’ belief that patient attributes unknown to the physician would lead to better-than-predicted outcomes.” Admittedly, a patient may also make decisions in this way, but substituted judgment doesn’t bind the surrogate the same decision-making process the patient might have used, only that they arrive at the outcome the patient would have chosen, regardless of the process. In fact, the process a surrogate uses to arrive at a decision is going to be different than what the patient would have used, just given that they’re different people.
So, it turns out surrogate decision-makers are human. As they receive prognostic information, some might be open to it, but others deflect or reject it. They’ll also express their own emotions, consider their role as a surrogate, and attempt to mobilize support for themselves in the midst of this decision. This is complicated work, involving not only the consideration of the patient’s wishes, but also advocacy for the patient, for other caregivers, and for themselves. They draw on previous experiences and struggle to accept the burden of the role. Many things help or hinder surrogate decision-making, like life circumstances of the surrogate and their own particular characteristics, their social networks, their relationship with the patient, and the quality of the relationship with the patient’s clinicians. They themselves identify that they need more preparation for their role, particularly in figuring out how to make informed surrogate decisions, how to initiate conversations with their loved ones, how to learn about their love one’s values and preferences, and how to communicate with clinicians.
Substituted judgment is the standard for surrogate decision-making and it’s not relied upon perfectly. Not only that, but it’s really only formally honored with matters requiring signature consent. We consider “what he would want” regarding a surgery. But if someone is psychotic in the hospital, it doesn’t matter that, when they were lucid, they told everyone they hate the way D2 antagonists make them feel and to never give them one. If their psychosis won’t improve without one, they’ll get a D2 antagonist in the hospital for lack of other options. Or if someone was ambivalent about taking their insulin, resulting in only 50% adherence at home, they will enjoy 100% adherence while they’re delirious in the hospital. If a patient would rather not be in the hospital but instead at home using cocaine, a surrogate can’t consent for their discharge to expedite the return to their habit even if that’s what they would have chosen when they had the capacity to choose it. There are limits to substituted judgment.
If Not Substituted Judgment, Then What?
There are good reasons to rely on substituted judgment. As David Wendler argues, five values are promoted through respecting the decisions of competent adults (and therefore using substituted judgment when they lose the capacity to make a decision): non-intrusion, sovereignty, authenticity, best interests, and consistency with one’s moral community. These are really important things to honor. When someone has the capacity to make medical decisions, we defer to them because we believe they are in the best position to reconcile a medical recommendation with their personal values and the narrative arc of their life. Absent that capacity, we hope someone else can help accomplish that same goal in decision-making.
What this requires, though, is intimate knowledge of “what they would want.” A truly substituted judgment is one in which a surrogate decision-maker authorizes a decision that we all know the patient would make for themselves. This is an uncommon occurrence, both because surrogates don’t often make decisions in this way, people so rarely know what patients want, and people’s preferences may change in light of new circumstances (affective forecasting errors). Maybe if we improved the quality of our advance care planning conversations and our advance directive forms? But we also can’t predict every single scenario that might befall someone. Would he want ECMO? Well, we never thought to ask that question. Surrogate decision-making will almost always involve some extrapolation from the surrogate’s knowledge of the patient’s values, rather than knowledge of individual preferences.
We end up doing something else in practice because we (both clinicians and surrogates) lack this detailed, intimate knowledge of a patient’s preferences. Sulmasy and Snyder describe it this way and call it substituted interests:
“Rather than interpreting a text or making a substituted judgment about what the patient might have wanted in imperfectly foreseen circumstances, the surrogate is asked to apply the patient’s authentic values and real interests, including the patient’s known preferences. The key question under best interests is, “What do you think is best for your mother?” Under substituted judgment it is, “what would your mother choose if she could tell us?” The substituted interests model says, “Tell us about your mother.” A good surrogate can articulate the patient’s authentic values by describing the patient’s loves, beliefs, and fundamental moral commitments rather than just specific preferences. Surrogates, ideally formally designated, know the patient as a person even if they do not know his or her precise wishes. This is what they substitute for the patient.”
Brudney, although still calling it substituted judgment, helpfully reframes the question, “What would she choose” into a different question that fits in the paradigm of substituted interests: “What could she choose, given her commitments?” In Brudney’s estimation, it’s the surrogate’s task to decide what is best for the patient, and a large part of what’s best for the patient is furthering the patient’s goals, values, and projects.
The standard for surrogate decision-making in the USA is still substituted judgment. The reality, though, is that surrogates often don’t have access to the kind of information they need to make a true substituted judgment which, by legal default, would push us into the territory of best interests. Suggestions from Sulmasy, Snyder, and Brudney help to guard against the overt paternalism that so often lurks around best interest decision-making.
