Clinical Epidemiology

Racial/Ethnic Differences in Dialysis Discontinuation and Survival after Hospitalization for Serious Conditions among Patients on Maintenance Dialysis

Abdulkareem Agunbiade, Abhijit Dasgupta and Michael M. Ward
JASN January 2020, 31 (1) 149-160; DOI: https://doi.org/10.1681/ASN.2019020122

Significance Statement

Racial and ethnic minorities, particularly blacks, survive longer on maintenance dialysis than whites. Minorities are also less likely to discontinue dialysis, which affects relative survival. Both differences have been attributed by some clinicians to better health among minorities on dialysis. To test if health differences explain racial and ethnic differences in dialysis discontinuation, the authors compared the frequency of dialysis discontinuation among racial and ethnic groups after hospitalization for one of four serious medical conditions. Despite hospitalization for the same condition, racial and ethnic minorities were less likely than whites to discontinue dialysis. Using a simulation that accounted for differences in dialysis discontinuation substantially reduced the survival advantage of racial and ethnic minorities. The study suggests that racial and ethnic differences in survival were largely determined by differential discontinuation of dialysis.

Abstract

Background Racial and ethnic minorities on dialysis survive longer than whites, and are less likely to discontinue dialysis. Both differences have been attributed by some clinicians to better health among minorities on dialysis.

Methods To test if racial and ethnic differences in dialysis discontinuation reflected better health, we conducted a retrospective cohort study of survival and dialysis discontinuation among patients on maintenance dialysis in the US Renal Data System after hospitalization for either stroke (n=60,734), lung cancer (n=4100), dementia (n=40,084), or failure to thrive (n=42,950) between 2003 and 2014. We examined the frequency of discontinuation of dialysis and used simulations to estimate survival in minorities relative to whites if minorities had the same pattern of dialysis discontinuation as whites.

Results Blacks, Hispanics, and Asians had substantially lower frequencies of dialysis discontinuation than whites in each hospitalization cohort. Observed risks of mortality were also lower for blacks, Hispanics, and Asians. In simulations that assigned discontinuation patterns similar to those found among whites across racial and ethnic groups, differences in survival were markedly attenuated and hazard ratios approached 1.0. Survival and dialysis discontinuation frequencies among American Indians and Alaska Natives were close to those of whites.

Conclusions Racial and ethnic differences in dialysis discontinuation were present among patients hospitalized with similar health events. Among these patients, survival differences between racial and ethnic minorities and whites were largely attributable to differences in the frequency of discontinuation of dialysis.

Maintenance dialysis is life-sustaining for patients with ESRD, but can be physically and psychologically challenging to maintain long term. In the United States, 20%–25% of patients who begin maintenance dialysis elect to discontinue dialysis because of complications of treatment or the burden of chronic illness.1,2 Although the decision to discontinue dialysis is complex and individual, some patient groups are more likely to discontinue dialysis than others. Among elderly patients, the frequency of dialysis discontinuation is 30% or higher.1 Apart from age, the strongest predictor of discontinuation is race/ethnicity. Whites are more than twice as likely to discontinue dialysis as patients of other races, and discontinuation is more common among non-Hispanics than Hispanics.1–5

It has been hypothesized that the lower frequency of dialysis discontinuation among blacks, compared with whites, is because blacks have better health-related quality of life on dialysis.6,7 Blacks on maintenance dialysis have been reported to have better physical and mental health status, greater energy, and less burden and interference from ESRD than whites,7–12 although a study of patients with incident ESRD did not find such differences.13 Having better health-related quality of life has been proposed to result in less motivation to discontinue dialysis among blacks. Consistent with this hypothesis, poorer physical health status has a graded association with the likelihood of discontinuation.14 Whether similar associations account for the lower frequency of dialysis discontinuation among Asian-Americans and Hispanics is unclear, as few studies have examined quality of life on dialysis in these patient groups.12

One approach to investigate if differences in health status account for racial/ethnic differences in the likelihood of discontinuation of dialysis is to compare patients who experienced the same major health event. Persistence of racial/ethnic differences in dialysis discontinuation in the setting of a new serious comorbidity would suggest that factors other than health status are a more important influence on the decision to discontinue dialysis.

Differences in the frequency of dialysis discontinuation by race/ethnicity may also distort comparisons of survival. If whites discontinue dialysis more often than patients of other races, their survival would be lower, not because of more severe disease or an intrinsically poorer prognosis, but because of selection to dialysis discontinuation and therefore earlier death.15–17 Conversely, longer survival among ethnic minorities compared with whites may be a consequence of lower rates of dialysis discontinuation. The extent to which racial differences in dialysis discontinuation may contribute to differences in survival has not been investigated.18

In this national, population-based, observational study, we compared the likelihood of discontinuation of dialysis by race/ethnicity among patients on maintenance dialysis who were hospitalized with one of four serious comorbid conditions: stroke, lung cancer, dementia, and failure to thrive. These conditions represent major acute or chronic diseases of functional and prognostic importance.19 We also examined survival after the hospitalization, and determined the extent to which differential discontinuation of dialysis accounted for racial/ethnic differences in survival. We hypothesized that racial/ethnic differences in the likelihood of dialysis discontinuation would be present among patients hospitalized with these conditions, and that posthospitalization survival differences would be largely accounted for by differences in the frequency of dialysis discontinuation.

