Racial Disparities in Health Beliefs and Advance Care Planning Among Patients Receiving Maintenance Dialysis

Abstract

Context.

Among people receiving maintenance dialysis, little is known about racial disparities in the occurrence of prognostic discussions, beliefs about future health, and completion of advance care planning (ACP) documents.

Objectives.

We examined whether Black patients receiving maintenance dialysis differ from White patients in prognostic discussions, beliefs about future health, and completion of ACP-related documents.

Methods.

We surveyed adult patients receiving maintenance dialysis from seven dialysis units in Cleveland, Ohio, and hospitalized patients at a tertiary care hospital in Cleveland. Of the 450 patients who were asked to participate in the study, 423 (94%) agreed. We restricted the current secondary analyses to include only Black (n=285) and White (n=114) patients. The survey assessed patients’ knowledge of their kidney disease, attitudes toward chronic kidney disease (CKD) treatment, preferences for end-of-life (EoL) care, the patient-reported occurrence of prognostic discussions, experiences with kidney therapy decision making, sentiments of dialysis regret, beliefs about health over the next 12 months, and advance care planning. We used stepwise logistic regression to determine if race was associated with the occurrence of prognostic discussions, beliefs about future health, and completion of an ACP-related document, while controlling for potential confounders.

Results.

We found no significant difference in the frequency of prognostic discussions between Black (11.9%) versus White patients (7%) (P=0.15). However, Black patients (19%) had lower odds of believing that their health would worsen over the next 12 months (OR 0.22, CI 0.12, 0.44) and reporting completion of any ACP-related document (OR 0.5, CI 0.32, 0.81) compared to White patients

Conclusion.

Racial differences exist in beliefs about future health and completion of ACP-related documents. Systemic efforts to investigate differences in health beliefs and address racial disparities in the completion of ACP-related documents are needed.

Introduction

Dialysis is often marked by high mortality and medicalized death.^1,2 Currently, 20–25% of persons receiving maintenance dialysis die each year¹ and many of these people’s end-of-life (EoL) course is marked by intensive treatments such as cardiopulmonary resuscitation and low rates of palliative care and hospice services.² Engagement in EoL planning will likely improve the EoL course for patients on dialysis,³ but effective EoL planning and discussions remain elusive. Currently, patient-provider prognostic and EoL discussions occur rarely,^4,5 and the resulting knowledge gaps can contribute to misunderstandings about the future illness trajectory and subsequent prognostic overestimation.^6,7 In sum, these factors may lead to increased use of potentially burdensome treatments.^6,8,7 In response to critical deficits in EoL planning, the National Academy of Medicine, has called for interventions to improve end-of-life (EoL) care for seriously ill people such as those who are receiving dialysis.⁹ Similarly, the Renal Physicians Association has urged nephrologists to engage in early prognostic discussions with patients and families.¹⁰ These discussions should encompass conversations about life expectancy and illness trajectory, and document EoL preferences by aiding patients in completing advance directives.¹⁰

Yet, as important—and valid—as these initiatives are, we need to consider who is most disproportionately impacted by end stage kidney disease (ESKD) and who has access to quality EoL care. For instance, Black patients experience ESKD more often than White patients;^11,12 and are more likely than White patients to receive potentially burdensome, goal discordant treatments near death.^13–16 These more intensive EoL utilization patterns among Black patients may be explained by the fact that Black patients continue to experience both interpersonal and structural racism.^17–20 Further, cross-racial patient-provider communication barriers and cultural differences are common.^21–25 Accordingly due to past and current health care injustices, Black patients may mistrust doctors, which can impact health care proxy decisions and end-of-life care preferences.^26–28 Improved EoL care is important for all people, but specifically within the Black community, the intersection of racial justice with patient-centered care is especially salient.

Given the increasing call for improved EoL care and communication for patients receiving maintenance hemodialysis and the vital association of prognostic understanding and ACP with EoL course,⁷ we conducted post-hoc analyses of a survey examining ESKD treatment preferences and EoL care goals and attitudes in a racially diverse cohort of hemodialysis patients.⁵ We hypothesized that Black patients receiving maintenance dialysis would be less likely than White patients to: (1) engage in prognostic discussions, (2) believe that their health will worsen over the next 12 months, and (3) complete advance directives. The study findings may help practitioners gain an understanding into racial differences in beliefs and behaviors that may influence patients’ end of life care.