Helping Surrogates
Given all this, I have a few suggestions about how we might do a better job of helping surrogates decide. This expands on what I had written earlier about “building” the goals of care, much of which applies to surrogate decision-making too. Let’s also keep in mind that when we’re helping surrogates, we’re also helping patients.
Review advance directives.
I’ll be honest: it’s not common when someone’s preferences outlined in an advance directive provide clear and convincing evidence of what they would want in this particular scenario. There are just too many extenuating circumstances. Oh yes, I know he wrote that he wouldn’t want to be sustained on machines if he had dementia, but in this case the doctors are offering a tracheostomy and he might be weaned off the ventilator in the next three or six months. I know she wrote that she wants to prioritize symptom management over longevity-focused care if she had a disease that was going to end her life soon, but she also really loves her grandchildren and it’ll be a several months before she’s able to see her new grandchild. In theory, an advance directive that is clear and convincing about a course of action can overrule a surrogate who says otherwise, but I’ve just not seen it play out that way, for better or worse. Nevertheless, there is benefit in reviewing advance directives because they might shed light on how this patient thinks about limitations on their care, and, who knows, might speak to the very scenario in which they now find themselves.
Bring the patient in.
Some patients are completely incapable of participating in these conversations and voicing any preference. But sometimes patients are capable of expressing some kind of preference, even if it isn’t fully informed. This information can be helpful for surrogates. For example, Navin and colleagues argue that there are three types of capacity for decision-making capacity for refusing medical interventions. The first is the one with which clinicians are most familiar: comparative refusal, formally meeting criteria set forth by Appelbaum and Grisso. Patients can also have burdens-based refusals (they understand their condition and appreciate the intervention carries a burden they reject), as well as goals-based refusals (they understand their condition and appreciate the intervention interferes with an important goal they could pursue without the intervention). The legal standard remains comparative refusal, but these other forms can help support surrogates in their decision-making if a patient is able to express preferences with some degree of appreciation.
Bring the surrogate in.
When someone has the capacity to make important medical decisions, clinicians often don’t involve the surrogate unless they happen to be present at an appointment or in the hospital room. It’s not part of the routine workflow in either setting to reach out to family, for example, to provide an update, even when a patient is planning to undergo a major procedure.3 Legally, the team is under no obligation to reach out, and the patient isn’t obligated to provide permission; they may even withhold permission. For an otherwise healthy person who’s prescribed antibiotics for sinusitis, that seems fine. For another person who has repeated bouts of hepatic encephalopathy, hepatocellular carcinoma, diabetes, and recurrent hospitalizations, the fact that they have the capacity to make a decision right this moment is likely balanced on the head of a pin. They could lose that capacity quickly and soon and, worse, fall into a life-threatening situation. In those cases, I recommend we involve the surrogate early and often in conversations, so that they’re up to date on health information, emotionally prepared to enter the surrogate role when needed, and are able to participate in important goals of care conversations. This way I can discern if there’s disparity between patient and surrogate in the perception of prognosis and GOC. For other patients, although they formally have the capacity to make a decision, they’re better equipped to make a good decision if they’re supported by their surrogate.
Ask, “What could the patient choose, given her commitments?”
I doubt you’d ask that exact question, but some variation of it. Questions that might help shine light on this could be:
“If she were able to talk about all this, what do you think she’d say?” Elevate the patient’s voice, even and especially if the patient can’t speak.
“Has she ever faced something like this before?” Try to discern not only how she may have decided when faced with similar hardship, but draw out themes of courage, resilience, and hope. What’s different, what’s similar?
“What did she enjoy doing when things were going well?” This will help you to contextualize your prognosis later in the conversation. Is she going to get back to rock climbing? Is she going to get back to sitting on the porch? Is she going to get back home?
“What are/were her most meaningful activities?” This question pulls for something more than simple enjoyment, though it can certainly include that. This question also helps to shape prognosis.
“What would she be willing to give up for a shot at getting X?” This highlights the trade-offs inherent in serious illness decision-making. Would she be willing to suffer more pain, or more time away from home, for a shot at living longer? Would she be willing to be sleepier, even unconscious, as long as her symptoms were well managed?
Don’t ask, “What would they want?”
Schwarze and colleagues have identified several problems with this innocuous-seeming question. First, it narrows attention onto specific preferences rather than goals. We don’t put the cart before the horse in this way when talking with patients, and neither should we with surrogates. Second, it limits discussion and bars the clinician from a surrogate’s deliberative process. Third, it sets surrogates up to choose fantasies, and also leads clinicians to assume a particular course of action without reckoning with trade-offs. The word “want” is unwieldy; use it with care.