Methods

Data Source

We used data from the US Renal Data System (USRDS), which is a comprehensive national database of all United States citizens or legal permanent residents who are receiving treatment for ESRD under Medicare.20 The USRDS includes longitudinal information on patient characteristics, comorbidities, ESRD treatment modalities, hospitalizations, and outcomes, which are derived from ESRD Evidence reports, Medicare billing data, and the Organ Procurement and Transplantation Network. Data on hospitalizations are obtained from Medicare institutional claims. Patients are followed from the date of first ESRD service until death, recovery of renal function, or 3 years after successful renal transplantation. Deaths are determined on the basis of the Centers for Medicare and Medicaid Services (CMS) death notification forms. Information on whether dialysis had been discontinued before death was obtained from CMS Form 2746.21

Data were provided through a data use agreement with the National Institute of Diabetes and Digestive and Kidney Diseases. This study was exempted from human subjects review by the National Institutes of Health Office of Human Subjects Research Protection.

Study Cohorts

We used as our source group patients registered in USRDS who were aged ≥18 years and who began maintenance dialysis by any modality between January 1, 2003 and December 31, 2014. From this group, we identified four inception cohorts of patients hospitalized with either stroke, lung cancer, dementia, or failure to thrive. We identified the first hospitalization for these conditions that occurred after ESRD onset, on the basis of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9) code of the primary discharge diagnosis (stroke, ICD-9 codes 430–434; lung cancer, ICD-9 code 162; dementia, ICD-9 codes 290, 294.1, 331.0, 331.1, and 331.2; failure to thrive, ICD-9 codes 783.7, 783.3, and 783.2). Each cohort was analyzed separately.

Race/Ethnicity and Covariates

We used data from the Medical Evidence Form CMS-2728 to create four non-Hispanic groups (white, black, Asian/Pacific Islander [hereafter Asian], American Indian/Alaska Native (AI/AN)), and a separate Hispanic group. We excluded the small number of patients whose race was listed as “other.”

Covariates included age at hospitalization (18–49, 50–59, 60–69, 70–79, and ≥80 years), sex, geographic region of residence (Northeast, South, Midwest, West), socioeconomic score, primary cause of ESRD, comorbidity index, and time from onset of dialysis to hospitalization. The socioeconomic score was a validated, census-based composite measure of seven indicators of income, education level, and housing costs in the patient’s ZIP code of residence, with higher scores indicating a higher socioeconomic area.22 The comorbidity index was a weighted sum of 11 conditions that was developed specifically for Medicare patients on dialysis, and validated to predict mortality.23 We computed the comorbidity index on the basis of diagnoses recorded on the Medical Evidence Form at the onset of dialysis or any inpatient claims through the index hospitalization, to provide a comorbidity index that was contemporary with the hospitalization.

Statistical Analyses

The two study outcomes were discontinuation of dialysis and survival after the index hospitalization. We used survival analysis to compare the likelihood of discontinuation of dialysis among racial/ethnic groups, and Cox regression models to estimate the hazard of discontinuation among blacks, Hispanics, Asians, and AIs/ANs relative to whites. In multivariable models, we adjusted for age group, sex, region of residence, socioeconomic score, comorbidity index, primary cause of ESRD, and time on dialysis. Patients were censored at the time of renal transplantation or the end of study follow-up (December 31, 2015). The same methods were used to compare the time from the index hospitalization to death among racial/ethnic groups. Cox–Snell residual plots confirmed the adequacy of model fit (see Supplemental Appendix 1).

To investigate the extent to which overall survival differences were attributable to racial/ethnic differences in discontinuation of dialysis, we developed a simulation model of survival after hospitalization that imposed the same risk of dialysis discontinuation found among whites on the other four racial/ethnic groups. We used simulations because the thresholds for discontinuation may be very different for patients of different races/ethnicities, and the simulations allowed us to estimate differences in survival among groups where everyone used the same decision threshold. By simulating the discontinuation patterns in whites and studying this counterfactual process in the other race/ethnicity groups, we directly addressed the question of what the survival differences would be if all patients shared the same discontinuation process.

First, for each hospitalization cohort, we modeled the dialysis discontinuation process among whites using a Weibull regression model (using survreg function in R). We chose a Weibull model because the Weibull plot (log[−log(S(t)] versus log[t]) is linear along most of the domain. The model included age group, sex, time on dialysis when index hospitalization occurred, region, socioeconomic score, and comorbidity index as covariates. This analysis provided a predicted Weibull model for the time to dialysis discontinuation, including the shape and scale parameters of the Weibull distribution, with the scale parameter depending on the linear prediction model from the Weibull regression.

Second, we applied these Weibull parameters to obtain a predicted distribution of the discontinuation time for blacks, Hispanics, Asians, and AIs/ANs that followed the discontinuation pattern among whites. We simulated the model using a Weibull random number generator with appropriate parameters. Third, we computed a simulated survival time for each patient on the basis of comparing the times of death, censoring, and simulated discontinuation for each individual. For example, the times of discontinuation for black men aged 70–79 years with high comorbidity were simulated to be similar to white men with similar characteristics, and then their survival times were determined by whether their recorded time of death, time of censoring, or simulated time of discontinuation was earliest; the corresponding event was used as the type of event for analyses. For patients who were assigned to have discontinued dialysis in the simulations, we assumed death would occur 7 days after discontinuation.24,25 Fourth, we then used this recomputed survival time to compare the relative hazards of death among ethnic groups using a Cox regression model, with the same covariates as the base analysis. This process was repeated separately for each of the four hospitalization conditions to allow for different patterns of dialysis discontinuation. Further details are provided in Supplemental Appendix 1.

