Attitudes of Minority Patients with End-Stage Renal Disease Regarding ABO-Incompatible List-Paired Exchanges


*Corresponding author: L. F. Ross,


A few transplant centers in the United States have implemented list-paired exchange programs that include both ABO-compatible and ABO-incompatible living-donor recipients. ABO-incompatible list-paired exchanges raise ethical concerns because they increase the total number of organs available but increase the waiting time for wait-list candidates of blood type O. In this manuscript, we explore attitudes of a convenience sample of minority patients with end-stage renal disease (ESRD) regarding living paired exchanges, ABO-compatible and ABO-incompatible list-paired exchanges and ABO-incompatible direct transplants. Data from 87 minority respondents were analyzed. Eighty-seven (100%) supported living paired exchanges and ABO-compatible list-paired exchanges. In contrast, only 50 of 85 (59%) respondents supported ABO-incompatible list-paired exchanges (p < 0.001), including half (12 of 24) of those with blood type O. Subjects were asked how much additional time it would be fair to ask wait-list candidates of blood type O to wait to implement ABO-incompatible list-paired exchanges. Forty percent (35 of 87) responded ‘no additional time’ and another 10% (9 of 87) responded ‘one month or shorter’. Minority dialysis patients hold mixed opinions about the fairness of ABO-incompatible list-paired exchanges. If our findings are confirmed in a more diverse randomly selected sample, then the UNOS variances that permit these exchanges should be reconsidered.


Living donor kidney paired exchanges involve two donor-recipient pairs in which Donor A cannot donate directly to Recipient A and Donor B cannot donate directly to Recipient B, but Donor A can donate to Recipient B and Donor B can donate to Recipient A. Despite heightened international interest in performing living donor kidney paired exchanges after the publication of a research protocol by Ross and colleagues in 1997 (1), only a few hundred have been performed worldwide (2). The major obstacle is that many individuals in end-stage renal disease (ESRD) are of blood type O and can only receive an organ from a donor of blood type O, whereas blood type O donors are ‘universal donors’ and will be able to donate directly with an intended recipient of any blood type unless there is a positive crossmatch.

Numerous variations on living donor kidney paired exchanges have been proposed (3,4). One such variation is a list-paired exchange in which Donor A, who cannot donate directly to recipient A, donates a kidney to a random individual on the deceased donor wait list and recipient A in return receives high priority for the next available ABO-compatible kidney. Ross and colleagues have argued that one must distinguish between ABO-compatible and ABO-incompatible list-paired exchanges because the ethical consequences of these two variations differ dramatically (3,4). If the members of a living donor-recipient pair have the same blood type but a positive crossmatch prevents a direct donation, then the exchange can be understood as simply replacing a deceased donor kidney of a particular blood type with a living kidney of the same blood-type. ABO-compatible list-paired exchanges are morally justified on the grounds that one can assume that wait-list candidates would prefer a living donor kidney to a deceased donor kidney because of the likelihood of improved graft function and survival. However, in ABO-incompatible list-paired exchanges, the living donors tend to donate a kidney of blood type A, B or AB (non-O) to the wait list and their intended recipient removes a kidney of blood type O. Non-O blood type wait-list candidates may be advantaged (i.e. shorter wait time) and wait-list candidates of blood type O may be disadvantaged (i.e. longer wait time) (4,5). Since wait-list candidates of blood type O already have a longer than average wait time, these exchanges will disadvantage further those who are already worse off.

In 2001, Zenios et al. modeled a solution that could avoid the O-disadvantage from list-paired exchanges (5). It required transplant teams to preferentially select the emotionally related donors of blood type O rather than to select the donor who appears most willing. Zenios et al. acknowledged that this solution raised ethical issues—specifically, it raised serious concerns about coercion (5). The proposal was not adopted by the transplant community, and in 2004, Ross and Zenios argued that it would be more ethically acceptable to restrict list-paired exchanges to ABO-compatible list-paired exchanges (6). Avoiding the ABO-incompatible list-paired exchanges would lead to a smaller overall increase of organs available for transplantation, but it would eliminate the problem of harming those on the deceased donor wait list with blood type O (5) and the problem of coercing potential paired exchange donors of blood type O (6).