Methods

Participants

In the previously published work, we designed a cross-sectional study exploring attitudes and beliefs about dialysis decision-making and EoL care among patients receiving hemodialysis.⁵ Using a convenience sample methodology, we surveyed adult patients receiving maintenance dialysis from seven dialysis units in Cleveland, Ohio, and hospitalized patients at a tertiary care hospital in Cleveland between November 1, 2014 and December 1, 2014.⁵ Patients unable to speak English and those lacking capacity, as assessed clinically (e.g., what would you do if you see a house on fire?) by a study team member, were excluded. Of the 450 patients invited, 423 (94%) agreed to participate in the study. The current secondary analyses derives data from our prior published work.⁵ Only Black and White patients were included in the current analyses (N=399) given the current analyses focused on racial disparities and the original study contained only a few (n=24) patients from ‘Other’ races. The Cleveland Clinic Institutional Board Review approved the study and all participants proceeding to complete the survey were deemed to verbally assent to participate.

Data Collection

Enrolled subjects completed a paper questionnaire assessing dialysis decision-making and EoL care. A study team member was available to clarify questions if they were not clear, and/or orally administer the survey to subjects with vision difficulties and/or low print literacy.

Measures

Adapted from the prior literature⁴ and as published in our previous work,⁵ we asked about prognostic discussion: “Has your doctor talked to you about how much time you will live?” (a) Yes (b) No.⁵ We asked about the quality of prognostic discussion: “How do you see your health in the next 12 months? (a) getting worse (b) worsening slightly (c) no change (d) improving (e) no response.”⁵ Last, we asked about completion of ACP-related documents: “Have you completed any of the following? (a) Living will (b) Personal directive/Medical Order for Life-Sustaining Treatment (MOLST) form (c) Healthcare proxy document (d) Enduring power of attorney(e) None of the above/Don’t know.”⁵ An affirmative response (“yes”) to any these items a - d was coded as completion of ACP, while a “None of the above or don’t know” response was coded as “No ACP”. In addition, the survey recorded (1) patient demographics (age, gender, marital status, race, education level) and religious affiliation (2) duration of dialysis dependency (3) attitudes toward dialysis decision-making and dialysis care, assessed by questions on the process of dialysis decision-making and importance of knowing detailed medical information (4) attitudes toward future medical care, and (5) engagement in advance care planning (Table 1).⁵

Table 1.

Demographic Characteristics of Respondents.

Participant Characteristics	N (%) Unless Otherwise Specified	Black Race	White Race	N Missing	Chi Square P Value
N of subjects	399	285 (71%)	114 (29%)	0
Age (years)				15
	Mean age (range)	101 (35.4)	45 (39.5)		0.45
>65 Years
≤ 65 Years (ref)	59 (21–91)	184 (64.6)	69 (60.5)
Gender				0
Female	175 (44)	135 (47.4)	40 (35.1)		0.03
Male (ref)	224 (56)	150 (52.6)	74 (64.9)
Marital status				1
Married	154 (39)	94 (33.0)	60 (52.6)		0.0003
Not married (ref)	244 (61)	191 (67.0)	54 (47.4)
Educational attainment				3
College level or higher	192 (48)	140 (49.1)	52 (45.6)		0.53
High school or less(ref)	204 (52)	145 (50.9)	62 (54.4)
Religious affiliation				12
Yes	366(95)	261 (91.6)	105 (92.1)		0.86
No (ref)	21 (5)	24 (8.4)	9 (7.9)
Mode of dialysis				16
Hemodialysis	366 (96)	260 (95.6)	106 (95.5)		0.85
Peritoneal dialysis	17 (4)	12 (4.4)	5 (4.5)
Dialysis duration of < 1 Year	Mean duration, in months (range)			18
	51 (0 –336)	226 (79.2)	77 (67.5)		0.01
Currently hospitalized	89 (22.3)	46 (16.1)	43 (37.7)	0	<0.0001