We repeated this process 1000 times to accommodate uncertainty in the Weibull random number generation that provided the times to dialysis discontinuation. Thus, for each hospitalization condition and each ethnic group, we generated a distribution of hazard ratios (HRs) of death in that ethnic group compared with whites in which the pattern of dialysis discontinuation among ethnic minorities was simulated to match that of whites hospitalized with the same condition. Because discontinuation of dialysis is more common among elderly patients, we repeated the analyses in two patient subgroups (age 18–69 years and ≥70 years) to determine if the associations were limited to older patients.

Results

Patient Cohorts

We studied 60,734 patients on their first hospitalization with a stroke, 4100 patients on their first hospitalization with lung cancer, 40,084 patients on their first hospitalization with dementia, and 42,950 patients on their first hospitalization with failure to thrive. The racial/ethnic composition of the cohort with stroke was 51.1% white, 30.8% black, 13.8% Hispanic, 3.5% Asian, and 0.8% AI/AN (Table 1). Similar diversity was present among patients with lung cancer, dementia, and failure to thrive. Median follow-up was 374, 80, 266, and 160 days after the stroke, lung cancer, dementia, and failure to thrive hospitalizations, respectively.

View this table:
Table 1.

Patient characteristics at the index hospitalization for either stroke, lung cancer, dementia, or failure to thrive

Likelihood of Discontinuation of Dialysis

Among patients hospitalized with stroke, 28.3% of whites discontinued dialysis, compared with 13.9% of blacks, 14.6% of Hispanics, 14.1% of Asians, and 21.6% of AIs/ANs (Table 2). Whites also discontinued dialysis sooner than patients of other ethnicities (Figure 1). The adjusted HR for discontinuation of dialysis relative to whites was 0.49 (95% confidence interval [95% CI], 0.47 to 0.52) for blacks, 0.59 (95% CI, 0.54 to 0.63) for Hispanics, and 0.55 (95% CI, 0.47 to 0.64) for Asians, indicating that the likelihood of discontinuation of dialysis after hospitalization for stroke among these groups was approximately one half that of whites (Table 2). The adjusted HR for AIs/ANs relative to whites was 0.79 (95% CI, 0.64 to 0.98). Results were similar for discontinuation of dialysis after hospitalization for lung cancer, dementia, and failure to thrive, with blacks, Hispanics, and Asians having substantially lower likelihoods of discontinuation than whites (Figure 1, Table 2). The likelihood of discontinuation was also lower among AIs/ANs relative to whites after hospitalization for lung cancer, dementia, and failure to thrive, but because of the small number of AIs/ANs, these differences were not statistically significant. In each hospitalization cohort, discontinuation of dialysis was also more likely among older patients, those with more comorbidity, and those living outside the Northeast (Supplemental Table 1).

View this table:
Table 2.

Risks of discontinuation of dialysis and mortality by race/ethnicity among patients hospitalized with stroke, lung cancer, dementia, or failure to thrive

Figure 1.
Figure 1.

Time to discontinuation of dialysis and survival by race/ethnicity after hospitalization for stroke, lung cancer, dementia, or failure to thrive. Figures were truncated at 750 days.

Survival

Survival after hospitalization for stroke was longer among ethnic minorities compared with whites (Figure 1). Adjusted HRs indicated that mortality risks were significantly lower among blacks, Hispanics, and Asians after hospitalization for a stroke compared with whites, although it was not different between AIs/ANs and whites (Table 2). Risk of death after hospitalization for stroke was also higher among older patients, women, those living outside the Northeast, and those with higher comorbidity scores, but the risk of death was lower among those living in higher socioeconomic areas (Supplemental Table 2). Similar racial/ethnic differences in survival were present after hospitalization for dementia and failure to thrive, with blacks, Hispanics, and Asians having lower mortality risks than whites, and survival among AIs/ANs no different from that of whites (Table 2). Adjusted risks of mortality after hospitalization for lung cancer were not different between blacks and Hispanics compared with whites, but Asians had lower mortality risk than whites (Table 2).

After the index hospitalization, 2.1%, 0.3%, 0.2%, and 1.4% of the patients with stroke, lung cancer, dementia, and failure to thrive, respectively, received a renal transplant. Results were unchanged if these patients were followed rather than censored at the time of transplantation (Supplemental Table 3).

Survival in Dialysis Discontinuation Simulations

In the simulation analysis, which assigned a risk of dialysis discontinuation (and consequently a risk of death) to ethnic minorities that matched the pattern of discontinuation observed among whites, ethnic differences in mortality risk after hospitalization for stroke were greatly attenuated (Figure 2). In the simulations, the median adjusted HR among blacks relative to whites was 0.97 (compared with 0.83 in the base analysis), whereas for Asians and AIs/ANs, the estimates clustered around 1.00. Among Hispanics, the risk of mortality after stroke remained somewhat lower than that of whites, with the HRs clustered at 0.90.

Figure 2.
Figure 2.

