Ambivalence in Clinical Decision-Making
Or, Having Your Cake and Eating it Too
Silence hung in the air for a moment. John could feel his heart beating. Was this the end?
“So, we could focus on your comfort,” Dr. Harris said. Had they not been doing that already? What does that mean? “There’s also a phase 1 trial that we could consider…”
John tried to keep up. He confronted a stark decision: either travel for hours to where this trial was being conducted for a shot at living, or enroll in hospice to live out the rest of his days at home. He didn’t consciously appraise all the burdens and benefits, but he felt these were two different ways of living his remaining life. His wife was home with a cold; he wished she were here.
“I don’t know. I don’t know. Let’s just take it one day a time and we’ll see.” John’s mouth was dry.
Dr. Harris nodded. “Okay.” She sounded like she was hesitating. “That’s just fine. We’ll check back in a month. One step at a time.”
There are some papers that aren’t just practice-changing but paradigm-shifting. I feel that way about this paper on ambivalence by Bryanna Moore and colleagues.
I help people make decisions about their health. That’s not as straight-forward as it sounds (as I’ve written about). What impairs decision-making? What impairs good decision-making? For that matter, what is a good decision? Unfortunately, the milieu of medical ethics for the past several decades has so desiccated the response to these questions that many clinicians are satisfied with a patient who has decision-making capacity and can provide “informed consent” (understood in a rigidly procedural way - sign the form that explains to you the “risks and benefits” of this intervention).
The most common reasons clinicians assume someone can’t make a decision are either that the patient lacks information (so more explaining is needed), or they lack the capacity to make the decision (so a surrogate decision-maker is needed). Moore and colleagues help us to appreciate that there are many other reasons why decision-making might stall out. As they write, “In healthcare, ambivalence related mental states can pose a significant challenge to the timely provision of potentially beneficial medical treatment. Trying to decode patients’ seemingly contradictory attitudes or statements can be time and resource intensive, and cause confusion and frustration on the part of those who care for such patients.”
They take ambivalence and explode it into nine concepts:
If you assume that someone is uninformed when in fact their thinking is in evolution, then providing them with more information might annoy or confuse them, undermining decision-making (beware toxic knowledge). What they might need instead is a deadline. This might seem counterintuitive as a deadline appears to constrain autonomy, but in this case it could provide the necessary structure to support good decision-making.
Alternatively, if someone is apathetic, more information will also frustrate them. A clinician could instead make a strong recommendation and even seek informed assent. The patient’s engagement with these maneuvers provides insight into their decision-making processes, goals, and values. For example, if they adamantly refuse what’s recommended, then they’re no longer demonstrating apathy and that provides grist for the mill of the conversation.
Identifying the mental state that contributes to impaired decision-making is critical. Moore and colleagues offer some immensely helpful guidance on how to respond to each of these states:
They also provide some ethical guidance for anyone who assesses decision-making capacity and helps patients make decisions (i.e., almost any clinician):
What To Do with Ambivalence?
In my experience, I find true ambivalence to plague many people in decisions about their health. It’s not a comfortable place to be. They have a terrible decision to face and the decision itself can feel like an affliction. They want to live longer and they want to keep their leg. They want to spend more time at home and also receive the care only the hospital can provide. They want to die and they want to live. It’s not only choice A and choice B that cause distress, but the “and” too.
Frank van Harreveld and colleagues provide a helpful review of the concept of ambivalence and how people respond to it. There are many reasons why someone might become ambivalent:
Conflict between thoughts or between emotions
Conflict between thoughts and emotions
Conflict between/among goals (including conflict between short- and long-term goals)
Conflict with other people (which, although an interpersonal conflict, can manifest as what Moore and colleagues call “vacillation” as one attempts to satisfy multiple stakeholders)
This is in addition to other uncomfortable internal states like dissonance (the discomfort one experiences when their own behavior is in conflict with their attitude). We also can’t forget about internal states not directly addressed in either of these papers (e.g., delirium, dementia, depression, psychosis, etc.). But let’s stick with ambivalence for right now: you have a choice and the options embroil you in an internal conflict.
Ambivalent choosers have at least two distinct goals: the goal related to the content of the choice (e.g., “I want to live longer. How can we treat this disease?”) and also the goal of integrating their motivation and resolving their ambivalence. This means people may make choices (or avoid making choices - more on that in a moment) simply to resolve ambivalence. In so doing, they may choose against their values; that is, they may make a bad decision even if, in the clinician’s eyes, it was an informed, autonomous decision. Ironically, the clinician may have contributed to that bad decision in ways I’ll discuss shortly.
