Secondary Logo

A Paired Survival Analysis Comparing Hemodialysis and Kidney Transplantation From Deceased Elderly Donors Older Than 65 Years

Lloveras, Josep1; Arcos, Emma2; Comas, Jordi2; Crespo, Marta1; Pascual, Julio1

doi: 10.1097/TP.0000000000000474
Original Clinical Science

Background Kidney transplantation from deceased donors aged 65 years or older is associated with suboptimal patient and graft survival. In large registries, survival is longer after kidney transplantation than when remaining on dialysis. However, whether recipients of these old grafts survive longer than their dialysis counterparts is unknown.

Methods We retrospectively assessed the outcomes of 5,230 recipients of first deceased donor grafts transplanted during the period of 1990 to 2010 in Catalonia, 915 of whom received grafts from donors 65 years or older. In a match-pair analysis, we aimed to pair each of 915 eligible cases with one control (1:1 ratio). Each pair had the same characteristics at the time of entering dialysis program: age, sex, primary renal disease, period of dialysis onset, and cardiovascular comorbidities. We found 823 pairs.

Results Patient survival of 823 recipients of elderly donors was significantly higher than that of their 823 matched dialysis waitlisted nontransplanted partners (91.6%, 74.5%, and 55.5% vs. 88.8%, 44.2%, and 18.1%, respectively at 1, 5, and 10 years; P<0.001). The probability of death after the first year was similar (8.1% transplant vs 10.3% dialysis; P=0.137); however, analyzing the whole period, the adjusted proportional risk of death was 2.66 (95% confidence interval, 2.21–3.20) times higher for patients remaining on dialysis than for transplanted patients (P<0.001).

Conclusion Our study demonstrates that despite the fact that kidney transplantation from elderly deceased donors is associated with reduced graft and patient survival, their paired counterpart patients remaining on dialysis have a risk of death 2.66 times higher.

This match-pair analysis from Catalonia compares survival after receiving a renal transplant from deceased donor over 65 years old to remaining on dialysis. The long-term results show a clear and dramatic patient survival advantage to receiving the transplant.

1 Department of Nephrology, Hospital del Mar, Passeig Maritim, Barcelona, Spain.

2 Catalan Renal Registry, Catalan Transplant Organization, Health Department, Generalitat of Catalonia, Passeig Taulat, Barcelona, Spain.

Received 20 June 2014. Revision requested 8 July 2014.

Accepted 26 August 2014.

This study was funded by Catalan Transplant Organization, Health Department, Generalitat of Catalonia (to E.A. and J.C.) and Institute Carlos III FIS-FEDER PI10-1370 and PI13-00598 (to M.C. and J.P.) and REDINREN 12/0021/0024.

J.P. is the Chief of Nephrology at Hospital del Mar, Barcelona, and his department has received grants from pharmaceutical companies that market immunosuppressive therapy, Novartis, Astellas and Roche, and other companies that market dialysis therapy and products for chronic kidney disease patients, including Fresenius, Amgen and Abbvie. J.P. has received travel and accommodation costs from Novartis to attend scientific meetings and sporadic consultation honoraria for scientific education materials. J.L.L. has received travel and accommodation costs from Novartis to attend scientific meetings. M.C. has received travel and accommodation costs from Novartis and Astellas to attend scientific meetings. E.A. and J.C. declare that they have no conflicts of interest.

All authors contributed to the study design and data interpretation. J.L.L., E.A., J.C., and J.P. performed the analysis and validation of the data. J.L.L. and J.P. drafted the initial report, whereas all authors contributed to the final draft.

Data were provided for the Catalonian Renal Registry by staff from the Catalonian renal replacement therapy centres, and compiled from the registry by E.A. and JC.

