Soft Codes and the Limits of Medicine

Photo by Joshua Chehov on Unsplash Meet Mr. Jones Mr. Jones was a 92-year-old man I encountered during my fourth-year Internal Medicine rotation.[1] He was a retired literature professor with a warm, didactic presence and a love for opera. Even in his hospital gown, he was never without his plaid baker boy hat. He had metastatic melanoma for two years, undergoing multiple rounds of chemotherapy that ultimately failed to control his disease. He was admitted with severe pneumonia requiring ventilatory support. Over the days that followed, I saw him deteriorate: already thin, he became almost translucent. His breathing grew increasingly labored, and he drifted in and out of consciousness as carbon dioxide accumulated in his blood. One afternoon, when he was briefly lucid, our attending asked the resident and me to establish his resuscitation preferences before handover to the night team. We entered his room to find his two daughters beside him, one knitting, the other reading aloud from a book he had grown too weak to hold. My colleague explained carefully that his condition was serious and that if his heart were to stop, we would need to know whether he wished to receive CPR and potentially be transferred to the intensive care unit. Before he could respond, his eldest daughter interjected: “We would want everything, right dad? You have always been a fighter.” The resident added what had become a familiar script: that CPR is not gentle, that ribs are often broken. “Does that sound like something you would want?” Mr. Jones was quiet for a moment. Then he said simply, “I would like you to do everything that you can.” His daughter looked at him with something like relief, even pride. I left the room with a sense that he had chosen wrong. Two nights later, a code blue was called on his floor. As we gathered outside Mr. Jones’ room, the senior resident turned to the team and said, “He is for full resuscitation, but he is very sick. It will be a soft code.” Everyone understood. They began chest compressions on his frail body, careful and controlled. His ribs did not break. One of his daughters stood in the corner, tears falling quietly, as the blood pressure cuff sighed. His time of death was called fifteen minutes later. Soft Codes as a Symptom This practice has a name, or rather, several. Documented in bioethics and clinical ethics literature under terms including “slow code”, “show code”, and “sham code”, it refers to cardiopulmonary resuscitation (CPR) that is deliberately limited or symbolic. It typically arises when a patient is officially designated as a full code, whether by default or by expressed preference, despite an internal clinical judgement that resuscitation is unlikely to be beneficial. A common ethical objection to soft codes concerns honesty. The American Heart Association’s guidelines describe deliberately low-quality CPR of this kind as “tantamount to deliberate deception” and therefore not an ethically sound approach.[2] This reflection is less concerned with delivering a verdict on whether soft codes are themselves ethical, and more with what their very existence reveals. The dominant critique places them within a framework of individual conduct: a failure of physician honesty to be corrected by following a fair process, for instance, by arranging a do-not-resuscitate (DNR) order when CPR is deemed non-beneficial. This framing does not address why the practice arises across different clinical settings, or why it can feel legible to clinicians at the bedside, as it did to those outside Mr. Jones’s room. Soft codes act as a pressure valve: a symptom of unresolved ethical tensions in how contemporary medicine approaches death and dying. Unexamined Assumptions “A Longer Life is Inherently Better” The World Health Organisation (WHO) defines the goal of public health to be prevention of “disease, promote health, and prolong life.”[3] This formulation assumes these aims are aligned and mutually reinforcing. It implies that prolonging life is itself a good, one that inherently occurs alongside, rather than in tension with, the promotion of health, which the WHO defines as a complete state of well-being.[4] But as our capacity for medical intervention has grown, this implicit assumption becomes harder to sustain. As Gawande observes, “technology can sustain our organs until we are well past awareness and coherence.”[5] The question increasingly becomes not whether we can prolong life, but when the prolongation of life begins to undermine well-being. Research from the Coping with Cancer project found that terminally ill patients who received aggressive life-prolonging interventions had a substantially worse quality of life in their final week, and their caregivers were three times more likely to suffer major depression six months later.