Statistical Analyses

We used SAS (Cary, NC version 9.4) to conduct statistical analyses. In addition to descriptive analyses, we carried out bivariate analyses to examine the differences between Black and White patients for the variables listed in Table 2. We also assessed for collinearity between the variables. We then used a dynamic, iterative stepwise regression process to fit a parsimonious multivariable adjusted model, after first selecting a smaller set of plausible variables based on the bi-variate results. We then conducted a multivariate analysis to ascertain if the Black-White differences attenuate after adjusting for demographics and clinical characteristics. The final stepwise regression model (N = 399) reports adjusted odd ratios [AOR along with 95% Confidence Intervals [AOR (95%CI)] for three main outcomes: (1) Occurrence of Prognostic discussion (2) Patient’s belief that health will worsen in the next year (3) ACP documentation. The P-value for significance was set at < 0.05. We conducted two sensitivity analyses. First, to account for missing values (n=19), we ran a fully efficient fractional imputation,²⁹ that utilized multiple non-missing values to calculate the most probable values for the following variables: age, marital status, educational attainment, religious affiliation, mode of dialysis, and dialysis duration. The imputation procedure achieved full convergence. Second, we ran an additional logistic regression model that included all variables with P <0.1 in bivariate analyses.

Table 2.

Bivariate Analyses of Patient Attitudes/Beliefs About Dialysis Decision-making and End-of-Life Care.

	Black N=285	White N=114	Chi Square p Value
Is there any ACP document present			0.01
Yes	118 (41.4%)	63 (55.3%)
No	167 (58.6%)	51 (44.7%)
Occurrence of conversation(s) about choices regarding EOL care in the last 12 months?	284 (99.6)	114 (100.0)	0.53
How do you see your health in the next 12 months?			<0.0001
Health will worsen/worsen slightly	19 (6.7%)	28 (24.6%)
No change/improving/no response*	266 (93.3%)	86 (75.4%)
Importance of QOL considerations about future care:			0.61
Important	241 (84.6)	94 (82.5)
Not important/unsure	44 (15.4)	20 (17.5)
Importance of being informed about prognosis:			0.70
Important	225 (78.9)	92 (80.7)
Not important/unsure	60 (21.1)	22 (19.3)
Importance of detailed info about medical condition:			0.11
Important	228 (80.0)	99 (86.8)
Not important/unsure	57 (20.0)	15 (13.2)
Importance of being prepared and planning in case of death:			0.86
Important	215 (75.4)	85 (74.6)
Not important/unsure	70 (24.6)	29 (25.4)
Importance of active family involvement in decision-making:			0.63
Important	219 (76.8)	85 (74.6)
Not important/unsure	66 (23.2)	29 (25.4)
Chose dialysis by your own choice	115 (40.4)	51 (44.7)	0.42
Regret the decision to start dialysis	62 (21.8)	14 (12.3)	0.03
Knowledgeable about palliative care	72 (27.7)	19 (16.7)	0.065
Knowledgeable about hospice	234 (82.1)	104 (91.2)	0.022
I have considered what might happen with your illness in the future?	190 (66.7)	80 (70.2)	0.50
Has your doctor talked to you about how much time you will live?			0.15
Yes	34 (11.9%)	8 (7.0%)
No	251 (88.1%)	106 3.0%)

‘No response’ option selected by only 3 patients.

Results

Table 1 shows patient demographics both descriptive and stratified by race. The study sample (n=399) consisted of 71% (n=285) Black and 29% White patients (n=114). The mean age of respondents was 59 years with 56% (n=224) men and 44% (n=175) women. Nearly 52% (n=204) reported an education level of high school or below. A majority (95%, n=366) reported that they were affiliated with a major religion and their mean duration of dialysis dependency was 51 months. The proportion of: women, unmarried persons, and people who had been on dialysis for <1 year was higher among Black patients versus White patients, while the proportion of patients who were hospitalized while the proportion of patients who were hospitalized was lower among Black individuals

Table 2 shows results of the bivariate analyses of patient attitudes and beliefs about dialysis decision-making and end-of-life care stratified by race and the Table 3 shows multivariate adjusted association of Black versus White race with: (1) the occurrence of prognostic discussion (2) patient’s belief that health will worsen in the next year (3) ACP-related documentation. In the bivariate analyses, notably, a higher proportion of Black patients regretted their decision to initiate dialysis (21.8% vs 12.3%, P=0.03) and fewer reported having knowledge of hospice (82.1% vs 91.2%, P=0.02) in comparison to White patients.

