[Start with Part I and Part II]
What is an acceptable outcome after critical illness?
It’s one thing to survive, but another to return to a good life. In many cases, we feel confident in predicting a poor prognosis less because we’re certain about the time remaining, and more because of the projected quality of that time.
We reason that:
- The patient with end-stage dementia might survive this bout of pneumonia, but even in the best case, will merely return to their previous state of being bed-bound and non-verbal. That’s no way to live.
- The patient with multi-organ dysfunction might make it out of the ICU, but only with a trach and PEG, and will be spending the rest of their life in a nursing facility. That’s no way to live.
- The young man who suffered a catastrophic intracranial hemorrhage with herniation might survive hospitalization, but will never again enjoy his hobbies, pursue his career, or do the things that make life meaningful to him. That’s no way to live.
It’s usually easier to describe outcomes qualitatively than quantitatively: for instance, we might not know the odds of a patient walking again, but we know that they’re being discharged to a nursing home and won’t be breathing off the ventilator any time soon. In short, prognostication—and hence decision-making—seems simplest when the projected quality of life appears poor.
But what is quality of life?
A good life
To limit life-prolonging care based on a patient’s projected quality of life (QOL) requires the accuracy of several nested predictions:
- First, Function: after stabilization of their acute illness, they will retain a degree of physical, cognitive, or emotional disability
- Then, Quality of life: their disability will result in a QOL they find unacceptable
- Then, the Determination that with such a QOL, their life would not be worth living
- Finally, an additional layer of subjectivity: that our assessment of the above will match the patient’s.
Each of these predictions will need to hold up to allow prognostication that leads us to appropriately limit life-prolonging care. This is a narrow needle to thread.
First, their function might end up better than we expected, and our ability to predict function is inconsistent at best.
Next, although their function is poor, their impression of it might be sanguine; not all disability is a disaster. Indeed, psychological and economic studies of happiness have repeatedly found evidence for a strong (although not completely immutable) hedonistic set-point: even after great setbacks in life, including new disability, most people’s overall happiness eventually tends to regress back toward their baseline. This appears to hold true in our patients as well: Detsky et al. found that six months after critical illness, patients were just as likely to report their QOL as being better, worse, or the same as before illness. How could it be better? Some patients may find it has given their lives new meaning or perspective, while others are surprised at how well they could tolerate the new changes. (We might be unable to imagine how we’d live a satisfying life with a tracheostomy, for instance, but that is no proof that we cannot; and even the avid tennis-player who will never run again might find new pleasure in playing music.) Of particular note, patients who were able to self-report QOL consistently graded it better than when surrogates reported it on their behalf—and across the board, clinicians were inaccurately pessimistic at the time of admission in predicting eventual QOL. People are more resilient than we think.
(Mitigating this somewhat, Detsky did find that better functional outcomes—particularly a normal neurologic status and the ability to return to one’s previous residence and pastimes—was associated with a better QOL. Happiness is not completely independent of function, it’s just not as dependent upon it as we might think.)
Finally, even if their QOL is greatly degraded, they might still find their life preferable to death. After all, most living people do, even unhappy ones.
What makes a life worth living?
Compared to the other layers, which are at least somewhat quantifiable, this final question is entirely subjective. Indeed, it lies more in the realm of philosophy than that of medicine, because it does not require illness to be relevant. When does a patient consider their life preferable to death after critical illness? Never mind that; when does any person consider their life preferable to death?
Camus famously described “the problem of suicide,” writing: “Deciding whether or not life is worth living is to answer the fundamental question in philosophy. All other questions follow from that.” Life is a random, perhaps even absurd series of events. Without a structuring perspective such as religion (which we generally try to separate from modern medicine), there is little intrinsic reason why anybody should prefer to be living instead of not. Yet we generally do.
There is undoubtedly a biological impetus here; a living organism without a natural urge to survive would be evolutionarily disadvantaged. But never mind that; there are biological underpinnings to most of our choices, but that doesn’t remove their meaning. Meaning doesn’t arise from biology, but from our perception of it.
