Share this article on:

Effects of Kidney Transplantation on Labor Market Outcomes in Sweden

Jarl, Johan, PhD1; Gerdtham, Ulf-G., PhD1,2; Desatnik, Peter, MD3; Prütz, Karl-Göran, MD4,5

doi: 10.1097/TP.0000000000002228
Original Clinical Science—General

Background Kidney transplantation is considered a superior treatment for end-stage renal disease compared with dialysis although little is known about the wider effects, especially on labor market outcomes. The objective is to estimate the treatment effect of kidney transplantation compared with dialysis on labor market outcomes, controlling for the nonrandom selection into treatment.

Methods The average treatment effect is estimated using an inverse-probability weighting regression adjustment approach on all patients in renal replacement therapy 1995 to 2012.

Results Kidney transplantation is associated with a treatment advantage over dialysis on employment, labor force participation, early retirement, and labor income. The probability of being employed 1 year after treatment is 21 (95% confidence interval, 16-25) percentage points higher for transplantation. The positive effect increases to 38 (95% confidence interval, 30-46) percentage points after 5 years, mainly due to worsening outcomes on dialysis. The effect on labor income is mainly mediated through employment probability. The productivity gains of transplantation compared to dialysis amounts to €33 000 over 5 years.

Conclusions Transplantation is superior to dialysis in terms of potential to return to work as well as in terms of labor income and risk of early retirement, after controlling for treatment selection. This positive effect increases over time after transplantation.

Based on Swedish registries' data between 1995 and 2012, the authors compare the labour market outcomes between patients treated by dialysis or kidney transplantation controlling for treatment selection bias. These outcomes are far better when patients receive a kidney graft and this superiority still grows with time.

1 Health Economics, Department of Clinical Sciences, Malmö, Lund University, Lund, Sweden.

2 Department of Economics, Lund University, Lund, Sweden.

3 Anesthesia and Intensive Care, Helsingborg Hospital, Helsingborg, Sweden.

4 Department of Internal Medicine, Helsingborg Hospital, Helsingborg, Sweden.

5 Swedish Renal Registry, Ryhov Hospital, Jönköping, Sweden.

Received 28 September 2017. Revision received 16 March 2018.

Accepted 21 March 2018.

Financial support from Region Skåne, Sweden (regional public healthcare provider), and the Government Grant for Clinical Research (“ALF”) is greatly acknowledged.

The authors declare no conflicts of interest. The results presented in this article have not been published previously in whole or part.

All authors participated in the design of the study, interpretation of the results, and write-up. J.J. conducted the analyses and drafted the first version. All authors have approved the final version.

Correspondence: Johan Jarl, PhD, Box 117 221 00 Lund, Sweden. (

Supplemental digital content (SDC) is available for this article. Direct URL citations appear in the printed text, and links to the digital files are provided in the HTML text of this article on the journal’s Web site (

Little is known about the effects of transplantation on labor market outcomes compared with dialysis treatment. There are a number of treatment-related complications that might hinder posttransplant employment.1 Comparing before and after transplantation, employment and participation in unpaid activities have been shown to both decrease1 and increase2 posttransplant. These ambiguous results could partially be explained by different definitions of employment.1 However, the employment rate after kidney transplantation has been found to vary between 29% and 59% in later studies.3,4 Thus, kidney transplantation cannot be expected to, on average, lead to a work status similar to the general population.5,6

There are a number of factors that have been shown to be related to the likelihood of posttransplant employment. Positive associations have been found with being young,4,7–9 male,3,4,7 married,4 having a positive perception of capability to work,4 employment before transplantation,3,4,7,8 quality of life,3,10 higher education,1,3,7,8,10 better health,1,6 having a living donor transplantation,7,8,10 preemptive transplantation,8 and less than 1 year of dialysis.8 Negative associations have been found with being diabetic,3,9 single,1 having dependents,7 and waiting time for a transplant.1,7 Better graft function has in some studies been associated with increased probability of employment,6 whereas other studies have failed to find such relationship.3 However, complications and side effects of transplantation and/or immunosuppressant medication are associated with a reduced employment probability.6,10 Healthcare insurance schemes have also been shown to affect the likelihood of employment after kidney transplantation.11 Thus, employment status after kidney transplantation depends perhaps more on contextual factors (eg, health and insurance systems) than the patient's clinical factors.3

