The ethics of organ donation after cardiocirculatory death: Do the guidelines of the Canadian Council for Donation and Transplantation measure up?

The practice of transplantation the world over is governed by the dead donor rule: non-paired vital organs can be retrieved only from patients who are dead. It is therefore important to have clear criteria for the determination of death. Most transplantable organs come from patients who are declared dead by neurologic criteria. These patients are called heart-beating donors; they have suffered a catastrophic brain injury, have been ventilated, and have had their vital functions maintained mechanically up to the point at which death is declared and then beyond, until their organs are retrieved. The donor pool can be expanded by permitting an alternative form of donation known as donation after cardiocirculatory death (DCD), or non-heart-beating organ transplantation. The donors in this case are patients for whom there is no hope of recovery but who are not dying because of a brain injury and hence will not suffer the neurologic death necessary for them to become heart-beating donors.

DCD has been an accepted medical practice internationally for 15 years1 but came to Canada only in 2006, when the Canadian Council for Donation and Transplantation (CCDT) published its national recommendations.2 Countries that permit DCD have had variable success in increasing the supply of transplantable organs. DCD of kidneys, the organ with which the CCDT recommends that Canadian centres begin their DCD programs,2 serves as a good example. In 2003, kidneys donated after cardiocirculatory death accounted for only 4% of the total kidneys transplanted in both Spain, which has the highest per capita rate of transplantation of any organs, and the United States, which has the highest total number of transplants. By contrast, in the Netherlands, which was then the world leader in DCD, kidneys donated after cardiocirculatory death accounted for 39% of the country’s kidney transplants.3* The effect of introducing DCD into Canada is uncertain, but Doig and Rocker estimate that DCD “could contribute to an increase in the number of solid organs for transplantation including 20% or more to the supply of kidneys, and increasing the supply of other solid organs including liver, pancreas or pancreatic islet cells.”4

The CCDT recommends that, for the purposes of DCD, death be diagnosed after 5 minutes of continuously observed absence of pulse, blood pressure and respiration after life support has been discontinued.2 This recommendation of a 5-minute interval coincides with that of the Institute of Medicine5 and is commonly accepted in North America and the United Kingdom. However, it falls short of the time frame of “at least 10 minutes of proven lack of circulation to the brain” estimated for the determination of neurologic death in a report by the US President’s Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research.6 It is also not a universally accepted interval for DCD, and lies between alternatives that include a 75-second interval advocated by Boucek and colleagues,7 a 2-minute interval used in Pittsburgh and at some other US and UK centres2 and a 10-minute interval used at still other US and UK centres and at St. Michael’s Hospital in Toronto.8

In its guidelines, the CCDT aimed to “promote patient-care-based principles for providing the option of donation within a sound ethical framework and provide guidance to individual programs in developing parameters for safe practice in this field.”2 We can measure the success of the CCDT in achieving this aim by considering two of the questions most often asked by families thinking about DCD: “Is my loved one really dead?” and “Will he or she feel any pain?”9 It is reasonable to suppose that any acceptable DCD program will either enable a health care professional to answer those questions with an unequivocal “Yes” in the first instance and “No” in the second or, if not, to give families the information they need to make an informed decision about donation. In this article, I will argue that a program that follows the recommendations of the CCDT will do neither of these things and hence will not satisfy ordinary consent requirements.

Consent and death

The CCDT guidelines assume that once patients or their families consent to organ retrieval at death and understand that death will be declared by cardiocirculatory criteria (Box 1), they have consented to organ retrieval at death as determined by those criteria. There would be no problem with this assumption if the condition that patients are in when they are diagnosed as dead by cardiocirculatory criteria were recognized as death by ordinary persons, for then there would be a clear match between what patients or families believed they were consenting to and what they were actually being asked to agree to. However, the condition that patients are in when they are diagnosed as dead by cardiocirculatory criteria does not fit any of the common understandings of the word “death.” Some people understand the word to refer to a permanent physical state of a patient, characterized by irreversible coma, the absence of spontaneous respiration and heartbeat, the impossibility of spontaneous recovery of these functions (auto-resuscitation) and the impossibility of restoration of these functions by others through artificial resuscitation.† Others understand death to occur as soon as there is no longer any possibility of auto-resuscitation, even if artificial resuscitation could conceivably restore respiration and heartbeat.‡ Still others understand death to require the permanent absence of spontaneous respiration and heartbeat and that these functions are not artificially supported.10§Death determined by the CCDT’s criteria for cardiocirculatory death, however, is not consistent with any of these understandings.