Correspondence: Julio Pascual, MD, PhD, Department of Nephrology,Hospital del Mar, Passeig Maritim 25, 08003 Barcelona, Spain. (

Until the 1990s, deceased donors older than 55 years were usually discarded. It was not until the progressive improvement of the whole process of procurement and better organ assessment1,2 that it became evident that age itself should not be a limiting parameter.3,4 The need to accept elderly donors was emphasized by the increase in the waiting lists and the decreasing number of young donors.5 Currently, older age and stroke as the cause of death predominate among deceased donors.6,7

The use of elderly donors revealed that the increase in donor age was associated with reduced graft function as well as recipient and graft survival.6,8 To minimize this impact, age matching criteria between donor and recipient were adopted, reasoning that elderly recipients have shorter life expectancy, independent of the extended lifetime provided by the graft.9 At the same time, the age of waitlisted patients also rose as a consequence of the increasing age of incident dialysis patients and their improved survival rates.6,10

One unresolved question is whether the reduced survival of the recipients of kidneys from elderly donors provides longer survival than maintenance dialysis, as some reports may suggest.11-13 However, this has not yet been conclusively demonstrated. We performed and published a systematic review of the results achieved with kidney transplantation using expanded criteria donors.14 For the current report, we were only able to find one large observational study assessing a comparison between transplanting a marginal kidney and remaining on dialysis.13 No large study compared two matched populations in the way the present study shows.

We carried out a paired analysis, which is the most reliable method when a randomized trial is not feasible. In our study, we first describe the characteristics of a large series of recipients of deceased donors according to donor age and identify the associated variables that influence recipient and graft survival. We subsequently present a comparative survival analysis between recipients of kidneys from deceased donors older than 65 years with that of their paired patients of the same characteristics who remained on maintenance dialysis.

Back to Top | Article Outline


The cohort of recipients of single first kidney transplantation (KT) from deceased donors during the period of 1990 to 2010 comprised 5,230 patients: 3,300 (63.1%) were males and 1,930 (36.9%) females. Recipient and donor mean age (standard deviation was 49.8 (15.2) years and 46.6 (18.6) years, respectively.

Back to Top | Article Outline

Characteristics According to Donor Age

In Table 1, we describe the donor characteristics according to their age: the percentage of males decreases with advancing age, with female predominance in the older groups (P<0.001). Cranioencephalic trauma remains the main cause of death younger than 45 years (54.4%) but stroke predominates over 45 years ranging from 63.4% for ages 45 to 54 to over 70% for donors 65 years or older (P<0.001). The percentage of donors with high glomerular filtration rate (GFR) is progressively reduced with increasing age and lower GFRs progressively rise with increasing age (P<0.001).



In Table 2, we describe the characteristics of the recipients and the KT according to donor age. Mean recipient age correlates with donor age, by recipient age segments (80.1% of the kidneys from donors ≥65 years went to recipients ≥60 years). Male recipients predominate (from 64.3% for donors aged 45 to 54 years to 57.1% for donors ≥75 years). Primary renal disease distribution is statistically different between the donor age groups (P<0.001). Mean time on dialysis before KT is significantly shorter for those younger than 45 years (P<0.001), although none of the donor age segments exceed a mean time of 3 years. Kidney transplantation recipients of older donors show lower panel reactive antibody (PRA) (P=0.03), lower mean human leukocyte antigen (HLA)-A, HLA-B, and HLA-DR matching (P<0.001) and shorter mean cold ischemia time (P=0.046). Delayed graft function (DGF) rate is lower in recipients of donors younger than 45 years compared to all other age groups (P<0.001).



Back to Top | Article Outline

Survival and Associated Variables

Patient survival of the 915 recipients of donors 65 years or older at 1, 5, 10, and 15 years was 91.8%, 74.9%, 54.3%, and 37.6%, respectively, significantly lower than that of the 4,139 recipients of younger donors (96.4%, 89.7%, 77.1% and 64.8%, respectively) (P<0.001). Graft survival was also significantly lower (84.4%, 61.0%, 40.1%, and 27.4% vs. 90.1%, 77.6%, 58.9%. and 42.7%) (P<0.001).

