Ask any intern about one of their hardest days in medical training, and they will probably tell you a story about the ICU. Many of our patients die here and so often, alongside the emotional pain of witnessing a death, we also bear witness to the pain of protracted dying. As trainees, we start to understand which patients are unlikely to survive and those who might have a real, fighting chance. But rarely are these discussions of prognosis made explicit to patients and their families. Importantly, the myriad of considerations that constitute a thoughtful prognostic estimate are rarely made explicit to trainees. We internalize a sense that prognosticating is an elusive proficiency.
Despite this, in academic medical centers, trainees are often the team members faced with family questions about prognosis. We are the ones asked to “update the family” or even lead discussions about code status or goals of care. This practice inadvertently signals that communication about prognosis is intuitive, straightforward, and harmless. When really, being unprepared for these discussions is a major source of moral distress for trainees.
On the other side of the ICU, as the family member of an ICU patient recently, I can say that this uneasiness with discussions around prognosis is harmful. I’ve heard the party line and so have you: “Well, I don’t have a crystal ball. I can’t predict the future.”
Prognosticating doesn’t imply certainty, so why do we preemptively remind families of something they already know? A prognosis is a forecast. It was pouring a beautiful, summer rain the moment my family member finally died in the ICU. This got me thinking about weather as a more helpful metaphor for what we are aiming to do with the messy task of prognosis. If the weatherman decided forecasting was too uncertain, how would we know to plan for rain?
We all unconsciously acknowledge the unpredictability of future events. Apparently, this is known as aleatory uncertainty. But there is another type of uncertainty: epistemic uncertainty, that which arises from limitations in our knowledge and our models of understanding. As physicians, I think we conflate these two types of uncertainty. Sometimes we get it wrong because of these epistemic limitations: our data isn’t as good as it could be, our reasoning isn’t as rigorous as it ought to be, our knowledge is still so limited, and all too often, system problems interfere. Sometimes though, we get it wrong because sometimes we are going to get it wrong.
Should we never prognosticate because sometimes we get it wrong? Of course not. There is uncertainty without prognosis. But what I learned on the other side of the ICU is that without clarifying the probabilities of what the future might hold, we face definite uncertainty. How much is too much? How little is too little? What are we working toward? When patients and families ask about prognosis, often, they are asking for “permission to grieve or permission to hope.” We want to know if we should pack an umbrella.
We don’t teach medical trainees to be weatherwomen and men. This is where I see a critical gap in medical education. We need a deliberate strategy to fill this gap – to build a dedicated curriculum on how to develop prognostic estimates (particularly for those patients who are critically ill) and how to communicate them to families.
To start building this curriculum, we have to reposition the patient rather than the pathology at the center of our planning. We have to incorporate the knowledge that comes from looking at our patients as people: their pre-illness functional status, their support systems, their ability and willingness to participate in physical therapy, their spiritual life.
Second, we need to enlighten trainees to the ever-growing repository of resources that can help us understand the impact of severity and type of pathology on prognosis. Trainees need to know where to look for this data. Electives spent with specialists could have dedicated attention to the prognosis of certain conditions. In collaboration with palliative trained specialists, we need instruction and practice inquiring about what type of prognostic information families want: time, functional status, or unpredictability.
Finally, learning about the psychosocial characteristics that impact functional survival will humble and motivate a new generation of physicians to meet the challenge of our patients’ realities. Ignoring data about how insurance status, geographic location, and socioeconomic status could impact recovery from critical illness doesn’t make it go away. We can’t aspire to health equity without the deliberate practice of trying to understand where our patients are coming from.
As physicians, we have the great honor of caring for patients and their families. This honor requires courage to take on the uncomfortable responsibility of offering prognostic estimates to patients and families despite uncertainty. We need to see our task not as fortune-telling but as forecasting. We need to support trainees in figuring out how to help patients and families plan for rain.
Blythe Fiscella is is a cat-mom who is allergic to cats. She is an emergency and internal medicine resident at ChristianaCare hospital in Delaware. Her interests include medical education, communication, prognosis, and critical care.
Image by Imagezoo / Gettyimages
