clinician–patient interactions and value. Current communication and practice near the end of life can be improved. For instance, a landmark study found that do-not-resuscitate (DNR) orders are written an average of 2 days before death of the patient (SUPPORT Principal Investigators, 1995), and many patients with metastatic cancer are given new active treatment regimens within 2 weeks of death (Harrington and Smith, 2008). While costs increase and treatment too often intensifies at the end of life, Dr. Back explained, what families remember is hearing the message that “There is nothing more to be done” and feeling a sense of abandonment. The clinical reality is one in which doctors are hesitant to be frank with patients and many clinicians behave as though patient well-being will take care of itself if only the right drug is given, with or without discussion. In fact, medical outcomes are highly influenced by communication. Greater communication about transitions to end-of-life care appears to direct subsequent medical care and correlates with lower odds of intensive care unit (ICU) admission, ventilator use, and attempted resuscitation (Wright et al., 2008), as shown in Table 3-1. Without causing increased distress or depression, advance care planning discussions also correlate with greater patient acceptance of terminal illness (Table 3-1), as well as with improved caregiver quality of life, preparation for death, and reduced feelings of regret (Wright et al., 2008).

Despite the importance of these discussions, physicians feel ambivalent toward having them. When asked, doctors feel as though they have no good communication options—they fear taking away hope from patients and families if they do discuss prognosis, but they also know that crucial opportunities to improve care will be missed if these discussions are avoided. Often clinicians have had traumatic experiences in previous discussions of prognosis. This leads to negative emotions toward explicitly discussing prognosis or other difficult topics, and a type of collusion develops between the patient and clinician (The et al., 2000), which Dr. Back phrased “Don’t offer, don’t dwell.”

There is limited data on whether physician discomfort translates into fewer end-of-life discussions, but one study found that medical students maintained greater positive affect1 if they concealed bad news in simulated encounters with standardized patients compared to those who disclosed


Positive affect was measured using the Positive and Negative Affect Scale (Watson et al., 1988). Positive affect is “the conscious subjective aspect of an emotion considered apart from bodily changes” (Affect, 2009).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement