Original ArticlesTransplantation

Quantifying Donor Effects on Transplant Outcomes Using Kidney Pairs from Deceased Donors

Kathleen F. Kerr, Eric R. Morenz, Heather Thiessen-Philbrook, Steven G. Coca, F. Perry Wilson, Peter P. Reese and Chirag R. Parikh
CJASN December 2019, 14 (12) 1781-1787; DOI: https://doi.org/10.2215/CJN.03810319

Abstract

Background and objectives In kidney transplantation, the relative contribution of donor versus other factors on clinical outcomes is unknown. We sought to quantify overall donor effects on transplant outcomes for kidney donations from deceased donors.

Design, setting, participants, & measurements For paired donations from deceased donors resulting in transplants to different recipients, the magnitude of donor effects can be quantified by examining the excess of concordant outcomes within kidney pairs beyond chance concordance. Using data from the Organ Procurement and Transplantation Network between the years 2013 and 2017, we examined concordance measures for delayed graft function, death-censored 1-year graft failure, and death-censored 3-year graft failure. The concordance measures were excess relative risk, excess absolute risk, and the fixation index (where zero is no concordance and one is perfect concordance). We further examined concordance in strata of kidneys with similar values of the Kidney Donor Profile Index, a common metric of organ quality.

Results If the transplant of the kidney mate resulted in delayed graft function, risk for delayed graft function was 19% higher (95% confidence interval [95% CI], 18% to 20%), or 1.76-fold higher (95% CI, 1.73- to 1.80-fold), than baseline. If a kidney graft failed within 1 year, then the kidney mate’s risk of failure was 6% higher (95% CI, 4% to 9%), or 2.85-fold higher (95% CI, 2.25- to 3.48-fold), than baseline. For 3-year graft failure, the excess absolute risk was 7% (95% CI, 4% to 10%) but excess relative risk was smaller, 1.91-fold (95% CI, 1.56- to 2.28-fold). Fixation indices were 0.25 for delayed graft function (95% CI, 0.24 to 0.27), 0.07 for 1-year graft failure (95% CI, 0.04 to 0.09), and 0.07 for 3-year graft failure (95% CI, 0.04 to 0.10). Results were similar in strata of kidneys with a similar Kidney Donor Profile Index.

Conclusions Overall results indicated that the donor constitution has small or moderate effect on post-transplant clinical outcomes.

Introduction

In kidney transplantation, graft outcomes are influenced by an interplay of donor effects and nondonor effects. Nondonor effects include recipient effects and differences among care protocols after donor nephrectomy. The relative contributions for the risk of graft failure derived from the donor kidney versus the new immunologic milieu of the recipient (or other nondonor effects) are unclear. Leveraging data from pairs of kidney transplants from deceased donors enables data analyses to parse the contribution of donors on transplant outcomes because the kidneys are transplanted into two different recipient environments. If a pair of kidneys from the same donor has strongly concordant recipient outcomes, this concordance is evidence that the donor’s genetic and biologic constitution strongly influences the outcome. Conversely, if the concordance for pairs of kidneys is close to the level expected by chance, this provides evidence of minimal donor effects.

Previous publications have leveraged pairs of transplanted kidneys to quantify donor effects using a variety of approaches but had important limitations such as study size or metrics used to quantify donor effects (1–4). In a small study, Johnson et al. (3) found that if a donor kidney had delayed graft function (DGF), its mate tended to have a larger decline in function 1-year post-transplant compared with organs with similar kidney function at baseline. Traynor et al. (4) examined Spearman correlations and adjusted odds ratios, concluding that there is a “significant degree of relationship” within kidney pairs for serum creatinine and the occurrence of DGF. Gourishankar et al. (1) concluded that the similarity in function between kidney mates is “both powerful and sustained,” but that donor effects on rejection were weak. Louvar et al. (2) examined pairs of donated kidneys using a large data set, but the authors primarily relied on odds ratios to summarize associations, concluding that unmeasured donor characteristics contribute to the risk of DGF and to allograft failure.