The results of the proportional risk Cox model applied to the analysis of the variables associated to patient survival, graft survival, and death-censored graft survival are shown in Table 3. Recipients of donors 65 years or older showed an increased risk of death 1.3 times higher than recipients of younger donors. The donor cause of death was also statistically related to recipient survival. In addition, the recipient sex, age, primary renal disease, and hepatitis C virus (HCV) significantly influenced recipient survival. Finally, DGF was also significantly associated with greater recipient mortality. Related to the graft survival multivariate analysis, we observed a significant increased risk of graft failure in recipients of donors aged 65 years or older both censoring and noncensoring for death. Lower graft survival was also significantly associated with recipient age, positive HCV, diabetes as primary renal disease, DGF, and high PRA level. The negative impact of advanced donor age was more evident after censoring for death (hazard ratio, 1.34 vs. 1.77, respectively). Contrarily, the negative impact of advanced recipient age disappeared, and female sex appeared as a risk factor for graft loss after censoring for death.



Back to Top | Article Outline

Comparison Between Kidney Transplanted Patients and Patients That Remained on Dialysis

Table 4 shows the characteristics of both cohorts. The distribution of matching variables was homogeneous in both groups. There were no statistically significant differences in cardiovascular comorbidity events (the sum of all comorbidities) or the presence of malignant tumors.



The whole cohort of 1,646 patients (823 KT and 823 dialysis controls) was followed-up for a maximum of 21 years, with a median of 3.2 years. During the whole observation period, there were 270 deaths (33.2%) among the 823 KT and 424 (52.1%) among the 823 controls. The survival of the KT group was significantly higher (P<0.001) than that of the dialysis group: 91.6%, 85,9%, 74.5% and 55.5 and 88.8%, 63,8%, 44.2%, and 18.1% at 1, 3, 5, and 10 years, respectively (P<0.001) (Fig. 1). Figure 1 shows adjusted survival rates taking into account all potential confounding aforementioned variables. For KT patients, uncensored-for-death graft survival at 1, 3, 5, and 10 years were 85.0%, 75.2%, 61.4%, and 40.5%, respectively. The probability of death at the end of the first year did not reach statistically significant differences between the two groups: 8.1% for KT and 10.3% for dialysis (P=0.137). When analyzing the whole period, the proportional risk of death was 2.66 (95% confidence interval [95% CI], 2.21–3.20) times higher for patients remaining on dialysis than that for KT patients. Sensitivity analysis, including stratification for age and first vascular access, yielded similar estimates. In all different age groups, the dialysis group had higher adjusted proportional risk of death, although there was a decreasing trend: 4.14 (95% CI, 2.48–6.90) in patients younger than 50 years; 2.96 (95% CI, 2.27–3.86) aged 51 to 64 years; 2.20 (95% CI, 1.60–3.09) aged 65 to 69 years; 1.86 (95% CI, 1.11–3.11) older than 70 years. The dialysis group had higher and similar adjusted proportional risk of death in both groups of first vascular access: 2.63 (95% CI, 1.79–3.88) in patients initiating chronic hemodiañysis treatment by catheter, and 2.65 (95% CI, 1.94–3.62) in patients initiating by arteriovenous fistula.



Patient survival of the 92 KT unpaired cases excluded from the study was not significantly lower compared to the 823 KT paired cases (94.2%, 77.9%, and 43.8% vs 91.6%, 74.5%, and 55.5%, respectively, at 1, 5, and 10 years, P=0.607).

Back to Top | Article Outline


This study analyzes one of the largest series of renal allograft recipients from deceased donors aged 65 years or older ever reported. It gathers information on the activity of the six adults transplant programs, which serve the 7.5 million Catalan population. One of the values of this registry is the homogeneity of the results among the centers, a consequence of sharing consensus protocols.