Box 1

Criteria used for death in organ donation

The possibility of auto-resuscitation after a 5-minute interval of continuously observed absence of pulse, blood pressure and respiration after life support has been discontinued has never been definitively ruled out.11 In addition, we do not know precisely when successful artificial resuscitation is no longer possible. The Institute of Medicine claims that “existing empirical data cannot confirm or disprove a specific interval at which the cessation of cardiopulmonary function becomes irreversible.”12 The CCDT comments in its guidelines that it was unable to identify in its literature review any evidence that either auto-resuscitation can occur or artificial resuscitation can succeed after the 5-minute interval.2 However, given the limited number of studies that have been published on these issues, this alleged absence of evidence does not show that those things are impossible and hence does not contradict the Institute of Medicine’s view. Indeed, nothing in the CCDT’s guidelines document contradicts this view.2 It is also not clear that evidence is in fact absent, as there are reports of the so-called Lazarus phenomenon, wherein spontaneous recovery is claimed to have occurred well after the 5-minute interval.13 Furthermore, even if future empirical investigations were to determine that there is no possibility of auto-resuscitation or successful artificial resuscitation after a 5-minute interval, this does not change the fact that we do not know this now.

Once we see that the ordinary sense of death differs from the CCDT’s understanding of this state, we can also see that most patients or families who have consented to having death determined by cardiocirculatory criteria have not realized that this consent entails death being declared when the CCDT wants to declare it. The natural way for patients or families to understand an invitation to have death declared by cardiocirculatory criteria is to assume that they are being asked to consent to a special way of determining when death in the ordinary sense occurs, that is, to determine the presence of an irreversible state by cardiocirculatory instead of neurologic means. However, this is not the way the CCDT understands consent to have death determined by cardiocirculatory criteria. It takes it, rather, as consent to have death declared as soon as the cardiocirculatory criteria for its determination are satisfied, whatever state the patient may be in at that time, which in this case is to have death declared when the patient is not known to be in an irreversible state. It is plain, however, that unless patients or families are told that this is the consequence of their consent, the CCDT’s assumption about what their consent means is not correct: in other words, consent by patients or families to death being declared by cardiocirculatory criteria does not mean that they have consented to death being declared as soon as the cardiocirculatory criteria for its determination are satisfied.¶

Accordingly, if the CCDT is to provide guidelines for DCD that satisfy consent requirements, it must do one of two things. It must either recommend that the interval be increased from 5 minutes to at least 20 minutes14** so that death determined by cardiocirculatory criteria coincides with death in the ordinary sense of the term†† (it is sometimes argued that this would not greatly reduce the number of usable organs),15 or it must recommend that health care professionals disclose that death determined by cardiocirculatory criteria differs from death as ordinarily understood and explain the difference. Undertaking the latter option would certainly complicate the recruitment of organ donors and may diminish the number of donors. We cannot be sure that fewer donated organs would be available for transplantation, as the public may respond positively to a fully transparent DCD program. Nonetheless, regardless of its effect on the number of donated organs, full transparency must be pursued because organ retrieval requires consent, and consent requires that patients and their families truly understand what they are consenting to. One might object that the information that health care professionals would have to give to patients or their families to fully explain death by cardiocirculatory criteria would be overly complicated. However, physicians provide complicated information to patients and families at times of grief in other situations where consent is required (e.g., when there is a question of terminating life-sustaining treatment), and it is not clear why seeking consent to DCD should proceed on different principles.‡‡

Distress

The second way in which the CCDT’s recommendations fall short of meeting consent requirements relates to their silence concerning the possibility of pain and suffering. There is a significant possibility of distress to the patient during DCD. Controlled DCD (which occurs after the planned withdrawal of life-sustaining therapy) may involve interventions such as vessel cannulation before life-sustaining therapy is withdrawn and death is declared and it may involve post-mortem interventions such as in situ preservation.2 Uncontrolled DCD (which occurs after unanticipated cardiac arrest) may additionally involve chest compressions and mechanical ventilation both before and after consent for DCD is obtained2 and typically requires the withdrawal of life-sustaining treatment. We know that all of these interventions cause distress to conscious patients who are not taking palliative medications. Because patients who are candidates for DCD are not neurologically dead either before or shortly after they are declared dead by cardiocirculatory criteria, the possibility that they may experience distress cannot be ruled out in either the preparation for DCD or during organ retrieval.