Our investigation used a large data set of deceased donor kidney recipients who underwent transplantation in the contemporary era of organ allocation and clinical practice. Specifically, this investigation accounted for the fact that the quality of donor kidneys is currently assessed using the Kidney Donor Profile Index (KDPI); thus, we updated previous work on transplanted kidney pairs to the era of KDPI. We also focused on measures of concordance that are more interpretable for clinical practice, rather than traditional epidemiologic measures such as odds ratios. In particular, we estimated the additional risk of graft failure if the kidney mate fails.

Materials and Methods

Study Population and Outcomes

This study used data from the Organ Procurement and Transplantation Network (OPTN). The OPTN data system includes data on all donors, wait-listed candidates, and transplant recipients in the United States, submitted by the members of the OPTN, and has been described elsewhere (5). The Health Resources and Services Administration (HRSA), US Department of Health and Human Services provides oversight to the activities of the OPTN contractor. The analyses are based on OPTN data as of July 31, 2017 and may be subject to change due to future data submission or correction by transplant centers.

The analytic data set consists of data on kidney transplants occurring on or after January 1, 2013. We selected this date to correspond to the time period when KDPI was used to evaluate donor kidneys. We consider three outcomes: DGF, death-censored 1-year graft failure, and death-censored 3-year graft failure. DGF was defined as dialysis within the first 7 days after transplant. To achieve adequate follow-up, the last date of transplant was March 31, 2017 for DGF; June 30, 2016 for 1-year graft failure; and June 30, 2014 for 3-year graft failure.

We considered only pairs of kidneys in which both organs were transplanted and neither recipient died within a time frame determined by each outcome. We excluded kidneys with missing KDPI. Figure 1 details the numbers of kidneys excluded by each criterion. For DGF, the analytic data set included 25,263 kidney pairs; for 1-year graft failure, 20,254 pairs; for 3-year graft failure, 8206 pairs.

Figure 1.
Figure 1.

The analytic dataset for each outcome was determined by the time frame of each outcome, with minimal exclusions for recipient death or missing KDPI. KDPI, Kidney Donor Profile Index.

Defining Measures of Excess Concordance

To measure excess concordance in kidney pairs, one must account for the level of concordance expected by chance. One measure, f, comes from population genetics and is called the fixation coefficient or inbreeding measure (6,7). f has an interpretation as a correlation coefficient and a κ statistic (8), and values range from −1≤f≤1. If f=0, then the proportion of discordant pairs equals the proportion expected by chance under independence. If f=1, then pairs of kidneys are never discordant, which can only happen if there is complete statistical dependence between kidneys within a pair. Mendelian traits in homozygotic twins have f=1. Negative values of f can arise in population genetics if there is disassortive mating. In a fish population with a lefty/righty phenotype, most matings were observed to be between lefty and righty fish (rather than between fish with concordant phenotypes), producing f values approaching −1 (9).

Excess concordance can alternatively be framed in terms of excess risk: if we knew a kidney had an event, how much more likely is the event for the mate compared with the baseline event incidence? We can quantify excess risk on either the relative risk scale or differences in absolute risk. We define excess relative risk as the probability a kidney has the event given its mate has the event, divided by the baseline event incidence. Excess absolute risk is the probability a kidney has the event given its mate has the event, minus the baseline event incidence.

Additional details on f, excess relative risk, and excess absolute risk are available in Supplemental Appendix 1.

Kidney Donor Profile Index and Accounting for Confounding

Kidneys judged to be low quality may tend to go to patients who are sicker, after special informed consent (if the KDPI is >85%), because patients who are sicker may be counseled that reducing time on the waiting list and dialysis is the most important consideration for improving their health. In addition, under the current allocation system, kidneys judged to be the highest quality (KDPI<20%) are preferentially allocated to patients who are healthier to maximize quality life years of the recipient. The tendency for unhealthy (healthy) kidneys to go to recipients who are unhealthy (healthy) could tend to produce concordant outcomes, even if in truth there are no donor effects. Even if judgments of the health or quality of donor kidneys were arbitrary, the nonrandom assignment to recipients could create concordant outcomes for kidney pairs due to recipient effects. This study covers a time period when the KDPI has been the primary metric to summarize the quality of donor kidneys. Therefore, we consider strata of kidneys with similar values of KDPI to reduce the potential for confounding from recipient effects. Results can alternatively be interpreted as quantifying donor effects conditional on the current allocation process. We partitioned kidneys into four KDPI strata: 0–20, 20–50, 50–85, and 85–100.

