Effect of Collaborative Requesting on DCD Refusal Rates: Randomized Controlled Trial

In 2020, the DBD and DCD refusal rates in France were 33.1% and 37.7%, respectively, the corresponding absolute numbers being 972 DBD (among 2941 recorded potential donors) and 160 DCD (among 425 recorded potential donors). These refusal rates are far greater than the 25% target set by the French national organ and tissue donation agency (Agence de la Biomédecine, ABM) for 2017-2021. Moreover, refusal was the leading reason for not recovering organs from recorded potential donors (1). High refusal rates decrease the number of recovered organs and transplantation procedures in the overall donor population, thereby worsening the organ shortage. The ABM national plan for organ and tissue donation suggests several measures to increase organ donation rates. Among these measures, evaluation by the professionals involved of the reasons for absence of organ recovery aim to improve practices, while tighter collaboration between hospital OPC teams and the ICU and emergency-room physicians involved with organ recovery seek to decrease refusal by families (1).

The French decree issued on 16 August 2016 sets forth good-practice rules for discussing possible organ and tissue donation with the family and specifies the conditions under which acceptance or refusal of donation should be recorded (2). The conversation should be led at least by the ICU physician in charge of the patient and an OPC; whenever possible, a non-physician ICU staff member providing care to the patient should also be present. This recommendation is based in part on cohort studies of factors associated with refusing or accepting organ donation (3-6). Although some of these factors are not modifiable, such as patient ethnicity and age, cause of death, and beliefs, others are, such as family understanding of brain death, timing of the organ-donation request, and training of the requesting healthcare professionals. In a British cohort study that included 1741 potential donors with brain death, the consent rate was substantially higher when the family was approached by a transplant nurse (6). In Norway, after the introduction of a transplant nurse, the consent rate increased from 20% to 70% (7). Thus, specific training and the accumulation of experience in communicating about donation may considerably increase the likelihood of consent. However, in a before-after study done in Australia, neither the consent rates before vs. after the introduction of an OPC team nor consent rates for this team vs. the medical team during the intervention period differed significantly (8). These discrepancies are probably largely ascribable to the small sample sizes and many sources of bias due to the absence of control groups and of randomization (9). The only randomized controlled trial published to date compared donation requesting by the medical team only vs. the medical team and an OPC jointly (collaborative approach), in 201 families of patients with brain death (10). The consent rates were 61.4% and 57% for the medical team only and the collaborative approach, respectively, and no significant difference was found after adjustment on age, sex, and ethnicity of the potential donors. Possible explanations to this result put forward by the study investigators include use of the collaborative approach in only 73% of families assigned to this intervention; a larger proportion in the medical-team-only group of potential donors with a known positive view of donation; and the absence of standardization of the collaborative approach, with the two teams adjusting to individual circumstances and the OPC adopting a passive attitude in some situations. Other weaknesses of this trial are the small sample size and exclusion of centers that used a collaborative approach before the trial.

European and French recommendations for the management of patients with brain death include a conversation with the family about WLST led only by the medical team, separately from the donation-request conversation, the goal being to limit the risk of conflict of interest for the medical team (11-13). Nonetheless, in many cases the medical team touches on donation, or even suggests donation to the family, immediately after the WLST conversation, as part of the same meeting and in the absence of an OPC. This practice has been described as flowing more naturally and logically. In the above-mentioned British cohort study, of 2962 families of patients with brain death, 824 were approached about donation by the medical team only and 909 by an OPC only (6).

The objective of this study was to evaluate whether, compared to an approach by the medical team only, without an OPC, a collaborative approach by the medical team and an OPC together was associated with a lower rate of refusal of DCD.

The reasons for limiting the study to DCD (as opposed to all types of potential donors) are listed below.

1. Confining the trial to DCD produces a uniform population. DBD requests can be made either before the death of the patient (e.g., in the emergency room when death is predicted due to a severe stroke unamenable to treatment) (14) or after the patient has died, in the emergency room or after admission to an ICU or neuro-vascular unit. These heterogeneities in the timing of the request (before or after the death of the patient), location of the patient when the request is made, and specialist training of the healthcare professionals in charge of the patient (intensivist, emergency physician, or neurologist) may produce numerous biases. With DCD, the death always occurs in the ICU, the professionals involved are all trained in critical care (intensivist, anesthesiologist-intensivist, and ICU nursing staff), and the donation request is always made after the family has been informed of the WLST decision made by the ICU team. Only potential donors registered with the hospital OPC team according to the procedure required by the Agence de la Biomédecine will be eligible for trial inclusion.
2. The refusal rate is higher for DCD than for DBD (6,15,16). In the UK, over an 18-month period, consent rates for DBD and DCD were 68.9% and 56.5% (P<0.0001), respectively. In France, this difference was first noted in 2019, when consent rates were 69.5% and 64.1% for DBD and DCD, respectively. Similarly, in 2020, corresponding consent rates were 67.0% and 62.4%, respectively.
3. Collaborative requesting may have a larger effect in DCD than in DBD. In a cohort of 2962 potential donors after circulatory death, the odds ratio for consent in the absence of an OPC was 0.32 (95% confidence interval, 0.26-0.42; P<0.0001) and the consent rate was higher when the OPC made the donation request (Hulme W Anaesthesia 2016).

The main hypothesis for the PRODON trial is a 25% refusal rate in the collaborative-request group. Since the late 1990s, the refusal rate has remained above 30% in France, compared to 15.3% in Spain. The method used to compute the rate differs between the two countries, with Spain determining the ratio of refusals over the number of requests and France the ratio of potential donors with recorded refusals over the total number of potential donors. Nonetheless, a 25% refusal rate in the collaborative-request group seems a reasonable target, particularly given the similar organization of hospital OPC teams in the two countries, as France took inspiration from the Spanish model to set these teams up in 1990-2000 (17). An important difference with Spain is that the OPC team is less often involved in requesting donation in France. In 2020, of 3200 DBD requests in France, only 65% involved an OPC, potentially explaining the stagnation of the refusal rate at 30% (nationwide Cristal Action data collected by the Agence de la Biomédecine).

PRODON is a randomized, controlled, open-label, two-parallel-arm trial comparing a control group (medical-team-only approach to determine whether the relatives refuse DCD) to an intervention group (collaborative approach by an OPC and the medical team jointly to determine whether the relatives refuse DCD). Individual as opposed to cluster randomization of potential donors will be used given the small annual number of potential DCDs in each ICU. The likelihood that the same healthcare professionals (ICU staff and OPC) will be involved in the same potential DCDs in a same ICU is low, and the risk of contamination bias is therefore small.