One way to resolve ambivalence is to just pick something. Despite how uncomfortable ambivalence can be, that’s not what everyone does. Why? At the very least, it may be due to:
Uncertainty about the outcomes.
The threat of regret. Not only that the outcome will be bad, but they’ll be responsible for it (having chosen what led to it).
Avoidance of negative thoughts and feelings associated with considering future outcomes.
There are two big ways to deal with this problem: emotion-focused and problem-focused coping.
Emotion-focused coping
One way to resolve ambivalence about a decision is to avoid making the decision altogether. The tragic irony of asking, “Do you want surgery or management with medication alone for your problem” is that the patient likely wants an unchooseable third option: to not be in this situation in the first place. Some people will indeed abandon the situation, leaving that mass on the side of their neck alone until they can no longer turn their head, until it’s physically unavoidable. Others will postpone their decision - “I’ll take it one day at a time.” These are flavors of avoidance.
The thing with avoidance is that it works. It can help assuage the fear, anxiety, and other negative thoughts and emotions associated with a difficult decision. At least, to some degree; the feelings may not go away entirely. Of course, it also doesn’t work, because most clinical decisions are about problems that will recur or grow. You’ll be faced with the problem again and again, so you can’t truly avoid it. But avoidance at least provides some emotional relief. Clinicians can foster this by avoiding hard conversations. Clinicians may do this because they’re also coping with their own emotions (who likes to deliver bad news?), or because the tyranny of the urgent nudges these conversations out of reach.
Another way to deal with ambivalence is to redefine the situation, like by denying responsibility. If you’re not responsible, you can’t regret a choice (though there are may be other negative valences associated with a particular outcome). In fact, if you’re not responsible, you might not have any decision to make at all. A patient might blame the clinical team, or “put it in God’s hands.” Clinicians may in turn defer all responsibility for decision-making to patients as a way to resolve their own ambivalence about the decisions at hand. If clinicians continue to engage a patient about a decision when the patient doesn’t believe they actually have a choice, it will just inflame conflict between the clinician and patient.
If one party redefines the situation, it’s usually not the case that this new perspective holds with the other party. So, a patient may “put it in God’s hands,” but the clinicians still need to make day-to-day decisions about what to do with the ventilator and the medications. The patient’s acceptance of medical interventions even while hoping God is in control demonstrate that they’re offloading decisional responsibility to God, even as their hope is actually split between God and the medical interventions.1 Their faith, in addition to doing other things, becomes a means of resolving ambivalence stoked by their medical circumstances.
Problem-focused coping
Well, the emotion-focused coping strategy will eventually break down as one’s health disintegrates. Eventually someone will need to make the decision.
One way to do this is to change one’s attitude about the decision. For example, you could trivialize one of the options: “That’s ridiculous,” “That’s quackery,” “That’s murder.” Or you could bolster another of the options: “A miracle cure!” “It’s really not that bad,” “I have no other choice.” You could have an accurate understanding of what the options are, but this doesn’t require you to.
Another way, probably the one with which we’re most familiar and at which the informed consent process aims most directly, is the systematic processing of information. What are the burdens, benefits, and alternatives of the proposed intervention(s)? The hope is that by resolving uncertainty about the future, decision-makers can also resolve ambivalence. The most obvious way to resolve uncertainty is to know as much about the options as possible. This is also the most effortful path to resolving ambivalence, though, and therefore usually the path taken last (even if, to the clinician’s eyes, patients appear to take it first - i.e., we opened with a discussion about all this to allow for informed consent).
There are less arduous problem-focused ways to resolve ambivalence. For example, one could rely on biases. I covered some of this when I wrote about toxic knowledge. If the complexity of information is overwhelming, or if the volume is too great or too little, there’s an increasing likelihood someone will rely on a bias to help manage that information. If a highly biased decision results in a goal-concordant outcome, it may only be by chance that it occurred. But remember: there are multiple goals in decision-making, one of them being the goal to resolve the discomfort of ambivalence, so it may be with greater the feelings of ambivalence, the more highly a patient will prioritize resolving that ambivalence over making a good decision.
One can also rely on heuristics. For example, someone can go with the majority or rely on their physician’s recommendation. This also helps to off-load responsibility in decision-making onto a process or another person.
