Ethicists proclaim "slow codes" to be ethical
Can refusing to perform "futile" CPR be a form of civil disobedience?
The journal Bioethics has devoted an entire issue to the issue of so-called “slow codes,” wherein healthcare teams deliberately provide inadequate chest compressions during CPR or provide other half-hearted care intended to allow a patient to die more peacefully.
Slow codes are inherently deceptive and have been pronounced to be ethically wrong, but the bioethicist authors argue this societal debate was foreclosed prematurely (and that they will be the ones to reopen it now).
The conversation they have started dovetails well with the one we’ve been having here regarding decision-making as to the appropriate duration of CPR.
Some of the ethicists’ arguments can be summarized as follows:
1. CPR was originally intended to be used selectively, and its expansion was unjustified
Stuart McLennan et al cite a 1965 monograph on the then-innovative practice of CPR, which delimited its indications to patients with acute, reversible causes of cardiac arrest:
“Resuscitation of the dying patient with irreparable damage to the heart, lungs, brain, or any other vital system of the body has no medical, ethical, or moral justification. The techniques described in this monograph are designed to resuscitate the victims of acute insult, whether it be from drowning, electrical shock, untoward effect of drugs, anaesthetic accident, heart block, acute myocardial infarction or surgery.
Resuscitative measures on terminal patients will, at best, return them to the dying state. The physician should concentrate on resuscitating patients who were in good health preceding the arrest, and who are likely to resume a normal existence.
In the authors’ telling, the subsequent expansion of the use of CPR to the default (for all patients without an advance directive) occurred ad hoc, without the necessary measures of consideration and debate, and itself rests on ethically questionable ground.
2. CPR as the default treatment for all cardiac arrest patients is based on flawed reasoning
The authors repeat arguments that will sound familiar to many physician trained by attendings of a certain age: CPR is a treatment like any other; the physician has the role and responsibility of determining whether it is indicated before providing it. (So let’s see you refuse to provide it, I always wanted to reply to these rants, which never took place within earshot of a patient or her family.)
To shift this accepted paradigm for medical treatment to one in which CPR must be provided to any patient who has not explicitly rejected it compels physicians and care teams to provide a futile treatment; this itself is unethical.
The “full code” default, they argue, has been largely based on
The elevation of patient autonomy above other legitimate concerns;
The assumption that fallible physicians should not make value judgments (in this case, deciding a patient should not receive CPR).
But, the authors point out, patients are denied other treatments deemed futile, and physicians are authorized to make explicit value judgments in other contexts. These value judgments frequently incorporate the probability of success, as does the decision to perform CPR.
Of course, CPR is fundamentally different than other medical treatments, from both the psychological and consequentialist perspectives, and so it’s not surprising it has come to be handled differently.
3. Default CPR is unethical because most patients do not consent to it
A novel argument (to me) is that default CPR is ethically inappropriate because most patients do not consent to it (as they are unconscious).
Treatments on unconscious patients must clearly be in their best interests. For many patients undergoing CPR (such as for dying patients, the authors argue), this test is not met. The actual performance of unconsented CPR is therefore often unethical, and so is compelling clinicians to perform it.
To McLennan et al, the solution is straightforward:
“Ethically, CPR should be removed as the required default position for all patients regardless of their individual situation, and hospital policies should be modified to allow clinicians to consider whether CPR is appropriate at the time of arrest if there is no patient-initiated DNR order in place. The prolongation of life at all costs is not an ethically appropriate goal of medicine[.]”
There you go—no more slow codes!
Remember, these are bioethicists, not physicians practicing in the consumer- and corporate-driven U.S. healthcare system. (Two of the contributing authors are physicians.)
And more, for later …
Further arguments by the various authors include:
a formal normative analysis of slow codes, which concludes they are sometimes permissible, according to the logic of bioethicists;
the inherent deceptiveness of slow codes is not conclusive proof of their wrongness, because there are “other justifiable acts of deception in medicine”—we’ll explore this intriguing assertion later;
recruitment of the famous (and tired) “trolley problem” to justify slow codes (i.e., a reductionistic thought experiment that justifies a consequentialist approach of sacrificing one to save many);
that slow codes are a defensible form of civil disobedience in which physicians prevent harm to patients by defying rules and norms created or enforced by ethically illegitimate entities (i.e., legislators and hospital administrators).
Discussion
There is no question that physicians are frequently put into morally complex and distressing situations when providing CPR to patients at the end of life. Providing CPR to dying patients feels wrong, and often probably is.
This group of bioethicists have made some interesting and provocative arguments that half-hearted CPR is a possible solution in some cases.
However, slow codes are themselves problematic in many ways, a sample of which includes:
coercion of less-powerful team members to participate in an action they find professionally or morally repugnant, or legally risky;
placing life-and-death ad hoc decisional authority on CPR back in the hands of physicians, which was deemed unacceptable in the 1980s due to its inconsistency and vulnerability to biased or incorrect judgment calls;
patients’ inconvenient tendency to die incompletely or awkwardly slowly after a cardiac arrest, with agonal breathing, intermittent resumption of organized cardiac rhythms, etc. Once the “slow code” has begun, the team enters profoundly morally compromised territory in which they may feel forced to abandon a dying patient, constrained from treating any ongoing suffering—or worse, compelled to declare that still-living patient dead prematurely.
And much more.
In a brief email conversation, two of the lead authors expressed an openness to engage in conversation with the critical care community regarding their provocative assertions on the ethics of slow codes.
Stay tuned for that.
While I generally agree that North American physicians should have a freer hand to not provide non-indicated treatment (specifically full-bore CPR, my sense is that Canada and the U.S. both vest this decision with patients/families) from my perspective as an EM/CCM physician and medical educator, slow codes are worse. As leaders of a resuscitation team, we should bear the weight of the decision making that says "No - we're not doing this" and "we're stopping now."
1 - We practice how we play. Ideally, IHCA should be rare. Each time the resuscitation team gathers and performs ACLS, we should be doing it in a standardized way (allowing for variation in treatment based on most likely cause of the arrest). Every time we deviate from this practice we are showing that standardization doesn't matter. Like Yoda said, "Do, or do not. There is no try."
2 - Slow codes are basically lying to the family. We're trying to show them "hey, we did everything" when, in fact, we did not. We should have the courage of our convictions to say we did not because, from a physiologic perspective, it would not have changed anything.
Finally, I have found utility (and healthcare system support from risk management/legal) in focusing on the concept of "physiologic futility", at least for maximally supported ICU patients. If a patient is already intubated, on high dose pressors, and is going to arrest from a cause that is not immediately reversible (like, we've already verified there's no pneumothorax or cardiac tamponade, etc.), my system has supported documentation in that moment of 1) the patient was already being resuscitated with the support they were receiving, 2) the cause of the arrest would not be addressed in the timeframe afforded by ACLS, and 3) that CPR would be merely performative. This started during COVID but has persisted and has been helpful.
“So let’s see you refuse to provide it”… highlights another difference between countries, as we can do just that in Australia.