Trends in the incidence, demographics, and outcomes of end-stage renal disease due to lupus nephritis in the US from 1995 to 2006


  • Data for these analyses were provided by the US Renal Data System (USRDS), but the analysis and conclusions are those of the authors and do not represent USRDS or the National Institute of Diabetes and Digestive and Kidney Diseases, NIH.



This study was undertaken to investigate whether recent advances in lupus nephritis treatment have led to changes in the incidence of end-stage renal disease (ESRD) secondary to lupus nephritis, or in the characteristics, treatments, and outcomes of patients with lupus nephritis ESRD.


Patients with incident lupus nephritis ESRD (1995–2006) were identified in the US Renal Data System. Trends in sociodemographic and clinical characteristics were assessed. We tested for temporal changes in standardized incidence rates (SIRs) for sociodemographic groups using Poisson regression. Changes in rates of waitlisting for kidney transplant, kidney transplantation, and all-cause mortality were examined using crude and adjusted time-to-event analyses.


We identified 12,344 incident cases of lupus nephritis ESRD. Mean age at ESRD onset was 41 years; 81.6% of the patients were women and 49.5% were African American. SIRs for lupus nephritis ESRD among those who were ages 5–39 years, African American, or lived in the southeastern US increased significantly from 1995 to 2006. Increases in body mass index and in the prevalence of both diabetes mellitus and hypertension were detected. Mean serum hemoglobin level at ESRD onset increased, while that of serum creatinine decreased over time. More patients received hemodialysis and fewer received peritoneal dialysis. There was a slight increase in the frequency of preemptive kidney transplantation at ESRD onset, but kidney transplantation rates within the first 3 years of ESRD declined. Mortality did not change over the 12 years of study.


Our findings indicate that the characteristics of patients with lupus nephritis ESRD and initial therapies have changed in recent years. While SIRs rose in younger patients, among African Americans, and in the South, outcomes did not improve in over a decade of evaluation.

Up to 60% of adults and 80% of children with systemic lupus erythematosus (SLE) develop nephritis (1, 2). In 10–30% of these patients, disease progresses to end-stage renal disease (ESRD) within 15 years of diagnosis, even with aggressive treatment (3–5). Renal damage is the most important predictor of mortality for patients with SLE, and 5-year survival is significantly worse among individuals with lupus nephritis than among those without nephritis (5–8).

Advances in the treatment of lupus nephritis have been made within the past 2 decades, with the identification of effective immunosuppressive treatments, including cyclophosphamide, azathioprine, and mycophenolate mofetil (9–13). The incidence of ESRD due to lupus nephritis, however, rose from 1.13 cases per million in 1982 to 3.20 cases per million in 1995 (14), and showed no reduction between 1996 and 2004 (15). Advances in the treatment of ESRD have also been made during these years. Kidney transplantation is now accepted as optimal therapy for the long-term treatment of most patients with ESRD (16, 17). Living kidney transplantation has been demonstrated to offer all ESRD patients the best chance of dialysis-free survival (18). Early evaluation for kidney transplantation and placement on the waiting list for allografts also has proven benefits, as rates of graft failure and mortality increase with dialysis wait times (19, 20). For patients with lupus nephritis, as for other causes of ESRD, kidney transplantation has been associated with decreased mortality and improved quality of life (17, 21).

We examined recent trends in the incidence and outcomes of ESRD due to lupus nephritis in the entire US population and in different sociodemographic groups, defined by age, sex, race, ethnicity, medical insurance type, and geographic region of the country, between 1995 and 2006. The goals of these analyses were to investigate changes during these years in the populations affected by lupus ESRD, their clinical characteristics, treatments, and rates of kidney transplantation and survival.


Data sources.

