Sensitivity analyses were conducted estimating the association between income 5 years before dialysis and the average income over 5 years before dialysis compared with 1 year before dialysis in the main estimates (models 1-4, supplemental digital content [SDC], Table S1, http://links.lww.com/TXD/A60). We also ran the full models for men and women separately to investigate whether the effects of income and education differ by sex. Moreover, we reran our analysis excluding patients with negative as well as very low disposable income (cutoff, < SEK 70 000) to study if potential misclassification of income threatens to affect the results.
Statistical significance was assumed for P values less than 0.05. All statistical analyses were performed using STATA software, version 14.0 (Stata Corporation, College Station, TX). The study has been approved by Lund Regional Ethical Review Board (Dnr: 2014/144).
The final sample included 13 982 adult patients on dialysis, 2694 (19.3%) of whom were placed on the waitlist. Among patients on the waitlist, 2164 (80.3%) received a KTx during the study period. Patient characteristics are shown in Table 1. The mean age at starting dialysis was 63.7 years (SD, 15.1), 65.6% was males. According to the univariate analysis, patients on the waitlist were younger, more educated, had higher income, fewer comorbidities, and were more likely to live closer to a transplantation centre compared with patients not on the waitlist (ie, dialysis patients) (P < 0.001). However, there was no sex difference between patients on the waitlist and not on the waitlist (P = 0.48).
Effect of SES on Access to the Kidney Waitlist
Table 2 shows the association between income and access to the waitlist in 4 Cox models. The results from model 1 found a U-shaped relationship between income and access to the waitlist while also showing that patients in quintile 5 had 1.66 times higher likelihood of accessing the waitlist compared with patients in quintile 1 (reference group). The effect of high income decreased substantially when simultaneously adjusting for education (model 2) although remaining positive and significant. Adjusting for demographic variables (model 3) dramatically increased the effect of income while further adjusting for clinical factors (model 4) did not influence the effect of income. The full model (model 4) showed a clear positive association between income and access to the waitlist, as well as removing the U-shaped relationship.
Table 3 shows the association between education and access to the waitlist in 3 Cox models. The results from model 1 found that patients with higher education had more than 3 times the likelihood of placement on the waitlist compared with patients with mandatory education. Although adjusting for other covariates reduced the effect, education was still significantly positively associated with the likelihood of placement on the waitlist in the fully adjusted model (model 3).
Younger age, being married, and having Swedish citizenship was found to increase the likelihood of placement on the waitlist. In the income model, neither sex nor living in the county where the transplantation centre was located was found to have any effect although a small positive effect of male sex was noted in the education model.
Effect of SES on Access to KTx Conditional on Waitlist Placement
Table 4 shows the association between income and access to KTx for waitlisted patients in 4 Cox models. The results from model 1 found that patients in quintile 5 had 1.27 times higher likelihood of receiving KTx compared with patients in quintile 1 (reference group). The effect of high income decreased marginally when simultaneously adjusting for education (model 2). Adjusting also for demographic variables (model 3) and clinical factors (model 4) slightly increased the effect size and showed a clear significant positive association between income and access to KTx. However, the effect seemed to be mostly isolated to the lowest income quintile.
Table 5 shows the association between education and access to KTx in 3 Cox models. The results from model 1 found that higher education had no effect on access to KTx compared with mandatory education. However, after adjusting for demographic variables (model 2) and clinical factors (model 3), a small, significantly positive association between education and access to KTx could be noted.
A positive effect on the likelihood of receiving KTx given waitlist was found for younger age, being married (versus being widowed), and having A, B, or AB blood type. Sex, living in the county in which the transplantation centre was located, and having Swedish citizenship were not found to have an effect on access to KTx.
Access to the Kidney Waitlist
When using income 5 years before dialysis, the effect of income decreased compared with the main estimates (model 1 in SDC, Table S1, http://links.lww.com/TXD/A60). The effect of high income increased when using the average income over 5 years before dialysis (model 2, SDC, Table S1, http://links.lww.com/TXD/A60). Both changes were relatively small.
Access to KTx Conditional on Waitlisting
When using income 5 years before dialysis, the effect of income decreased and became nonsignificant compared with the main estimates (model 3, SDC, Table S1, http://links.lww.com/TXD/A60). Using the average income over 5 years before dialysis, the effect of high income decreased compared with the main estimates but remained significant (model 4, SDC, Table S1, http://links.lww.com/TXD/A60).
Men and Women Separately
For both men and women, the effects of income and education on access to the waitlist/KTx were similar to the main estimates (SDC, Table S2 and S3, http://links.lww.com/TXD/A60), as were the effects of other factors (eg, citizenship).
Excluding Patients With Negative and Very Low Disposable Income
Excluding patients with negative and very low disposable income (< SEK 70,000) (SDC, Table S4, http://links.lww.com/TXD/A60), did not change the results compared to the main analysis in any meaningful way.