In 2004, Delmonico and colleagues described the list-paired exchange protocol that was developed in New England for which United Network for Organ Sharing (UNOS) authorized a variance (7). Their data confirmed the outcomes anticipated by Ross and colleagues in 2000 (4). Of the first 17 list-paired exchanges performed, all the recipients except one (who received two kidneys) were of blood type O, and all but one of their paired donors were non-O blood types (7). Recently, the regional consortium in Washington D.C. published their data (8). Their first 10 list-paired living donors had non-O blood types; the recipients' blood types were not reported (8).

In numerous writings, Veatch has argued in favor of an egalitarian theory of justice which treats all human beings as equally valuable (9,10). In transplantation, egalitarian justice theories place moral priority on how the organs are distributed (justice as fairness) in contrast with an utilitarian theory that seeks to maximize the number of organs transplanted. The most widely-accepted egalitarian theory was developed by Rawls (11). Rawlsian justice would permit policies that increase organ transplants if the policies are not harmful to those who are already worst off (11). As such, Rawlsian justice would not permit ABO-incompatible list-paired exchanges.

Veatch, however, argues that inequalities that do not benefit the worst off can be ethical if the least well-off consent to waive the requirements of Rawlsian justice (9). If wait-list candidates of blood type O support ABO-incompatible list-paired exchanges despite the fact that they would have to wait longer, then such exchanges could be ethical. We hypothesized that individuals on dialysis of blood type O would not support ABO-incompatible list-paired exchanges because they would be harmed, but that a significant majority of individuals of other blood types would be supportive as they stand to benefit from the policy's implementation.

Alternatively, Steinberg has suggested that transplant candidates who do not know their blood types can be considered to be behind a Rawlsian ‘veil of ignorance’ (12), and it would be useful to know whether they would accept these exchanges because they do not know whether they would benefit or be harmed by such a policy (12). Although we examined the perspective of this group as well, we believe that transplant candidates of blood group O are the appropriate group to study because they are the ones who will be harmed, and it is their consent that is necessary to waive the requirements that Rawlsian justice demands.

In this manuscript, we present the attitudes of minority individuals with ESRD on dialysis regarding the various transplant options.


A convenience sample of adults between 18 and 80 years with ESRD at three dialysis centers affiliated with the University of Chicago were approached after consultation with the head nurse on the unit. Excluded from the study were dialysis patients who did not speak English or those who had neurocognitive deficits that would not allow them to understand the intricacies of the exchange protocols. In total, 100 potential subjects were approached.

The survey consisted of 7 questions based on 4 transplant scenarios and 10 demographic questions (see Supplementary Material available for download as part of the full-text version of the article from The survey was piloted twice on non-transplant health care professionals and previous transplant recipients at the University of Chicago. To ensure respondent understanding and minimize interviewer bias, one individual administered all of the surveys (PDA).

Two separate surveys were distributed at the dialysis centers depending on whether the subject was of non-O or unknown blood types or of blood type O. The researcher (PDA) asked the subjects to read each question and then he verbally explained the scenario and the possible survey responses. The first two scenarios described living paired kidney exchanges and list-paired exchanges involving an ABO-compatible living donor-recipient pair. The subjects were asked if they approved of these exchanges and how they thought the first two programs would affect them as a transplant candidate on the wait list. The third scenario explained that most list-paired exchanges would involve ABO-incompatible living donor recipient pair (most commonly with the paired recipient being of blood type O and the living donor having a non-O blood type). Regarding ABO-incompatible list-paired exchanges, subjects were asked if they supported this type of exchange. Subjects who were of non-O or unknown blood types were then asked: ‘How much longer is it fair to ask wait-list candidates of blood type O to wait in order to increase the number of kidneys for wait-list candidates of other blood types?’ By contrast, blood type O individuals were asked, ‘How much longer would you be willing to wait in order to increase the number of kidneys for wait-list candidates of other blood types?’ Respondents were given six options: up to 1 week; up to 1 month, up to 6 months, up to 1 year; up to 2 years or more; or no additional time. The last scenario explained that ABO-incompatible direct donations are possible and could obviate the need for most exchanges. Subjects were asked which options should be offered to donor-recipient pairs that are ABO-incompatible and were unable to participate in a living donor exchange. They were given four options: (i) list-paired exchanges only; (ii) ABO-incompatible direct donation only; (iii) both or (iv) neither. The researcher (PDA) spent approximately 45 min with each respondent explaining the transplant options and documenting their responses.

