Bioethics Forum Essay
Whispered Compassion: Slow Codes, Ritual, and the Moral Limits of Transparency
A patient is dying of cancer. She insists that she wants everything done, including CPR. The doctors fear that honoring her wishes will be futile and lead to multiple invasive, painful procedures. What if she tells the doctors, in private, that she doesn’t really want CPR, but that her family is pressuring her not to give up, and so, when they are present, she accedes to their wishes? Would it ever be acceptable to agree not to withhold CPR, yet when the moment comes, either forgo it altogether or perform it less vigorously than usual?
Ongoing controversy about these situations, and about “slow codes” (or “show codes,” or “faux codes”) is surprising because the ethical arguments against deception seem so clear and the opinion of professional bodies unified in opposition to the practice. A recent essay in the Hastings Center Report highlights some of the complexities that keep controversy alive. The article describes an elaborate deception that involves putting false information in medical records—the patient’s chart indicates that the patient is full code, but there is an unwritten agreement between the patient and the medical team to withhold CPR if she has a heart attack. The article also reviews recent evidence suggesting that slow codes and their many variants are becoming more common. Their persistence suggests that something is missing from many analyses.
More than a decade ago, Bill Meadow and I defended slow codes in certain very restricted circumstances. Much has changed since then. Shared decision-making begins earlier, goals-of-care conversations are more robust, and palliative care has been more fully integrated into the care of patients with serious illnesses. One might have expected such developments to diminish the need for symbolic or attenuated resuscitation. Yet recent data cited in the Hastings Center Report article tell a different story. These practices not only persist; in some settings, they appear to be becoming more openly acknowledged.
Surveys and qualitative accounts show that many clinicians have participated in symbolic resuscitative efforts. Clinicians say that these were undertaken, at least in part, to meet the emotional needs of families and the legal demands of hospital policies, rather than to benefit the patients. (Many hospital policies require CPR for all patients having a heart attack who do not have a DNR.) Symbolic resuscitations are rarely experienced by clinicians as casual or self-serving lies. They are more often the moral residue of situations in which every available choice feels inadequate, and the only choices are among equally troubling options.
This trajectory suggests that slow codes respond to pressures that our ethical discourse has not yet learned how to relieve. Perhaps they should be understood not as evidence of moral laxity, but as expressions of a complex moral response to unresolved structural and emotional pressures that shape end-of-life decisions.
Biomedical ethics tends to treat deception as a moral bright line. To mislead patients or families—even for benevolent reasons—is framed as a violation of autonomy and a threat to trust. These concerns are real and serious. A medical culture that is comfortable with deception risks undermining the trust that allows competent and compassionate clinical care.
Public intuitions further complicate the picture. When members of the public are asked about slow or faux resuscitation, many say they would find such practices acceptable under certain circumstances. These views reveal an understanding of the emotional complexity that surrounds the withdrawal of life support and the quiet, mutually protective pretense that often shapes family conversations. If deception were as intuitively abhorrent as ethical theory presumes, these responses would be puzzling. Instead, they suggest that many people value presence, acknowledgment, and moral flexibility over rigid procedural purity at the end of life.
To acknowledge the persistence of a morally troubling practice is not to endorse it. Deception, as a rule, should be condemned. But when a practice survives despite decades of condemnation, morality may require more than louder denunciations. It may require understanding. Moral clarity does not arise from denying that transgressions occur but from honestly confronting the needs they express. As one recent commentator observed, slow codes may reflect “an act of disobedience against an unethical system.”
The responses to our earlier essay revealed this tension vividly. Many commentators condemned slow codes unequivocally, insisting that deception has no place in medical ethics. Yet almost all added an important qualification: Not every resuscitation must be conducted with identical vigor. It is appropriate, many argued, to “tailor” efforts to clinical realities and the psychological needs of families.
What was striking was how closely this proposed alternative resembled the very practice being disavowed—differing mainly in vocabulary rather than substance. Moral purity was preserved at the level of principle while discretion quietly reappeared at the bedside. The ethical problem of deception was not solved. It was just reconceptualized as discretion.
Slow codes are neither aberrations nor surprises. They persist because unresolved moral conflicts about end-of-life decisions persist. They are the predictable outcome of a system in which resuscitation remains the institutional default for patients without a DNR and clinicians have no sanctioned way to say, “This will not help, and we will not do it.” That such practices may be increasing should concern ethicists—not because clinicians are becoming more deceptive, but because of what they tell us about the legal and institutional pressures that create no-win choices for clinicians who want to provide compassionate care for dying patients.
If bioethics is to remain faithful to the realities of care, it must move beyond condemnation toward understanding, crafting moral spaces capacious enough to hold the emotional and spiritual complexities of dying. Resuscitation is not only a medical intervention but also a symbolic end-of-life ritual that theatrically expresses hope. Symbolic resuscitations attempt to preserve that ritual value. Many people don’t need that. Some do. Symbolic expressions, by their very nature, seldom comport with the demands of straightforward honest and total transparency. They are more complex morally.
Slow codes do not reflect a collapse of professional integrity. They gesture at the persistence of an unresolved moral conflict: an ethical framework fluent in the language of autonomy and agency, yet poorly equipped to respond when families cannot yet relinquish hope. Ethicists should be troubled not by the supposed deceptiveness of clinicians, but by the inadequacy of our moral and institutional frameworks to recognize wide variation in people’s emotional and relational needs, or the importance—for some—of the ritual dimensions of care at the end of life.
John D. Lantos, MD, is President of JDL Bioethics Consulting and a Hastings Center Fellow.