The US Renal Data System (USRDS) is the US national registry of patients with ESRD (22). The USRDS database includes ∼94% of the patients in the US who receive renal replacement therapy in the form of dialysis or kidney transplantation, and in 2006 included information on 1.6 million individuals with ESRD since 1988 (22). For each new patient, the attending nephrologist is required to complete the Medical Evidence Report form (CMS-2728) at enrollment. It serves to establish Medicare eligibility for those patients who were not previously Medicare eligible, reclassify previously eligible Medicare beneficiaries as ESRD patients, and provide demographic and diagnostic information, including the etiology of ESRD, on all new ESRD patients according to International Classification of Diseases, Ninth Revision (ICD-9) codes. The date of first service is derived from the earliest of dialysis start dates reported on the medical evidence form, the date of kidney transplant as reported on a Centers for Medicare and Medicaid Services (CMS) or Organ Procurement Transplant Network form, the Medical Evidence Report, a hospital inpatient claim, or the date of the first Medicare dialysis claim. Patients who received transient dialysis for acute renal failure, died before being enrolled in the database, or refused renal replacement therapy might not be included. Completion of the CMS ESRD Death Notification Form (CMS-2746) is also mandatory and enforced by CMS. Providers have 45 days to return the completed form to the ESRD networks and USRDS Coordinating Center (22).

From the US Census Bureau (23), we obtained age-, sex-, race-, ethnicity- and state-specific annual population estimates (from the 2000 US Census, as well as annual intercensal population estimates for July of each year). Estimates of the Hispanic and non-Hispanic US population were available only with the 2000 US Census and thereafter.

Study population.

We identified all individuals ages 5–100 years with SLE (ICD-9 code 710.0) indicated as the cause of ESRD at enrollment in the USRDS from January 1, 1995 to December 31, 2006. From the USRDS, we obtained information concerning patient demographics, including age, sex, race (white, African American, Asian/Pacific Islander, or American Indian), Hispanic ethnicity, and US state or territory of residence at the time of initiation of ESRD treatment. The following data for each patient at ESRD onset were extracted: body mass index (BMI); serum levels of creatinine, hemoglobin, and albumin; use of an erythropoiesis-stimulating agent; concomitant comorbidities (diabetes mellitus, hypertension, coronary artery disease, congestive heart failure, chronic obstructive pulmonary disease, peripheral vascular disease, cerebrovascular disease, inability to ambulate, and inability to transfer from a wheelchair); type of medical insurance prior to ESRD (Medicare, Medicaid, employer group, other, or none); current employment status; and type of initial renal replacement (hemodialysis, peritoneal dialysis, or preemptive kidney transplantation). Updated USRDS data concerning dates of kidney transplant waitlisting, kidney transplantation, and death were extracted. Patients with missing data on age or sex were excluded from all analyses.

Statistical analysis.

A range of clinical and demographic characteristics were examined for lupus nephritis patients with incident ESRD, in 3-year intervals. Two-sided Cochran-Armitage tests for categorical variables and general linear models for continuous variables were used to detect significant trends over time.

To account for changes in the underlying population composition of the US, we calculated standardized incidence rates (SIRs; per million individuals per year), using age-, sex-, race-, ethnicity- and geography-specific population estimates for July of each year obtained from the US Census, averaged in 3-year intervals. Poisson regression models were used to test whether there were significant changes in SIRs, and linear trends of SIRs over time were assessed in the models. The 95% confidence intervals (95% CIs) for lupus nephritis ESRD SIRs were calculated. We investigated changes in incidence according to geographic region, classifying the US states according to Federal Information Processing Standards codes into 4 geographic regions (Northeast, Midwest, South, and West), as designated by the US Census (24). (Data from US islands and territories, including Puerto Rico, the US Virgin Islands, American Samoa, and Guam, were excluded from SIR calculations since we were unable to find intercensal population estimates for each year.)