Medical directors and head nurses of the dialysis units associated with the University of Chicago reviewed the surveys and agreed to allow the investigator to interview dialysis patients. University of Chicago IRB approved the study and waived the need for written informed consent. Oral consent was procured. Data were analyzed by chi-square 2 × 2 and 2 × 3 tables with p < 0.05. Cross-tabulations were done to examine whether blood type, knowledge of blood type, age, education, gender or whether the individual is currently on the wait list had any impact on the subject's attitude about these transplant options.


One hundred individuals were approached. Ninety agreed to be interviewed and completed surveys. Our analysis is limited to the 87 respondents who self-identified as non-Caucasian.

Of the 87 subjects eligible for analysis, one subject identified himself as Hispanic and another identified herself as black-Hispanic. The rest self-identified as African American or black. Demographics are given in Table 1. It is noteworthy that 36 individuals (41%) did not know their blood type. Knowledge of blood type did not correlate with gender, age or education.

Table 1.  Subject population demographics

Number (%)
of subjects
(n = 87)
  1. *Subject identified herself as a black-Hispanic.

Blood type
 A9 (10%)
 B15 (17%)
 AB3 (4%)
 O24 (28%)
 Do not know36 (41%)
 Male55 (63%)
 Female32 (37%)
 African American85 (98%)
 Hispanic1 (1%)
 Mixed*1 (1%)
 Under 5038 (44%)
 50 and over49 (56%)
 High school diploma or less53 (61%)
 Some schooling beyond high school34 (39%)
On deceased donor waiting list?
 Yes50 (57%)
 No27 (31%)
 No, but trying to get on10 (12%)
How long on waiting list?
 Less than 1 year22 (25%)
 More than 1 year28 (32%)
 Not on waiting list37 (43%)
Any potential living donors?
 Yes43 (49%)
 Yes, but refused the offer10 (12%)
 No33 (38%)
 Do not know1 (1%)

Just over 60 percent of the respondents (53 patients) indicated that they had at least one potential living donor. Of those 53 patients, 10 also mentioned that they had refused the offer for various reasons. Only 16 (18 %) of the subjects failed to complete high school while 34 (39%) of the subjects had some post-high school education. Thirty-eight (44%) patients were under 50 years old.

All 87 survey respondents expressed their support for the living donor paired exchange. They understood that this would not affect their priority on the wait list, but would prevent some individuals from having to be placed on the wait list. All 87 survey respondents also expressed their support for an ABO-compatible list-paired exchange. They understood that this would not affect their priority on the wait list, but that it meant a wait-list candidate would be offered a kidney from a living donor rather than from a deceased donor. By contrast, respondents were divided about their support for ABO-incompatible list-paired exchanges (see Table 2). Overall, 50 of 85 (59%) respondents supported ABO-incompatible list-paired exchanges which is significantly less support than they express for ABO-compatible list-paired exchanges (p < 0.001). Overall, half of respondents who knew their blood type, whether O or non-O, supported these programs. Although not statistically different, 24 of 34 (71%) of those who did not know their blood type supported the exchange. Neither gender, education, nor whether the subject was actively listed for a transplant predicted support for these exchanges.

Table 2.  Support for ABO-incompatible list-paired kidney exchange
Blood type O
(n = 24)
N (%)

Blood types A,
B, AB (n = 27)
N (%)
blood type*
(n = 34)
N (%)

(n = 85)
N (%)
  1. *Two subjects (both of whom did not know their blood type) answered ‘not sure’ and were excluded.