Despite the many studies on employment rates and determinants of employment participation after kidney transplantation, little is known about the actual effect of transplantation on employment and other labor market outcomes. To study this, we need to compare to the relevant treatment alternative, which is dialysis. This is rarely done in prior studies as the comparisons generally are between renal transplant recipients (RTR) and the general population.6,8,12 The only study, to the authors' knowledge, that compares employment participation after kidney transplantation to dialysis finds that transplantation, automated peritoneal dialysis (PD), and home haemodialysis (HD) is associated with higher employment rates compared with in-center HD.9 However, no study has of yet controlled for treatment selection effect, that is, that a factor might influence both treatment choice and the outcome under study. The purpose of this study was therefore to estimate the treatment effect of kidney transplantation compared with dialysis on labor market outcomes in Sweden, controlling for treatment selection.

Back to Top | Article Outline


Patients in renal replacement therapy (RRT) were identified through the Swedish Renal Registry which has a coverage rate above 95%13 and almost 100% for RTR.14 The register includes information on treatment and treatment time, survival, comorbidities, and primary disease. Linked to these data, using unique personal identification numbers, was the Longitudinal Integration Database for Health Insurance and Labor Market Studies database from Statistics Sweden which holds annual registers of all individuals older than 16 years registered in Sweden. Education and all labor outcomes were taken from Longitudinal Integration Database for Health Insurance and Labor Market Studies which is mainly based on administrative registers from the tax office and the social insurance agency. As such, there is hardly any missing information for the labor market outcomes in the current study. The study included all patients 20 to 60 years of age who started RRT during the years 1995 to 2012 in Sweden. Individuals who recovered before day 91 or where treatment is unknown were excluded. The investigated labor market outcomes were employment (full and part time), labor force participation, labor income, early retirement, months in early retirement, and participation in educational activities (see Materials and Methods, SDC, for details).

Observational studies of treatment effects generally have the problem that treatment assignment is not random and factors that influence the likelihood of receiving a specific treatment are related to the outcome of the treatment.15 This is apparent in the field of RRT where the likelihood of receiving a transplant is higher for those that are expected to have the best outcomes of the treatment. Comparing outcomes for transplanted with those that remain on dialysis is expected to create a bias in the favor of transplantation.16 Thus, it is crucial to account for nonrandom selection into treatment which was undertaken in the current study by using the inverse probability weighting regression adjustment (IPWRA) approach.16 The IPWRA first estimates a treatment model where observed treatment assignment is a function of individuals' characteristics. The inverse probability of being in the observed treatment group for each individual is then used as weight when estimating the outcome models for each treatment. These outcome models are then used to predict treatment-specific outcomes for each individual. Thus, the inverse probability weighting part of the approach gives more weight to those least likely to be in the observed group, whereas the RA part uses linear regression models to predict the outcomes of each treatment for each individual. This ideally leads to a balanced sample where the distributions of the covariates do not vary over treatment choice. Finally, the average treatment effect (ATE) is the difference between the averages of the treatment-specific outcomes, whereas the ATE of the treated is the difference for RTRs. All analyses were done in Stata 14.1 using the “teffects” command,17 and the study was approved by Lund Regional Ethical Review Board (dnr: 2014/144). Details of the model specifications and the estimation method are reported in Materials and Methods, SDC (

It is impossible to determine if all relevant factors were controlled for and the estimated effects might therefore still be biased, especially given that individual preferences play an important role in RRT treatment choice.18 Nevertheless, applying the IPWRA approach, thereby controlling for known observed factors, should be helpful in reducing the potential bias and a number of checks were performed to investigate this (see Materials and Methods, SDC, We also limited our sample in a separate analysis to only those patients who at any time have been on the waiting list. This was expected to reduce the variance between patients and increase the likelihood of achieving comparable treatment groups and thereby reduce estimation bias.

For the effect of the different treatments to be comparable in terms of time on dialysis, the start of dialysis was adjusted for those that do not receive transplantation based on the average waiting time for RTRs. Thus, “treatment” refers to the comparison between transplantation and dialysis after 593 days in RRT (the average number of days in RRT before transplantation in the sample). Due to this the time between start of RRT and start of treatment was constant for dialysis patients while it varied between 0 and 4533 days for transplanted. Two sensitivity analyses were therefore conducted: (a) only including transplanted who waited more than 1 year for a transplant, and (b) only including transplanted who waited less than 1 year. See Materials and Methods, SDC ( for all performed sensitivity analyses.

Back to Top | Article Outline


Renal transplant recipient patients constitute three quarters of all RRT patients (Table 1) and hemodialysis patients constitute about two thirds of the dialysis population. Table 1 shows that there are differences between the treatment groups, especially in terms of education, risk factors, comorbidity, and history of labor market outcomes. All differences found in the univariate analysis favor RTR.



There are large differences between the 2 groups in terms of the outcome variables (Table 2). Renal transplant recipient, compared with patients on dialysis, have higher employment rate (61% vs. 18 %), income and fewer months in early retirement, 1 year after treatment. This was expected due to treatment selection. Attempting to control for this, Figure 1 shows the standardized difference between groups before and after weighting based on the inverse probability of being in the observed treatment group. Indications are that a balanced distribution is achieved of the observed variables as no covariate exceeds the |0.1| value after weighting. The distribution in terms of variance ratios is also acceptable and the treatment model passes the overidentification test for covariate balance (see Materials and Methods, SDC, We find that the likelihood of receiving transplantation compared with dialysis is related to all demographic, socioeconomic, and medical factors specified in the logistic treatment model (Materials and Methods, SDC,, pseudo-R 2 = 0.26) with the exception of sex (results not shown).





Table 3 shows the estimates of average outcomes of treatments and ATE of all labor market outcomes under study. The results are stable to separating dialysis into HD and PD (Table S1, SDC, If all individuals in the sample received dialysis, 33% is expected to be employed the year after. However, if everyone instead received a transplant, the employment rate the year after is expected to be 54%. Thus, transplantation is associated with a 21-percentage point increase on the likelihood of employment compared to dialysis. Labor force participation is even more positively associated with transplantation, with a probability increase by 24 percentage points compared with dialysis. Transplantation is associated with a reduction in risk of early retirement by 12 percentage points, compared with dialysis, while the average time on early retirement decreases by 0.8 month. Educational activities are unaffected by choice of treatment. The ATE of the treated shows similar but stronger results than the ATE. The estimations based on the waiting list sample give similar results for employment and labor force participation while the results for early retirement are statistically insignificant. For labor income, the unconditioned estimation shows a positive association between transplantation and income. Labor income is estimated to be about 2 times higher after transplantation compared with being on dialysis but income levels are low as it includes many nonworking individuals. However, when conditioning on being employed, the treatment effect is substantially smaller, although still in favor of transplantation. We also condition on earning Swedish kronor (SEK) 100 000 (€11561) which further reduces the associated ATE of transplantation to 26 percentage points. The ATE of the treated and the waiting list sample are similar.



In Table 4, the estimated effect of transplantation compared to dialysis on employment 1 to 5 years after treatment is shown. Transplantation is associated with a 38-percentage point increase in employment after 5 years. This increase is mainly an effect of a fall in the estimated average outcome of dialysis over time. That is, the likelihood of being employed falls over time on dialysis while being stable over time with a transplant. The results for the treated are similar but stronger, whereas the results based on the waiting list sample are considerable lower years 3 and 4.



Several factors are associated with the likelihood of employment after treatment, as estimated in the outcome model. Higher education and working the year before start of RRT are associated with an increase in the likelihood of working after both treatments. Female sex, age, and number of risk factors are associated with a reduced chance of working after transplantation but are not associated with the likelihood of working after dialysis. Disposable income is associated with an increased chance of working after transplantation but is (borderline) insignificant for dialysis. Finally, civil status only seems to be related to the likelihood of employment when in dialysis and not after transplantation.

The reduced labor income losses of transplantation the year after treatment can be calculated, based on the estimates, by comparing the expected labor income (probability of employment × labor income conditioned on employment) after transplantation and dialysis. The difference between the 2 treatments sums to, on average, SEK 60 200 (€ 7000) in year 2013 prices. Discounted reduced labor income losses from transplantation compared to dialysis amounts to almost SEK 283 800 (€ 32 800) over 5 years. See Materials and Methods, SDC ( for all sensitivity analyses.