Incident lupus nephritis ESRD cases were stratified by year of ESRD onset into four 3-year periods, and we examined trends over time in the following 3 outcomes: kidney transplant waitlisting, kidney transplantation, and overall survival. Kaplan-Meier survival analyses and log rank tests were used to test for differences in the rates of these 3 outcomes during the first 3 years after ESRD onset. The referent group for each pairwise comparison was composed of individuals with ESRD onset in 1995–1997. We also calculated the proportion of individuals who had each outcome in each time period with 95% CIs by the exact method. (These proportions could not be calculated for the final period since fewer than 3 years of followup were captured in the available data.)

Age-adjusted and multivariable-adjusted Cox proportional hazards models were developed to account for factors contributing to variation in successive time periods in rates of the 3 outcomes within 3 years of ESRD onset. Again, onset of incident ESRD within the years 1995–1997 was the comparison group for all subsequent 3-year periods. Multivariable Cox models were based upon the results of univariable trend and Kaplan-Meier survival analyses, additionally adjusting for potential changes over time in the distributions of sociodemographic and clinical factors in incident patients. The final multivariable model included age, sex, race, ethnicity, medical insurance type, and geographic region, as well as the presence of diabetes mellitus, hypertension, smoking, and the type of initial renal replacement therapy. Adjustment for other clinical factors, including other comorbidities, serum laboratory values, and use of erythropoiesis-stimulating agents, did not influence hazard ratios and thus they were not included. Tests for proportionality of hazards over time were performed using interaction terms between the covariates and time, for each of the 3 outcomes of interest. (The final 3-year block 2004–2006 was not included in Cox proportional hazards analyses since the proportional hazards assumption did not hold for this incidence stratum with limited followup.) Tests for significant trends in the hazard ratios for these outcomes over successive 3-year blocks were assessed in separate models.

Individuals who received a kidney transplant before initiation of dialysis were excluded from waitlisting and kidney transplantation analyses, as were those who had been placed on the waiting list >90 days prior to enrollment in the USRDS. However, we did include individuals who were placed on the waiting list within up to 90 days prior to the date of first service and imputed their date of waitlisting as day 1. Followup was censored at 3 years (or on December 31, 2006) for all subjects. Sensitivity analyses were performed in which all of the above survival analyses were repeated with the ESRD onset date delayed 90 days to assess for potential instability in ESRD reporting during that initial time. SAS, version 9.2 was used for all analyses. Data were obtained from the USRDS through a data use agreement, and data are shown in accordance with their policy. The Partners' Healthcare Institutional Review Board reviewed this study protocol and granted it a waiver as non–human subjects research since no identifiable data were used.


We identified a total of 12,344 incident cases of ESRD due to lupus nephritis in the US from 1995 through 2006. Changes in the sociodemographic characteristics of individuals with incident ESRD due to lupus nephritis over this time period are shown in Table 1. Mean age at onset was relatively stable at ∼40 years, as was the proportion of women, representing slightly more than 80% of all new patients. We did not observe any changes in the distribution of incident cases according to medical insurance type at ESRD onset. The proportion of individuals who were reported to be currently employed at the onset of ESRD due to lupus nephritis, however, rose over these years. The mean BMI of patients with ESRD due to lupus nephritis increased over time, while serum creatinine concentrations at the initiation of renal replacement therapy decreased. Serum albumin levels decreased slightly among patients with new-onset ESRD due to lupus nephritis during these years. Serum hemoglobin concentrations at onset of ESRD, in contrast, rose steadily during this time, with a corresponding increase in the use of erythropoiesis-stimulating agents prior to initiation of dialysis. The prevalence of both concomitant diabetes mellitus and hypertension among incident ESRD patients increased steadily during these years, while the proportion of patients reported to have coronary artery disease and the proportion of patients who were current smokers remained unchanged. The proportions of patients who were reported to be unable to ambulate and unable to transfer from a wheelchair at ESRD onset were small, but did increase over these years.

Table 1. Baseline demographic and clinical characteristics of the 12,344 patients with ESRD due to lupus nephritis*
 Year of ESRD onsetP for trend
  • *

    Except where indicated otherwise, values are the mean ± SD.

  • By Cochran-Armitage 2-sided test for trend for categorical variables and by general linear models for continuous variables.

  • Data on type of medical insurance were not available for 174 patients.

  • §

    Data on body mass index (BMI) were not available for 445 patients.

  • Data on creatinine level were not available for 199 patients.

  • #

    Data on hemoglobin level were not available for 1,424 patients.

  • **

    Data on albumin level were not available for 2,944 patients.

  • ††

    Data on erythropoietin use were not available for 267 patients.

  • ‡‡

    Data on initial end-stage renal disease (ESRD) treatment were not available for 608 patients.

  • §§

    Of the 12,027 patients who did not receive a preemptive transplant.

No. (%) new cases2,600 (21.1)3,115 (25.23)3,232 (26.2)3,397 (27.5)
Age, years40.5 ± 15.240.6 ± 15.440.6 ± 15.640.1 ± 15.30.25
No. (%) women2,131 (82.0)2,552 (81.9)2,637 (81.6)2,750 (81.0)0.29
Type of medical insurance, no. (%)     
 Medicare659 (26.5)795 (25.8)802 (24.9)848 (25.1)0.16
 Medicaid571 (23.0)715 (23.2)729 (22.6)844 (25.0)0.10
 Private1,012 (40.7)1,263 (41.0)1,331 (41.3)1,351 (40.0)0.56
 Uninsured243 (9.8)307 (10.0)361 (11.2)339 (10.0)0.49
Employed at ESRD onset, no. (%)549 (21.1)757 (24.3)793 (24.5)889 (26.2)<0.0001
BMI, kg/m2§23.5 ± 7.225.0 ± 7.226.0 ± 6.926.5 ± 7.1<0.0001
Creatinine, mg/dl8.6 ± 3.58.0 ± 3.57.6 ± 3.57.1 ± 3.4<0.0001
Hemoglobin, gm/dl#8.9 ± 1.89.2 ± 1.99.5 ± 1.89.6 ± 1.8<0.0001
Albumin, gm/dl**3.0 ± 0.82.9 ± 0.82.9 ± 0.82.9 ± 0.80.03
Erythropoietin use at ESRD onset, no. (%)††688 (28.5)983 (32.2)1,184 (36.7)1,221 (35.9)<0.0001
Comorbid conditions, no. (%)     
 Diabetes mellitus139 (5.4)199 (6.4)225 (7.0)308 (9.1)<0.0001
 Hypertension1,724 (66.3)2,244 (72.0)2,435 (75.3)2,735 (80.5)<0.0001
 Coronary artery disease132 (5.1)175 (5.6)201 (6.2)191 (5.6)0.28
 Congestive heart failure439 (16.9)457 (14.7)506 (15.7)476 (14.0)0.01
 Chronic obstructive pulmonary disease59 (2.3)75 (2.4)71 (2.2)66 (1.9)0.29
 Peripheral vascular disease75 (2.9)102 (3.3)115 (3.6)108 (3.2)0.48
 Cerebrovascular disease122 (4.7)164 (5.3)170 (5.3)169 (5.0)0.71
 History of malignancy32 (1.2)39 (1.3)58 (1.8)51 (1.5)0.18
 Current smoking101 (3.9)108 (3.5)128 (4.1)122 (3.6)0.82
 Current alcohol abuse16 (0.6)10 (0.3)18 (0.5)12 (0.4)0.76
 Current drug abuse14 (0.5)19 (0.6)23 (0.7)27 (0.8)0.19
 Unable to ambulate51 (2.0)61 (2.0)68 (2.1)90 (2.7)0.05
 Unable to transfer from a wheelchair15 (0.6)23 (0.7)29 (0.9)39 (1.2)0.01
Initial ESRD treatment, %‡‡     
 Hemodialysis1,974 (75.9)2,482 (79.7)2,625 (81.2)2,850 (83.9)<0.001
 Peritoneal dialysis426 (16.4)370 (11.9)355 (11.0)337 (9.9)<0.001
 Preemptive transplant42 (1.6)74 (2.4)84 (2.6)117 (3.4)<0.001
 On waiting list for renal transplant at ESRD onset date§§37 (1.5)42 (1.4)50 (1.6)72 (2.2)<0.001