Prepare surrogates to be holistic advocates.
Many surrogates are good at advocating for patients to get the care they need. It’s also the job of surrogates, though overlooked, to help patients avoid interventions that might harm them, given their goals and values. Surrogates need to advocate just as carefully against certain types of intervention as they should advocate for other types of intervention. This is a major frame shift for some surrogates and it requires emotional preparation.
Praise.
Being a surrogate is really hard. Many surrogates are also the primary caregivers and advocates for the patient, which are exhausting roles as well. Part of supporting a surrogate decision-maker is praising them for their effort and care. Now, if you’re speaking with a court-appointed guardian, this is probably less relevant, though I have encountered some who experience moral distress in making decisions at the end of life for patients, so it may still apply.
Make a recommendation.
Some of the least helpful conversations I’ve seen with surrogates have been when clinicians have offered choice A or B, tell the surrogate that they should choose what the patient would, and leave it at that. Instead, clinicians could offer a maternalistic recommendation: that is, a recommendation grounded in the values of the patient. “Given everything we know about her health right now and everything you’ve told me about her, I’d recommend B.”
There’s more to be said about pediatrics, about guardians, about patients who have no one to speak for them. I’ll leave those topics for another day. How else can we help surrogate decision-makers?
Trajectories
Following a meandering reading-path, sharing some brief commentary along the way.
“C. Thi Nguyen - Value Collapse”
Nguyen provides an important perspective on the concept of value collapse. The argument is that, as values become more explicit, they lose nuance and lead to harm because they distract us from things that fall outside their rigid purview. In medicine, this occurs when surrogate end-points become more important than the pursuit health. I realize, though, that I also do this when I clarify the goals of care: I try to make values more concrete and explicit, so we know priorities and can thus make decisions. If values remain expansive or embedded in narrative, they’re less actionable in a medical context. But what’s the trade-off here? Maybe that the goals of care can medicalize one’s life. I’ve argued that, while the GOC are ways medicine intervenes to support and restore health, they don’t exhaust what health is. But the more frail someone becomes, the more resources they’ll need to pour into pursuing their goal of care and the more comprehensive medical intervention will become. So what? If this is what I need to do to stay alive, get stronger, deal with this pain, so be it. But Nguyen argues that such value collapse inhibits “error metabolism,” the ability to manage and learn from mistakes. In medicine, we might miss that there are ways to flourish outside of a medical context, there is such a thing as dying, you can harm a dying person, medical technology has limits, etc. Ultimately, the human condition is tragic: we can’t have everything, sometimes having one thing means giving up another, and eventually everything is taken from us. Seriously ill people must prioritize their GOC. If they don’t, their GOC will be prioritized for them, usually in favor of longevity. One helpful way to foster better attention amidst these challenges is to check an explicit value against a more subtle, qualitative value - e.g., checking the GOC against the often more nebulous concepts of health and flourishing.
There’s something about digital culture that numbs us to wonder. That’s a big problem in medicine because without wonder, we risk losing curiosity. Without curiosity, compassion soon withers. Perhaps the way to cultivate curiosity and wonder is not to add something (a perennial quest in medical education), but rather to quiet down: draw oneself away from the screens, away from the din of medical technology, and deeper into the human experience of compassion and suffering.
Closing Thoughts
“Working with dying people asks you for a certain kind of soft focus, a slow gait in your thinking and your speech.
…
When a hard time is upon us, being hard is not required. Being supple in your understanding, keeping your willingness close to your memories and your skills, asking your eyes to stay open and wondering what is needed of you in that time is some of what is required. Whatever side of dying we are on, being willing to see things for what they ask of us is something to work at. That is a human skill.“
Stephen Jenkinson, Die Wise
Surrogate decision-making laws vary significantly from state to state within the USA (including the Veterans Affairs hospital system), and other countries have their own laws as well. What I offer here is, I hope, a broad framework for supporting surrogate decision-makers regardless of the particularities of the law, but you’ll need to consider it within your legal framework. Summarizing the history of the legal developments in this area goes way beyond what I’m trying to do here, though that history is important and interesting.
This says nothing about the accuracy of either surrogate or patient beliefs regarding CPR. Their agreement may also be incidental - e.g., just by chance, more people want CPR than not, and so surrogates have a better-than-chance possibility of guessing a patient’s preference correctly. I raise these concerns given how little agreement there was about the actual goals of care, and the fact that they hadn’t discussed limits on dialysis itself.
Exceptions to this include surgeries like transplant.