Back to Top | Article Outline


This is the first study, to the authors’ knowledge, of the effect of kidney transplantation on labor market outcomes that controls for treatment selection and compares with the alternative treatment, dialysis. We find that transplantation not only has a strong positive effect but also that it is essential to control for treatment selection. If everyone in the sample had received transplantation, 54% would be employed 1 year after treatment. However, if the same patients instead would have received dialysis, 33% would be employed, a number that would fall over time on dialysis increasing the positive treatment effect of transplantation. This shows a positive societal and individual effect of kidney transplantation. The positive effect of transplantation is also evident in labor force participation, early retirement, and labor income. The current study supports the positive association of transplantation compared with in-center HD as found in Helanterä et al9 while also showing that a large part of the association is due to the treatment and not due to selection into treatment based on employment status.

The direct unadjusted employment rate 1 year after transplantation in the current study (61%) is high compared with the other studies in the field3,4 and can be compared with the general Swedish population (75-79%19). However, these figures themselves say little about the effectiveness of kidney transplantation in Sweden because high rates could be due However, as the estimate of average outcome of transplantation is as high as 54 % this indicates that the treatment and structures to promote return to work is relatively successful in Sweden, although there is a treatment selection to the most appropriate candidates. Compared to the general population however, a considerable room for improvement is noted.

The likelihood of early retirement decreases as well as the average time on early retirement the year after transplantation compared with dialysis. Labor income is substantially affected although mainly mediated through employment probability. Conditioned on being employed, transplantation is associated with a doubling of average labor income compared with dialysis. This could be mediated through increased work hours and/or a wage effect by gaining access to more advanced employment possibilities but is unlikely to be due to an increased compensation for the same work. Each transplantation is on average associated with a reduction in labor income losses (lost productivity) the year after treatment of €6600. Over 5 years after transplantation, the discounted reduction is around €33 400 which roughly corresponds to the cost of a transplantation.20

The results of the current study can be used directly in health economic evaluations to estimate the productivity gain of interventions/policies that are effective in increasing the number of performed kidney transplantation. However, the appropriateness of using these results in economic evaluations depends on whether the results can be generalized to other settings. Being based on the total population of working-age patients in RRT, the results are valid for the whole of Sweden and may be generalizable to comparable countries, at least in relative terms. The results of the current study indicate that comparable countries can expect 64% higher employment rate after kidney transplantation compared with dialysis. However, a number of factors are expected to influence the employment rate after transplantation (see Background) and thereby hamper transfer of the results between different contexts. For example, type of insurance (system) has been shown to be associated with transplant outcomes.11,21,22 To investigate what particular factors and characteristics of different insurance systems encourage return to work is left to future research. However, insurance systems that are flexible and where patients do not perceive a risk of losing their chronic disease status if they return to work (for example, by being allowed to work and collect benefits at the same time) is expected to better promote return to work compared with inflexible insurance systems that risk create disability traps. Compared with the Swedish system with universal healthcare and social insurance with negligible private insurance coverage, countries with insurance systems that are less well suited to accommodate the particular needs of RTR are expected to have a lower effect on labor market outcomes, whereas countries with better suited systems are expected to have an even larger effect.

The current study shows that the beneficial effect of kidney transplantation is not due to selection of the most appropriate transplantation candidates, but that the positive effect can be expected also when expanding the treatment. Given the substantial positive effect on labor market outcome and the reduction in healthcare costs associated with kidney transplantation compared to dialysis,20 interventions can potentially be rather costly and still be cost-effective. For example, increased resources to the treatment of critical ill patients are expected to improve health outcomes but is associated with very high cost (for example, in terms of an increase in number of beds in intensive care units). If the increased treatment turn out not to be beneficial to the patient, an increased number of organs could be retrieved, a side effect that could be expected to offset part of the increased cost.

Prior studies have found that the majority (71-100%) that return to work after a kidney transplantation do so within 1 year,1,8,23 but that the employment rate can continue to increase in the following years.6 This is consistent with the findings in the current study. It is interesting to note that the treatment effect of transplantation compared to dialysis increases over time after transplantation, mainly due to worsening outcome in dialysis. It is therefore important to consider additional years posttreatment when discussing the effect of transplantation.

Several factors besides treatment are associated with labor market outcomes in the sample. Education and being employed the year before RRT are associated with increased probability of employment posttreatment for both treatments, the latter being the strongest predictor. However, disposable income, age, and sex are only relevant for employment probability after transplantation and not dialysis. It has been speculated that higher education might lead to increased employment possibilities after transplantation.10 Working the year before RRT could likewise increase employment possibilities through strong labor market attachment. It is also possible that working the year before RRT indicates better general health status which can remain posttreatment. Indeed, it has been shown that renal failure affects labor market outcomes before start of dialysis6 which is a cause for concern in the current study, especially as a reliable measure of general health status is lacking. Likely, employment before treatment captures both labor market attachment and health status, and it is left for future studies to disentangle these effects.

