Relationship Between Patients' Perceptions of Disadvantage and Discrimination and Listing for Kidney Transplantation

Abstract

Objectives. This study explored wait-listing decisions among African American and White men and women eligible for kidney transplants, focusing on lifetime experiences of race and sex discrimination as a possible influence.

Methods. Patient records from 3 Baltimore-area hemodialysis units were reviewed, and semistructured face-to-face interviews were conducted with transplant-eligible patients and with unit staff members.

Results. African American patients reported more racial discrimination, and women reported more sex discrimination. Women and older patients were less likely to be placed on the waiting list, as were patients with previous experiences of racial discrimination. Discrimination measures predicted list access more strongly than patient race.

Conclusions. Lifetime experience of and response to discrimination may contribute to race and sex differences in access to care and should be included in research on health care disparities.

There is a growing body of research in the United States examining cultural and social disparities in medical care and health. Sex and ethnic disparities in diagnosis and treatment, in particular, have been extensively documented across a broad spectrum of clinical conditions, including cardiac and cerebrovascular disease, cancer, HIV/AIDS, and such common conditions as asthma and pneumonia.1,2 To date, much of this work is descriptive, documenting who is disadvantaged under what circumstances.3–5

Although identification of disparities is a necessary first step, it leaves unanswered the question of how such disparities occur and neither contributes new theory to advance research nor suggests interventions to reduce gaps. Furthermore, while there is ample evidence that physician bias affects clinical decisionmaking, less detailed attention has been paid to factors related to patients' choices.

In this article, we report the results of a study examining access to the cadaveric transplant waiting list among patients with end-stage renal disease (ESRD). We used semistructured surveys to investigate how patients make the decision to enter the waiting list while receiving center-based hemodialysis. The goal was to identify psychosocial factors contributing to nationally observed sociodemographic differences in waiting-list access: women, older patients, and African American ESRD patients are placed on the waiting list at lower rates than younger White men.6–14

Patients, families, and health care providers must consider both medical and social needs in choosing from the variety of treatment modalities available to individuals with ESRD. Successful transplantation is an improvement over center-based hemodialysis in terms of quality of life, long-term survival, and health care costs.15–17 Transplantation failure and return to dialysis, however, have negative effects on patients and waste scarce resources. Therefore, transplantation involves risk for the medical system, the family, and the patient. Despite trends toward liberalizing criteria for patient selection,18 many patients may prefer dialysis.

Conversely, access among older, lowincome, or minority patients may be biased as a result of compliance criteria, social support, or ability to return to employment.19 Little is known about decisionmaking on the part of patients with ESRD or the impact of the decisionmaking process on long-term success and satisfaction with treatment modality.

In 1997, among US ESRD patients aged 20 to 64 years, African Americans made up 46% of the center-based hemodialysis population but only 18% of those living with functioning transplants.18 This difference in the prevalence of successful transplantation is caused by multiple factors: lower rates of first transplant, higher rates of graft failure, and lower rates of retransplantation.6

Previous research 6–14 provides evidence that female, African American, and older patients are significantly less likely to be listed for or to receive a kidney transplant after factors such as income, referring unit size, geography, and underlying health have been controlled. Most studies have involved large national databases, which, although comprehensive in terms of number of patients included, contain only limited psychosocial data.

Recent surveys20–23 add information on patients' beliefs and preferences regarding treatment. These surveys suggest that when attitudes are considered, race and sex patterns change, indicating the importance of patient attitudes as a source of listing differences.

To date, researchers have inadequately addressed the issue of comorbidity as a barrier to transplantation. Analyses often include all patients in hemodialysis units, regardless of comorbidity, although many would not meet medical criteria at any transplant center. As a means of controlling for comorbidity, patients with selected illnesses are grouped together for analyses, despite differences in disease severity.21 Important contraindications to transplantation, such as HIV/AIDS, mental illness, and substance use, are often not measured.

Much of the existing research has examined cohorts of patients within set time periods after ESRD onset, thus measuring incidence of first transplant listing. A complementary question would be the following: Within the dialysis population, what is the prevalence of listing among patients with different characteristics? Such a measure would provide information about the entire population of ESRD patients: long-term dialysis patients who may not have been considered for transplantation under previous, more restrictive criteria; patients who return to dialysis after transplant failure; and patients who enter the waiting list later after ESRD onset. However, prevalence studies undersample patients who move out of dialysis populations rapidly as a result of either death or transplant.24

It is difficult, but ultimately most relevant, to study access to transplantation among all patients who meet current transplant program criteria. Our study was designed to examine the decisionmaking process among a prevalence sample of transplant-eligible hemodialysis patients within a single but representative geographic area, thus controlling for geographic variations in access to transplant programs. We studied units with significant African American ESRD populations and chose complementary quantitative and qualitative survey methods to fully explore sensitive issues such as disadvantage.

METHODS

Study Design

During 1996 and 1997, we recruited all transplant-eligible African American and White patients aged 21 to 70 years from 3 Maryland hemodialysis centers. A transplant nephrologist (David K. Klassen) reviewed medical records for all unit patients. Following nationally accepted guidelines for transplantation candidates,25 patients were excluded who had contraindicating diseases such as HIV, metastatic cancer, or osteomyelitis; severe psychiatric illnesses or psychosocial barriers to compliance such as homelessness, drug or alcohol addiction, or cognitive limitations; and comorbid conditions introducing unacceptable surgical risk, including severe cardiac, vascular, or pulmonary diseases. Rating surgical risk required the most extensive evaluation of medical history, including hospital and cardiologist reports to supplement the dialysis unit medical record.

Of 297 patients receiving hemodialysis in the 3 units, 152 (51%) were eligible for our study. One hundred forty-five patients were ineligible owing to age (35%), race (2%), or medical ineligibility for transplantation (63%). Of the 91 medically ineligible patients, 51% had comorbid contraindications to surgery, 23% had infectious or immunologic diseases, and 26% had psychosocial barriers to transplantation.

The first 3 authors interviewed patients, nurses, technicians, social workers, and physicians in the units with survey instruments that included both closed- and open-ended questions. After chart reviews identified studyeligible patients, patients were recruited for the face-to-face, audiotaped 1.5- to 2-hour interview. Of 152 initially eligible patients, 3 died and 17 changed modality or treatment setting before the interview. Of the remaining 132 patients, 114 completed the interview (an 86% response rate). Nonrespondents did not differ from respondents in regard to race, sex, or waiting-list status. We conducted 34 staff and 3 physician interviews to inform our interpretation of patient data. All respondents were offered $15 in appreciation of their time.

Measures

Patient race in medical records was confirmed by means of direct patient questioning. Patients' listing status was verified by monitoring orders for the monthly antibody screening required by all transplant programs. Patients' preferred modality was measured by asking “If you could choose between staying on dialysis or getting a transplant right away, which would you choose?”

We used measures of lifetime occurrence of sex and racial discrimination developed by Krieger.26 For example, patients were asked “Have you ever experienced discrimination, been prevented from doing something, been hassled or made to feel inferior in any of the following situations because of your race or color/because you are a woman (or a man): at school, getting a job, at work, getting housing, getting medical care, from the police or in the courts?” We created a 4-category variable indicating whether respondents had experienced both race and sex discrimination, only one of these 2 forms of discrimination, or neither form.

We also included measures of respondents' coping strategies. During interviews, patients were asked “If you feel you have been treated unfairly, how do you usually respond? Do you accept it as a fact of life, or do you try to do something about it?”

Four-point Likert scale items were developed by means of which patients could indicate whether they strongly agreed, somewhat agreed, somewhat disagreed, or strongly disagreed with 3 statements regarding transplantation. The first item addressed perceived costs and benefits of transplantation: “If a person gets a kidney and keeps it for only a year, they would have been better off if they had not gotten a transplant, because of all the suffering they went through.” The second item measured inequality in the waiting-list process: “If a person has lots of money or knows some important people, he or she could get a kidney more quickly on the waiting list than someone who is just an average person.” The final item measured acceptance of the status quo in terms of transplantation access: “The system for giving kidneys to the people who are on the waiting list is probably as fair as possible.”

Quantitative and Qualitative Data Analyses

We used SPSS statistical software in conducting analyses.27 Analysis of variance and χ² statistics, respectively, were used to test for significant differences in continuous and categorical bivariate relationships. We used multiple logistic regression analyses to predict the likelihood of being listed and the likelihood of wanting a transplant according to various patient characteristics.

Model building included testing all variables from bivariate analyses in preliminary multivariate models. The final models presented here included only those variables marginally significant at the P < .10 level. Similarly, we tested all levels of categorical variables (e.g., experience with discrimination) but combined categories that did not differ significantly to create the most parsimonious final model.28

The texts of all open-ended responses were transcribed verbatim from audiotapes. We include selected examples from these open-ended responses to inform interpretation of our primarily quantitative findings.29

RESULTS

Dialysis Unit and Patient Population Characteristics

In 1998, 7.1% of patients in study units left these units to receive transplants; the overall Maryland rate was 5.7% (M. Turner, Mid Atlantic Renal Coalition; written communication; June 2000). Maryland's 1997 standardized transplantation ratio, an ageadjusted measure of access to first transplant, was 1.09.18 Nationally, state rates ranged from 0.52 to 2.18, in comparison with a national (reference) rate of 1.0.

Study participants and transplant-eligible patients in the study units were younger and slightly more likely to be African American than the general unit population. Both study and unit populations were predominantly African American (71% and 68%, respectively), while White women represented the smallest group, both in the unit populations and in the transplant-eligible group.

The ratio of patients on the waiting list to list-eligible patients in the study population closely reflected the listed/eligible ratio in the units overall. Of 297 patients, 90 (30%) were on the waiting list; 159 of these 297 patients (54%) were found to be transplantation eligible. Thus, 90 of 159 (57%) eligible patients were listed. Among study participants, 61% of transplant-eligible patients were listed.

Table 1 ▶ reports social characteristics of the study participants by race and sex, as well as by waiting-list status and desire for transplant. There was not complete agreement between patients' wishes and their list status. Twenty patients who desired a transplant were not on the waiting list, and 4 patients who were on the list reported that they would prefer to remain on dialysis. Of the entire study population, 21% did not want the treatment modality they were currently receiving or in the process of obtaining.

TABLE 1—

Characteristics of Study Participants, by Race, Sex, and List Status: Maryland, 1996–1997

	African American		White			Not Listed
	Men (n = 40)	Women (n = 41)	Men (n = 24)	Women (n = 9)	Listed (n = 70)	Wants Transplant (n = 20)	Wants Hemodialysis (n = 24)
Mean age, y	51	53	47	58*	49	54	57**
Currently married, %	48	29	37	22	31	50	42
Working full time, %	15	15	17	11	17	15	8
Household income ≤ $30 000, %	70	83	57	89*	75	75	73
Less than high school education, %	37	29	42	44	46	25	17**
Average time with ESRD, y	6.9	6.4	4.3	5.0	5.9	4.8	7.4
Male, %	…	…	…	…	67	50	34**
White, %	…	…	…	…	30	35	21
Other modalities tried, %
Peritoneal dialysis	20	12	13	22	21	5	8
Home hemodialysis	15	7	4	11	11	5	8
Previous transplant	25	17	17	22	20	25	13
Cause of ESRD, %
Diabetes	6	25	25	33	14	30	27
Hypertension	60	30	29	33	41	50	23
Other	34	45	46	33	45	20	50
Mean self-rated healtha	2.5	2.5	2.5	2.5	2.6	2.4	2.3
Attitudes toward transplant,b mean
1 year with transplant not worth it	2.4	2.3	2.0	2.7	2.0	2.5	3.0***
Rich, important go more quickly	2.9	2.8	2.6	3.4	2.8	2.7	3.1
System as fair as possible	3.2	3.1	2.9	2.6	3.0	2.9	3.4
Experience with racial discrimination, %
Any setting	63	59	21	22***	47	35	68*
Obtaining medical care	10	15	4	11	9	0	8*
Sex discrimination, %
Any setting	33	37	21	44	4	5	33**
Obtaining medical care	10	7	0	11	13	5	8
Response to unfairness
Accepts as fact of life	20	15	17	56**	14	35	25
Tries to do something	80	85	83	44	86	70	75

Note. ESRD = end stage renal disease.

^a1 (poor) to 4 (excellent).

^b1 (strongly disagree) to 4 (strongly agree).

*P < .10; **P < .05; ***P < .01 (ANOVA and χ² statistics for comparisons between race/sex groups and among list status groups).

Bivariate comparisons also showed substantial differences between groups. Women and African Americans in this population appeared to be disadvantaged relative to White men in several areas that could potentially influence listing. White men represented the youngest group and the group least likely to have annual household incomes of $30 000 or less. Listed patients were younger and more highly educated than nonlisted patients. There were no statistically significant bivariate differences in disease cause, time since onset, modalities tried, or self-assessed physical health status.

Attitudes toward transplantation varied according to list status. The more strongly patients felt that losing a transplant would not be worth the “suffering they went through,” the less likely they were to be waiting for a cadaveric transplant. Patients' comments explaining their disagreement included greater perceived benefits to costs (“No, even a day off this machine would be worth it to me”) and acceptance of uncertainty in transplant (“There's no way to tell how long you'll have it; you have to be able to accept that and take your chances”).

Across all categories, African Americans were more likely than Whites to report racial discrimination, and women were more likely than men to report sex discrimination. These trends were weaker in analyses examining discrimination in obtaining medical care; White and African American women and African American men were similar in regard to reported rates of race and sex discrimination. White women were most likely to report that they accepted these experiences as a fact of life.

Qualitative comments indicated that some African American patients believed that they were more vulnerable to discrimination when they sought care in unfamiliar settings. One older African American man reported difficulty visiting his relatives because treatment required a 3-hour drive from his family's home, and treatment slots were often not available for him. At one center, his wife, who drove him, was not allowed inside and waited in the car during his treatment. Although many patients reported difficulties with treatment while traveling, this particular individual perceived that his problems were caused in part by his race and lack of financial resources.

Some patients viewed racial relations as more positive in the dialysis units than in other settings, reporting that the shared experience of disease and uncertain health, as well as sitting side by side for many hours every week, allowed patients of diverse backgrounds to relate to each other as equals. One patient reported that ESRD had exposed him to a greater number of African American health care professionals, motivating him to take more responsibility for his health. However, one patient perceived discrimination on the part of unit staff members.

Overall, more patients made positive than negative comments regarding racial relations within their unit, although many described incidents of discrimination in other medical settings. It is possible that positive views of racial relations within the unit, coupled with overall experiences of discrimination, resulted in patients being more reluctant to seek care in new settings or modalities, such as transplant.

Multivariate Models

Table 2 ▶ reports the results of a multivariate model assessing significant independent predictors of a patient's being currently listed for cadaveric transplant. Age was highly negatively predictive of transplant listing, with a 6% decrease in likelihood of being listed for each 1-year increase in age. After adjustment for age, health, and attitudinal differences, women were more than 3 times less likely than men to be listed. Full-time employment was negatively associated with being listed, perhaps identifying patients who had successfully combined center-based dialysis and employment.

TABLE 2—

Odds of Waiting List Inclusion, by Significant Patient Characteristics: Maryland, 1996–1997 (n = 114)

	Odds Ratio	P
Age	0.94	.02
Sex
Male	1.00a
Female	0.31	.02
Employed full time	0.18	.04
Years with end stage renal disease	0.92	.08
Previous peritoneal dialysis treatment	6.49	.04
Self-assessed health	2.51	.007
Attitude: 1 year not worth it	0.47	.005
Self-reported experiences of discrimination
None, only sex, or both sex and race	1.00a
Only race	0.15	.003
Response to unfairness
Try to do something	1.00a
Accept as fact of life	0.34	.07

Note. Other variables that were tested and removed when nonsignificant included marital status, household income, education, race, previous transplant, and cause of end stage renal disease.

^aReference.

After adjustment for sex effects, there were no significant differences in list status according to race. African American and White men were both more likely to be listed than female patients; the sex effect was consistent across racial groups, and there was not a significant interaction effect between race and sex.

Three health-related measures were significant predictors of list status. First, the more years since a patient had developed ESRD, the less likely he or she was to be listed. Second, patients who had previously received peritoneal dialysis treatment were more likely to be listed. Third, better self-rated health was a positive predictor of list status. These 3 measures were stronger predictors of list status than cause of renal failure or previous transplant.

The differences in attitudes toward the trade-off between time with a transplanted kidney and the process of transplantation itself seen in the bivariate analyses were also significant in the multivariate model. Patients who agreed that a year with a transplant would not be worth the suffering were less likely to be listed. Of 3 possible discrimination experiences entered into the preliminary models (sex only, race only, and both sex and race), patients reporting only the experience of racial discrimination were significantly less likely to be listed.

Therefore, apparently it was not one's race per se that influenced access to transplantation, but rather the subjective disadvantage experienced as a result of one's race, most commonly reported by African American patients. In addition, overall, regardless of race or sex, patients whose usual strategy was to accept discrimination as a fact of life were slightly less likely to be listed than patients who attempted to change such situations.

To explore reasons for nonlisting, we created a multivariate model predicting, among those patients not currently listed, characteristics of those wishing to have a transplant (Table 3 ▶). Although this analysis was limited by the small numbers of individuals not listed for transplantation, several relationships emerged that were consistent with the findings shown in Table 2 ▶.

TABLE 3—

—Odds of Desire for Transplant, by Significant Patient Characteristics: Maryland, 1996–1997 (Among Patients Not on List, n = 44)

	Odds Ratio	P
Age, y
30 to 44	1.00a
45 to 69	0.10	.04
Self-reported experiences of discrimination
None	1.00a
Only sex, only race, both	0.16	.03
Attitude: rich, important go quickly
Disagree	1.00a
Agree	0.16	.03
Attitude: system as fair as possible
Disagree	1.00a
Agree	0.13	.04

Note. Other variables that were tested and removed when nonsignificant included sex, race, marital status, employment, income, education, years with end stage renal disease (ESRD), previous treatment modalities, cause of ESRD, and self-rated health.

^aReference.

Patients who were 45 years or older were 10 times less likely to desire a transplant than were younger patients. No other sociodemographic characteristic (e.g., sex, race, education, income, cause of ESRD) was predictive of whether the patient would prefer transplantation or remaining on dialysis.

Lifetime experience with any type of discrimination—sex, race, or both—was negatively related to desire for a kidney transplant. Two of the attitudinal measures were also predictive of whether or not a patient wished to have a transplant. Patients who believed that the system for allocating organs was “probably as fair as possible” and patients who believed that those who were well off financially or had connections received organs more rapidly were both less likely to want a transplant. We interpret these seemingly contradictory results as representing fatalism and a belief that a fairer system of organ allocation is not achievable. These views predicted transplantation wishes after adjustment for personal experience with discrimination.

DISCUSSION

Demographic Trends in Listing

In terms of general demographic trends, this study confirmed the sex differences previously observed in modality of ESRD treatment; women in our study population were less likely than men to be listed. Overall, the study women were slightly older than the men, but sex differences in listing remained in the multivariate model after adjustment for age and physical health status.

In contrast to the case with sex differences, however, the national patterns of racial differences in listing were not seen in this population. There are several possible explanations. The first is that transplant centers have become more aggressive in their pursuit of transplantation candidates so as to build their waiting-list populations and compete for the increasingly inadequate supply of organs. In a geographic area with many transplant programs and a large African American ESRD population, it is possible that African Americans' access to the waiting list, although not necessarily to transplant, has improved to match that of White patients.

A second consideration is that our study may have been able to more successfully eliminate patients who were medically ineligible for transplant. Most previous work considered all patients and adjusted only for broad groups of diseases that can, but do not always, rule out transplant. Almost half of our unit population was medically ineligible for transplant. Third, we did find that the experience of racial discrimination significantly reduced a patient's likelihood of listing. Perceived discrimination, as a variable, appears to have more fully captured the experience of racial disadvantage in access to transplantation than has patients' actual race.

How might perceived discrimination influence waiting-list status? One possible pathway is direct: it could be that our patients' answers reflected specific experiences of discrimination in referral to, or acceptance by, transplant programs. However, we found greater list-status differences when we examined discrimination experiences in all aspects of life than when we examined only medical care experiences. Therefore, we theorize that global discrimination has a more powerful cumulative influence on waiting-list access.

Repeated experiences of disadvantage leave victims with fewer social and economic resources across their entire lives, from education, housing, and employment to neighborhood safety and quality of community life. The accompanying psychological effect, in our opinion, is that victims of discrimination learn to anticipate poorer outcomes than advantaged members of society and thus become reluctant to enter into situations in which they expect to be treated unfairly. From our findings, we theorize that patients with greater exposure to perceived discrimination in their pasts do not want to risk new treatment situations, such as transplantation, because they have a lower expectation of successful outcomes and do not believe that there is anything they can do to improve their chances.

This finding is consistent with a growing body of literature identifying the physical26,30,31 and psychological32,33 health effects of lifetime experiences of discrimination. It also supports the argument that many of the health consequences of one's ethnic identity are socially rather than biologically driven. More work is needed to develop explanatory models for how these subjective assessments of both race- and sex-based discrimination influence health care choices and how the effects of such experiences can be recognized and addressed by health care providers. Although the federally funded ESRD program pays for and mandates access to transplantation for all ESRD patients, the reality is that this disease-specific model of “equal access” health care is shaped by the inherently unequal society in which it exists.

Mismatches in Desire for Transplantation and List Status

A second group of conclusions from these analyses focuses on patients who may not have been receiving the modality of their choice. There were 24 such patients, or 21% of the overall study population. How much of this “mismatching” is avoidable, and what can be done to reduce it?

The 6% of listed patients in the study units who had a desire to remain on dialysis are important for several reasons. On a larger level, if 6% of nationally listed patients would not accept an organ offer, this represents potential waste and inefficiency. Moreover, some patients are listed but are never psychologically ready for a transplant. One respondent had turned down 3 offered kidneys in the previous year, telling us that although he was willing to have a transplant someday, he was not going to have it unless he was sure everything would be just right.

Other patients, often those whose bodies have rejected previous transplants and become highly immunologically sensitized, wait for many years with little hope of finding another cadaveric match. This is especially prevalent among African Americans, whose lower graft survival rate often necessitates retransplantation. According to one older respondent who had been on the waiting list for 20 years after a graft rejection: “If they called me tomorrow, I wouldn't go after all this time. I think they've just forgotten about me.” Although his situation may represent an unfortunate medical reality, it has become a source of alienation. One potential strategy to reduce disadvantage would be to ensure that such patients are offered living donations through newer approaches to desensitization based on plasmapheresis.34,35

At present, there is no incentive to remove patients from the list on the part of transplant centers, dialysis providers, or even patients themselves. If a patient has been removed from the list, the waiting time accrued cannot be regained in the event that he or she needs transplantation rapidly. On a policy level, if there is interest in reducing the active waiting list to only those patients who would actually accept an organ if offered, then we should encourage centers to more frequently classify ambivalent patients as “inactive.”

In the case of most of the patients in our study, transplantation was the desired modality of treatment. However, of 86 patients wishing to have a transplant, 20 were not listed. Almost half of the medically eligible patients who were not listed desired a transplant. Why are these patients not on the transplantation waiting list? Some unlisted patients do not understand the process and believe that they are on the list, althought they have only discussed transplantation with medical staff. Other patients may want transplantation but are avoiding the evaluation or waiting-list experience.

Finally, examination of which patients in this population do not in fact desire transplantation indicates room for improvement in this area as well. Patients wishing to remain on dialysis reported feeling as they do in part because they are satisfied with dialysis but also in part because they are not attracted to transplantation. If this reluctance to consider transplantation stems from social inequities, those issues should be addressed. Apparently, patients who avoid transplantation are also more passive about racial disadvantage; thus, it is important that they have access to culturally competent providers with which they can discuss transplantation.36

Our research was exploratory and was limited by small sample sizes, especially in analyses of sex patterns. Larger samples from more diverse settings are needed to further explore the issues raised by our findings.

Much of the commentary on discrimination in health care suggests that providers should attempt to combat disadvantage by changing their current behaviors and practices.37–39 Our results suggest that these changes, while critically important, may not be sufficient. We must acknowledge that the cumulative effect of negative experiences across the life course becomes the prism through which current choices are viewed. Caregivers cannot undo a patient's past experiences, but they can be aware of the influence of those experiences on current decisions and well-being.