In 1980, Catalonia pioneered a procurement model based on a comprehensive approach to the whole process15 which led to world leading figures in 1984 (19.1 deceased donors per million population (pmp) and 36.4 KT pmp) that have been maintained over time (30.1 deceased donors pmp and 54.4 KT pmp in 2012).16 This has been in part because of the increasing acceptance of expanded criteria deceased donors since the early 1990s.17,18 In 2012, deceased donors 60 years or older represented 45% of our total number of effective donors and stroke as cause of death among donors 65 years or older represent 73.2% of the cases,16 figures significantly higher than those reported by other registries like United Network for Organ Sharing (UNOS) or Eurotransplant.19,20 Donor kidney discarding (17.9%)16 is also lower in our series than in other reports.19,21 In particular, the kidney discarding rate from donors older than 65 years in the UNOS database is above 60%, a policy that probably does not allow access to KT for a significant number of elderly recipients who remain on dialysis.

We have also been proactive in including older patients on the waiting lists, as some other programs do.21-23 We give them priority for the allocation of kidneys from deceased donors older than 65 years so that 80.1% of those kidneys went to recipients older than 60 years. The Eurotransplant Senior Program that is a specific program of strict elderly donor-recipient age (≥65 years) matching allocation, reported a fivefold increase in this activity between 1991 (3.6%) and 2007 (19.7%).20,22

As with other reports,14 our series confirms that patient and graft survival is reduced with the increasing age of donor and recipient, particularly in those that receive a kidney from a donor aged 65 years or older. However, we also observed, as did others,24-27 that elderly recipients of these old kidneys have a reduced risk of death-censored graft failure compared to younger recipients of these grafts. On the other hand, it has been reported that old recipients of young donor kidneys show graft survival exceeding patient survival, which means a significant graft-years loss.28

The second part of our study contains the first report ever that demonstrates a threefold lower death risk for KT patients receiving a graft from a deceased donor older than 65 years compared to their control pairs remaining on dialysis. The analyses of large series have shown that transplantation is the preferred treatment option for end-stage renal disease.11-13 According to a seminal report from the UNOS Registry11 and a systematic review, including the meta-analysis of 110 studies,12 also concluded that KT is associated with reduced risk of mortality and cardiovascular events as well as with better quality of life than treatment with chronic dialysis. Nonetheless, the information on the benefits of KT using kidneys from deceased donors aged 65 years or older over staying on dialysis is limited. Ojo et al.,13 analyzing data from the UNOS registry, reported that recipients older than 65 years who received marginal donor kidneys lived an average of 3.8 years longer than their dialysis wait-listed counterparts. However, the value of their report is quite limited because their “marginal” donors included one of the following circumstances: age older than 55 years (36% were <55 years), hypertension longer than 10 years, diabetes longer than 10 years, cold ischemia time longer than 36 hr, and non–heart-beating donors. The specific and unique value of our study lies in the selection of the KT-dialysis pairs (one dialysis control pair (1:1 ratio) for 100% of the KT cases), patient by patient, with comparable characteristics that prevents any confusing effect. No previous reports are based on paired patients, the most reliable method for comparing groups and validating the results, particularly when true randomization is not feasible. In contrast, all previous reports are based on an intention-to-treat manner using the nonproportional Cox regression technique adjusted for some baseline variables.

We should mention, in agreement with others,11,12 that the probability of death at the end of the first year did not reach statistically significant differences (P=0.137) between KT and dialysis patients. Our survival curves began to significantly separate both populations after the first year, showing that the proportional risk of death was 2.66-fold higher for those that remained on dialysis than for those that received a kidney from a deceased donor older than 65 years.

In conclusion, our study shows that KT from deceased donors older than 65 years provides significantly better survival than chronic dialysis for comparable paired patients. Our findings demonstrate and emphasize the need for the implementation of active policies and strategies addressed to increase the number of elderly patients to be included in the waiting lists who may benefit from KT from elderly donors. To attain this objective, it is necessary to be proactive in accepting deceased donors older than 65 years and to review some current discard policies of potentially viable organs.

Back to Top | Article Outline


We analyzed the data provided by the Catalan Renal Registry, a mandatory population-based registry that collects information on all patients with end-stage renal disease requiring renal replacement therapy in Catalonia. All Dialysis and KT patients receive permanent full public medical coverage, including all medications.

Back to Top | Article Outline


A total of 23,923 had started renal replacement therapy until 2010 in Catalonia. For the initial descriptive analysis of recipients, we considered 5,230 patients who received a first KT from a deceased donor in any of the six adult transplant centres of Catalonia between 1990 and 2010. For the paired analysis, we considered 915 (17,4%) of the 5,230, who received a kidney from a deceased donor older than 65 years: 357 (39%) donors were aged 65 to 69 years, 376 (41.1%) between 70 and 75 years, and 182 (19.9%) were older than 75 years.

Back to Top | Article Outline

Variables Analyzed and Statistical Analysis

We first describe donor, recipient, and KT characteristics according to the different donor age groups: 0 to 45 years, 45 to 54 years, 55 to 64 years, 65 to 69 years, 70 to 75 years, and older than 75 years. Donor variables considered were sex, cause of death, and estimated GFR (calculated by the Cockroft-Gault formula). Recipient variables were: sex, age, primary renal disease, HCV serology, and time on dialysis before KT. Transplant variables were: cold ischemia time, DGF defined as need of dialysis the first week after KT, PRA level, and HLA-A, HLA-B and HLA-DR matching.

The differences between each donor age group were assessed using the chi-square test for categorical variables and the analysis of the variance for continuous variables. We identified the variables significantly associated with patient, graft, and graft-censored death survival by means of the Cox proportional hazard model. In the cases where the Cox regression model did not fit, we calculated the multivariate risk by means of a model of generalized parametric survival choosing the exponential distribution after testing the proportional hazard assumptions based on the Schoenfeld residuals. The exponential distribution was tested by plotting the cumulative hazard function at the mean value of the covariates, obtaining a line function Λ (t)=t. The final model was chosen using the Akaike Information Criterion29 which assumes that the lower the values the better the value of the model.

We then studied the 915 KT cases that received a first single kidney graft from deceased donors aged 65 years or older. Controls were defined as those patients that initiated dialysis until 2010 and were included in the waiting list for KT but were not transplanted during the whole study period. We paired each case with one control (1:1 ratio). Each pair had the same characteristics at the time of entering dialysis program: age, sex, primary renal disease, period of dialysis onset, and to have or not have at least one of the five cardiovascular comorbidities. The control patients were included in the waiting list with the same criteria as the transplanted patients, thus the final population of controls and transplant recipients was comparable. Of the 915 KT patients, we excluded 92 because we did not find a paired control. We finally chose 823 KT patients that we were able to pair with 823 patients who remained on dialysis.

The starting point for the calculation of patient survival was the transplant date for KT cases. Calculation of their controls’ survival was started at the date on which they had spent the same amount of time on dialysis as their respective KT pairs. The observation period was until death, date of the follow-up loss, or December 31, 2010. Survival time for both groups was calculated and compared by the Kaplan-Meier curve and the log-rank test. The risk for death onset was estimated using the Cox proportional hazard model, making a cluster by pairs and calculating the robust estimation of the variance. Potential confounding baseline variables were included: to have chronic liver and respiratory diseases, malignant tumors and diabetes not considered as primary kidney disease. Sensitivity analysis included stratification of final models by (i) all different age groups and (ii) first vascular access for hemodialysis. The statistical analysis was carried out by means of the STATA version 11 software.

Back to Top | Article Outline


The authors thank all the staff working at the Catalan Renal Registry and all the health professionals involved in data managing in the nephrology and kidney transplantation units in Catalonia.

Back to Top | Article Outline


1. Matesanz R, Miranda B, Felipe C, et al. Organ procurement in Spain: the national organization on transplants. In: Organ shortage: The solutions. Touraine JL, et al. (Eds). Dordrecht: Kluwer Academic Publishers, 1995: 167.
2. Wight C, Cohen B. Shortage of organs for transplantation. Crisis measures must include better detection and maintenance of donors. Br Med J 1996; 312: 989.
3. Jacobbi LM, McBride VA, Etheredge EE, et al. The risks, benefits and costs of expanding donor criteria. A collaborative prospective three-year study. Transplantation 1995; 60: 1491.
4. Light JA, Kowalski AE, Ritchie WO, et al. New profile of cadaveric donors: what are the donor limits? Transplant Proc. 1996; 28: 17.
5. Shackford SR, Mackersie RC, Holbrook TL, et al. The epidemiology of traumatic death. A population based analysis. Arch Surg 1993; 128: 571.
6. Catalan renal registry. Statistical report 2010. Catalan Transplant Organization, Health Department: Barcelona, Spain: 2012. Available at:
7. Organ Procurement and Transplantation Network. Available at
8. Mezrich JD, Pirsch JD, Fernandez LA, et al. Differential outcomes of expanded-criteria donor renal allografts according to recipient age. Clin J Am Soc Nephrol 2012; 7: 1163.
9. Waiser J, Schreiber M, Budde K, et al. Age-matching in renal transplantation. Nephrol Dial Transplant 2000; 15: 696.
10. Himmelfarb J, Ikizler TA. Hemodialysis. N Engl J Med 2010; 363: 1833.
11. 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 1999; 341: 1725.
12. Tonelli M, Wiebe N, Knoll G, et al. Systematic review: kidney transplantation compared with dialysis in clinically relevant outcomes. Am J Transplant 2011; 11: 2093.
13. Ojo AO, Hanson JA, Meier-Kriesche H, et al. Survival in recipients of marginal cadaveric donor kidneys compared with other recipients and wait-listed transplant candidates. J Am Soc Nephrol 2001; 12: 589.
14. Pascual J, Zamora J, Pirsch JD. A systematic review of kidney transplantation from expanded criteria donors. Am J Kidney Dis 2008; 52: 553.
15. Espinel E, Deulofeu R, Sabater R, et al. The capacity for organ generation of hospitals in Catalonia: a multicenter study. Transplant Proc. 1989; 21: 1419.
16. Donation and transplantation activity in Catalonia. 2012 Report. Servei Català de la Salut Organització Catalana de Trasplantaments. Barcelona, 2013. Available at:
17. Puig JM, Solà R, Vela E, et al., the Committee of the Renal Patients Registry of Catalonia. Renal transplantation using kidneys from elderly donors. Transplant Proc 2001; 33: 1141.
18. Miranda B, Vilardell J, Grinyó JM. Optimizing cadaveric organ procurement: the Catalan and Spanish experience. Am J Transplant 2003; 3: 1189.
19. Organ Procurement and Transplant Network. Deceased Donors Recovered in the U.S. by Donor Age. Accessed at:
20. Eurotransplant International Foundation. Annual Report 2011. Accessed at:
21. Cecka JM, Cohen B, Rosendale J, et al. Could more effective use of kidneys recovered from older deceased donors result in more kidney transplants for older patients? Transplantation 2006; 81: 966.
22. Giessing M, Fuller TF, Friedersdorff F, et al. Outcomes of transplanting deceased-donor kidneys between elderly donors and recipients. J Am Soc Nephrol 2009; 20: 37.
23. Fijter JW. Counselling the elderly between hope and reality. Nephrol Dial Transplant 2011; 26: 2079.
24. Dahmane D, Audard V, Hiesse C, et al. Retrospective follow-up of transplantation of kidneys from “marginal” donors. Kidney Int 2006; 69: 546.
25. Remuzzi G, Cravedi P, Perna A, et al. Long term outcome of renal transplantation from older donors. N Engl J Med 2006; 354: 343.
26. Swanson SJ, Hypolite IO, Agodoa LY, et al. Effect of donor factors on early graft survival in adult cadaveric renal transplantation. Am J Transplant 2002; 2: 68.
27. Andrés A, Morales JM, Herrero JC, et al. Double versus single renal allografts from aged donors. Transplantation. 2000; 69: 2060.
28. Meier-Kriesche HU, Schold JD, Gaston RS, et al. Kidneys from deceased donors: Maximizing the value of a scarce source. Am J Transplant 2005; 5: 1725.
29. Burnham KP, Anderson DR. Model Selection and Multimodel Inference: A Practical Information-Theoretic Approach. 2nd ed. Springer-Verlag, 2002.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.