Statistical Methods

Using the analytic data set for each outcome, we estimated the event incidence, denoted as P, and measured excess concordance using f and excess relative and absolute risk. We constructed confidence intervals using a bootstrap percentile method, which performed well in a simulation study evaluation (data not shown). Because sample sizes are large and confidence intervals tend to be narrow, we report confidence intervals in tables of results but not in the text for ease of presentation. We also partitioned kidney pairs within each analytic data set into strata according to KDPI. Within strata we estimated the event incidence, fixation index, and excess relative and absolute risk. We conducted additional supplemental analyses on subgroups of kidney pairs based on the following factors: recipient diabetes status, number of HLA mismatches, transplant center volume, and recipient dialysis time.

We computed incidence of DGF, 1-year graft failure, and 3-year graft failure for kidneys whose mate was not transplanted. It is not possible to compute f or measures of excess risk for such kidneys. We further examined the number of discarded kidneys whose mate was transplanted over the time period of this study.

All analysis was performed on the R statistical computing platform.

Sensitivity Analysis for Death Censoring

For 1-year (and 3-year) graft failure, a recipient death before 1 year (and subsequently, before 3 years) without graft failure removed the kidney pair from the analytic data set. This approach implicitly assumes that death is noninformative for graft failure. As a secondary analysis to explore sensitivity to this assumption, we analyzed a composite outcome of death and graft failure. As a second sensitivity analysis, we simulated graft failure outcomes for kidneys whose recipients died, assuming that kidneys transplanted to recipients who died were twice as likely to experience graft failure as other kidneys at risk of graft failure. We examined the effect of such informative censoring on results. We conducted an additional exploratory analysis examining whether we could distinguish the incidence of graft failure in right kidneys depending on whether the corresponding left kidney’s recipient had lived or died.

Institutional Review Board Approval

We adhered to the ethics principles of the Declaration of Helsinki. The study was approved by HRSA and the institutional review board at participating institutions. The clinical research activities being reported are consistent with the Principles of the Declaration of Istanbul as outlined in the “Declaration of Istanbul on Organ Trafficking and Transplant Tourism.”

Results

Tables 1 and 2 describe donor and recipient characteristics in the analytic data set for the DGF analysis. The mean (SD) age of donors was 36 (15) years with a median KDPI score of 38. The mean (SD) age of recipients was 50 (15) years with a mean (SD) of 55 (42) months of dialysis before transplantation. There were notable differences among recipients receiving kidneys with different KDPIs (Supplemental Table 1). For example, the mean (SD) recipient age was 42 (16) years for a kidney with KDPI 0–20; 51 (14) years for a kidney with KDPI 20–50; 55 (12) years for a kidney with KDPI 50–85; and 62 (9) years for KDPI >85. Such differences were expected and motivated the stratified analyses.

View this table:
Table 1.

Characteristics of 25,169 deceased kidney donors in the United States from whom two kidneys were transplanted

View this table:
Table 2.

Characteristics of 50,134 kidney transplant recipients (delayed graft failure analysis)

On examining DGF overall and in KDPI strata (Table 3), the incidence of DGF was highest in the two strata with highest KDPI, with about a third of transplanted kidneys having DGF. The trend of higher DGF incidence in higher KDPI strata could result from donor effects, nondonor effects, or both. As measured by f, donor effects are similar in the four KDPI strata, although smallest in the lowest and highest KDPI strata. On the other hand, the excess relative risk was highest in the lowest KDPI stratum, with double the risk of DGF for a kidney in this stratum if its mate had DGF. The excess absolute risk, which is arguably more relevant to patients, is roughly comparable across the strata, with 12%–19% additional risk of DGF if the kidney-mate graft experiences DGF. In totality, there is evidence of moderate donor effects on DGF.