Hemodialysis was the initial type of dialysis offered to most patients during these years, but a marked decline in the proportion of patients initially begun on peritoneal dialysis was observed (from 16.4% to 9.9%). Three hundred seventeen patients (2.6%) underwent elective kidney transplantation as their initial form of ESRD treatment. This small percentage did grow over the years of the study, as did the small numbers of those who were already on the waiting list for kidney transplant at the time of ESRD onset.

Relative to the population at risk, we found a significant increase in the SIRs of lupus ESRD among younger patients (ages 5–19 and 20–39 years), whereas the SIRs among patients age ≥40 years did not change (Table 2). The SIRs also increased among African Americans and American Indians. African Americans had an SIR of 6–7 times that of white subjects and represented 49% of all incident cases. The absolute incidence rate of ESRD due to lupus nephritis has been higher among African American subjects than among white subjects in every year since 1997 (data not shown). There has also been a significant increase in new cases occurring in the southern US.

Table 2. SIRs for ESRD due to lupus nephritis in the US, 1995–2006*
 No. of patients (n = 12,198)Year of ESRD onsetP for trend
  • *

    Values are the standardized incidence rate (SIR; per million population per year) (95% confidence interval) determined with Poisson regression models using the US Census population estimates for each particular demographic group in July of the indicated years. ESRD = end-stage renal disease.

  • By Poisson regression.

  • Data on race were not available for 174 patients.

  • §

    SIRs for the Hispanic population were calculated for the year 2000 and later, since estimates of the US Hispanic population first became available with the 2000 US Census.

  • Intercensal population estimates were not available for US island territories; therefore, these individuals (n = 146) were excluded from this analysis.

Age at ESRD onset, years      
 5–198100.84 (0.66–1.06)1.05 (0.85–1.28)1.26 (1.04–1.51)1.35 (1.13–1.62)<0.0001
 20–395,6485.06 (4.67–5.48)5.68 (5.19–6.44)5.86 (6.18–7.21)6.31 (5.87–6.78)0.005
 40–594,3094.55 (4.13–5.00)5.08 (4.66–5.53)5.00 (4.61–5.43)4.88 (4.50–5.30)0.38
 ≥601,4312.35 (2.00–2.76)2.66 (2.29–3.09)2.75 (2.38–3.17)2.53 (2.19–2.93)0.45
 Female9,9515.54 (5.19–5.92)6.40 (6.03–6.80)6.42 (6.05–6.81)6.52 (6.15–6.91)0.007
 Male2,2471.28 (1.11–1.48)1.48 (1.30–1.68)1.51 (1.33–1.71)1.59 (1.41–1.79)0.007
 White5,3441.96 (1.83–2.10)2.15 (2.01–2.30)2.12 (1.99–2.27)2.11 (1.98–2.25)0.22
 African American5,92312.80 (11.92–13.74)15.04 (14.12–16.03)15.24 (14.33–16.21)15.55 (14.64–16.51)0.008
 Asian6294.51 (3.62–5.62)4.74 (3.88–5.80)4.39 (3.61–5.35)5.52 (4.67–6.52)0.09
 American Indian1282.46 (1.36–4.46)3.99 (2.57–6.20)4.16 (2.75–6.30)5.10 (3.53–7.35)0.002
Hispanic ethnicity§1,9185.55 (4.57–6.74)5.21 (4.66–5.83)5.47 (4.92–6.08)0.84
Geographic region      
 Northeast2,0172.98 (2.63–3.39)3.46 (3.08–3.88)3.51 (3.14–3.94)3.41 (3.04–3.83)0.15
 Midwest2,3192.69 (2.38–3.03)3.48 (3.13–3.87)3.32 (2.99–3.70)3.34 (3.00–3.72)0.09
 South5,3234.19 (3.87–4.53)4.71 (4.38–5.07)4.77 (4.44–5.12)5.13 (4.79–5.48)0.002
 West2,5393.39 (3.03–3.79)3.61 (3.25–4.01)3.73 (3.38–4.13)3.56 (3.22–3.94)0.21