Table 3.

Multivariable Adjusted Aassociation of Black versus White Race With Prognostic Discussion, Future Health Beliefs and Advance Care Planning.

Participant Characteristic	Occurrence of Prognostic Discussion	Patient Believes That Health Will Worsen in the Next 12 Months OR (95% CI)	ACP-Related Documents OR (95% CI)
Black patients (ref = White patients)	1.928(0.834, 4.460)	0.22 (0.12, 0.44)	0.50 (0.32, 0.81)
Female (ref=male)	0.807(0.412, 1.581)	1.70 (0.87,3.34)	1.16 (0.76, 7.76)
Married (ref=not married)	0.804 (0.398, 1.624)	0.89 (0.46,1.72)	0.85 (0.55,1.31)
Affiliation with religion (ref=no religious affiliation)	0.952 (0.311,2.919)	0.74 (0.26, 2.14)	3.31 (1.38, 7.96)
Currently in the hospital (ref=not hospitalized’)	1.462 (0.681,3.136)	1.54 (0.77, 3.07)	0.75 (0.45,1.24)

Model adjusted for all variables shown. “race, gender, marital status, affiliation with religion, and hospitalization status

A total of 42 patients (10.5%) reported having discussions about life expectancy. In bivariate analyses, there were no differences between Black (n=34,11.9%) and White individuals (n=8, 7%) in reporting the occurrence of a prognostic discussion (P=0.15). These findings did not change in the final stepwise multivariable model which retained gender, marital status, religious affiliation, and hospitalization status. (Table 3).

A total of 47 patients responded “yes” to the question that their health would worsen in the next 12 months. In bivariate analyses, Black patients (n=19, 6.7%) were statistically significantly less likely than Whites (n=28, 24.6%) to believe that their health would worsen over the next 12 months (P=<0.0001). In the stepwise regression model, gender, marital status, religious affiliation, and hospitalization status were retained. Black patients had lower odds {OR 0.22, 95% (CI 0.12,0.44)} of believing that their health would worsen in the next 12 months.

A total of 181 participants completed an ACP–related document. In bivariate analyses, Black individuals (n=118,41.4%) were statistically significantly less likely than Whites (n=63,55.3%) to complete any ACP-related document (P=0.01). These findings persisted in the final model where Black patients had lower odds of completing any ACP-related document {OR 0.50, 95% CI (0.32, 0.81)}.

Results of two additional sensitivity analyses were broadly consistent with those reported in Table 3. We did not detect any collinearity between the variables.

Discussion

In this study of people receiving maintenance dialysis, we found that few patients reported having prognostic discussions with their doctor, with no significant differences between Black versus White patients. What potentially highlights an important health disparity—is the difference in Black patients’ understanding of their illness trajectories and lower completion of any ACP documentation.

In our study, we observed statistically significant differences in health beliefs about future illness trajectories among Black and White patients; Black patients were less likely to believe that their health would worsen in the next 12 months. These beliefs could stem from the observation that older Black patients receiving dialysis (but not younger Black patients) have a lower death rate than White patients³⁰ or due to higher rates of kidney transplantation among Whites leaving less healthy Whites on dialysis.³¹ Another recent study suggests that most of the difference in survival outcomes between Black and White patients is due to the fact that Black patients discontinue dialysis at rates lower than their White counterparts³² and are transplanted at lower rates resulting in selection bias among dialysis patients. It’s also possible that Black patients may have a more optimistic outlook on their disease trajectories compared to White patients given previously reported higher quality of life in Black patients.^33,34 It is still unclear what specifically drives racial differences in prognostic outlook. Future perspective studies of racially diverse patients receiving dialysis to explore patient views about health beliefs and future illness trajectory may be informative.

Our second major finding was that the Black patients receiving dialysis had lower odds of having completed any ACP-related documents. This result is consistent with previous reports^35,36 and likely reflects multiple factors. First, Black patients’ positive beliefs about their future health may influence their future expectations which in turn may influence their perceived need for ACP.³⁷ Second, literature from the general population and from other patients with a serious illness suggests that White patients are 2 to 3 times more likely to complete ACP, compared to Black patients.²⁶ Tamura et al also reported that racial minorities receiving outpatient dialysis were less likely to document their EoL wishes.³⁸ It is possible that Black patients may be reluctant to complete advance directives due to fear that these documents may deprive them of appropriate treatments near their end of life.^26,38,39 However, clinician bias may also contribute to the lower prevalence of advance care planning among racial and ethnic minority patients.¹⁹ Eneanya et al in a sample of 152 patients (41% Black individuals) with advanced chronic kidney disease previously reported that Black patients were less likely to discuss EoL preferences with their family members.⁴⁰ It is unclear how racial differences in family dynamics and social support structures influences ACP completion.

Our study has several research, clinical, and policy implications. Broadly, we need to understand why Black patients receiving dialysis are less likely to complete ACP documentation—to this end, qualitative studies may be useful as the current study explores prognostic discussions and ACP completion in a somewhat limited fashion. Future qualitative studies may inform interventions to improve physician-patient communication around prognostic discussions, meaningful EoL planning where patients and family members understand what is at stake, what are patient wishes, and how to prepare surrogate decision-makers for in-the-moment decision making.⁴¹ A trusting relationship of clinicians, patients and families is also critical to fostering a psychologically and culturally safe environment where all stakeholders can effectively engage in emotionally challenging conversations.⁴² At the clinician level, training in primary palliative care skills and culturally competent care⁴³ to provide anticipatory guidance about estimates of life expectancy and future quality of life, illness trajectory, and to document EoL wishes may improve EoL planning.^44,45 Clinicians need to be reminded that each medical event (hospitalization etc.) can be an opportunity to discuss the anticipated future health and complete advance directives.⁴⁶ Additionally, nephrologists and dialysis facilities should offer discussions of prognosis and anticipated illness trajectory at regular intervals with patients. Currently, only a few dialysis units have ACP policies and training opportunities⁴⁷ despite the literature demonstrating that the comfort level of dialysis staff when discussing ACP affects completion of ACP-related documentation.⁴⁸ A policy level mandate to add ACP to the form 2728 (Medicare form to register the ESRD status) completed at the initiation of dialysis may potentially offer some solution. At the patient level, interventions incorporating storytelling and building trust may hold promise in improving the understanding of future health and ACP completion rates.^49,50 Efforts from community organizations, community health workers and other stakeholders such as family members may be needed. For example, peer mentoring and interventions involving community health workers on ACP have been shown to be effective in increasing the completion of ACP-related documents.^50,51,52 Furthermore, faith-based interventions where church leaders partner with physicians to address ACP needs and health misconceptions may be helpful.^53,54

The current study has several strengths and limitations. The major strength is the study of an understudied yet important topic of racial disparities in ACP and prognostic discussion and inclusion of a large number of Black patients receiving dialysis consistent with the local demographics of Cleveland.⁵⁵ Broadly, we need to understand why Black patients receiving dialysis are less likely to complete ACP documentation—to this end– qualitative studies may be useful as the current study explores prognostic discussions and ACP completion in a somewhat limited fashion. For example, the prognosis question only explored life expectancy. Ideally, prognostic discussions should also include estimates of expected quality of life. However, our outcomes (e.g., ACP completion) are based on patient self-reports from a relatively older pre-COVID era dataset. We also lack data to assess health literacy, quality of life, comorbidity burden, transplant eligibility status and income level, all of which may potentially have affected the results. We acknowledge that patients’ responses may be prone to reporting bias due to misunderstanding of questions. Additionally, since we only included patients with clinical capacity to answer questions, results may not be generalizable to patients with advanced dementia or those lacking capacity. Further, we surveyed using a convenience sample methodology and could not assess for a response bias. Finally, these findings in an urban population may be less generalizable to other settings.

In summary, compared to White patients, Black patients reported similarly low engagement in prognostic discussions, more optimistic views of their future health, and less frequent documentation of ACP. Systematic efforts to improve prognostic communication and ACP completion for persons receiving dialysis are needed.

Key Message.

In this cross-sectional post-hoc study, we found that despite no differences in occurrence of prognostic discussions, Black people were more optimistic about their future health than White patients and were less likely to complete advance directives. Development of culturally relevant interventions to improve anticipatory guidance and end-of-life planning are needed.