We systematically reviewed the literature for original, peer reviewed, articles on the ethical issues associated with uDCD.

MEDLINE via Ovid, Embase via Ovid, CINAHL via Ebsco, Scopus, PsycInfo via Ebsco, LexisNexis, WestLaw and Web of Science Core Collection were searched using the following MeSH terms or other subject terms, and synonyms. The full search strategy can be found in the Supplementary Appendix, but can be summarized as: (1) out of hospital cardiac arrest terms OR uncontrolled AND (2) donation terms (adj5) AND (3) ethical OR legal terms The search was designed by a health librarian (IK) in collaboration with authors AG and ZF. Searches were run from 1st January 1997 until 1st April 2022 and re-run 30th May 2023 and again Sept 19th 2024. See Figure 2 for the PRISMA flow diagram.

9,920 papers in total (7,605 in 2022, 1,035 in 2023 and 1,280 in 2024) were screened for inclusion/ exclusion, using the criteria shown in Table 1. Three authors, AG, MI and ZF, performed screening. Each paper was blindly screened by two of the three screening authors using Rayyan. The inclusion and exclusion criteria were applied to the title and abstract, and where the outcome was ambiguous, the full publication was read in full by all three members of the screening team and discussed until conflict was resolved.

We included two kinds of papers:

1. Papers which presented original peer reviewed ethical analysis (where ethical analysis is defined as: identifying questions regarding what is the “right” thing to do (or what we ought not to do); critically and reflectively examining [48, 49] different viewpoints; and presenting a reasoned argument, ideally with a normative conclusion of issues relating to uDCD) [50, 51].

Systematic reviews were excluded from the review, although these papers were read as a further way of identifying relevant papers. Reviews, commentaries and opinion pieces were excluded. Papers which reported protocols, outcomes or processes of donation without ethical analysis were excluded. The nature of the included papers can be seen in Table 2.

All papers included on the basis of title and abstract were read in full by authors AG, MI and ZF. The inclusion/ exclusion criteria were applied blindly by each member. A full team meeting was held to agree final includes.

The search was re-run in May 2023 using the same method as above, and again in October 2024 following peer review. A fourth member of the research team, author WT, read all papers included from both searches in full to ensure consistency of application of inclusion and exclusion criteria. Any conflicts were discussed and resolved as a team.

In alignment with the PRISMA guidelines for scoping reviews, an assessment of whether quality assessment would be appropriate was undertaken [52]. Given the wide range and nature of articles identified, a formal quality assessment of included studies was not performed [53].

The final included papers were read in full by authors AG and ZF and data was extracted.

A excel spreadsheet was created for extraction of data relating to both publication characteristics (title, date, author(s), country, article type, participants, limitations and content. Framework analysis was undertaken [54]: an initial coding framework was created based on themes identified in the background literature (see Supplementary Appendix for data extraction proforma).

Having familiarized themselves with the data in each paper, and iterated the initial coding themes further, authors AG and ZF independently coded each paper.

Further themes were added upon data extraction and discussion among the authors, who together charted, mapped (see Figure 1) and interpreted the data.

Some themes were grouped together for ease of understanding, with sub-themes being created. For example, the themes of “consent” and “‘relatives” were brought together under one heading, “consent and involvement of next of kin,” with two sub-themes, “informed consent to the use of organ preserving techniques” and “the approach to discussion of organ donation with next of kin” having emerged. In another example, “optimisation and termination of resuscitation” emerged as an important theme.

The final coding framework comprised seven broad ethical themes, along with subthemes. The frequency with which each ethical theme was raised in the literature was documented.

33 papers were included. Table 2 shows publication breakdown by country of origin, region, publication type and date.

7 broad ethical themes, along with subthemes, were identified. Table 3 shows the proportion of papers addressing each issue. We explore each of these themes below and summarize the results in Table 4, highlighting areas of broad consensus and areas of ongoing ethical tension.

An overarching ethical challenge that was identified throughout the literature was the perceived tension between maximizing the chances of successful resuscitation for the patient who has arrested and maintaining organ viability if resuscitative attempts were to fail.

Once death has been declared, and a hands-off period observed, organ preserving techniques (OPTs) are instigated. Broad ethical issues include risk of resuming brain circulation [8, 63, 65, 69, 73] and retroactively negating declaration of death [8, 68, 70], violation of bodily integrity [8, 58, 62, 72, 74], resource cost [8, 72] and stress for the treating physicians [8].

Violation of bodily integrity was the most cited with a common theme: there was consensus that the least invasive methods possible should be used to preserve the opportunity to donate [59, 62, 74–76]. Bruce et al’s study of emergency department (ED) patients and relatives found that a majority felt that insertion of groin tubes, CPR and ventilation were acceptable as OPTs, if there was as little invasion of the body as possible [75] and Goudet et al’s survey found that majority of respondents did not consider cannulation as a “bodily integrity alteration” [58]. Volk et al also found support for OPTs; 80% of their participants expressed support for a rapid organ recovery where they live [57].