Adjusted HRs for survival after hospitalization for stroke, lung cancer, dementia, or failure to thrive for blacks, Hispanics, Asians, and AIs/ANs relative to whites in the simulation analysis. The histograms represent the distribution of HRs from 1000 simulations that assigned the discontinuation patterns of whites to each ethnic group. An HR of 1.00 indicates mortality risk identical to that of whites. The arrow marks the observed adjusted HR relative to whites.

In the simulations for survival after hospitalization for lung cancer, blacks and Hispanics had a slightly higher risk of mortality than whites, with median estimated HRs clustered at 1.1, whereas in the base analysis their adjusted HRs were 0.92 and 0.99, respectively (Figure 2). In the simulations, the mortality risk in Asians overlapped that of whites, with a median HR of 0.95, compared with an adjusted HR of 0.57 relative to whites in the base analysis. Risk in AIs/ANs was not different from whites.

In the simulations for survival after hospitalizations for dementia, blacks and Hispanics had mortality risks similar to that of whites, compared with adjusted HRs of 0.76 and 0.82 in the base analysis (Figure 2). In Asians, the simulated HRs clustered at 0.90, and AIs/ANs had a slightly higher risk of mortality relative to whites. In the simulations for survival after hospitalization for failure to thrive, the estimated HRs for blacks, Hispanics, and Asians clustered between 0.90 and 0.95, whereas in the base analysis, the adjusted HRs for these groups ranged from 0.76 to 0.81 (Figure 2).

Crude mortality rates, stratified by whether the patient discontinued dialysis or not, are presented in Supplemental Table 4.

Age-Stratified Results

Because discontinuation of dialysis is more common among older patients, we repeated the analysis in two age strata to determine if associations were present in both younger and older patients. Although the frequency of discontinuation of dialysis was lower among patients aged 18–69 years, differences in dialysis discontinuation by race/ethnicity were similar in this group and in those aged ≥70 years (Table 3). The pattern of racial/ethnic differences in survival after hospitalization with each condition, and the attenuation of HRs in the simulations, were similar in the two age groups. The attenuation was more evident in blacks and Hispanics in the 18–69 year age group because of the larger numbers of patients of these ethnicities.

View this table:
Table 3.

Risks of discontinuation of dialysis and mortality by race/ethnicity among patients hospitalized with stroke, lung cancer, dementia, or failure to thrive, by age stratum

Discussion

Our study has three main findings. First, the likelihood of discontinuation of dialysis was substantially lower among blacks, Hispanics, and Asians than whites even among patients hospitalized with the same serious health condition. Second, apparent risks of mortality were 15%–25% lower among blacks, Hispanics, and Asians compared with whites after hospitalization for stroke, dementia, and failure to thrive, and the apparent risk of mortality was 40% lower among Asians compared with whites after hospitalization for lung cancer. Third, accounting for differential discontinuation of dialysis greatly attenuated the differences in mortality risk after hospitalization between whites and ethnic minorities. In the simulations, survival after stroke was not different between whites and blacks, Asians, and AIs/ANs, nor was survival after hospitalization for dementia in blacks and Hispanics. Mortality risks after hospitalization for lung cancer were higher in blacks and Hispanics, relative to whites, in the simulations, and the apparent survival benefit of Asian ethnicity was greatly reduced.

The presence of substantial differences in the likelihood of dialysis discontinuation between whites and ethnic minorities among patients hospitalized with the same major, potentially life-threatening condition indicates that differences in health are likely not the explanation for racial/ethnic differences in dialysis discontinuation. We would have anticipated the frequency of dialysis discontinuation to be generally similar among patients with the same condition if health was the determinative factor. Instead, we found major differences in risk of discontinuation across racial/ethnic groups, in patterns similar to those among unselected patients with ESRD.1–3 These patterns replicated in each of the four conditions studied, despite differences in prognosis, effect on functioning, and potential treatment. We studied hospitalized patients to provide additional standardization for the severity of each of the four conditions. Adjustment for comorbid conditions of prognostic importance should have further reduced differences in health. Although thresholds for hospitalization may vary by ethnicity and health may vary among patients with the same condition, it is unlikely that residual racial/ethnic differences in health would be so large as to result in a two-fold difference in dialysis discontinuation.

These findings counter the hypothesis that ethnic differences in health lead to ethnic differences in the propensity to discontinue dialysis. The underlying evidence that ethnic minorities have better health status on dialysis than whites is also quite limited. Two small, early studies suggested that blacks on dialysis had fewer symptoms and better quality of life than whites, but a third study found no differences.8,9,11 In two larger studies, blacks reported better health status than whites on the Short Form-36.7,12 However, mean scores differed by <1.5 points on scales, where a difference of 3–5 points is considered clinically meaningful.26,27 In the Dialysis Outcomes and Practice Patterns Study, although poorer health status was significantly associated with survival, health status did not confound the associations between ethnicity and survival.12 Blacks have also been reported to experience ESRD as less burdensome than whites.7,9,12,13 However, this may reflect coping responses or expectations more than health status. These studies may also be biased in that they examined prevalent patients on maintenance dialysis. Patients who had received renal transplants, who are disproportionately white and healthy, were not studied. Differential access to transplantation may contribute to racial/ethnic differences in survival.28 Indeed, among incident patients in the Dialysis Morbidity and Mortality Study Wave 2, there were no differences in health status between blacks and whites.13 Additionally, studies of quality of life have focused on black/white comparisons, yet lower frequencies of dialysis discontinuation are also seen among Asians and Hispanics. Few studies have reported on quality of life in Asian-Americans, Hispanics, or AI/ANs on dialysis, leaving open the question of whether higher quality of life in these groups relative to whites results in the observed pattern of dialysis discontinuation.12