Given this possibility of pain and suffering, there are three approaches that can be taken: (1) provide palliative medications when there are physical signs compatible with distress; (2) withhold all such medications on the ground that even if signs of distress are occurring, the patient does not have sufficient cognition to interpret any sensations as noxious; or (3) provide palliative medications prophylactically to prevent any possible distress.16 The CCDT does not recommend any one of these approaches in particular. Instead, it proposes that the management of the dying process, including procedures to withdraw life-sustaining treatment, sedation, analgesia and comfort care, should proceed according to the existing practices of individual intensive care units.2 This would be appropriate if the 3 approaches were equally acceptable.§§ That, however, is not so.

The distress protocol for DCD that is perhaps the most common is based on the first approach, in which medication is provided contingent on signs compatible with distress. The Pittsburgh protocol, for example, stipulates that “if narcotics and sedatives are administered, these drugs must be titrated to the patient’s need for provision of comfort. The administration of clinically appropriate medications in appropriate doses to prevent discomfort is acceptable, with titration of medication predicated on signs compatible with distress.”17 The problem with such protocols is that they do not guarantee that organ donors will not experience distress. To have a distress protocol at all is to grant that DCD may involve distress to the patient, and to medicate only on signs suggestive of distress is to expose patients to the possibility of experiencing it. However, if there is a possibility of distress (however slight or transitory) this must be disclosed, for it is surely information that anyone would want to have before agreeing to DCD. It is fair to assume that if families perceived that their family member could suffer during organ harvesting, they would be less likely to consent to the procedure.

The second approach is no more acceptable than the first. It is highly speculative to claim that patients declared dead by the CCDT’s cardiocirculatory criteria are in such a debilitated condition that they would not be able to experience distress. It may be true that when such patients exhibit signs compatible with distress they feel nothing objectionable. However, no one can know this, and many health care professionals—all those who subscribe to protocols based on the first approach, for instance—do not think it is true. Thus, centres that adopt the first or second approach will be faced with the invidious alternatives of either disclosing the possibility of distress (and thereby deterring donation) or not disclosing that possibility (and thereby violating consent requirements).

If the first or second approach is used to address the possibility of pain and suffering, consent problems will still exist even if there is full disclosure at the time that consent is sought. Patients themselves can choose to run the risk of distress. However, in most cases consent for DCD will be sought from families, and it is not clear that a family can consent on behalf of a loved one to a procedure that might cause their loved one distress. It is not enough that the family thinks that running the risk is acceptable: they must have some reason to think that their loved one would be willing to run the risk. The decision about whether to grant consent would be straightforward for the family if their loved one had directly communicated their willingness to be exposed to distress to them. However, in the absence of that surely rare event it is not at all straightforward, because then the family must infer the loved one’s willingness from something else and it is not clear what that could be. They cannot infer it from the loved one’s desire to be an organ donor, for organ donation is usually believed to be a procedure that does not involve any distress and they cannot infer the desire to undergo a procedure that carries the risk of distress from the desire to undergo one that is not believed to carry such a risk. It is also not easy to identify anything else that would enable the family to confidently say that their loved one would want to undertake that risk.

The conclusion that must be reached is that any protocol that does not entirely eliminate the possibility of distress will raise seemingly insurmountable ethical problems whether or not that possibility is fully disclosed. This points us toward the third approach (the prophylactic provision of palliative medications to prevent pain) as the protocol of choice.

Concluding remarks

From an ethical viewpoint, the CCDT’s recommendations for DCD are highly problematic with regard to defining death and obtaining consent. Centres planning to set up DCD programs that will satisfy ethical requirements will thus have to modify the CCDT’s recommendations. Specifically, given the foregoing discussion, they must first ensure that patients or families interested in DCD have no misapprehensions about when organs will be retrieved. This can be achieved either by disclosing that death determined by the CCDT’s cardiocirculatory criteria differs from death in the ordinary sense and explaining the difference, or by altering those criteria so that the two definitions of death are equivalent. Second, they must adopt a policy of providing palliative medications prophylactically to ensure that organ retrieval does not carry any risk whatsoever of distress. Only once these things are done will there be a DCD program that makes patients and families full partners in the enterprise to close the gap between the supply and demand of transplantable organs.