Overall, 2,167 lupus ESRD patients underwent kidney transplantation during the first 3 years of ESRD, 1,298 (59.9%) of which were kidneys from living donors. A substantial increase in the proportion of kidneys being donated by living rather than deceased donors was observed. In the 1995–1997 incidence cohort, approximately half of all transplants were obtained from living donors, which increased to >70% of all kidney transplants in patients who had ESRD in 2004–2006 (P for trend <0.0001). In Kaplan-Meier survival analyses, we found that rates of waitlisting for kidney transplant within 3 years of ESRD onset increased slightly, but rates of kidney transplantation have significantly declined (Table 3). Overall survival did not improve over the 12 years studied.

Table 3. Three-year outcomes among individuals with incident ESRD due to lupus nephritis, 1995–2006*
Year of ESRD onsetWaitlisting for kidney transplant (n = 4,237)PKidney transplantation (n = 2,167)PMortality (n = 4,901)P
  • *

    Values are the percent of patients who had the indicated outcome within 3 years (95% confidence interval [95% CI]), determined by Kaplan-Meier analysis; 95% CIs were determined by exact method. All incident dialysis patients (n = 12,027) were included in the calculations of waitlisting for kidney transplant and kidney transplantation, and all patients with incident end-stage renal disease (ESRD) (including 317 patients who received preemptive kidney transplants) (n = 12,344) were included in the calculation of mortality. All followup was censored on December 31, 2006.

  • By pairwise log rank test.

  • Estimates of 3-year failure rates are not shown due to <3 years of followup.

  • §

    By overall log rank test.

1995–199735.7 (33.8–37.6)Reference21.5 (19.9–23.1)Reference27.6 (25.9–29.3)Reference
1998–200036.2 (34.5–37.9)0.6020.5 (19.1–22.0)0.7027.6 (26.0–29.2)0.65
2001–200337.2 (35.5–39.0)0.0918.0 (16.7–19.4)0.00227.1 (25.6–28.7)0.79
P for 1995–2006§0.11 0.0008 0.35 

Age-adjusted and multivariable-adjusted Cox proportional hazards models for the 3 outcomes are shown in Table 4. The results of these models were similar to those of unadjusted Kaplan-Meier analyses. Compared to patients with ESRD onset in 1995–1997, there was an indication of slightly increased waitlisting rates among patients with ESRD onset after 2000. However, there was a significant decrease in the rates of actual kidney transplantation within the first 3 years of ESRD and there was no change in survival rates. The results were similar in age-adjusted and multivariable-adjusted models. In sensitivity analyses in which the ESRD onset date for all survival analyses was delayed by 90 days, results were also unchanged.

Table 4. Cox proportional HRs for outcomes within 3 years of ESRD onset according to incident year, 1995–2003*
Year of ESRD onsetWaitlisting for kidney transplant (n = 3,405)Kidney transplantation (n = 1,859)Mortality (n = 4,390)
Age-adjusted HR (95% CI)Multivariable-adjusted HR (95% CI)Age-adjusted HR (95% CI)Multivariable-adjusted HR (95% CI)Age-adjusted HR (95% CI)Multivariable-adjusted HR (95% CI)
  • *