Yes12 (50)14 (52)24 (71)50 (59)
No12 (50)13 (48)10 (29)35 (41)

Subjects were asked a follow-up question regarding how much additional time blood type O transplant candidates could ethically be asked to wait in order to increase the overall number of kidney transplants (see Table 3). All of those who objected to ABO-incompatible list-paired exchange asserted that the transplant community should not ask blood type O candidates to wait any additional time. Of the patients who supported the exchange, they were almost evenly split between those who thought that 6 months should be the maximum acceptable increase in waiting time and those who were willing to permit an additional wait of a year or longer.

Table 3.  How much time should wait-list candidates of blood type O be asked to wait?

Blood type
Additional time N (%)#
No additional

1 week

1 month
Up to
6 months
Up to
1 year
2 years
  1. *Three subjects (one of whom was blood type O and two of whom did not know their blood type) answered ‘not sure’ and were excluded.

  2. #Total adds up to greater than 100% due to rounding errors.

O (n = 23*)12 (52)0 (0)2 (9)3 (13)4 (17)2 (9)
A, B and AB (n = 27)13 (48)0 (0)1 (4)5 (19)3 (11)5 (19)
Uncertain (n = 34*)10 (29)3 (9)3 (9)6 (18)6 (18)6 (18)
Total (n = 84)35 (42)3 (4)6 (7)14 (17)13 (15)13 (15)

Respondents were then informed that ABO-incompatible direct donation, while not routinely done at many transplant centers, is a clinically valid option. They were informed that it entails pre-transplant desensitization and a more intensive post-transplant immunosuppression regimen that increases the incidence of complications to the recipient, and that the risk of graft failure may be increased. They were asked which of the following treatment alternatives physicians should present to ABO-incompatible donor-recipient pairs: only the list-paired exchange; only the direct donation; both or neither (see Table 4). In this scenario, responses differed significantly depending on the candidate's blood type. Only one of the 24 blood type O patients (4%) responded that transplant centers should only present the list-paired exchange to incompatible donor-recipient pairs compared to those of uncertain blood type (41%) and those of non-O blood types (45%) (p < 0.01). Rather, blood type O respondents preferred providing incompatible donor-recipient pairs with both options (46%) more often than did individuals of uncertain blood type (29%) and those of non-O blood types (11%) (p < 0.05).

Table 4.  Which transplant options should be offered to patients with an ABO-incompatible living donor?
 Blood type O
(n = 24)
N (%)
Blood types A,
B, AB (n = 27)
N (%)
Uncertain blood
Type* (n = 34)
N (%)
(n = 85)
N (%)
  1. *Two subjects (both of whom did not know their blood type) answered ‘not sure’ and were excluded.

ABO-incompatible list-paired only1 (4)12 (45)14 (41)27 (32)
ABO-incompatible direct donation only8 (33)6 (22)7 (21)21 (25)
Both11 (46)3 (11)10 (29)24 (28)
Neither4 (17)6 (22)3 (9)13 (15)


Individuals with ESRD requiring dialysis generally support methods to increase the supply of organs. All of our respondents favored living donor paired exchanges and ABO-compatible list-paired exchanges.

Given that non-O blood type patients would all stand to benefit from the implementation of the ABO-incompatible list-paired exchanges, we hypothesized that these individuals would support ABO-incompatible exchanges. By contrast, we hypothesized that blood type O transplant candidates, whose wait times would increase as a result of adopting the protocol, would overwhelmingly disapprove of these exchanges and resist their implementation. Moreover, we expected the majority of those patients uncertain of their blood type to express serious reservations about the implementation of the exchange because both the survey form and the interviewer explained that the majority of people are of blood type O. That is, we expected those behind the veil of ignorance to hold beliefs similar to individuals of blood type O.

Our hypotheses were refuted by our data. We found that respondents were divided in their willingness to make sacrifices or to ask others to make sacrifices. However, for those who want to argue for proceeding with ABO-incompatible list-paired exchanges, our data must give pause. Half of those who would be harmed by the implementation of such a system did not support the exchange and argued that they should not be asked to wait any additional time. Many of those of non-O blood types also rejected these exchanges because they did not believe it would be fair to ask their blood type O counterparts to wait any additional time despite the fact that they would benefit from its implementation.