Family relationships, interpersonal support, and religious beliefs influence the decision to discontinue dialysis.11,29–32 These sociocultural factors also vary among racial/ethnic groups, and could be alternative explanations for differences in dialysis discontinuation.30–32 Blacks with ESRD report having generally greater social support and religiosity than whites, and blacks often endorse a high level of religiosity or spirituality.11,33,34 Social support and spirituality have in turn been associated with longer survival, which may be mediated by better mental health or greater adherence to treatment.35–37 Religious beliefs also have an important influence on end-of-life care considerations.38,39 Social support and spirituality among Asian-Americans and Hispanics on dialysis has not been studied, although these cultures traditionally place strong emphasis on family ties, including an important role of family members in health care decision making.40–43 Some evidence indicates that blacks with ESRD are less likely to receive end-of-life counseling, complete advanced directives, and be referred to hospice care than whites, although limited data suggest no differences between Asian-Americans and whites.5,19,21,44 These racial/ethnic differences in family support, religious beliefs, education regarding treatment options, and trust in providers may be responsible for the observed differences in dialysis discontinuation.

Survival after hospitalization also varied by race/ethnicity. Within each diagnosis, the survival advantage relative to whites was greatest in the ethnic group that had the lowest risk of discontinuation of dialysis. This association was most striking in the case of lung cancer, where Asians had a markedly lower risk of discontinuation (adjusted HR, 0.37) and the lowest mortality risk (adjusted HR, 0.57) relative to whites. In each of the four diagnosis groups, AIs/ANs had risks of discontinuation close to that of whites, and also had mortality risks most similar to whites. These associations suggested that the apparent mortality risks were linked with the likelihood of dialysis discontinuation, an association supported in the simulation analyses.

In the simulations, which constructed hypothetical cohorts that differed from the observed cohort only in that ethnic minority patients were modeled to have the discontinuation patterns of whites, the survival advantage of ethnic minorities relative to whites was greatly attenuated. With the exception of Hispanics with stroke, Asians with dementia, and blacks and Asians with failure to thrive, survival was similar between ethnic minorities and whites. Among those hospitalized with lung cancer, the survival advantage of blacks and Hispanics was reversed. These findings indicate that the observed survival differences were biased by differential selection to dialysis discontinuation. Similar associations were present in both younger and elderly patients, despite lower frequencies of discontinuation in the younger group. These results indicate that the associations were not limited to patients with a high baseline risk of dialysis discontinuation. The simulation results were also more consistent with clinical expectations for survival among patients with similar health than the observed results. Among patients with substantial comorbidity who are hospitalized with a serious illness, differences in discontinuation of dialysis may account for ethnic differences in survival, whereas other factors, including differences in nutritional status, inflammation, and dialysis intensity, may play a lesser role in this group.16,17

The strengths of this study include the population-based sample, examination of four different diagnosis groups, and use of simulations to provide robust survival estimates under the counterfactual of similar risks of discontinuation among all ethnicities. The validity of the simulations is supported by accurate modeling of the discontinuation process among whites. However, our study is limited in that we did not have patient-reported outcomes of health status or quality of life. It is possible that perceived quality of life could differ by ethnicity despite similar health, and that these perceptions influenced decisions regarding discontinuation of dialysis. We also did not have measures of social support or religiosity, and could not directly test if these factors mediated the association between ethnicity and dialysis discontinuation. Future studies should examine the role of sociocultural factors in racial/ethnic differences in discontinuation. The information on Form 2746 may underestimate discontinuations of dialysis, although we are not aware of any data suggesting this differs by race/ethnicity.45 The small number of AIs/ANs resulted in imprecise estimates for this group, although their rates of discontinuation and survival were close to those of whites.

In conclusion, racial/ethnic differences in the frequency of dialysis discontinuation were present among patients hospitalized with the same serious health condition, suggesting that lower rates of discontinuation among ethnic minorities may not be because of their having better health on dialysis than whites. Additionally, most of the survival difference between ethnic minorities and whites was attributable to more frequent (and earlier) discontinuation of dialysis by whites. Although our results are limited to patients hospitalized with one of these four illnesses, similar processes may occur in the wider ESRD population. The notion that blacks (or other ethnic minorities) fair better on dialysis than whites is made on the basis of two independent and reinforcing lines of evidence: better health status, which is inferred from lower rates of discontinuation of dialysis, and longer survival. Our results do not support either line of evidence.

Disclosures

None.

Funding

This work was supported by the Intramural Research Program, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health (NIH) (grant ZIA-AR-041153). Dr. Agunbiade was supported by the NIH Medical Research Scholars Program.

Supplemental Material

This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/doi:10.1681/ASN.2019020122/-/DCSupplemental.

Supplemental Appendix 1. Methodology.

Supplemental Table 1. Risks of discontinuation of dialysis by race/ethnicity among patients hospitalized with stroke, lung cancer, dementia, or failure to thrive.