    All incident dialysis patients (n = 8,830) were included in the calculations of waitlisting for kidney transplant and kidney transplantation, and all patients with incident end-stage renal disease (ESRD) (including patients who received preemptive kidney transplants) (n = 8,947) were included in the calculation of mortality. All followup was censored on December 31, 2006. The multivariable model consisted of age, sex, race (white, African American, Asian, American Indian, or other/missing), ethnicity (Hispanic or non-Hispanic), medical insurance type (Medicaid, Medicare, private, none, or other/missing), region (Northeast, Midwest, South, West, Puerto Rico and US islands, or missing), diabetes mellitus, hypertension, smoking, and initial ESRD therapy (hemodialysis, peritoneal dialysis, other/missing, and preemptive transplant for mortality analyses only). The 95% confidence intervals (95% CIs) were determined by Wald's test. Parameter estimates indicated positive trends for kidney transplant waiting list hazard ratios (HRs) over time intervals but negative trends for kidney transplantation and mortality HRs.

  • By Cox regression analysis.

1995–19971.0 (reference)1.0 (reference)1.0 (reference)1.0 (reference)1.0 (reference)1.0 (reference)
1998–20001.02 (0.94–1.12)1.05 (0.96–1.15)0.97 (0.86–1.10)1.01 (0.89–1.15)1.03 (0.92–1.14)0.94 (0.84–1.07)
2001–20031.07 (0.98–1.17)1.09 (1.00–1.20)0.82 (0.72–0.93)0.87 (0.76–0.99)1.02 (0.91–1.13)0.95 (0.85–1.07)
P for trend0.120.050.0010.030.800.44


Several important changes in the sociodemographics, clinical characteristics, and outcomes of US patients with ESRD due to lupus nephritis have occurred between 1995 and 2006. The SIRs of lupus nephritis ESRD among young patients, African Americans, and patients in the southern US have increased. Although there have been increases in preemptive kidney transplantation at ESRD onset and waitlisting for kidney transplants since 2000, overall rates of kidney transplantation in the first 3 years of ESRD onset have declined. Despite advances in the treatment of lupus nephritis and ESRD in recent years, survival for lupus nephritis patients during the first 3 years of ESRD has not improved.

We found a significant and steady increase in the BMIs of patients, which is likely reflective of increasing obesity in the US population, and not necessarily overall health. We observed an increase in the proportion of concomitant diabetes mellitus and hypertension, and some of these increases could be related to obesity. The increase in BMI is also important given recent findings that predictors of graft failure among lupus nephritis transplant recipients included increasing weight, as well as African American compared with white race, and increasing Charlson comorbidity index (25).

Serum creatinine level in more recent years has been slightly lower at the initiation of ESRD therapy, suggesting earlier start of kidney replacement therapy (i.e., with more preserved kidney function), while hemoglobin level was higher, reflecting an increasing use of injected erythropoiesis-stimulating agents, higher residual kidney function, and perhaps improved management of lupus nephritis. The growing use of erythropoiesis-stimulating agents has been observed for all types of ESRD in the past 15 years (26), but has since declined in response to several trials that raised concerns about the safety of erythropoiesis-stimulating agents in patients with chronic kidney disease, and corresponding changes in drug indications (27). In our multivariable models, use of erythropoiesis-stimulating agents at ESRD onset was not associated with changes in survival over time.

Our findings of increases in the SIRs of ESRD due to lupus nephritis among individuals younger than age 40, among African Americans and American Indians, and in the southern US are slightly different from those of Ward (15), who reported relative stability through 2004. That analysis, however, included only patients ages 15 years and older and did not investigate changes by US region or adjust for annual population changes. We also found a growing absolute incidence of ESRD due to lupus nephritis among African Americans. Unfortunately, recent nationwide data on the incidence of SLE, both overall and in specific racial and ethnic groups, are lacking (28). Thus, it is not known whether the observed changes are due to changes in the incidence of SLE or of lupus nephritis. It is possible that observed increases in SIRs for ESRD due to lupus nephritis, particularly among children, reflect a decrease in early mortality due to SLE.