[6] I am aware that in reaching for this data, I am doing something worth pausing on: feeling the need to legitimate the value-based claim that how we die matters with empirical evidence. Part of what sustains our tendency to value longevity over the experience of dying is that the length of life is measurable in a way in which the quality of dying is not. “If We Can, We Should” If prolonging life is treated as a default good, a further assumption follows: that the capacity to intervene entails an obligation to do so. The emergence of CPR as common practice illustrates how this assumption became institutionalized. Modern CPR was first presented as a context-specific intervention for immediate reversal of unexpected cardiac arrest in otherwise healthy patients.[7] However, the practice quickly escaped this narrow indication and became, in effect, a requirement for every in-hospital death. The frictions arising from this blanket application eventually prompted the development of formal DNR orders, making it legally possible to withhold CPR while simultaneously cementing it as an operational default.[8] This default is not value neutral. It reflects a commitment to technological maximalism, the belief that more intervention is always better.[9] This is evident in our language itself: framing maximal resuscitation as “doing everything” makes it seem the most generous option.[10] But as Hume established, one cannot derive an ought from an is.[11] The fact that an intervention is technically possible cannot tell us whether it should be pursued. There will always be a value-laden premise between what we can do and what we should do, whether we acknowledge it or not. “Science Can Define What is Beneficial and What is Not” Physician integrity entails an obligation not to provide interventions that are unlikely to benefit the patient.[12] But can medical science alone determine what constitutes benefit to a patient, beyond the physiological effects of an intervention? Drawing again on Hume’s warning against deriving an ought from an is, I would argue that it cannot. Science can describe survival probabilities, prognosis, and complications, but it cannot determine under what conditions a life ought to be prolonged. The line between beneficial and futile is not simply discovered; it is drawn, informed by clinical evidence but ultimately requiring a value judgement. Just as drawing that line is value-laden, not drawing it is no more neutral. Medicine appears uncomfortable acknowledging its value-laden nature. It tends to avoid what has been described as “the inconvenient truth about the real limits of medical science as the basis for decisions.”[13] The evolution of the physician’s role, in part, reflects this discomfort. Where the paternalistic “Doctor Knows Best” model made value judgments explicit, for better or worse, the pendulum has shifted toward an informative model, in which the clinician presents options, and the patient decides.[14] I have encountered this model in practice. The hospitals in which I trained used a framework of levels of care (LOC). On admission, patients would be walked through a menu of interventions (CPR, intubation, dialysis) and assigned a label from A to D, where A implied “doing everything” and D implied no life-prolonging therapy. As a junior trainee, I was often tasked to go “get an LOC” on a patient. I would introduce myself and ask whether, if their heart were to stop, they would want the listed interventions. When a patient asked what dialysis was, I would explain that kidneys filter the body’s toxins, and that dialysis meant a machine doing that work in their place. They would then check yes or no in the box. Autonomy requires informed consent. It is worth asking whether a checklist, administered in this way, genuinely enables informed decisions. This brings me back to Mr. Jones. At the time, I found myself thinking he had chosen wrong. In retrospect, my reaction reflects the very problem with the informative model. A patient reasonably assumes that the options presented are ones their doctor believes may be beneficial. Yet under the informative model, we present options expecting patients to choose in accordance with our unspoken clinical judgment, thereby asking patients to read our minds under the guise of respecting their autonomy. But autonomy requires honesty, not unlimited menus.[15] We should therefore ask who these checklists are designed to serve. Their appeal can be understood through defensive practice:[16] for institutions, documenting patient choices reduces legal exposure; for clinicians, it can limit the extent to which they must defend explicit value judgements about futility. “Action Means Exerting Physiological Control” On my clinical teaching unit, patients who were no longer designated as LOC A were commonly referred to as “inactive”. This terminology carries a revealing assumption: that action, in clinical practice, is defined as intervention aimed at measurable physiological control, its ultimate expression being the commitment to stave off death. Meanwhile, interventions no longer curative in intent, including managin