While many transplant surgeons and nephrologists offer the utilitarian argument that the benefits derived by transplanting additional patients justifies ABO-incompatible list-paired exchanges (12), Rawlsian justice, even with Veatch's modification (9), would hold that these exchanges can only be justified if those of blood type O are willing to be made worse off. Ideally, one would require universal support to justify harming those who are ‘worst off’. Since only half are willing to wait any additional time, ABO-incompatible list-paired exchanges cannot be ethically justified on the basis of respondents' professed willingness to waive their right to be harmed (i.e. to wait additional time on the deceased donor wait list).

Those who did not know their blood type and were behind the Rawlsian veil of ignorance appeared to be more accepting of list-paired exchanges, although the difference was not statistically significant. This may be due to our small sample size and needs to be investigated further. Despite their increased support, however, over one-quarter did not support these exchanges.

Even those respondents who supported ABO-incompatible list-paired exchanges expressed reservations concerning justice issues. Approximately half of those who supported ABO-incompatible exchanges (23 of 49) stated that transplant candidates of blood type O should not wait more than 6 additional months. Again this 50% split was true for all respondents regardless of blood type or knowledge of blood type.

Respondents were interested in the possibility of an ABO-incompatible direct exchange. Forty-five (53%) stated that this option should be offered to individuals with ESRD who have an ABO-incompatible donor, half of whom believed it should be the only option offered. Despite good success rates reported in the literature (13–17), relatively few US transplant centers perform ABO-incompatible direct donations (13,15). Given the ethical controversy, it may be that U.S. transplant centers should pursue both this option and the option of living donor paired exchanges with wider sharing pools (2) rather than ABO-incompatible list-paired exchanges. While there may be some conditions under which certain individual ABO-incompatible list-paired exchanges could fulfill the criteria of Rawlsian justice, increasing the waiting time of O recipients to benefit those with shorter waiting times is not ethically justifiable.

There are several limitations to our study. First, our data represent the attitudes of a convenience sample of minorities from the Southside of Chicago. It is not known whether our data are generalizable to other ethnic communities, geographic regions, rural or suburban communities or those of a different socioeconomic or religious backgrounds. Second, in our surveys, names were assigned to the donors and recipients to facilitate discussion about the impact of the donation on each of the parties. This was a frequent suggestion in our first pilot study. In the diagram for the ABO-incompatible list-paired exchange (scenario 3), there is a typographical error and Mr. Harris is referred to as a ‘she’. Since Mr. Harris does not have a paired recipient, this should not have had serious impact, and in fact, the typographical error was not noted until after survey administration was complete. Third, the scenarios in the survey involved complex concepts. To promote understanding and to minimize bias, one interviewer was responsible for the administration of all the surveys. One reason to believe that the respondents understood the survey was that all were able to articulate that living paired exchanges and ABO-compatible list-paired exchanges would not affect their status on the wait list. The fact that the data refute our hypotheses suggests that his presentation was not overly colored by our expectations. Still, one cannot be certain that respondents would have given the same responses to another interviewer.


A convenience sample of minority dialysis patients expressed mixed opinions regarding the fairness of ABO-incompatible list-paired exchanges. Although these exchanges would increase the number of organs available overall, they do so by making wait-list candidates of blood type O worse off. Without widespread support of the ESRD community, particularly those of blood type O, ABO-incompatible list-paired exchanges that utilize kidneys from deceased blood type O donors in exchange for a living kidney of another blood type cannot be morally justified unless algorithms are developed that eliminate additional disadvantages to those who are already disadvantaged in terms of waiting times.

Our pilot data found only half of the ESRD candidates of blood type O supported ABO-incompatible list-paired exchanges. Our data need to be confirmed or overturned by studying transplant candidates from other ethnic communities, geographic regions and other socioeconomic and religious backgrounds. If our data are generalizable, UNOS should reconsider its authorization of variances that allow ABO-incompatible list-paired exchanges to be performed.


We would like to thank Stefanos Zenios, Ph.D., Carol Stocking Ph.D. and Michele Josephson M.D. for their comments on the survey. We would also like to thank the health-care professionals and transplant recipients who helped us pilot our survey tool twice, the dialysis center physicians and nurses who allowed us access to their units to conduct the survey and the individuals in end-stage renal disease who agreed to participate as research subjects.