Another strong trend observed was a decline in the initial use of peritoneal dialysis, already used in a minority in the earlier years of our study. Limited geographic access to peritoneal dialysis services and a national transition to chain-affiliated hemodialysis facilities are thought to be responsible for increasing the proportion of all ESRD patients started on hemodialysis in the US in the past 2 decades (29). Pretransplant use of peritoneal dialysis compared with hemodialysis has been associated with 51% better allograft survival among recipients with lupus nephritis ESRD (25). Further understanding of the initial choice of dialysis modality and associated outcomes in ESRD patients with lupus is becoming increasingly relevant given the relatively lower cost of peritoneal dialysis and the new capitated reimbursement program for dialysis services which starts in 2011 (30, 31).

Our analyses detected only slight recent increases in kidney transplant waitlisting for lupus nephritis ESRD patients, but kidney transplant rates declined. This decrease may be due to changes in the demographics of the lupus nephritis ESRD population, or to the increase in living (versus deceased) donor kidney transplants and a relative shortage of organs for transplant. In an analysis of access to kidney transplantation among adult individuals with lupus nephritis ESRD from 1987 to 1995, younger age, white race, and living in the Midwest were all significant predictors of receiving a living transplant (32). At that time, however, only 11% of patients who received transplants received a living donor kidney, compared with a much higher proportion in recent years.

The limitations of this study include the use of data reported by the attending nephrologist and staff on the USRDS Medical Evidence Report at the onset of ESRD for many baseline variables. Consistent completion of this form is expected, as it is a US government document establishing the onset of ESRD and qualification for Medicare coverage, but missing data do exist for some of the baseline clinical characteristics, including albumin level, hemoglobin level, and BMI, and some underreporting of comorbidities is also known to occur (33). Furthermore, the validity of the cause of ESRD as reported on the Medical Evidence Report has not been studied. In contrast, the outcomes studied, i.e., waitlisting for kidney transplant, transplantation, and death, are well documented in the national databases that serve as the basis for the USRDS data, including the United Network of Organ Sharing. Data regarding lupus disease activity and damage in other organ systems are not available for these patients, limiting our interpretations of the observed changes in standardized incidence, transplantation rates, and mortality rates in different population subgroups. Lastly, patients who received transient dialysis for acute renal failure, who died before being enrolled in the database, or who refused renal replacement therapy may not be included in USRDS data.

Since this study included virtually all new-onset ESRD cases in the US for a span of 12 years, these analyses of changes over time in incidence, therapies, and outcomes in lupus nephritis have important strengths. The slightly growing incidence of ESRD and lack of improvements in important outcomes among patients with ESRD due to lupus nephritis over the past 12 years are disappointing. Despite slightly increased waitlisting for kidney transplant, kidney transplantation rates have fallen and survival has not changed. This may reflect the changing sociodemographics of lupus nephritis with more cases occurring among minority populations, in younger individuals, and in the southern US, where waitlisting and kidney transplantation rates have been lower in recent years (34). It may also reflect limited use or effectiveness of current treatments, nonadherence to therapies, barriers to health care access, or the increasing demand for and shrinking supply of donor organs. Taken together with other recent findings of increased risks of graft failure, recurrent lupus nephritis, and death after kidney transplant (25, 35), the changing sociodemographics and the lack of improvement in survival in ESRD due to lupus nephritis underscore the ongoing challenge of caring for these patients. There is a pressing need to identify the potentially modifiable contributing factors and to intervene to improve both the incidence and the outcomes of ESRD due to lupus nephritis.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Costenbader had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Costenbader, Desai, Brookhart, Massarotti, Lu, Solomon, Winkelmayer.

Acquisition of data. Costenbader, Shaykevich, Brookhart, Lu, Winkelmayer.

Analysis and interpretation of data. Costenbader, Desai, Alarcón, Hiraki, Shaykevich, Brookhart, Massarotti, Lu, Solomon, Winkelmayer.