We each define our life by the stories we tell, the structures of significance we create, the labels we assign, and the places and people and things we choose to hold dear. None of it is writ in stone; the molecules around us don’t care. None of it is objectively right or wrong. They’re all merely decisions, determinations about the type of reality we prefer to inhabit.
Many of us choose to value human life very highly, especially our own. That unique, impossibly complex network of neurons and the magical supervenience of consciousness is perhaps the most beautiful thing in the universe—from our perspective.
Or mine, anyway. But that’s the point. Throughout our lives, each of us has the highly personal task of deciding what’s important to us. Nobody can decide it for us. Nobody can tell us whether our life is better than our death. Critical illness changes the balance, giving as additional burdens, but still the calculation can only be performed by the patient, and perhaps not even by them. (As Odermatt and Stutzer found, even our own predictions of well-being after life-altering events are often inaccurate. The only way to know how we’ll tolerate disability is to try.)
I don’t know whether your life is worth living, you can’t tell me, and neither of us can tell the patient who is radically disabled after a severe stroke. His life is now radically altered, but is his new reality worth perpetuating? Who can say?
Where do we go from here?
We’ve shown now that what we previously considered straightforward, the “goals of care” discussion to “determine the patient’s wishes,” is in fact an attempt to answer a nigh-unanswerable question, one stymied by difficulties in prognostication, psychological smokescreens, and philosophical dead-ends. And yet, nearly every day in the ICU we are faced with addressing it, and avoiding it is no solution, simply an answer by default. Where do we go from here?
There’s no clear solution. But we might do a little better if remember a few guiding lights.
For all our training, we’re no closer to being expert on these issues than a lay-person. At most, we can provide some cautious guidance. Inevitably, though, there comes a point where the facts end, and subjectivity begins. If you think that the answers are clear, you are misunderstanding the questions.
Hard as it is for us, it’s even harder from the other side.
Wanting to live, even when the odds are against it, is not stupid. Wanting the same for our loved ones is not cruel. When it’s your turn in the crucible, you might just want it, too.
Acknowledge your biases
By recognizing and accepting our own perspective at the table, we can take steps to minimize its impact. When we begin to engage a family in an effort to “get the DNR,” we can pause, acknowledge to ourselves that we prefer to provide certain types of care over others, but assert that our preferences have no bearing on what goals are best for the patient.
Life-prolonging care, even low-yield, desperate attempts at it, is not evil.
Palliative end-of-life care is not noble.
Each is merely a tool, with a correct time and place to be wielded. The application of either one should inspire in us neither pleasure or shame—only a small whisper of doubt as to whether we chose correctly.
Hand in hand with the idea of humility is the precautionary principle. When in doubt, it may be wiser and sounder to err on the side of caution, and that may mean prolonging life.
When faced with this impossibly difficult question, often nobody will know the answer, not even the patient. But prolonging life at least temporarily (and it is always temporary) offers more time for clarity to emerge, and in the long run, more opportunities to adapt and discover happiness—as John Stuart Mills liked to say, more opportunities for “experiments in living.” At the peak of acuity during an ICU admission is the most difficult time to decide whether to live or die, yet that’s often when we ask.
Medical care that brings about comfort is just is important as that which seeks to prolong life. Yet when it’s unclear which is most appropriate, it takes greater hubris, with greater potential for error, to recommend accepting death than to postpone it.
These are difficult, troubling questions. Difficult because we often lack the answer, but also because there may be no answer; troubling because they reach beyond medicine, to the heart of what life is really about.
There’s no way to simplify the equation, and maybe that’s the real lesson here. We need to stop believing we have the right answer, and stop judging patients and their advocates who fail to “get it.” Instead, we should sit with them, offer whatever facts and guidance we can, and struggle alongside them to reach an outcome as good as we can endeavor to make it.
That may or may not be enough. But it’s all we’ve got.