This study indicates that, in Sweden, differences linked to patients' SES exist in the likelihood of being placed on the kidney waitlist and getting a KTx. After multivariate adjustment, patients in the highest income group had a more than 1.7 times and 1.3 times increased chance of access to the waitlist and KTx, respectively, compared with patients in the lowest income group. Patients with higher education had more than 2 times and 1.15 times higher chance of access to the waitlist and KTx, respectively, compared with patients with only mandatory education.
Separate Cox models were constructed, using different groups of covariates to assess their relative contribution on the likelihood to be waitlisted/get a transplant, especially how the effect of income and education changed when adding more covariates. When income was used as the measure of SES, education seemed to be a modifier especially for the difference between the highest and lowest income group in the probability to be waitlist. In particular, the likelihood of listing for the highest income group was reduced, but still statistically significant, after adjustment for education. However, education did not have much effect on the difference between the highest and lowest income group as to access to KTx. Retrospectively, when education was used as a measure of SES, the likelihood of waitlisting for the higher education group decreased, whereas the likelihood of getting a transplant slightly increased when adding covariates gradually.
The socioeconomic gradient was found to be stronger for placement on the waitlist compared with receiving a transplant once on the waitlist. The decision to put a patient on the waitlist is probably more subjective and more vulnerable to inequality because this entails a closer relationship between the treating physician and the patient. The transplantation decision once on the waitlist is more objectively based on medical factors without the patient necessarily meeting the transplantation surgeon.
Moreover, the association between education and access to waitlist is stronger than the association between income and access to waitlist. Because of the more subjective decision to waitlist compared with KTx, this could be explained by education potentially capturing other aspects, such as knowledge and attitudes to disease and treatment, compliance and/or communication skills. Education could therefore be expected to be more related to the likelihood of being waitlisted compared to income, even when income potentially captures aspects of general health (see below).
Income 1 year before dialysis runs the risk of capturing patients’ general health status as a result of the kidney disease as well as their SES status. We therefore conducted sensitivity analyses using both income 5 years and the average income over the 5 years before dialysis. These measures should be less influenced by the patients’ kidney-related health status and should therefore be purer SES measures. The downside of using them is that income levels may have changed over the period and that the income 5 years before dialysis may not be a good indicator of current SES. The results of the sensitivity analysis using average income over 5 years before dialysis were similar but somewhat lower compared with the main analysis on access to KTx given on the waitlist. When income 5 years before dialysis was used as income measurement, the effect of income was further reduced (hazard ratio, 1.15) and only significant on the 5% level. This indicates that income 1 year before dialysis captures aspects of general health, and given a positive effect of health on likelihood of access to KTx, that the estimated association of SES on access to KTx is potentially overestimated. This could also explain why the association between income and KTx is stronger compared with the association between education and KTx, given that education is not affected by current health status. However, to disentangle the health and income (SES) effects, we would need to adjust further for general health, which unfortunately is not available in the current data set.
The data include patients with negative or very low disposable income, who accordingly are categorized in income quintile 1. There are several potential reasons for such low income level despite that the Swedish social insurance systems should guarantee everyone a certain “liveable” income level. The main concern is that these individuals might belong to very rich households that, for example, evade tax or live on savings during a particular year. This potential misclassification could introduce an underestimation of the effect of income. However, excluding patients with a disposable income below SEK 70, 000 resulted only in a very small reduction of the association between income and likelihood of getting a transplant compared with the baseline results. This indicates that the results are not sensitive to potential misclassification of rich individuals as poor due to negative or very low disposable income during a particular year.
We could also show that A, B, and AB blood types were all associated with a higher probability of receiving KTx compared with O blood type, whereas 1 study found the positive association only for A and AB blood types.9 This discrepancy could potentially be explained by the relative proportion of different blood types in different countries.21 In addition, a small positive effect of male sex was shown when estimating the effect of education on access to the waitlist. The possible reason for this may be that disposable income is lower for women than men, which is captured by the sex variable when income is not adjusted for.
The current results are consistent with studies from the United States,5,7 showing that higher education was associated with greater likelihood of being placed on the waitlist and undergoing KTx and the International Dialysis Outcomes and Practice Patterns Study,22 which found that education was not associated with access to KTx when adjusting for income. The current results regarding income are also consistent with studies from the United States,6,10 showing an association between living in lower SES neighborhoods and decreased likelihood to complete steps to KTx. Contrary to this, other studies from France9 and the United Kingdom8 found no effect of SES neighborhoods on likelihood of being waitlisted or receiving KTx. In the current study, though the effect size of SES was lower for getting a transplant compared with being placed on the waitlist, it was still significant. These conflicting results between the current study and prior studies may be due to use of different SES measures: individual-level versus area-level SES (eg, neighborhood deprivation, degree of urbanization,9 and Carstairs score to assess social deprivation8). In addition, differences in healthcare systems between countries might be another possible reason for these conflicting results.
Potential reasons for SES discrepancies in access to KTx lie with both the patients and the healthcare provider. From the healthcare provider side, studies have found bias in identifying potential transplant candidates.23 From the patient side, SES-disadvantaged patients may have more, and more severe, comorbidities and worse adherence compared with SES-advantaged patients.5,24 Lower education is associated with factors, such as smoking, less exercise, and overweight, by themselves relative contraindications to transplantation or factors with impact on comorbidities that are contraindications.25 Differences in knowledge, attitudes to disease and treatment, and preference for transplantation may lead to different treatment choices by SES-disadvantaged patients compared to SES-advantaged patients.8 Hence, SES may have both a direct effect on access to KTx process (eg, through discrimination) and an indirect effect (operating through patients’ preference).26 The SES inequalities shown above are therefore not necessarily due to discrimination. However, they will still contribute to societal inequalities in health and wealth, and as such, it is of interest to mitigate them.
A limitation of this study is that although we controlled for many important confounding factors, we lack information on other unobserved factors (eg, more measures of general health, adherence, race/ethnicity, patients’ preference, and physician bias) and other biochemical data (eg, serum albumin level and other measures of inflammation; levels of parathyroid hormone) which also are associated with access to KTx.9 Additionally, we have information on comorbidities but not about their severity, nor about changes during follow-up.
In addition to the limitations above, there are important strengths. Instead of geographically defined SES, 2 classic individual-level SES indicators, education and income, were used and were expected to better capture the SES and thereby give more accurate effect estimates. In addition, population based national register data with almost 100% coverage and a data reporting incidence of 95%17 gives the study high power and excellent generalizability to ESRD patients in Sweden.
Individual-level low income and education both are associated with a reduced chance of access to the waitlist and KTx among Swedish ESRD patients. However, the factors behind the observed SES inequalities in Swedish Health Care System, which aims to provide good and equal healthcare for all Swedish citizens, are unknown. To this end, further studies are needed to identify the mechanisms behind these inequalities to construct interventions to reduce SES barriers and to assess if these inequalities are unfair inequalities.
This study was based on data and received (nonfinancial) support from the Swedish Renal Registry, which is gratefully acknowledged.
1. Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med
2. Neipp M, Karavul B, Jackobs S, et al. Quality of life in adult transplant recipients more than 15 years after kidney transplantation. Transplantation
3. Eggers P. Comparison of treatment costs between dialysis and transplantation. Semin Nephrol
4. Jarl J, Desatnik P, Peetz Hansson U, et al. Do kidney transplantations save money? A study using a before–after design and multiple register-based data from Sweden. [published online August 7, 2017]. Clin Kidney J
5. Schaeffner ES, Mehta J, Winkelmayer WC. Educational level as a determinant of access to and outcomes after kidney transplantation in the United States. Am J Kidney Dis
6. Schold JD, Gregg JA, Harman JS, et al. Barriers to evaluation and wait listing for kidney transplantation. Clin J Am Soc Nephrol
7. Khattak MW, Sandhu GS, Woodward R, et al. Association of marital status with access to renal transplantation. Am J Transplant
8. Oniscu GC, Schalkwijk AA, Johnson RJ, et al. Equity of access to renal transplant waiting list and renal transplantation in Scotland: cohort study. BMJ
9. Kihal-Talantikite W, Vigneau C, Deguen S, et al. Influence of socio-economic inequalities on access to renal transplantation and survival of patients with end-stage renal disease. PLoS One
10. Saunders MR, Cagney KA, Ross LF, et al. Neighborhood poverty, racial composition and renal transplant waitlist. Am J Transplant
11. Qureshi AR, Evans M, Stendahl M, et al. The increase in renal replacement therapy (RRT) incidence has come to an end in Sweden-analysis of variations by region over the period 1991-2010. Clin Kidney J
12. Zhang Y, Jarl J, Gerdtham UG. Are there inequities in treatment of end-stage renal disease in Sweden? A longitudinal register-based study on socioeconomic status-related access to kidney transplantation. Int J Environ Res Public Health
17. Schon S, Ekberg H, Wikstrom B, et al. Renal replacement therapy in Sweden. Scand J Urol Nephrol
21. Christian Baudelot YC, Olivier Godechot, Sylvie Mercier. Renal diseases and social inequalities in access to transplantation in France. Population (Engl Ed)
22. Satayathum S, Pisoni RL, McCullough KP, et al. Kidney transplantation and wait-listing rates from the international Dialysis Outcomes and Practice Patterns Study (DOPPS). Kidney Int
23. Kasiske BL, London W, Ellison MD. Race and socioeconomic factors influencing early placement on the kidney transplant waiting list. J Am Soc Nephrol
24. Kasiske BL, Ramos EL, Gaston RS, et al. The evaluation of renal transplant candidates: clinical practice guidelines. Patient Care and Education Committee of the American Society of Transplant Physicians. J Am Soc Nephrol
25. Cowell AJ. The relationship between education and health behavior: some empirical evidence. Health Econ
26. Moscelli G, Siciliani L, Gutacker N, et al. Socioeconomic inequality of access to healthcare: does choice explain the gradient? [published online June 23 2017]. J Health Econ
. 2017; doi: 10.1016/j.jhealeco.2017.06.005.
Supplemental Digital Content
© 2018 The Authors. Published by Wolters Kluwer Health, Inc.