Whether you employ emotion- or problem-focused coping to deal with ambivalence depends, in part, on your motivation (e.g., for consistency, accuracy) and ability. Having made a decision in one direction, for example, a patient may continue to make decisions in that direction to maintain consistency even if they continue to suffer harm (i.e., sunk costs). Ability can encompass things like health literacy, cognitive impairment (e.g., delirium, dementia), or, as the authors cite, ego depletion (that is, the fatigue that comes from making one weighty decision after another).
What’s a Clinician To Do (and Not Do)?
Amelie Rorty claims that “the best policy in the face of ambivalence involves a persistent, imaginative, and responsible attempt to understand and evaluate its sources and grounds. If it is discovered to be appropriate and justified, the best strategy involves a fertile and epistemically responsible use of the imagination to find ways of preserving the terms of both commitments.”
That is, figure out a way to have your cake and eat it too! Or as Harreveld and colleagues call it, redefine the situation. Rorty argues that “reframing the situation in this way opens the possibility of engaging like-minded members of the community in enlarging the range of acceptable solutions.”
Before we get to what that looks like in the clinical encounter, let’s consider some pitfalls.
Structuring malformed choices. Maybe a patient is ambivalent about something that is a non-issue or could have been avoided altogether. For example, armed with knowledge about a patient’s physiology but not about the patient, a well-intended clinician may present two courses of action and make a disease-centered recommendation based on the data they collected and what they know from trials. This may stoke ambivalence in the patient as part of them wants to pursue the treatment that is best for their problem and part of them wants to pursue the treatment that is best in some other way for them (and they don’t want to reject their physician’s recommendation). Similarly, a clinician may ask a patient “what they want,” which unintentionally pulls for fantasies that conflict with reality. So, too, discussing options that aren’t possible can impair decision-making by making a patient feel ambivalent about a choice they in fact do not have but wish they did (e.g., if your cancer had this mutation, we could use this therapy. Alas, it doesn’t have this mutation, so we can’t use this therapy).
Resolving ambivalence too quickly. Ambivalence tells us there’s a major conflict. Maybe it’s a conflict of emotion, maybe a conflict of goals. Who knows. If you try to resolve ambivalence too quickly, you might rush past important work that needs to be done to make a good decision. You might miss the deep emotional and existential elements below the surface. This means you may not have helped your patient make a good decision, ambivalence may flare again later about something you thought was settled, and trust and rapport might diminish as patients sense you’re dealing with them in a cursory way.
Failing to see the conflict. As Moore and colleagues showed us, if we can’t properly identify someone’s mental state, we’ll fail to respond to it appropriately. In the case of ambivalence, we need to understand what’s conflicted: Goals? Emotions? Thoughts? Where is there a conflict and why?
The gamified cup-and-ball technique. One way to dispel ambivalence, or avoid it altogether, is to displace the goals and values so that they no longer feel like they’re in conflict. We’re not talking about the existential threat of renal failure; we’re talking about creatinine. We’re not talking about when we run out of cancer-directed therapies; we’re talking about what cancer mutations would qualify you for trial enrollment. I call this technique “gamified” because it allows clinicians and patients alike to collapse values so that they’re easily digestible and comparable similar to how we experience games.
Death by a thousand decisions. Decision-making is hard. Resolving ambivalence is hard. Sometimes clinicians make this harder. When they don’t anchor clinical decision-making on the goals of care, clinicians are liable to turn the clinical encounter into many, possibly dozens, of decisions, rather than a decision for a care plan aimed at a prioritized goal. As alluded to earlier, people will fatigue after one or more weighty decisions and start to stall out in their decision-making or rely on unreliable methods to resolve ambivalence and make a decision. Clinicians undermine patients’ agency when they do this.
Taking it one day at a time. As I mentioned earlier, patients who report they’re “taking it one day a time” may be procrastinating as a means of emotionally coping with ambivalence. Clinicians can conspire along with patients in doing this as they, too, take it one day at a time. They lose sight of the bigger goals and prognosis of this person’s health. Sometimes “taking it one day a time” can be a helpful way of remaining grounded in the present moment, particularly if there are no decisions to make but rather you’re following through on a decision already made. But sometimes clinicians and patients delay decision-making together by kicking hard conversations further down the road.
What would you do? Patients can pack a lot in this question. It can be a problem-focused way of relieving ambivalence: they can rely on your decision instead of their own. It can also be an appeal: are you really offering me what you think is best? But it’s not about what you, the clinician, would do if you were in their shoes because decision-making is so sensitive to individual goals and values.
I agree with Rorty. In cases of true ambivalence, reframing may offer the best possibility of moving beyond it. This requires a lot of effort from both the clinician and patient. I think it’s also very context sensitive - not only depending on what the choices are but what kind of person the patient is. For example, a chaplain may help a patient understand how their faith interacts with their health such that the decision is seen in a new light.
Here are a few thoughts in response to the pitfalls mentioned above:
Structure choices properly. If you offer choices before understanding a patient’s goals and values, you’ve put the cart before the horse. In order to explore someone’s goals and values, you also need a shared understanding of what’s going on. In brief, you move from the news to the goals to the planning. At least then if a patient experiences ambivalence, it’s about real choices oriented around real values.
Stick with it. Don’t be afraid of ambivalence; venture deeper in. Consider asking open-ended questions to explore further: “I see you’re really struggling with this decision. What’s coming up for you right now?” You might make an inference: “After talking to you for a while, it seems like you’re stuck between wishing you could live on your own but recognizing you need to be in a facility. Do I have that right?” You might ask other questions about their experiences with ambivalence: “There’s a lot going on here. Has there ever been a time when you felt this way? What did you do?”
Look for it. Of course, the eyes don’t see what the mind doesn’t know. Just like you should expect emotion after giving hard news, you should expect ambivalence (or one of Moore and colleagues’ ambivalence-like mental states) after offering hard choices. You’re doing your patients a disservice if you present this like picking an ice cream flavor and find yourself surprised when they’re ambivalent.
Keep your eye on the ball. It’s really easy to lose sight of what we’re doing in medicine: helping to restore and sustain someone’s health by intervening on their body with three goals (longevity, function, and comfort). How we measure and monitor our work is important and often involves surrogate markers. That’s okay, as long as we see them for what they are, and don’t allow them to displace harder decisions about more concrete realities.
Know the goals of care. It’s worth repeating. Knowing the goals of care also helps you to make a recommendation for a particular path, rather than subjecting someone to several different decisions viewed in isolation. Sometimes you will have legitimate reasons for making several decisions. Often, though,
Avoid myopic procrastination. As I mentioned, “taking it one day a time” can be a helpful coping strategy to keep people grounded in the present moment when a decision has already been made or there’s truly no decision to make. But when it’s used to procrastinate about an important decision, it can lead people to wander into a crisis unprepared. Exploring hopes and worries is a gentle way to help people peek beyond today without pushing them to confront the essence of their ambivalence. Often they’ll say they’re hoping to get better or some variation thereof; don’t stop there - ask for other hopes. Sometimes you can also reframe the future not in terms of what someone might lose, but what they’ll need: “Is it okay if I talk some about what you’ll need when you get home?” It’s a gentler way of looking at a hard future. Another strategy would be to use a time-limited trial. It allows a patient to choose something, to try out a choice, but with a time limit and other boundaries.
Make a recommendation. A proper recommendation incorporates not only your knowledge of the relevant medical literature and your anecdotal experience, but the patient’s goals and values. Lacking that latter information, you’re just being paternalistic. The challenge with ambivalence and recommendations, though, is that those things that are meant to inform the recommendation are in conflict. Tricky! The recommendation itself may resolve the ambivalence or at least inform a discussion about the conflict that could eventually lead to resolution.
One more thought. “Respect for autonomy” is a bedrock principle of medical ethics.2 Folks typically take this to mean we shouldn’t interfere with an autonomous person’s choices as long as they’re not going to hurt someone. However, promoting liberty (freedom from) is only part of the picture. We should also aim to promote agency (freedom for). We want to not only help protect our patient’s from undue influence that might inhibit their liberty to make a decision, but also empower their agency so they can make a good decision.
Ambivalence reveals how disempowering the clinical encounter can be for patients. If we think that by leaving patients alone to make a hard decision we’re respecting autonomy, we may actually be undermining their agency. There’s more to it than offering information and asking for a decision. We should want to help our patients make good decisions about their health, which means we need to learn how to do that. It’s not easy (neither was learning about antibiotics, cancer, or endocrinological disorders), but it’s worth it.
This is the most useful piece I have read on this topic in my 40+ years of working in palliative care. Congratulations and thank you to the authors of the original paper and this superb commentary. Susan Block, MD
This is a keeper! As someone who is asked to participate in ordinary bedside ethics, that are never really ordinary, I will refer to this essay agin and again. Thank you for sharing.