Authors

Publication Details

Journal
Voices in Bioethics
Published
2026-11-14
DOI
https://doi.org/10.52214/c4dx6v04
Primary Topic
Palliative Care and End-of-Life Issues
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Soft Codes and the Limits of Medicine

Anna Elizabeth Droppert
Voices in Bioethics
Palliative Care and End-of-Life Issues
article

Soft Codes and the Limits of Medicine

Anna Elizabeth Droppert
article en

Abstract

Photo by Joshua Chehov on Unsplash Meet Mr. Jones Mr. Jones was a 92-year-old man I encountered during my fourth-year Internal Medicine rotation.[1] He was a retired literature professor with a warm, didactic presence and a love for opera. Even in his hospital gown, he was never without his plaid baker boy hat. He had metastatic melanoma for two years, undergoing multiple rounds of chemotherapy that ultimately failed to control his disease. He was admitted with severe pneumonia requiring ventilatory support. Over the days that followed, I saw him deteriorate: already thin, he became almost translucent. His breathing grew increasingly labored, and he drifted in and out of consciousness as carbon dioxide accumulated in his blood. One afternoon, when he was briefly lucid, our attending asked the resident and me to establish his resuscitation preferences before handover to the night team. We entered his room to find his two daughters beside him, one knitting, the other reading aloud from a book he had grown too weak to hold. My colleague explained carefully that his condition was serious and that if his heart were to stop, we would need to know whether he wished to receive CPR and potentially be transferred to the intensive care unit. Before he could respond, his eldest daughter interjected: “We would want everything, right dad? You have always been a fighter.” The resident added what had become a familiar script: that CPR is not gentle, that ribs are often broken. “Does that sound like something you would want?” Mr. Jones was quiet for a moment. Then he said simply, “I would like you to do everything that you can.” His daughter looked at him with something like relief, even pride. I left the room with a sense that he had chosen wrong. Two nights later, a code blue was called on his floor. As we gathered outside Mr. Jones’ room, the senior resident turned to the team and said, “He is for full resuscitation, but he is very sick. It will be a soft code.” Everyone understood. They began chest compressions on his frail body, careful and controlled. His ribs did not break. One of his daughters stood in the corner, tears falling quietly, as the blood pressure cuff sighed. His time of death was called fifteen minutes later. Soft Codes as a Symptom This practice has a name, or rather, several. Documented in bioethics and clinical ethics literature under terms including “slow code”, “show code”, and “sham code”, it refers to cardiopulmonary resuscitation (CPR) that is deliberately limited or symbolic. It typically arises when a patient is officially designated as a full code, whether by default or by expressed preference, despite an internal clinical judgement that resuscitation is unlikely to be beneficial. A common ethical objection to soft codes concerns honesty. The American Heart Association’s guidelines describe deliberately low-quality CPR of this kind as “tantamount to deliberate deception” and therefore not an ethically sound approach.[2] This reflection is less concerned with delivering a verdict on whether soft codes are themselves ethical, and more with what their very existence reveals. The dominant critique places them within a framework of individual conduct: a failure of physician honesty to be corrected by following a fair process, for instance, by arranging a do-not-resuscitate (DNR) order when CPR is deemed non-beneficial. This framing does not address why the practice arises across different clinical settings, or why it can feel legible to clinicians at the bedside, as it did to those outside Mr. Jones’s room. Soft codes act as a pressure valve: a symptom of unresolved ethical tensions in how contemporary medicine approaches death and dying. Unexamined Assumptions “A Longer Life is Inherently Better” The World Health Organisation (WHO) defines the goal of public health to be prevention of “disease, promote health, and prolong life.”[3] This formulation assumes these aims are aligned and mutually reinforcing. It implies that prolonging life is itself a good, one that inherently occurs alongside, rather than in tension with, the promotion of health, which the WHO defines as a complete state of well-being.[4] But as our capacity for medical intervention has grown, this implicit assumption becomes harder to sustain. As Gawande observes, “technology can sustain our organs until we are well past awareness and coherence.”[5] The question increasingly becomes not whether we can prolong life, but when the prolongation of life begins to undermine well-being. Research from the Coping with Cancer project found that terminally ill patients who received aggressive life-prolonging interventions had a substantially worse quality of life in their final week, and their caregivers were three times more likely to suffer major depression six months later.