View this table:
Table 3.

Summary of outcomes for pairs of donor kidneys in groups with similar Kidney Donor Profile Index

The incidence of 1-year death-censored graft failure was 2% in the strata of kidneys with lowest KDPI, and increased to 8% in the stratum with highest KDPI (Table 3). In the lowest KDPI stratum, there was only marginal evidence of donor effects on 1-year graft failure; kidney pairs with KDPI≤20 had 1-year graft failure very close to the expected rates for independent kidneys. The data demonstrated minimal donor effects in the second-lowest KDPI stratum of kidneys, with f=0.05. Although the excess relative risk is 2.6, this is in the context of a small event incidence of 3%. The excess absolute risk in this stratum is 5%. The data suggest greater donor effects in the two higher KDPI strata, with excess absolute risk of 7%–8%.

For 3-year death-censored graft failure (Table 3), incidence rates increase from 6% in the lowest KDPI stratum to 15% in the highest KDPI stratum. Estimated donor effects are highest in the two middle strata of kidneys (KDPI, 20–50 or 50–85), with excess relative risk of 2.0 and 1.8 and excess absolute risk of 7%–8%. Estimated donor effects in the lowest and highest KDPI strata are smaller, although the sample size in the highest stratum is relatively small and so estimates are imprecise.

Figure 2 displays results for all three outcomes, all three summary measures, and the overall event incidence.

Figure 2.
Figure 2.

Donor effects are moderate for DGF and more modest for graft failure outcomes. Relatively large excess relative risks for 1-year graft failure is against a background of low incidence. For all three outcomes, donor effects in strata of kidneys with similar KDPI are similar to overall results. Concordance metrics for delayed graft function, 1-year death-censored graft failure, and 3-year death-censored graft failure among pairs of donor kidneys overall and in strata of similar KDPI. Three metrics are shown: the fixation index f, excess relative risk, and excess absolute risk.

The secondary analysis of the composite outcome of death and graft failure produced smaller estimates of donor effects compared with death-censored graft failure (Supplemental Appendix 2 and Table 2). The sensitivity and exploratory analyses did not challenge the decision to censor graft failure outcomes for death (Supplemental Table 3 and Figures 1 and 2). The supplementary analyses considering different subgroups of paired transplants did not reveal substantively different magnitudes of donor effects (Supplemental Tables 4–7).

Over the 4-year period covered in this investigation, there were 3190 kidney pairs in which one kidney was transplanted and one was discarded (Table 4). The majority of discarded kidneys had KDPI in the range of 20–85, and the number of discards has increased in recent years (Supplemental Figure 3). When comparing organs of a similar KDPI, event rates for kidneys whose mates were discarded were similar to kidneys from pairs in which both were transplanted (Table 4).

View this table:
Table 4.

Rates of outcomes for donor kidneys whose mate was discarded, compared with kidneys from pairs in which both organs were transplanted

Discussion

In kidney transplantation, the relative contributions for the risk for graft outcomes derived from the donor kidney or the recipient protoplasm is challenging to quantify. In both unadjusted analyses and adjusting for KDPI, we found evidence of small to moderate donor effects for DGF and graft failure in kidney transplantation. For DGF, donor effects were moderate across KDPI strata. For 1- and 3-year graft failure, donor effects were more modest, with 4%–8% excess absolute risk over baseline for a graft if the mate graft failed.

For all three outcomes, there were higher event rates in higher KDPI strata, likely due to a combination of donor effects and recipient effects, because high-KDPI kidneys tend to go to recipients who are sicker. To account for this, we examined concordance metrics in strata of kidneys with similar KDPI. These stratified analyses are analyses that adjust for KDPI. One way to interpret the stratified results is as a quantification of donor effects given the current allocation system.