As shown in Materials and Methods, SDC (, the current inverse probability weighting approach is successful in balancing the 2 treatment groups in terms of covariates and thereby adjusting for treatment selection making the groups comparable to reduce estimation bias. However, there is some concern regarding the extent the 2 groups overlap, especially whether the estimation of the effect of receiving transplantation for patients on dialysis with a low probability of transplantation can be correctly interpreted (Materials and Methods, SDC, We therefore also estimated the ATE of the treated, that is, those that actually get transplantation, as well as the ATE for a limited sample consisting only of those that at any time were on the waiting list for transplantation. These approaches are less likely to violate the overlap assumption (Materials and Methods, SDC,, but introduce other complications. The ATE of the treated shows the effectiveness of the performed transplantations but cannot answer what the effect will be from an increase in the supply of kidney available for transplantation. Likewise, the approach of only including patients on the waiting list will only show the effectiveness using the current patient selection rules that, to some extent, are influence by the scarcity of organs. In addition, most working aged patients on the waiting list do receive a transplant during the study period, resulting in a low sample size in the dialysis group (25 % of baseline). We are therefore using these 2 additional approaches mainly to identify any potential problems with overlap in the baseline estimations. However, interpretation should be cautious as large variations between estimates could be due to either overlap issues or the low sample size in the waiting list sample.

As expected, the ATE of the treated is somewhat higher than the baseline estimate, and these patients would do relatively better in both treatments (results not shown). However, the effects are similar compared with baseline and the stability of the result over different specifications indicate that the overlap assumption holds for the baseline estimates. This conclusion is further strengthened by the results based on the waiting list sample for employment, work force participation, and labor income showing similar results. The estimates related to early retirement based on the waiting list sample, however, differs substantially compared with the baseline estimates and should therefore be interpreted with caution. All in all, these estimates are a strong confirmation that the IPWRA approach has been effective in reducing selection bias.

Controlling for all factors influencing the treatment selection would allow for causal interpretation of the estimated treatment effects. Unfortunately, there is no way of knowing to what extent this has been achieved. Indeed, prior research have suggested adherence to be related to the likelihood of receiving a kidney transplantation,24 information that we, unfortunately, lack. Other factors, such as ethnicity and area of residence influences treatment choice,24,25 but had no influence in this Swedish sample. The results should therefore be interpreted as estimates that reduce selection bias. However, the factors shown to be associated with an increased chance of transplantation are also, in general, considered to be associated with improved labor market outcomes. This means that the reduced bias estimates can be interpreted as an upper bound of the causal ATE of kidney transplantation compared with dialysis.

The strengths of the current study are that (i) a number of known factors that affect the treatment selection are controlled for when estimating the effect of transplantation (ii) compared with the relevant alternative treatment (dialysis). Comparing to the general population, as often done in prior studies, tells us very little about what would happen in a policy relevant scenario, such as an increase or decrease in transplantation rate. This is not the case in the current study because the average outcomes of both treatments are estimated. Indeed, the individual's probability of a specific outcome can be calculated using the estimated coefficients in the outcome models. In addition, the data material used is of very high quality, being a population sample with comprehensive register information pretreatment and posttreatment.

In conclusion, kidney transplantation is a preferred treatment choice compared to dialysis and rightly should be. The potential to return to work is substantially higher with transplantation after controlling for treatment selection, and the same positive effect will thus be evident if the treatment were to be expanded to patients currently on dialysis. Also, in terms of labor income and early retirement, transplantation is superior to dialysis. The advantageous effect of transplantation increases over time. Future research should quantify the positive effects in an economic evaluation to establish the value of increasing the supply of donated kidneys.

Back to Top | Article Outline


This study was conducted on data from, and with the (nonfinancial) support of, the Swedish Renal Registry which is gratefully acknowledged. The authors also thank several anonymous reviewers for informed and helpful comments.

