Providers are often afraid to discuss accurate prognostic information with patients and their families for fear of removing all hope or causing sadness. A study of hospitalized elderly patients showed that patient preferences for end-of-life care were not documented or incorrectly documented in the medical record 70% of the time.7 Most studies, however, indicate that discussions of prognosis lead to greater patient satisfaction, improved concordance between treatments wanted and those actually received, less aggressive care at the end of life, and improved psychological outcomes for patients and their families. In fact, inadequate preparation for death is associated with poor bereavement outcomes for survivors with complicated grief, post-traumatic stress disorder, and a sense of regret or guilt.
In providers who do discuss prognosis, survival is often overestimated. In one study physicians overestimated survival by 5-fold.8 In fact, the longer a physician knows a patient, the more likely they are to provide an overly optimistic survival rate. Therefore, providers should prepare for these discussions with tools that use evidence-based calculators to improve prognostication such as eprognosis.org.9
Despite prognostic communication, families may not “hear” what clinicians tell them. Surrogate decision-makers generally believe that their loved one has a better chance of survival than the odds being cited.10 For example, if a clinician says there is a 10% chance of survival, the family member may believe that their loved one has about a 50% chance. These optimistic biases may be due to the belief that the patient is stronger than most or that the provider is misinformed and does not have all the information.
A family member may say: “My mother is a fighter. She had bad pneumonia before and walked out of here just fine last month.” A reasonable response to this may be to highlight why this situation is different from before and the outcome will likely be different as well. One might say: “I understand your mom is a fighter and survived the bad pneumonia last month. This time is different though because your mom is on a life support machine now and the emphysema has progressed. I’m worried she won’t be able to breathe on her own again.”
Optimistic biases are further reinforced by television shows that depict unrealistic survival rates following cardiopulmonary resuscitation (CPR) and create an expectation by family members of overly hopeful results and superior functional outcomes than are actually possible.11 Here, the Ask–Tell–Ask framework of communication also allows for clarification.
A family member may ask a question: “How long does she have to live?” The provider should answer honestly if she knows. For example, one might give an accurate range and say: “Most people in her condition can live for days to a week.” A follow up might be: “Knowing that time is short, are there others we should call to be here now?”
It is equally important to discuss a patient’s likely functional outcome if they do survive their ICU or hospital course. This may significantly impact medical decisions. Uncertainty can be acknowledged with phrases such as:
I wish I could be more certain right now. Your mom is very sick and the stroke was quite large. We’ll have a better understanding of her chances of surviving this over the next two days as we follow her blood pressure. Right now it is dangerously low. In terms of knowing her outcome, it may take weeks to know if she survives.
We are going to do everything we can to help your daughter’s body recover from this serious infection. The kinds of things we’ll be looking for are improvement in blood pressure, the lung function and whether she wakes up in the next few days. We will communicate with you at each step along the way.
Finally, when decisions are made for patients to receive life-supporting technologies, they should include clearly defined markers of success or failure of the intervention, and an end-point if unsuccessful. This will facilitate discontinuation of unwanted life-sustaining therapies if the condition fails to improve or deteriorates. Time-limited trials may be warranted if there is uncertainty about the reversibility of the illness. For example, a 48-hour trial of mechanical ventilation and ICU care may be employed to assess whether the pneumonia resolves in an elderly patient with end-stage heart failure, and can be withdrawn if there is no clinical improvement or clinical decline.
It should be noted that the overall goal of effective communication strategies in the ICU is to align medically appropriate treatments with patient and family values. For some, every moment of life has value regardless of quality. For these patients, family members may believe that survival alone, despite the fact that the patient may be ventilator-dependent, comatose with a feeding tube, and discharged to a long-term care facility, is an acceptable outcome.