We estimated the excess relative risk and excess absolute risk that a kidney will have an outcome if its kidney mate has the outcome. Absolute risk is more relevant to patients because doubling or tripling a small event incidence may not result in a clinically meaningful absolute risk. Large values of excess absolute risk are potentially useful in patient care; if one kidney from a pair fails, the recipient of the mate may warrant additional surveillance if this information was shared with the transplant center (1,4). Depending on the cause of the failure, knowledge that the kidney mate failed might motivate a recipient toward greater adherence with immunosuppressant therapy or other disease-modifying therapies for underlying comorbidities such as diabetes or hypertension.

Louvar et al. (2) reported a similar study of DGF and 1-, 2-, and 3-year death-censored graft failure in 2009. Using different methodology, they also found evidence for donor effects on DGF and graft failure. Donor effects in (2) are summarized using adjusted odds ratios, whereas this report uses the fixation index, excess relative risk, and excess absolute risk. The choice to use odds ratios hinders interpretability of results because odds ratios cannot be interpreted as risk ratios (which are more readily understood) for nonrare outcomes. Moreover, the rationale for the choice of adjustment variables in (2) was not adequately described in the report.

This report updates the study of donated kidney pairs to the era of KDPI and the current allocation system. Strengths of this report are large samples sizes and the interpretable measures of excess risk. Because of the nature of the data, we could only quantify donor effects in kidney pairs in which both organs were transplanted, raising the possibility of selection bias. It would have been desirable to quantify recipient effects in an equivalent way. However, the equivalent analysis of recipient effects would require data on recipients who received two kidneys from two different donors. We further note that the analysis presented does not account for transport factors such as cold-ischemia time and pumping of the donor kidneys. These nondonor factors might be correlated within kidney pairs, and could explain some of the excess concordance we observed among kidney pairs and attributed to donor effects. Therefore, the estimates of donor effects we report may be overestimates.

Our findings may have implications for trials of interventions applied to potential donors aimed at reducing the incidence of poor outcomes in recipients. Trialists might choose to enroll only transplants of high-KDPI kidneys because the data show higher event rates with higher KDPI. Nominally, higher event rates imply that a smaller sample size suffices to conduct an adequately powered trial (10,11). However, the finding of small donor effects in this stratum casts doubt on this enrichment strategy. Consider the highest KDPI stratum of kidneys, which had 8% incidence of 1-year graft failure compared with 2%–3% in the lowest two KDPI strata. Because the higher incidence of 1-year graft failure in this stratum is likely a combination of donor and nondonor effects, it may be unrealistic to expect that a donor intervention could reduce the event incidence from 8% down to 2%–3%, because only a fraction of the higher incidence may be attributable to donor factors.

Currently, >700 kidneys are discarded annually where the kidney mate is transplanted, with the number of discards increasing in recent years. In light of our finding of small donor effects on graft failure, decisions to discard solely on the basis of donor quality could be questioned (12,13). One could argue that if one kidney from a pair is acceptable for transplant, then the other should be also, except in instances of a unilateral problem such as vessel damage, abnormal cyst, or similar reasons. The rates of graft failure (Table 4) were comparable in kidneys in which only one from a pair was transplanted, compared with the kidneys in our study (both donor organs were transplanted). In the highest KDPI stratum, we observed >85% chance of 3-year graft survival with small donor effects. Although some discarded kidneys undoubtedly had serious unilateral defects, these findings suggest too many missed opportunities for transplantation (13–16). Given the small effects of donor factors on ultimate graft performance, reliance on those factors for decision making regarding transplant may be misplaced, particularly in light of the risks of continued dialysis (13). Clinicians participating in the kidney transplant allocation process should take these results into consideration and educate their patients on the benefits of accepting an available donor kidney as opposed to the risks of waiting for a higher-quality organ.

Disclosures

Dr. Coca reports receiving personal fees from CHF Solutions, Goldfinch, Janssen, Quark, and Takeda; personal fees from and stock options in pulseData; and personal fees from and owning equity in RenalytixAI; all outside the submitted work. Dr. Parikh reports positions on the Data Safety and Monitoring Boards of Abbott and Genfit Biopharmaceutical Company, consulting fees from Akebia Therapeutics and Genfit Biopharmaceutical Company, and a position on the Advisory Board and owning equity in RenalytixAI, all outside of the submitted work. Dr. Reese reports investigator-initiated grants from AbbVie, CVS Caremark, and Merck; consulting fees from COHRDATA; and an Associate Editor position at the American Journal of Kidney Diseases; all outside of the submitted work. Dr. Kerr, Mr. Morenz, Ms. Thiessen-Philbrook, and Dr. Wilson have nothing to disclose.

Funding

Dr. Coca is supported by National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK) grants (R01-DK115562 and U01DK106962). Dr. Kerr is supported by a National Institutes of Health subcontract from Johns Hopkins University to University of Washington. Dr. Parikh is supported by National Heart, Lung, and Blood Institute and NIDDK grant (R01-DK 093770). Dr. Wilson is supported by NIDDK grants (R01-DK 113191 and U01DK079210).

Supplemental Material

This article contains the following supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.03810319/-/DCSupplemental.

Supplemental Table 1. Recipient characteristics stratified by KDPI group.

Supplemental Table 2. Results for all-cause graft failure.

Supplemental Table 3. Additional sensitivity analysis for death-censoring.

Supplemental Table 4. Donor effects in subgroups defined by diabetes.

Supplemental Table 5. Donor effects in subgroups defined by HLA status.

Supplemental Table 6. Donor effects in subgroups defined by time on dialysis.

Supplemental Table 7. Donor effects in subgroups defined by transplant center size.

Supplemental Figure 1. Exploring possible associations between recipient death and 1-year graft failure in the mate.

Supplemental Figure 2. Exploring possible associations between recipient death and 3-year graft failure in the mate.

Supplemental Figure 3. Kidney discard by year.

Supplemental Appendix 1. Summarizing concordance for kidney pairs.

Supplemental Appendix 2. Additional sensitivity analyses for censoring graft failure at death.

Acknowledgments

Dr. Parikh and Dr. Kerr conceived the study question. Dr. Parikh obtained the data and funding for the project and supported all phases of the project. All authors participated in the research design and development of the analytic plan. Mr. Morenz and Ms. Thiessen-Philbrook performed all data analysis and compilation of results. Dr. Kerr wrote the first draft of the paper and all authors reviewed and contributed to drafts.

The data reported here have been supplied by the United Network for Organ Sharing as the contractor for the OPTN. The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the OPTN or the US Government.

Footnotes

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

  • Received March 28, 2019.
  • Accepted October 7, 2019.

References

  1. ↵
    1. Gourishankar S,
    2. Jhangri GS,
    3. Cockfield SM,
    4. Halloran PF
    : Donor tissue characteristics influence cadaver kidney transplant function and graft survival but not rejection. J Am Soc Nephrol 14: 493–499, 2003pmid:12538752
    OpenUrlAbstract/FREE Full Text
  2. ↵
    1. Louvar DW,
    2. Li N,
    3. Snyder J,
    4. Peng Y,
    5. Kasiske BL,
    6. Israni AK
    : “Nature versus nurture” study of deceased-donor pairs in kidney transplantation. J Am Soc Nephrol 20: 1351–1358, 2009pmid:19389849
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Johnson JF,
    2. Jevnikar AM,
    3. Mahon JL,
    4. Muirhead N,
    5. House AA
    : Fate of the mate: The influence of delayed graft function in renal transplantation on the mate recipient. Am J Transplant 9: 1796–1801, 2009pmid:19519811
    OpenUrlCrossRefPubMed
  4. ↵
    1. Traynor C,
    2. O’Kelly P,
    3. Denton M,
    4. Magee C,
    5. Conlon PJ
    : Concordance of outcomes of pairs of kidneys transplanted into different recipients. Transpl Int 25: 918–924, 2012pmid:22775290
    OpenUrlCrossRefPubMed
  5. ↵
    1. Organ Procurement and Transplantation Network
    : About data. Available at: https://optn.transplant.hrsa.gov/data/about-data/. Accessed October 15, 2019
  6. ↵
    1. Wright S
    : The genetical structure of populations. Ann Eugen 15: 323–354, 1951pmid:24540312
    OpenUrlCrossRefPubMed
  7. ↵
    1. Cockerham CC
    : Variance of gene frequencies. Evolution 23: 72–84, 1969pmid:28562963
    OpenUrlCrossRefPubMed
  8. ↵
    1. Zou GY,
    2. Donner A
    : The merits of testing Hardy-Weinberg equilibrium in the analysis of unmatched case-control data: A cautionary note. Ann Hum Genet 70: 923–933, 2006pmid:17044866
    OpenUrlCrossRefPubMed
  9. ↵
    1. Takahashi T,
    2. Hori M
    : Evidence of disassortative mating in a Tanganyikan cichlid fish and its role in the maintenance of intrapopulation dimorphism. Biol Lett 4: 497–499, 2008pmid:18577501
    OpenUrlCrossRefPubMed
  10. ↵
    1. Temple R
    : Enrichment of clinical study populations. Clin Pharmacol Ther 88: 774–778, 2010pmid:20944560
    OpenUrlCrossRefPubMed
  11. ↵
    1. Kerr KF,
    2. Roth J,
    3. Zhu K,
    4. Thiessen-Philbrook H,
    5. Meisner A,
    6. Wilson FP,
    7. Coca S,
    8. Parikh CR
    : Evaluating biomarkers for prognostic enrichment of clinical trials. Clin Trials 14: 629–638, 2017pmid:28795578
    OpenUrlCrossRefPubMed
  12. ↵
    1. Tullius SG,
    2. Rabb H
    : Improving the supply and quality of deceased-donor organs for transplantation. N Engl J Med 378: 1920–1929, 2018pmid:29768153
    OpenUrlCrossRefPubMed
  13. ↵
    1. Reese PP,
    2. Harhay MN,
    3. Abt PL,
    4. Levine MH,
    5. Halpern SD
    : New solutions to reduce discard of kidneys donated for transplantation. J Am Soc Nephrol 27: 973–980, 2016pmid:26369343
    OpenUrlAbstract/FREE Full Text
    1. Marrero WJ,
    2. Naik AS,
    3. Friedewald JJ,
    4. Xu Y,
    5. Hutton DW,
    6. Lavieri MS,
    7. Parikh ND
    : Predictors of deceased donor kidney discard in the United States. Transplantation 101: 1690–1697, 2017pmid:27163541
    OpenUrlCrossRefPubMed
    1. Mohan S,
    2. Chiles MC,
    3. Patzer RE,
    4. Pastan SO,
    5. Husain SA,
    6. Carpenter DJ,
    7. Dube GK,
    8. Crew RJ,
    9. Ratner LE,
    10. Cohen DJ
    : Factors leading to the discard of deceased donor kidneys in the United States. Kidney Int 94: 187–198, 2018pmid:29735310
    OpenUrlCrossRefPubMed
  16. ↵
    1. Stewart DE,
    2. Garcia VC,
    3. Rosendale JD,
    4. Klassen DK,
    5. Carrico BJ
    : Diagnosing the decades-long rise in the deceased donor kidney discard rate in the United States. Transplantation 101: 575–587, 2017pmid:27764031
    OpenUrlCrossRefPubMed
Back to top

In this issue

View Selected Citations (0)
Print
Download PDF
Email Article
Citation Tools
Request Permissions
Quantifying Donor Effects on Transplant Outcomes Using Kidney Pairs from Deceased Donors
Kathleen F. Kerr, Eric R. Morenz, Heather Thiessen-Philbrook, Steven G. Coca, F. Perry Wilson, Peter P. Reese, Chirag R. Parikh
CJASN Dec 2019, 14 (12) 1781-1787; DOI: 10.2215/CJN.03810319
del.icio.us logo Digg logo Reddit logo Twitter logo Facebook logo Google logo Mendeley logo

Jump to section

Keywords