Many authors commented on the ethical issues raised by specific techniques. Use of NRP (ECMO to perfuse the organs only) is defended by some [65, 72] as an essential means of preserving donation opportunity with good outcomes for kidneys [65] but specifically opposed by others [8, 63, 68, 69]. Use of an aortic occlusion balloon was described by some as a responsible method of facilitating organ preservation whilst preventing perfusion of the brain [60, 65, 67, 77] while others were concerned that it might render a physician complicit in a patient’s death [8, 63, 68, 70]. Use of cold preservation solution was opposed by two authors for reasons of poorer outcomes [65] and interference with determination of death [8] but supported by others as it carries less risk of brain reanimation [68]. Dubois et al in their survey of 70 members of the public found that 72% expressed support of a law permitting organ cooling in order to preserve organs [78].

Given the time pressures of instigating OPTs, several authors explored the issues with consent, and attempted to determine the optimal timing, place, and content for conversations with relatives of the deceased which are both respectful of autonomy and compassionate. Volk et al found that hypothetical family consent to donation in uDCD settings was high [69% (95% CI 65%–73%)] when compared with cDCD [70% (95% CI 66%–75%)] and DBD [66% (95% CI 62–71)], however that participants were less confident in making donation decisions about a relative when compared with themselves (71% and 75% respectively) [57].

In uDCD protocols the patient will lack capacity and decisions must be made in their best interest [84, 85]. Best interests decisions must be made on a case-by-case basis [80, 81], but often little is known about the individual’s desires in the emergency situation: best interests decisions are therefore based in the first instance on population level knowledge.

If the concept of best interests is taken in its narrow, medical sense, uDCD becomes ethically challenging because OPTs are invasive and done beyond the point at which there will be medical benefit to the patient. Potential harms to the patient are physical [72, 73, 80], and non-physical e.g., treating the patient as a means to an end [73], violation of a deep desire to not donate [73], distress to the family [80], and impact on the dignity of a person [62]. Moorlock et al and De Lora et al questioned whether any harm can be inflicted on a nearly dead patient [80, 81] but concluded that there is a duty to treat cadavers with respect [80], and that respect for dead persons and posthumous wishes is an established ethical concept [81].

Several authors suggested that best interests should be interpreted more broadly than considering physical integrity [72]. Arguments included that best interests are now accepted as extending beyond the strictly clinical [80] and include fulfillment of wishes to donate [73, 80], promotion of dignity for example by “favoring the accomplishment of their life project” to donate [62] and permitting altruism in end of life planning [73]. OPT may preserve the family’s opportunity to make their own informed decision about donation [80] and preserve the autonomy of those patients who turn out to have expressed wishes to donate [72]. Without OPTs, the opportunity to donate is lost [72, 74, 78].

Several papers considered the ethical duty to consider responsibilities to society as well as to the individual.

Public and clinician trust in organ donation and in the wider medical system is imperative and links to our societal responsibilities [64]. If people do not trust the system they are less likely to donate and support transplantation as a whole [69].

Finally, two of the included papers raised the issue of how recipients of organs procured via uDCD should be consented for a transplant [8, 76] and how much information they should be given on the source of the organ. Data suggests that the long-term outcomes of organs transplanted in uDCD protocols are as good as those transplanted in DBD protocols [62] and in cDCD protocols [22, 59] however there is evidence to suggest that there is a higher rate of shorter term complication, i.e., delayed graft function in uDCD [8]. Transparency and informed consent are essential, particularly in areas where uDCD is in development.

In attempting CPR, the primary intention is to regain spontaneous circulation and neurological recovery for the patient [87]; a secondary effect is optimal perfusion of the organs for transplantation should CPR be stopped. In transferring a patient to hospital, the primary intention is to ensure a comprehensive assessment of the irreversibility of the condition; a secondary effect is ensuring efficient organ recovery should CPR be stopped. Therefore, by optimizing CPR and transferring an arrest patient to hospital, the treating team is both optimizing patient outcomes and organ viability; they are not choosing one over the other [88].

Further, once CPR is stopped, 5 min of hands-off time followed by 5 min of continuous monitoring while cannulating results in a total of 10 min before the aortic occlusion balloon is inflated and NRP started. While case studies of autoresuscitation have been reported after termination after CPR, most are associated with confounders and a recent systematic reviews showed that it is extremely rare for them to take place after 10 min; none have occurred in the situation being proposed, with continuous monitoring for 5 min [89–91].

Some caveats do remain. First, we found consensus that decisions about TOR and entry into a uDCD protocol should be made by different teams; research into the logistics and outcomes of this is needed. Second, we found disagreement over whether the level of resuscitation should extend beyond advanced life support to ECMO/E-CPR; more research is needed into the benefits of these techniques in the out-of-hospital cardiac arrest setting and the impact that having uDCD without ECMO/ E-CPR may have on outcomes and on community trust.

Authors of the included papers disagreed whether best interests in the uDCD context extend beyond the strictly clinical and beyond the individual’s lifetime. UK law supports a broad reading of best interests, and in the cDCD debate it has been argued that “where a patient would wish to donate, measures [that] are necessary for organ donation to proceed … serve, rather than deny, the best interests of a patient” [92] and are therefore autonomy respecting. The difficulty is that within an opt out system (and without Moorlock and Draper’s ambitious proposal of mandated anticipatory consent) [81], the specific wishes of most individuals are not known.

Although best interest decisions are, by definition, person specific, they are often initially made on population level knowledge. For example, a person found in cardiac arrest will be subject to CPR while further information on their wishes is sought [93]. This logic can reasonably be applied in the donation setting given that a majority of the population – with the information currently available to them – would like to donate [2]; while information is being sought about a person’s wishes, it may be in their best interests to cannulate and start NRP to preserve opportunity for donation.

There has been significant discussion about the role of the aortic occlusion balloon in All forms of DCD [8, 60, 63, 65, 67, 68, 70, 77, 94, 95]; this discussion is intimately associated with the definition of death [47, 96], and the philosophical debate around the ethical relationship between acts and omissions [97].

The device is required because a secondary effect of starting NRP is to resupply blood and oxygen to the brain to the same level of the attempted CPR. There is no evidence that this level of recirculation is likely to facilitate awareness or pain, but it is impossible to say for certain that there is no perception. Therefore, to avoid an unintended harm, the aortic occlusion balloon is inserted to prevent all circulation to the brain and maintain a peaceful death.

A recent prospective study by Royo-Villanova et al showed that when the thoracic aorta was blocked with an aortic occlusion balloon the mean intracranial arterial blood pressure at the circle of Willis was the same during circulatory arrest as it was following NRP being started, confirming that this technique works to stop brain perfusion [98]; this study should provide reassurances to those to those who were concerned about the efficacy of the aortic occlusion balloon.

Some have expressed concern that insertion of an aortic occlusion balloon in order to block circulation to the brain is itself an act which hastens death [8, 59, 63, 68]. In the cases we are considering, however, the patient has already died; their heart has stopped, there has been no responsiveness with CPR, and a multidisciplinary team has recognized the futility of further efforts. We agree with Schiff et al who say: “this is similar to ex vivo perfusion, in which perfusion is restored to the recovered organ to increase transplant viability, while the process towards loss of brain function in the donor body is allowed to continue.” [47] On this view, the aortic occlusion balloon is acting to minimize harm, while maximizing the individuals’ potential to donate.

The above conclusions - namely that the interests of both resuscitation and donation can be simultaneously respected; that best interests apply posthumously and can be read broadly; and that an aortic occlusion balloon is in a patient’s best interests – are contingent on transparency and public engagement [8, 47, 56, 58, 59, 63, 64, 69, 72, 80, 83, 99]. If population level data is to be used to inform initial presumptions about what is in a patient’s best interests, public attitudes must be regularly surveyed and assumptions cannot be made [8, 56, 62, 80, 83].

There is some nervousness surrounding public discussion of the details of uDCD. While public attitudes toward donation are predominantly positive, there is an awareness that one bad media story can change views and potentially cost lives if it results in people opting out [21, 22]. The risks are increased when - as needs to be done - relatives are being asked to consent not only for transplant but for research into a new way of undertaking transplant. This nervousness is justified given the stakes, but it is a reason for ensuring that information about transplantation is understandable and widely available; hiding information is much more likely to erode trust in the long term. We should borrow from the World Health Organization’s advice on transparency in public health emergencies: information must be “factually accurate, easily understood by the intended audience and presented in a manner that promotes adoption of the desired behaviors” and we must “promote trust by being forthcoming and open …, including the evidence and assumptions used by authorities in making decisions, the manner in which those decisions are being made and by whom.” [100]

Finally, if public engagement and trust are to be sought, an alternative name to “Uncontrolled Donation after Cardiac Death” should be considered. The name derives from differentiating it from the controlled setting of an intensive care unit with planned withdrawal of treatment, but to those who don’t know this history, the term “uncontrolled” implies chaos and lack of regulation. As O’Rourke et al state, “who would wish to be involved in an “uncontrolled process”?” [101] A name that clearly describes the practice could be considered: Organ Donation After Sudden Irreversible Cardiac Arrest (ODASICA), or some other clearly descriptive explanation, may go some way towards engaging the public.