Supplemental Table 2. Risks of mortality by race/ethnicity among patients hospitalized with stroke, lung cancer, dementia, or failure to thrive.

Supplemental Table 3. Risks of discontinuation of dialysis and mortality by race/ethnicity among patients hospitalized with stroke, lung cancer, dementia, or failure to thrive, in analyses that did not censor patients at the time of renal transplantation.

Supplemental Table 4. Crude mortality by race/ethnicity among patients who did or did not discontinue dialysis.

Acknowledgments

Dr. Agunbiade, Dr. Dasgupta, and Dr. Ward designed the study. Dr. Dasgupta performed the analyses. Dr. Ward drafted the paper, and all authors revised the paper and approved the final version of the manuscript.

The data reported here have been supplied by the US Renal Data System. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as official policy or interpretation of the US Government.

Footnotes

  • Published online ahead of print. Publication date available at www.jasn.org.

  • See related editorial, “Can Dialysis Withdrawal Explain Why White Patients Have Worse Survival than Black Patients?” on pages 2–4.

References

  1. ↵
    1. United States Renal Data System
    : 2017 USRDS Annual Data Report: Epidemiology of Kidney Disease in the United States, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2017
  2. ↵
    1. Foley RN,
    2. Sexton DJ,
    3. Drawz P,
    4. Ishani A,
    5. Reule S
    : Race, ethnicity, and end-of-life care in dialysis patients in the United States. J Am Soc Nephrol 29: 2387–2399, 2018pmid:30093455
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Leggat JE Jr.,
    2. Bloembergen WE,
    3. Levine G,
    4. Hulbert-Shearon TE,
    5. Port FK
    : An analysis of risk factors for withdrawal from dialysis before death. J Am Soc Nephrol 8: 1755–1763, 1997pmid:9355079
    OpenUrlAbstract
    1. Gessert CE,
    2. Haller IV,
    3. Johnson BP
    : Regional variation in care at the end of life: Discontinuation of dialysis. BMC Geriatr 13: 39, 2013pmid:23635315
  5. ↵
    1. Thomas BA,
    2. Rodriguez RA,
    3. Boyko EJ,
    4. Robinson-Cohen C,
    5. Fitzpatrick AL,
    6. O’Hare AM
    : Geographic variation in black-white differences in end-of-life care for patients with ESRD. Clin J Am Soc Nephrol 8: 1171–1178, 2013pmid:23580783
    OpenUrlAbstract/FREE Full Text
  6. ↵
    1. Price DA,
    2. Owen WF Jr.
    : African-Americans on maintenance dialysis: A review of racial differences in incidence, treatment, and survival. Adv Ren Replace Ther 4: 3–12, 1997pmid:8996615
    OpenUrlCrossRefPubMed
  7. ↵
    1. Unruh M,
    2. Miskulin D,
    3. Yan G,
    4. Hays RD,
    5. Benz R,
    6. Kusek JW, et al. HEMO Study Group
    : Racial differences in health-related quality of life among hemodialysis patients. Kidney Int 65: 1482–1491, 2004pmid:15086492
    OpenUrlCrossRefPubMed
  8. ↵
    1. Tell GS,
    2. Mittelmark MB,
    3. Hylander B,
    4. Shumaker SA,
    5. Russell G,
    6. Burkart JM
    : Social support and health-related quality of life in black and white dialysis patients. ANNA J 22: 301–308; discussion 309–310, 1995pmid:7786078
    OpenUrlPubMed
  9. ↵
    1. Kutner NG,
    2. Brogan D,
    3. Fielding B,
    4. Hall WD
    : Black/white differences in symptoms and health satisfaction reported by older hemodialysis patients. Ethn Dis 10: 328–333, 2000pmid:11110348
    OpenUrlPubMed
    1. Hicks LS,
    2. Cleary PD,
    3. Epstein AM,
    4. Ayanian JZ
    : Differences in health-related quality of life and treatment preferences among black and white patients with end-stage renal disease. Qual Life Res 13: 1129–1137, 2004pmid:15287279
    OpenUrlCrossRefPubMed
  11. ↵
    1. Kimmel PL,
    2. Emont SL,
    3. Newmann JM,
    4. Danko H,
    5. Moss AH
    : ESRD patient quality of life: Symptoms, spiritual beliefs, psychosocial factors, and ethnicity. Am J Kidney Dis 42: 713–721, 2003pmid:14520621
    OpenUrlCrossRefPubMed
  12. ↵
    1. Lopes AA,
    2. Bragg-Gresham JL,
    3. Satayathum S,
    4. McCullough K,
    5. Pifer T,
    6. Goodkin DA, et al.: Worldwide Dialysis Outcomes and Practice Patterns Study Committee
    : Health-related quality of life and associated outcomes among hemodialysis patients of different ethnicities in the United States: The Dialysis Outcomes and Practice Patterns Study (DOPPS). Am J Kidney Dis 41: 605–615, 2003pmid:12612984
    OpenUrlCrossRefPubMed