[6] I am aware that in reaching for this data, I am doing something worth pausing on: feeling the need to legitimate the value-based claim that how we die matters with empirical evidence. Part of what sustains our tendency to value longevity over the experience of dying is that the length of life is measurable in a way in which the quality of dying is not. “If We Can, We Should” If prolonging life is treated as a default good, a further assumption follows: that the capacity to intervene entails an obligation to do so. The emergence of CPR as common practice illustrates how this assumption became institutionalized. Modern CPR was first presented as a context-specific intervention for immediate reversal of unexpected cardiac arrest in otherwise healthy patients.[7] However, the practice quickly escaped this narrow indication and became, in effect, a requirement for every in-hospital death. The frictions arising from this blanket application eventually prompted the development of formal DNR orders, making it legally possible to withhold CPR while simultaneously cementing it as an operational default.[8] This default is not value neutral. It reflects a commitment to technological maximalism, the belief that more intervention is always better.[9] This is evident in our language itself: framing maximal resuscitation as “doing everything” makes it seem the most generous option.[10] But as Hume established, one cannot derive an ought from an is.[11] The fact that an intervention is technically possible cannot tell us whether it should be pursued. There will always be a value-laden premise between what we can do and what we should do, whether we acknowledge it or not. “Science Can Define What is Beneficial and What is Not” Physician integrity entails an obligation not to provide interventions that are unlikely to benefit the patient.[12] But can medical science alone determine what constitutes benefit to a patient, beyond the physiological effects of an intervention? Drawing again on Hume’s warning against deriving an ought from an is, I would argue that it cannot. Science can describe survival probabilities, prognosis, and complications, but it cannot determine under what conditions a life ought to be prolonged. The line between beneficial and futile is not simply discovered; it is drawn, informed by clinical evidence but ultimately requiring a value judgement. Just as drawing that line is value-laden, not drawing it is no more neutral. Medicine appears uncomfortable acknowledging its value-laden nature. It tends to avoid what has been described as “the inconvenient truth about the real limits of medical science as the basis for decisions.”[13] The evolution of the physician’s role, in part, reflects this discomfort. Where the paternalistic “Doctor Knows Best” model made value judgments explicit, for better or worse, the pendulum has shifted toward an informative model, in which the clinician presents options, and the patient decides.[14] I have encountered this model in practice. The hospitals in which I trained used a framework of levels of care (LOC). On admission, patients would be walked through a menu of interventions (CPR, intubation, dialysis) and assigned a label from A to D, where A implied “doing everything” and D implied no life-prolonging therapy. As a junior trainee, I was often tasked to go “get an LOC” on a patient. I would introduce myself and ask whether, if their heart were to stop, they would want the listed interventions. When a patient asked what dialysis was, I would explain that kidneys filter the body’s toxins, and that dialysis meant a machine doing that work in their place. They would then check yes or no in the box. Autonomy requires informed consent. It is worth asking whether a checklist, administered in this way, genuinely enables informed decisions. This brings me back to Mr. Jones. At the time, I found myself thinking he had chosen wrong. In retrospect, my reaction reflects the very problem with the informative model. A patient reasonably assumes that the options presented are ones their doctor believes may be beneficial. Yet under the informative model, we present options expecting patients to choose in accordance with our unspoken clinical judgment, thereby asking patients to read our minds under the guise of respecting their autonomy. But autonomy requires honesty, not unlimited menus.[15] We should therefore ask who these checklists are designed to serve. Their appeal can be understood through defensive practice:[16] for institutions, documenting patient choices reduces legal exposure; for clinicians, it can limit the extent to which they must defend explicit value judgements about futility. “Action Means Exerting Physiological Control” On my clinical teaching unit, patients who were no longer designated as LOC A were commonly referred to as “inactive”. This terminology carries a revealing assumption: that action, in clinical practice, is defined as intervention aimed at measurable physiological control, its ultimate expression being the commitment to stave off death. Meanwhile, interventions no longer curative in intent, including managin

Voices in Bioethics
Palliative Care and End-of-Life Issues
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon