Advance directive conversations still arrive too late for many patients
Advance directives are meant to do something simple and difficult at the same time: translate a patient’s values into medical instructions before a crisis strips away the chance to explain them. They include a living will, which covers preferences for CPR, mechanical ventilation, and artificial feeding, along with a durable power of attorney for health care. The problem is not the concept. It is the timing.
Only about 20% of adults have advance directives, and most people do not discuss goals of care until they are already hospitalized with serious illness. By then, the patient is often frightened, the family is under pressure, and the decision is being made in a room that was never designed for thoughtful planning. Patients who do have advance directives are less likely to be hospitalized in their final months and more likely to receive care that matches their wishes. The downstream effect is plain: when these conversations happen earlier, care tends to be better aligned with what patients actually want.
Code status is only one part of the broader discussion, but it often becomes the centerpiece because it forces a direct choice about whether to attempt resuscitation. In the emergency department, that question lands at the worst possible moment. A study of 187 high-risk patients and their surrogate decision makers found that 84% had never previously discussed code status, and 83% were unaware of their current code status. Many of these decisions, then, are being made in crisis rather than in clinic.
Code status remains one of the least taught clinical conversations
Physicians are not failing because they do not care. They are often failing because they were not trained to do this well.
Conversations about code status are challenging even for oncologists, and physician confidence in these discussions is lower than confidence in many other routine medical exchanges. The burden frequently lands on residents, who may be asked to handle the discussion before admission even though they have had little preparation. In one survey published in the Journal of Pain and Symptom Management, 68% of residents said they had not received code status training in medical school, 75% reported no formal training during residency, and only 13% felt competent to have the conversation on their first day of residency. Just 39% had ever had a code status discussion observed or critiqued by a supervising physician.
That training gap matters because the conversation is not a checkbox. It is a clinical skill that requires language, pacing, and the ability to respond to fear without turning the exchange into a lecture. A rushed question about resuscitation can sound like a mechanical prompt. A thoughtful conversation can uncover the patient’s actual goals, concerns, and limits.
Television has distorted expectations about CPR
Public understanding of resuscitation has been shaped by entertainment more than by bedside reality. A 1996 review of 96 episodes of “ER,” “Chicago Hope,” and “Rescue 911” found that 75% of patients shown receiving CPR were successfully resuscitated, and 67% of those were discharged from the hospital. A later review of 91 episodes of “Grey’s Anatomy” and “House” found a 69.6% CPR success rate, with 71.9% of those patients discharged.
Real-world outcomes are far less favorable. In-hospital CPR succeeds in about 44% of cases, and only about 17% of patients are ultimately discharged alive. For older adults and patients with serious illness, survival to discharge is less than 5%. Many who survive do so with substantial neurological and functional impairment. The gap between television and reality is wide enough to reshape expectations, and those expectations can drive decisions that are poorly matched to the likely outcome.
One sentence belongs in every serious discussion with patients and families: CPR is a choice to try to reverse death, not a neutral medical ritual.
Three questions can open a more useful conversation
A common emergency department prompt — “If your heart stops, do you want us to restart it?” — is technically direct, but it is often too abrupt to produce an informed answer. It invites a reflexive yes from people who have not yet been asked to think through the consequences. A more useful conversation starts earlier and builds from the patient’s own understanding.
A practical approach begins with a simple introduction: “I’d like to talk with you about a health care choice that I review with all patients.” From there, three questions can guide the discussion:
“What do you understand about your current health situation?”
“What are your health care goals for the future?”
“If you should die despite all our efforts, or in the event of your death, do you want heroic measures to try and bring you back?”
The wording matters. Using “die” or “death” makes the medical reality plain: CPR is an attempt to reverse death, not a vague promise of recovery. The conversation can then become more specific if the patient or family wants details. If someone chooses not to undergo CPR, that does not mean other treatments stop. Medications, antibiotics, fluids, and other supportive interventions can still continue for the current illness.
The bedside consequence is straightforward
When clinicians avoid these conversations until the emergency department, they often lose the chance to hear what the patient values before panic enters the room.
That is especially true for older adults and people with chronic or life-threatening disease, where the likelihood of meaningful recovery after CPR is low and the chance of surviving to discharge may be very small. In those settings, code status should be discussed as part of routine outpatient care, then revisited as illness evolves. The discussion needs repetition because goals can change, prognosis can change, and family members may later need to act on what the patient said months earlier.
The article’s central argument is not that every patient should refuse resuscitation. It is that every patient deserves a real conversation before a crisis forces a rushed one. Health care teams need better training, and patients need better information. Without both, code status decisions will continue to be made late, under stress, and with incomplete understanding of what CPR is likely to achieve.