  13. ↵
    1. Kutner NG,
    2. Zhang R,
    3. Brogan D
    : Race, gender, and incident dialysis patients’ reported health status and quality of life. J Am Soc Nephrol 16: 1440–1448, 2005pmid:15800127
    OpenUrlAbstract/FREE Full Text
  14. ↵
    1. Knight EL,
    2. Ofsthun N,
    3. Teng M,
    4. Lazarus JM,
    5. Curhan GC
    : The association between mental health, physical function, and hemodialysis mortality. Kidney Int 63: 1843–1851, 2003pmid:12675862
    OpenUrlCrossRefPubMed
  15. ↵
    1. Agodoa L,
    2. Eggers P
    : Racial and ethnic disparities in end-stage kidney failure-survival paradoxes in African-Americans. Semin Dial 20: 577–585, 2007pmid:17991208
    OpenUrlCrossRefPubMed
  16. ↵
    1. Robinson BM,
    2. Joffe MM,
    3. Pisoni RL,
    4. Port FK,
    5. Feldman HI
    : Revisiting survival differences by race and ethnicity among hemodialysis patients: The Dialysis Outcomes and Practice Patterns Study. J Am Soc Nephrol 17: 2910–2918, 2006pmid:16988065
    OpenUrlAbstract/FREE Full Text
  17. ↵
    1. Streja E,
    2. Kovesdy CP,
    3. Molnar MZ,
    4. Norris KC,
    5. Greenland S,
    6. Nissenson AR, et al
    .: Role of nutritional status and inflammation in higher survival of African American and Hispanic hemodialysis patients. Am J Kidney Dis 57: 883–893, 2011pmid:21239093
    OpenUrlPubMed
  18. ↵
    1. Port FK,
    2. Wolfe RA,
    3. Hawthorne VM,
    4. Ferguson CW
    : Discontinuation of dialysis therapy as a cause of death. Am J Nephrol 9: 145–149, 1989pmid:2741993
    OpenUrlCrossRefPubMed
  19. ↵
    1. Kurella Tamura M,
    2. Goldstein MK,
    3. Pérez-Stable EJ
    : Preferences for dialysis withdrawal and engagement in advance care planning within a diverse sample of dialysis patients. Nephrol Dial Transplant 25: 237–242, 2010pmid:19734137
    OpenUrlCrossRefPubMed
  20. ↵
    1. United States Renal Data System
    : 2017 Researchers Guide to the USRDS Database, Bethesda, MD, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2017
  21. ↵
    1. Murray AM,
    2. Arko C,
    3. Chen S-C,
    4. Gilbertson DT,
    5. Moss AH
    : Use of hospice in the United States dialysis population. Clin J Am Soc Nephrol 1: 1248–1255, 2006pmid:17699355
    OpenUrlAbstract/FREE Full Text
  22. ↵
    1. Ward MM
    : Socioeconomic status and the incidence of ESRD. Am J Kidney Dis 51: 563–572, 2008pmid:18371532
    OpenUrlCrossRefPubMed
  23. ↵
    1. Liu J,
    2. Huang Z,
    3. Gilbertson DT,
    4. Foley RN,
    5. Collins AJ
    : An improved comorbidity index for outcome analyses among dialysis patients. Kidney Int 77: 141–151, 2010pmid:19907414
    OpenUrlCrossRefPubMed
  24. ↵
    1. Neu S,
    2. Kjellstrand CM
    : Stopping long-term dialysis. An empirical study of withdrawal of life-supporting treatment. N Engl J Med 314: 14–20, 1986pmid:3079619
    OpenUrlCrossRefPubMed
  25. ↵
    1. Cohen LM,
    2. Germain MJ,
    3. Poppel DM,
    4. Woods AL,
    5. Pekow PS,
    6. Kjellstrand CM
    : Dying well after discontinuing the life-support treatment of dialysis. Arch Intern Med 160: 2513–2518, 2000pmid:10979064
    OpenUrlCrossRefPubMed
  26. ↵
    1. Kosinski M,
    2. Zhao SZ,
    3. Dedhiya S,
    4. Osterhaus JT,
    5. Ware JE Jr.
    : Determining minimally important changes in generic and disease-specific health-related quality of life questionnaires in clinical trials of rheumatoid arthritis. Arthritis Rheum 43: 1478–1487, 2000pmid:10902749
    OpenUrlCrossRefPubMed
  27. ↵
    1. Leaf DE,
    2. Goldfarb DS
    : Interpretation and review of health-related quality of life data in CKD patients receiving treatment for anemia. Kidney Int 75: 15–24, 2009pmid:18813284
    OpenUrlCrossRefPubMed
  28. ↵
    1. Kucirka LM,
    2. Grams ME,
    3. Lessler J,
    4. Hall EC,
    5. James N,
    6. Massie AB, et al
    .: Association of race and age with survival among patients undergoing dialysis. JAMA 306: 620–626, 2011pmid:21828325
    OpenUrlCrossRefPubMed
  29. ↵
    1. Murtagh F,
    2. Cohen LM,
    3. Germain MJ
    : Dialysis discontinuation: Quo vadis? Adv Chronic Kidney Dis 14: 379–401, 2007pmid:17904506
    OpenUrlCrossRefPubMed
  30. ↵
    1. Hussain JA,
    2. Flemming K,
    3. Murtagh FEM,
    4. Johnson MJ
    : Patient and health care professional decision-making to commence and withdraw from renal dialysis: A systematic review of qualitative research. Clin J Am Soc Nephrol 10: 1201–1215, 2015pmid:25943310