Back to Top | Article Outline


1. Nour N, Heck CS, Ross H. Factors related to participation in paid work after organ transplantation: perceptions of kidney transplant recipients. J Occup Rehabil. 2015;25:38–51.
2. Ostrowski M, Wesołowski T, Makar D, et al. Changes in patients’ quality of life after renal transplantation. Transplant Proc. 2000;32:1371–1374.
3. Böhlke M, Marini SS, Gomes RH, et al. Predictors of employment after successful kidney transplantation—a population-based study. Clin Transplant. 2008;22:405–410.
4. De Baere C, Delva D, Kloeck A, et al. Return to work and social participation: does type of organ transplantation matter? Transplantation. 2010;89:1009–1015.
5. Böhlke M. Dialysis and kidney transplantation: why have our rehabilitation hopes not been achieved fully? Am J Kidney Dis. 2012;59:598–600.
6. van der Mei SF, Kuiper D, Groothoff JW, et al. Long-term health and work outcomes of renal transplantation and patterns of work status during the end-stage renal disease trajectory. J Occup Rehabil. 2011;21:325–334.
7. Eng M, Zhang J, Cambon A, et al. Employment outcomes following successful renal transplantation. Clin Transplant. 2012;26:242–246.
8. Eppenberger L, Hirt-Minkowski P, Dickenmann M. Back to work? Socioeconomic status after kidney transplantation. Swiss Med Wkly. 2015;145:w14169.
9. Helanterä I, Haapio M, Koskinen P, et al. Employment of patients receiving maintenance dialysis and after kidney transplant: a cross-sectional study from Finland. Am J Kidney Dis. 2012;59:700–706.
10. Chisholm-Burns MA, Erickson SR, Spivey CA, et al. Health-related quality of life and employment among renal transplant recipients. Clin Transplant. 2012;26:411–417.
11. Sangalli V, Dukes J, Doppalapudi SB, et al. Work ability and labor supply after kidney transplantation. Am J Nephrol. 2014;40(4):353–361.
12. Ferrario A, Verga FC, Piolatto PG, et al. Return to work after organ transplantation: a cross-sectional study on working ability evaluation and employment status. Transplant Proc. 2014;46:3273–3277.
13. Schon S, Ekberg H, Wikström B, et al. Renal replacement therapy in Sweden. Scand J Urol Nephrol. 2004;38:332–339.
14. Dovärn A, Tidemalm D, Wahlström E, et al. Rapporteringen till nationella kvalitetsregister och hälsodataregistren—Jämförelser av täckningsgrader 2014 [Report to national quality registers and health data registers—comparison of coverage rates 2014]. Socialstyrelsen: Stockholm; 2014.
15. Garrido MM, Kelley AS, Paris J, et al. Methods for constructing and assessing propensity scores. Health Serv Res. 2014;49:1701–1720.
16. Wooldridge J. Econometric analysis of cross section and panel data. Cambridge: Massachusetts; 2002.
17. StatCorp. Stata treatment effects reference manual: Potential outcomes/counterfactual outcomes. College Station, Texas: Stata Press; 2015.
18. Ilori TO, Enofe N, Oommen A, et al. Factors affecting willingness to receive a kidney transplant among minority patients at an urban safety-net hospital: a cross-sectional survey. BMC Nephrol. 2015;16:191.
19. Ekonomifakta. Sysselsättningsgrad [Employment rate] Stockholm: Ekonomifakta; 2015 Available from:
20. Jarl JP, Desatnik P, Peetz-Hansson U, et al. Do kidney transplantations save money? A study using a before-after approach and linked multiple register data from Sweden. Clin Kidney J. 2017 DOI: 10.1093/ckj/sfx088.
21. Tumin D, Foraker RE, Smith S, et al. Health insurance trajectories and long-term survival after heart transplantation. Circ Cardiovasc Qual Outcomes. 2016;9:576–584.
22. Pagani FD. Insurance coverage and heart transplant outcomes. Circ Cardiovasc Qual Outcomes. 2016;9:501–503.
23. Matas AJ, Lawson W, McHugh L, et al. Employment patterns after successful kidney transplantation. Transplantation. 1996;61:729–733.
24. Hod T, Goldfarb-Rumyantzev AS. The role of disparities and socioeconomic factors in access to kidney transplantation and its outcome. Ren Fail. 2014;36:1193–1199.
25. Crews DC, Liu Y, Boulware LE. Disparities in the burden, outcomes and care of chronic kidney disease. Curr Opin Nephrol Hypertens. 2014;23:298–305.

Supplemental Digital Content

Back to Top | Article Outline
Copyright © 2018 Wolters Kluwer Health, Inc. All rights reserved.