    OpenUrlAbstract/FREE Full Text
    1. Song MK,
    2. Hanson LC
    : Relationships between psychosocial-spiritual well-being and end-of-life preferences and values in African American dialysis patients. J Pain Symptom Manage 38: 372–380, 2009pmid:19356896
    OpenUrlCrossRefPubMed
  32. ↵
    1. Levin JS,
    2. Taylor RJ,
    3. Chatters LM
    : Race and gender differences in religiosity among older adults: Findings from four national surveys. J Gerontol 49: S137–S145, 1994pmid:8169348
    OpenUrlCrossRefPubMed
  33. ↵
    1. Tanyi RA,
    2. Werner JS
    : Spirituality in African American and Caucasian women with end-stage renal disease on hemodialysis treatment. Health Care Women Int 28: 141–154, 2007pmid:17364977
    OpenUrlCrossRefPubMed
  34. ↵
    1. Ko B,
    2. Khurana A,
    3. Spencer J,
    4. Scott B,
    5. Hahn M,
    6. Hammes M
    : Religious beliefs and quality of life in an American inner-city haemodialysis population. Nephrol Dial Transplant 22: 2985–2990, 2007pmid:17597086
    OpenUrlCrossRefPubMed
  35. ↵
    1. Spinale J,
    2. Cohen SD,
    3. Khetpal P,
    4. Peterson RA,
    5. Clougherty B,
    6. Puchalski CM, et al
    .: Spirituality, social support, and survival in hemodialysis patients. Clin J Am Soc Nephrol 3: 1620–1627, 2008pmid:18922991
    OpenUrlAbstract/FREE Full Text
    1. Siegal BR,
    2. Calsyn RJ,
    3. Cuddihee RM
    : The relationship of social support to psychological adjustment in end-stage renal disease patients. J Chronic Dis 40: 337–344, 1987pmid:3558708
    OpenUrlCrossRefPubMed
  37. ↵
    1. Chan R,
    2. Steel Z,
    3. Brooks R,
    4. Heung T,
    5. Erlich J,
    6. Chow J, et al
    .: Psychosocial risk and protective factors for depression in the dialysis population: A systematic review and meta-regression analysis. J Psychosom Res 71: 300–310, 2011pmid:21999973
    OpenUrlCrossRefPubMed
  38. ↵
    1. Johnson KS,
    2. Elbert-Avila KI,
    3. Tulsky JA
    : The influence of spiritual beliefs and practices on the treatment preferences of African Americans: A review of the literature. J Am Geriatr Soc 53: 711–719, 2005pmid:15817022
    OpenUrlCrossRefPubMed
  39. ↵
    1. Balboni TA,
    2. Vanderwerker LC,
    3. Block SD,
    4. Paulk ME,
    5. Lathan CS,
    6. Peteet JR, et al
    .: Religiousness and spiritual support among advanced cancer patients and associations with end-of-life treatment preferences and quality of life. J Clin Oncol 25: 555–560, 2007pmid:17290065
    OpenUrlAbstract/FREE Full Text
  40. ↵
    1. Marin G,
    2. Marin BV
    : Research with Hispanic Populations, Newbury Park, CA, Sage Publications, 1991
    1. Smith AK,
    2. Sudore RL,
    3. Pérez-Stable EJ
    : Palliative care for Latino patients and their families: Whenever we prayed, she wept. JAMA 301: 1047–1057, E1, 2009pmid:19278947
    OpenUrlCrossRefPubMed
    1. Blackhall LJ,
    2. Frank G,
    3. Murphy ST,
    4. Michel V,
    5. Palmer JM,
    6. Azen SP
    : Ethnicity and attitudes towards life sustaining technology. Soc Sci Med 48: 1779–1789, 1999pmid:10405016
    OpenUrlCrossRefPubMed
  43. ↵
    1. Bowman KW,
    2. Singer PA
    : Chinese seniors’ perspectives on end-of-life decisions. Soc Sci Med 53: 455–464, 2001pmid:11459396
    OpenUrlCrossRefPubMed
  44. ↵
    1. Eneanya ND,
    2. Wenger JB,
    3. Waite K,
    4. Crittenden S,
    5. Hazar DB,
    6. Volandes A, et al
    .: Racial disparities in end-of-life communication and preferences among chronic kidney disease patients. Am J Nephrol 44: 46–53, 2016pmid:27351650
  45. ↵
    1. Chen JCY,
    2. Thorsteinsdottir B,
    3. Vaughan LE,
    4. Feely MA,
    5. Albright RC,
    6. Onuigbo M, et al
    .: End of life, withdrawal, and palliative care utilization among patients receiving maintenance hemodialysis therapy. Clin J Am Soc Nephrol 13: 1172–1179, 2018pmid:30026285
    OpenUrlAbstract/FREE Full Text
Back to top

In this issue

View Selected Citations (0)
Print
Download PDF
Email Article
Citation Tools
Request Permissions
Racial/Ethnic Differences in Dialysis Discontinuation and Survival after Hospitalization for Serious Conditions among Patients on Maintenance Dialysis
Abdulkareem Agunbiade, Abhijit Dasgupta, Michael M. Ward
JASN Jan 2020, 31 (1) 149-160; DOI: 10.1681/ASN.2019020122
del.icio.us logo Digg logo Reddit logo Twitter logo CiteULike logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords