Visual Abstract
Keywords: economic analysis, economic impact, health policy, kidney transplantation, outcomes, renal transplantation, survival, deceased donor organ transplantation
Abstract
Key Points
Health care utilization rates in elderly kidney transplant were no different between the high kidney donor profile index and low kidney donor profile index group.
Number of readmissions and surgical interventions in elderly kidney transplant was no different between the high kidney donor profile index and low kidney donor profile index group.
Death censored graft loss was higher in the high kidney donor profile index group, whereas the patient survival was no different between the high kidney donor profile index and low kidney donor profile index groups.
Background
High kidney donor profile index (KDPI) kidneys (KDPI >85%) are often associated with increased risks of post-kidney transplant (KT) complications and greater health care utilization (HCU), particularly in elderly recipients with multiple comorbidities. These concerns frequently lead to underutilization of high KDPI kidneys in older transplant candidates, thereby limiting their access to transplantation despite potential clinical benefits. The aim of this study was to compare post-KT HCU between elderly recipients of high versus low KDPI deceased donor kidneys. The primary outcome included differences in hospital readmissions, surgical interventions, and outpatient service use at multiple post-transplant time points. Secondary outcome included graft function, graft survival, and patient survival.
Methods
A retrospective analysis of elderly solitary KT recipients (age ≥65 years) between January 1, 2011, and December 31, 2020. Patients were stratified into high KDPI (>85%, n=160) and low KDPI (≤85%, n=453) groups. We assessed inpatient and outpatient HCU, including readmissions at 30, 90, and 365 days post-transplant, surgical interventions, and outpatient visits (emergency department, transplant clinic, endocrine clinic, and ambulatory infusion units).
Results
There were no statistically significant differences between high and low KDPI groups in rates of hospital readmissions, surgical interventions, or outpatient visits (all P > 0.05). Delayed graft function rates were also similar (P = 0.61). The low KDPI group had significantly higher eGFR at 4 months, 1, 2, and 3 years post-transplant (all P < 0.001). However, days alive and out of the hospital within 90 and 365 days post-transplant were comparable between the groups (all P > 0.05). Overall patient survival was similar (P = 0.08), although death-censored graft loss was significantly higher in the high KDPI group (P = 0.001).
Conclusions
Among elderly KT recipients, high KDPI kidney recipients had comparable post-transplant HCU and overall survival with those receiving low KDPI kidneys, despite lower long-term graft function. These findings suggest that high KDPI kidneys remain a viable option for older patients, offering timely access to transplantation.
Introduction
Elderly patients with CKD are progressing to ESKD at an alarming rate. The United States Renal Data System indicates that the incidence of ESKD is the highest among patients age older than 65 years compared with younger patients.1 Not surprisingly, this high incidence of elderly patients in the ESKD population has increased referrals to transplant programs. Patients age older than 65 years represent 26% of the kidney transplant (KT) waiting list in the United States.2 Comorbid conditions are common, and the typical survival benefit ascribed to the procedure may be abbreviated. Perhaps, one of the most significant challenges for elderly KT candidates is identifying the best graft option, either from living donor (LD) or deceased donor (DD). Frequently, these patients cannot identify a suitable LD and must consider DD options, which represents a challenging decision for both patients and transplant teams.
Rao et al. developed the kidney donor risk index to predict the risk of long-term graft loss of a particular donor kidney for the average KT recipient.3 By converting the donor risk index to kidney donor profile index (KDPI), it has been used in clinical decision making and US kidney allocation policy to match donor kidneys with recipients on the basis of projected lifespans from the Estimated Post-Transplant Survival (EPTS) model.4 Several studies have been published regarding the benefits of transplanting high KDPI kidneys (>85%). Specifically, long-term survival benefit was identified with these kidneys compared with waiting for a lower KDPI kidney.5–8 High KDPI KT is also cost-effective because it carries lower costs and higher quality adjusted life years compared with those waiting on dialysis for a low KDPI kidney.9 We have previously shown that high KDPI KT is associated with similar short-term and long-term clinical outcomes across age groups as low KDPI KT.10 These studies suggest that high KDPI kidneys may be ideally suited to elderly candidates who may face long waiting times, the risk of destabilization of health status with dialysis, and abatement of quality of life because of therapeutic obligations.
However, elderly KT candidates may never receive a high KDPI kidney offer despite their projected clinical benefit. High KDPI kidneys continue to be at a high risk of nonuse or even nonrecovery from extended criteria donors.11–13 Perceptions of poor intermediate outcomes from use of these kidneys14–17 may affect organ offer decisions, particularly for elderly KT candidates. At baseline, older recipient age is a known risk factor for higher health care utilization (HCU) including early readmissions and emergency department (ED) visits in KT recipients.18,19 When coupled with the higher allograft risks of delayed graft function (DGF), infections, surgical complications, graft loss, or patient death, transplant programs may be appropriate to decline high KDPI kidneys because these complications lead to high consumption of post-transplant hospital resources. However, there is little existing data on HCU patterns among elderly KT recipients on the basis of the KDPI of their donor kidney. Furthermore, it is not clear that post-KT complications in elderly KT recipients are specific to allograft quality alone—recipient comorbidities may also affect HCU.
In this study, we aimed to assess the effect of KDPI on HCU rates among elderly KT recipients. On the basis of previous data and clinical experience, we hypothesized that HCU would be similar in elderly KT recipients regardless of high KDPI or low KDPI status, as would overall patient and death-censored graft loss (DCGL).
Methods
Data Sources
This was a retrospective study of all the consecutive patients who underwent DD kidney transplantation at Mayo Clinic Arizona between January 1, 2011, and December 31, 2020. The last date of follow-up was December 31, 2023. Inclusion criteria included elderly transplant recipients 65 years or older who underwent solitary KT. Dual transplants were included. We excluded patients aged 65 years and older with the following transplant characteristics: LD KT, simultaneous multiorgan transplant recipients, any nonrenal transplant before or after a solitary KT, and redo KT. Two groups were identified on the basis of donor KDPI, i.e., KDPI ≤85% (low KDPI group) and KDPI >85% (high KDPI group). Mayo Clinic Institutional Review Board approved this study.
Recipient, donor, and transplant variables were collected from electronic health records. Organ Procurement and Transplantation Network database/Scientific Registry of Transplant Recipients was used to look at patient's survival and death censored graft survival. We predominantly used basiliximab as an induction agent in patients age 65 years and older. Standard maintenance therapy was with tacrolimus, mycophenolate mofetil, and prednisone. Tacrolimus was initiated on post-transplant day 1 irrespective of graft function, with trough goals of 8–10 ng/ml for first 30 days and 6–8 ng/ml after day 30. Mycophenolate mofetil was initiated at 2 g per day. We have previously published our approach for accepting high KDPI kidney.10 The 2021 CKD Epidemiology Collaboration equation was used to calculate the eGFR.20
We used the following HCU resources: length of stay (LOS) of index transplant surgery admission, DGF, defined as requirement for dialysis within 7 days post-transplant, post-transplant readmission events within the first 12 months, transfusion rates during index surgical admission, and surgical or interventional radiology interventions within the first 12 months. Please refer to Supplemental Methods for further details on outpatient visits and surgical interventions.
Approach to Accepting High KDPI Kidney Donor
Our center follows a standardized approach for high KDPI kidney acceptance in patients aged ≥65 years. All such candidates are routinely consented for high KDPI kidneys before listing unless they explicitly opt out. In addition to chronological age, we incorporate the EPTS score to further guide candidacy; e.g., at age 65 years, the EPTS score approximates a 40% survival estimate, increasing with age, dialysis duration, diabetes status, and previous transplant.
Candidate qualifying time on the waitlist, including preemptive, status is not factored into listing for high KDPI organs. Instead, we prioritize a multidisciplinary selection conference discussion, especially considering comorbidities such as diabetes and cardiovascular disease and the increased risk of waitlist mortality in this population.
All high KDPI organ offers undergo review of procurement biopsies—particularly for kidneys with AKI, donation after circulatory death status, or other high-risk features. These biopsies are interpreted in real-time by dedicated institutional transplant pathologists to aid clinical decision making.
Outcomes
The primary outcome of interest included HCU parameters between the high KDPI donors kidneys compared with low KDPI donor kidney. Secondary outcomes of interest included eGFR, death censored graft failure, and patient survival.
Statistical Methods
Descriptive statistics were used to summarize patient demographic and clinical characteristics. The comparison between high and low KDPI groups was performed by the Wilcoxon rank sum test for continuous variable or Chi-squared test for categorical variables. Overall survival was defined as time from transplant to death, and patients who were alive on the last date of follow-up period (December 31, 2023) were censored. DCGL was defined as the time from transplant to graft failure, and patients who were dead or who were alive on the last date of follow-up period were censored. Multivariable analysis for overall survival and death censored graft survival was done using the Cox proportional hazards model method. The Kaplan–Meier survival estimates were generated at 1, 3, 5, and 10 years after transplant. The comparison between the survival curves was performed by the log-rank test. All tests were two-sided, and a P value < 0.05 was considered statistically significant. The analysis was conducted using SAS 9.4 (SAS Institute) and R 4.2.2 (R Foundation for Statistical Computing).
Results
During the study period, 613 elderly KTs met our criteria. Among the 613 KT recipients, 160 (26.1%) received high KDPI (>85%) donor kidneys, including three dual transplants, and 453 (73.9%) received low KDPI (≤85%) donor kidneys. The median KDPI in the high KDPI group was 92 (interquartile range [IQR], 88–96) and in the low KDPI group was 51 (IQR, 35–69). The median follow-up was 5.04 years (IQR, 3.56–7.60) in the high KDPI group versus 5.1 years (IQR, 3.64–7.29) in the low KDPI group.
Recipient, Donor, and Transplant Characteristics
Recipient characteristics are shown in Table 1. High KDPI group were older (median age 71 versus 69 years; P < 0.001), more likely to be male (P = 0.02), and lower calculated panel reactive antibody at transplant (mean 6.23 versus 14.37; P = 0.01) compared with the low KDPI group. Pretransplant dialysis vintage was lower in the high KDPI group (median 998.5 versus 1152 days; P = 0.02) but no difference in wait time pretransplant (P = 0.72). There was no difference in recipient multiple pretransplant comorbidities including ESKD etiology, diabetes mellitus, cardiovascular disease including stroke, and peripheral vascular disease between the two groups.
Table 1.
Baseline recipient demographics
| Variables | High KDPI (>85%) (n=160) |
Low KDPI (≤85%) (n=453) |
Total (N=613) | P Value |
|---|---|---|---|---|
| Age at transplant (yr.), median (Q1–Q3) | 71.00 (68.00–74.00) | 69.00 (67.00–72.00) | 70.00 (67.00–73.00) | <0.001a |
| Race/ethnicity category | 0.86b | |||
| Black | 11 (6.9%) | 35 (7.7%) | 46 (7.5%) | |
| Hispanic | 27 (16.9%) | 84 (18.5%) | 111 (18.1%) | |
| Others | 19 (11.9%) | 60 (13.2%) | 79 (12.9%) | |
| White | 103 (64.4%) | 274 (60.5%) | 377 (61.5%) | |
| Sex, male | 114 (71.2%) | 276 (60.9%) | 390 (63.6%) | 0.02b |
| BMI at transplant, median (Q1–Q3) | 27.30 (23.61–29.98) | 27.47 (24.38–30.89) | 27.38 (24.26–30.59) | 0.14a |
| Pretransplant dialysis at transplant | 102 (63.8%) | 318 (70.5%) | 420 (68.7%) | 0.11b |
| UNOS wait time (d), median (Q1–Q3) | 392 (100–803) | 354 (75–855) | 362 (85.25–845) | 0.72a |
| Dialysis type | 0.74b | |||
| Hemodialysis | 70 (68.6%) | 205 (64.5%) | 275 (65.5%) | |
| PD | 15 (14.7%) | 52 (16.4%) | 67 (16.0%) | |
| Unknown | 17 (16.7%) | 61 (19.2%) | 78 (18.6%) | |
| Pretransplant dialysis vintage (d), median (Q1–Q3) | 998.5 (568.50–1451.75) | 1152 (712.00–1660.50) | 1117 (678.00–1610.25) | 0.02a |
| Primary diagnosis at transplant | 0.92b | |||
| Diabetes | 66 (41.2%) | 174 (38.4%) | 240 (39.2%) | |
| GN | 15 (9.4%) | 50 (11.0%) | 65 (10.6%) | |
| Hypertension | 22 (13.8%) | 65 (14.3%) | 87 (14.2%) | |
| Polycystic kidneys | 12 (7.5%) | 41 (9.1%) | 53 (8.6%) | |
| Others | 45 (28.1%) | 123 (27.2%) | 168 (27.4%) | |
| Comorbid conditions | ||||
| Diabetes | 76 (47.5%) | 225 (49.8%) | 301 (49.2%) | 0.62b |
| Hypertension | 152 (95.0%) | 424 (93.8%) | 576 (94.1%) | 0.58b |
| MGUS pretransplant | 6 (8.3%) | 12 (5.4%) | 18 (6.1%) | 0.57b |
| CAD/stent/CABG/ICM | 37 (23.1%) | 117 (25.9%) | 154 (25.2%) | 0.49b |
| Arrhythmia | 21 (13.1%) | 67 (14.8%) | 88 (14.4%) | 0.60b |
| Anticoagulation | 10 (6.2%) | 37 (8.2%) | 47 (7.7%) | 0.43b |
| NICM | 1 (0.6%) | 10 (2.2%) | 11 (1.8%) | 0.30c |
| Stroke | 11 (6.9%) | 53 (11.7%) | 64 (10.5%) | 0.09b |
| Liver disease | 7 (4.4%) | 41 (9.1%) | 48 (7.8%) | 0.06b |
| PVD | 11 (6.9%) | 44 (9.7%) | 55 (9.0%) | 0.28b |
| Cancer | 40 (25.0%) | 125 (27.6%) | 165 (26.9%) | 0.53b |
| Sleep apnea | 27 (16.9%) | 77 (17.0%) | 104 (17.0%) | 0.96b |
BMI, body mass index; CABG, coronary artery bypass surgery; CAD, coronary artery disease; ICM, ischemic cardiomyopathy; KDPI, kidney donor profile index; MGUS, monal gammopathy of unknown significance; NICM, nonischemic cardiomyopathy; PD, peritoneal dialysis; PVD, peripheral vascular disease; UNOS, United Network for Organ Sharing.
Wilcoxon rank sum test.
Chi-squared test.
Fisher exact test.
Donor characteristics are shown in Table 2. Donors of high KDPI kidneys were older (median age 65 versus 42 years; P < 0.001) and less likely to be male (43.8% versus 58.5%; P = 0.001). High KDPI donors were more likely to be diabetic (27.5% versus 6%; P < 0.001), hypertensive (75.3% versus 25.5%, P < 0.001), and die of cerebrovascular causes (55% versus 19.4%; P < 0.001) compared with low KDPI group. High KDPI group were less likely to receive an AKI donor kidney (20% versus 41.4%; P < 0.001). Donor hepatitis C (P = 0.16) and donation after circulatory death status (P = 0.12) were similar in both the groups.
Table 2.
Donor and transplant characteristics
| Variables | High KDPI (>85%) (n=160) |
Low KDPI (≤85%) (n=453) |
Total (N=613) | P Value |
|---|---|---|---|---|
| Donor age (yr), median (Q1–Q3) | 65.00 (60.00–69.00) | 42.00 (31.00–53.00) | 50.00 (35.00–60.00) | <0.001a |
| Donor race/ethnicity | 0.006b | |||
| Black | 15 (9.4%) | 26 (5.7%) | 41 (6.7%) | |
| Hispanic | 19 (11.9%) | 84 (18.5%) | 103 (16.8%) | |
| Others | 18 (11.2%) | 23 (5.1%) | 41 (6.7%) | |
| White | 108 (67.5%) | 320 (70.6%) | 428 (69.8%) | |
| Donor sex, male | 70 (43.8%) | 265 (58.5%) | 335 (54.6%) | 0.001b |
| Donor BMI, median (Q1–Q3) | 28.37 (23.98–33.52) | 28.14 (23.63–34.19) | 28.27 (23.67–34.13) | 0.57a |
| Donor cause of death | <0.001c | |||
| Anoxia | 53 (33.1%) | 241 (53.2%) | 294 (48.0%) | |
| Cerebrovascular/stroke | 88 (55.0%) | 88 (19.4%) | 176 (28.7%) | |
| CNS tumor | 1 (0.6%) | 0 (0.0%) | 1 (0.2%) | |
| Head trauma | 18 (11.2%) | 106 (23.4%) | 124 (20.2%) | |
| Other specify | 0 (0.0%) | 18 (4.0%) | 18 (2.9%) | |
| DCD donor | 29 (18.1%) | 109 (24.1%) | 138 (22.5%) | 0.20b |
| AKI donor | 32 (20.0%) | 187 (41.4%) | 219 (35.8%) | <0.001b |
| Donor hepatitis C | 4 (2.5%) | 23 (5.1%) | 27 (4.4%) | 0.27c |
| Donor hypertension | 119 (75.3%) | 113 (25.5%) | 232 (38.6%) | <0.001b |
| Donor history of diabetes | 44 (27.5%) | 27 (6.0%) | 71 (11.6%) | <0.001b |
| Share type—local/regional/national | 0.02b | |||
| Local | 53 (33.1%) | 209 (46.1%) | 262 (42.7%) | |
| Regional | 50 (31.2%) | 116 (25.6%) | 166 (27.1%) | |
| National | 57 (35.6%) | 128 (28.3%) | 185 (30.2%) | |
| HLA mismatch level | 0.02b | |||
| 0–3 | 27 (16.9%) | 119 (26.3%) | 146 (23.8%) | |
| 4–6 | 133 (83.1%) | 334(73.7%) | 467 (76.2%) | |
| Cold ischemic time (h), median (Q1–Q3) | 21.58 (16.91–24.85) | 21.35 (16.87–25.02) | 21.50 (16.87–25.00) | 0.89a |
| cPRA at transplant, mean (SD) | 6.23 (16.74) | 14.37 (28.22) | 12.25 (25.96) | 0.01a |
| Induction | 0.69c | |||
| None | 4 (2.5%) | 12 (2.6%) | 16 (2.6%) | |
| Alemtuzumab | 3 (1.9%) | 16 (3.5%) | 19 (3.1%) | |
| Basiliximab | 147 (91.9%) | 401 (88.5%) | 548 (89.4%) | |
| Thymoglobulin | 6 (3.8%) | 24 (5.3%) | 30 (4.9%) |
BMI, body mass index; CNS, central nervous system; cPRA, calculated panel reactive antibody; DCD, donation after circulatory death; KDPI, kidney donor profile index.
Wilcoxon rank sum test.
Chi-squared test.
Fisher exact test.
Transplant characteristics are shown in Table 2. High KDPI organs were more likely to be allocated nationally (35.6% versus 28.3%; P = 0.02) but no difference in cold ischemia time (median 21.58 versus 21.35 hours; P = 0.89). HLA mismatch was not different between the two groups (P = 0.06). Basiliximab was the predominant induction regimen with no difference between the two groups (P = 0.61).
HCU
The operative time was higher in the high KDPI group (median 148 versus 138 minutes; P < 0.001) compared with the low KDPI group (Tables 3 and 4). The median post-transplant LOS for the index hospital admission was 3 days in both the groups (P = 0.12). The rates of DGF were similar in both the groups (56.9% versus 54.6%; P = 0.61). The number of readmissions within 30 days (36.9% versus 37.6%, P = 0.87), 31–90 days (26.9% versus 25.9%, P = 0.81), and 91–365 days (31.4% versus 33.2%, P = 0.69) were similar in both the groups. Reasons for readmissions ranged from infection, gastrointestinal issues, AKI, intravascular volume depletion, and sometimes multifactorial etiologies. Supplemental Table 1 lists number of readmissions within 90 days and within 365 days. The number of surgical interventions within 30 days (9.4% versus 11.7%, P = 0.42), 31–90 days (6.2% versus 8.2%, P = 0.43) and 91–365 days (13.1% versus 14.6%, P = 0.65) were similar in both the groups. Supplemental Table 1 lists surgical intervention related to allograft and nonallograft at various time points. We also performed a subgroup analysis excluding preemptive transplants, and our results were comparable with the overall cohort (Supplemental Table 2). The number of days alive and outside hospital within 90 days (P = 0.57) and within 365 days (P = 0.71) were no different between the high KDPI and low KDPI group (Supplemental Table 3).
Table 3.
Health care utilization between the high kidney donor profile index and low kidney donor profile index group
| Variables | High KDPI (>85%) (n=160) |
Low KDPI (≤85%) (n=453) |
Total (N=613) | P Value |
|---|---|---|---|---|
| Operative time (min) | <0.001a | |||
| Median (Q1–Q3) | 148.00 (127.00–175.50) | 138.00 (118.00–159.50) | 140.00 (120.00–164.00) | |
| Transplant LOS (d) | 0.11a | |||
| Median (Q1–Q3) | 3.00 (2.00–4.00) | 3.00 (2.00–4.00) | 3.00 (2.00–4.00) | |
| DGF | 91 (56.9%) | 245 (54.6%) | 336 (55.2%) | 0.61b |
| Readmission <30 d | 59 (36.9%) | 170 (37.6%) | 229 (37.4%) | 0.87b |
| Readmission 31–90 d | 43 (26.9%) | 117 (25.9%) | 160 (26.1%) | 0.81b |
| Readmission 91–365 d | 50 (31.4%) | 150 (33.2%) | 200 (32.7%) | 0.69b |
| RBC during index transplant surgery admission | 43 (26.9%) | 125 (27.6%) | 168 (27.4%) | 0.86b |
| Plasma during index transplant surgery admission | 2 (1.2%) | 14 (3.1%) | 16 (2.6%) | 0.26c |
| Platelet during index transplant surgery admission | 0 (0.0%) | 7 (1.5%) | 7 (1.1%) | 0.20c |
| Surgery 0–30 d | 15 (9.4%) | 53 (11.7%) | 68 (11.1%) | 0.42b |
| Surgery 31–90 d | 10 (6.2%) | 37 (8.2%) | 47 (7.7%) | 0.43b |
| Surgery 91–365 d | 21 (13.1%) | 66 (14.6%) | 87 (14.2%) | 0.65b |
DGF, delayed graft function; KDPI, kidney donor profile index; LOS, length of stay; RBC, red blood cell.
Wilcoxon rank sum test.
Chi-squared test.
Fisher exact test.
Table 4.
Outpatient and emergency department utilization between the high kidney donor profile index and low kidney donor profile index group
| Variables | High KDPI (>85%) (n=160) |
Low KDPI (≤85%) (n=453) |
Total (N=613) | P Value |
|---|---|---|---|---|
| ED visits ≤90 d of discharge | 51 (31.9%) | 153 (33.8%) | 204 (33.3%) | 0.66a |
| No. of ED visits ≤90 d of discharge, mean (SD) | 0.42 (0.75) | 0.48 (0.85) | 0.46 (0.82) | 0.60b |
| Transplant clinic visits ≤90 d of discharge | 160 (100.0%) | 448 (98.9%) | 608 (99.2%) | 0.18a |
| No. of transplant clinic visits ≤90 d of discharge, mean (SD) | 16.09 (4.97) | 16.22 (4.91) | 16.18 (4.92) | 0.16b |
| Endocrine clinic visits ≤90 d of discharge | 76 (47.5%) | 225 (49.7%) | 301 (49.1%) | 0.63a |
| No. of endocrine clinic visits ≤90 d of discharge, mean (SD) | 0.75 (0.97) | 0.89 (1.08) | 0.86 (1.06) | 0.22b |
| AIC visits ≤90 d of discharge | 107 (66.9%) | 311 (68.7%) | 418 (68.2%) | 0.67a |
| No. of AIC visits ≤90 d of discharge, mean (SD) | 2.57 (5.00) | 2.19 (3.59) | 2.29 (4.00) | 0.97b |
AIC, ambulatory infusion clinic; ED, emergency department; KDPI, kidney donor profile index.
Chi-squared test.
Wilcoxon rank sum test.
Post-transplant outpatient visits within 90 days including ED visits (31.9% versus 33.8%, P = 0.66), transplant clinic visits (100% versus 98.9%, P = 0.18), endocrine clinic visits (47.5% versus 49.7%, P = 0.63), and ambulatory infusion center visits (66.9% versus 68.7%, P = 0.67) were similar in high and low KDPI groups as shown in Table 4. We also performed a subgroup analysis of outpatient HCU visits, excluding preemptive transplants, and found that the results were comparable with those of the overall cohort (Supplemental Table 4).
Allograft Function, DCGL, and Patient Survival
High KDPI group had lower eGFR at 4 months, 1, 2, and 3 years (P < 0.001; Figure 1 and Supplemental Table 5). We had 84 deaths (52.5%) in the high KDPI group and 193 deaths (42.7%) in the low KDPI group. Overall survival was similar between high and low KDPI groups (P = 0.08) as shown in Figure 2. Overall survival at 1, 3, 5, and 10 years was 94%, 81%, 65%, and 29% for the high KDPI group and 95%, 86%, 72%, and 35% for the low KDPI group, respectively. The median survival time was 7.22 years for high KDPI and 8.31 years for the low KDPI group. On multivariable analysis for significant factors on the basis of baseline differences plus diabetes, we found that older age was significantly associated with an increased risk of death (hazards ratio [HR]=1.37, P < 0.001; Table 5). Similarly, diabetes was also significantly associated with an increased risk of death (HR=1.45, P = 0.03).
Figure 1.
eGFR for elderly KT recipients by KDPI group. KDPI, kidney donor profile index; KT, kidney transplant.
Figure 2.
Overall patient survival for elderly KT recipients by KDPI group. CI, confidence interval.
Table 5.
Multivariable Cox model for survival
| Variable | Comparison | HR | 95% CI | P Value |
|---|---|---|---|---|
| Overall survival | ||||
| KDPI | High versus low | 1.18 | 0.91 to 1.53 | 0.22 |
| Recipient age at transplant | Per 5 yr older | 1.37 | 1.17 to 1.59 | <0.001 |
| Sex | Male versus female | 1.29 | 1.00 to 1.67 | 0.05 |
| Pretransplant dialysis vintage | Per 5 yr longer | 1.13 | 0.89 to 1.44 | 0.32 |
| Diabetes | Yes versus No | 1.45 | 1.13 to 1.84 | 0.03 |
| Death censored graft survival | ||||
| KDPI | High versus low | 2.26 | 1.39 to 3.68 | <0.001 |
| Recipient age at transplant | Per 5 yr older | 0.94 | 0.68 to 1.30 | 0.72 |
| Sex | Male versus female | 0.96 | 0.59 to 1.57 | 0.87 |
| Pretransplant dialysis vintage | Per 5 yr longer | 1.34 | 0.84 to 2.14 | 0.22 |
| Diabetes | Yes versus No | 1.41 | 0.88 to 2.28 | 0.15 |
CI, confidence interval; HR, hazards ratio; KDPI, kidney donor profile index.
Ninety-three patients (58%) in the high KDPI group and 204 patients (45%) in the low KDPI group had DCGL. The DCGL was significantly higher in the high KDPI group (P = 0.001; Figure 3). At 1, 3, 5, and 10 years post-transplant, the graft survival was 95%, 90%, 84%, and 67% for the high KDPI group and 97%, 95%, 92%, and 85% for the low KDPI group, respectively. On multivariable analysis on the basis of same factors used for overall survival, higher KDPI was significantly associated with an increased hazard for graft failure (HR=2.26, P < 0.001; Table 5). We did a subgroup analysis looking at DCGL and overall survival excluding preemptive transplants with no change in our results (results not shown).
Figure 3.
DCGL for elderly KT recipients by KDPI group. DCGL, death-censored graft loss.
Discussion
By providing a granular evaluation of clinical outcomes and hospital resource use, this analysis demonstrated that elderly patients who received DD in a high-volume center had similar patterns of HCU regardless of donor KDPI. In a population of more than 600 elderly DD recipients, we assessed multiple components of inpatient post-KT HCU and found no difference by the KDPI group in the initial LOS, readmission rates within 12 months, 12-month rate of invasive procedures related to complications, and blood product transfusion. Outpatient HCU patterns were also similar between high KDPI and low KDPI patients. As expected, post-KT eGFR was lower over time for high KDPI recipients versus low KDPI recipients. Although we observed a higher DCGL in the high KDPI group, patient survival remained similar between the groups. This finding is expected, as higher graft loss aligns with the known risk profile of high KDPI donors. Importantly, our study supports the use of high KDPI kidneys in elderly recipients, especially in light of their limited life expectancy, elevated waitlist mortality, and the potential benefit of reduced time to transplantation with similar HCU rates between the two groups.
High KDPI KT has been the subject of significant clinical research study, particularly related to outcomes, clinical benefit questions, and financial implications. Our previous work suggests that high KDPI kidneys are associated with similar HCU patterns as low KDPI kidneys across the recipient age spectrum in the first 90 days after KT. However, that study neither focused on specific subgroups in an ever growing population of KT candidates, the elderly, nor did it provide a broader view of 1-year HCU outcomes.10 HCU can be a surrogate for health care costs, which is well described, especially related to early hospitalization rates.21,22 Older age, recipient comorbidities, and longer initial hospital stays were risk factors for greater resource use, but these studies were not specific to elderly KT recipients or high KDPI kidneys, which are clinically significant. Our study aimed to fill this gap by capturing all health care events at our institution with greater granularity, allowing for a more comprehensive assessment of resource utilization among elderly transplant recipients. Notably, we found no significant difference in HCU rates by graft KDPI group; patient survival and graft loss rate for selected high KDPI kidneys were clinically acceptable. These findings are particularly relevant for the transplant community, suggesting that the use of high KDPI kidneys can be expanded without a significant increase in HCU rates or compromise in post-transplant outcomes, including 90-day and 1-year program metrics, as well as long-term outcomes. The transplant community may benefit from future studies assessing HCU across various age spectrums and time points up to 1-year post-transplantation, to more comprehensively understand the interplay between recipient age, KDPI, and health care resource use.
One of the notable strengths in this study was related to the study cohort. The study cohort is derived from a single high-volume KT center with long-established practice pattern focused on DD and utilization of suboptimal donor kidneys. This provides some homogeneity to interpretation of clinical outcomes, as a single care model and streamlined practice patterns are being used, with close linkages between inpatient and outpatient care. Our findings support the continued use of high KDPI allografts in older recipients as a rational and patient-centered strategy. This strength carries a key limitation as a counterweight. Although elderly transplant candidates presenting at our center may not differ from other centers, nuances in clinical approach during candidate evaluation, organ offer acceptance, post-KT care, and resources may limit generalizability. Our center relies heavily on clinical judgment for utilization of high KDPI kidneys, and there may be institutional bias in selecting what is usable for transplantation. Although our care model and data collection account for post-transplant care delivery outside health care facilities, it is possible that brief low acuity HCU events may not have been fully captured in these data. However, the use of health care facilities aside from the transplant center is a common occurrence in the US transplant environment and generalizable issue across transplant centers. The retrospective study design may also pose potential for study bias. However, this study is one of the largest cohorts looking at 1-year HCU in elderly DD including long-term graft and patient survival.
The use of high KDPI donor kidneys in elderly patients remains controversial because of concerns over intermediate outcomes and health care costs.14,23 Weeda et al.19 identified early hospital or ED visits, cardiac comorbidities, and sociodemographic factors as predictors of high post-transplant HCU. Current KT reimbursement models do not account for recipient or graft factors, potentially disincentivizing high KDPI kidney use despite benefits in elderly population with multiple comorbidities.24 Axelrod et al.9 showed survival benefits and cost-effectiveness across all KDPI categories and that high KDPI kidneys are associated with a $10,000 cost advantage over dialysis over 10 years and a 12.2% lower cost per quality adjusted life year.9 However, despite benefits to the payer environment at large, greater hospital resources as observed in other studies of high KDPI KT impacts hospital margin. Until now, current hospital and professional reimbursement of kidney transplantation lack clinical risk-adjustment or incentives for high KDPI KT use.25–28 Financial concerns, therefore, may contribute to escalating rate of nonuse of high KDPI kidneys, up to more than 50% in 2022.9,13,29 The data presented in this analysis and further implementation of value-based care models in KT, such as the Increase Organ Transplant Access model, may hold significant promise to encourage clinical programs to use high KDPI kidneys.
Our institution's high KDPI acceptance model—grounded in rigorous geriatric and cardiovascular evaluation, multidisciplinary review including transplant cardiologist, and strong post-transplant coordination—has allowed for safe and expanded access to KT in elderly recipients. Close monitoring of recipients and collaboration with referring nephrologists and outside facilities also allows for timely intervention in elderly KT patients. These practices are critical in high-risk populations and in the use of higher-risk kidneys, and dissemination of these practices is key in the current transplant environment.
Still, nearly one third of the elderly population are still unwilling to accept a high KDPI kidney29 and disproportionately underutilized11 despite benefits,5 as compared with being on waitlist. Our findings help dispel concerns that high KDPI kidneys lead to poor overall outcomes and support the case for their continued and thoughtful use in appropriate recipients. Our organ acceptance approach for high KDPI kidney has allowed us to maintain high organ acceptance rates while expanding access to transplantation in an aging ESKD population. By using this approach, we have been able to transplant more elderly patients who might otherwise face prolonged waitlist time or die waiting for transplant.
There are few limitations in our retrospective study especially regarding generalizability. This was a single-center study with sample size and may not apply to other centers with higher wait times, and underlying resources available at transplant center. Our study was focused on post-transplant outcomes rather than waitlist dynamics. We acknowledge that our study does not directly quantify the discard rate or offer-level decision making. However, our findings do contribute to the growing body of evidence advocating for broader acceptance and utilization of high KDPI kidneys, particularly in the context of an aging recipient population, a persistently limited organ supply, and hopefully decrease organ discard rate. We agree that although graft survival may be lower, however, the trade-off may be acceptable in populations with high waitlist mortality risk and limited life expectancy. However, we believe the elderly patient's characteristics may not be significantly different in other transplant centers as discussed above. Rigorous data collection was undertaken to assess our HCU rate, but there is always a question of missing data given the retrospective nature of this study. We chose 12 months as the cutoff as most health care events related to transplant including associated costs tend to happen more frequently in this time frame. We did not capture events which occurred more than 12 months after transplant or those events occurring at an outside institution within 12 months, which could have influenced our findings; however, most early post-transplant events from outside institutions are typically communicated to our center and were captured in this study. However, our analysis does not capture the long-term costs associated with earlier graft failure, which may limit cost equivalence comparisons between high and low KDPI organs. We hypothesize, however, that these costs may be partially offset by the quality of life and survival benefits of earlier dialysis cessation and reduced dialysis-associated costs resulting from receiving a transplant sooner with a high KDPI organ. Our rates of preemptive transplants are higher than national reporting of preemptive transplants which may affect the results, but on subanalysis, we did not find any difference in graft loss and survival, which remains a limitation. Pretransplant HCU patterns were not assessed, which may limit our understanding of their impact on post-transplant outcomes and cost-effectiveness. We acknowledge the potential for selection bias, and unmeasured confounders are an inherent limitation in retrospective studies.
In conclusion, the transplant community in the United States has an objective to expand DD transplantation. Organ offer acceptance metrics are monitored by the Organ Procurement & Transplantation Network and publicly reported by the Scientific Registry of Transplant Recipients. Achieving target financial incentives in the US Centers for Medicare & Medicaid Services Increase Organ Transplant Access model will require broader organ acceptance. Utilization of high KDPI kidneys is critical to the effort of expanding access to transplantation. This analysis should help provide reassurance to transplant programs evaluating their strategies to use suboptimal donor kidneys. Selection of high KDPI kidneys for elderly KT candidates is safe with regards to post-transplant outcomes with the appropriate clinical practices and guardrails in place. Additional studies should focus on how KDPI kidneys affect timeliness of transplant, their utility in various elderly subpopulations by diagnosis, and quality of life and patient-centered outcomes. As the population ages, the organ donation and transplant community are likely to see more elderly kidney donors and KT candidates, which will require more knowledge in this realm to guide practice and policy.
Supplementary Material
Acknowledgments
Part of these data were presented at the World Transplant Congress, August 2025.
Footnotes
See related editorial, “Measuring What Matters: Days Alive and Out of Hospital and Health Care Utilization in Kidney Transplantation,” on pages 473–474.
Disclosures
Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/KN9/B407.
Author Contributions
Conceptualization: Amit K. Mathur, Girish Mour.
Data curation: Bassam G. Abu Jawdeh, Yu-Hui Chang, Lavanya Kodali, Amit K. Mathur, Girish Mour, Sumi S. Nair, Rachana Punukollu.
Formal analysis: Yu-Hui Chang, Amit K. Mathur, Girish Mour.
Investigation: Girish Mour, Rachana Punukollu.
Methodology: Amit K. Mathur, Girish Mour.
Project administration: Girish Mour.
Supervision: Girish Mour.
Writing – original draft: Bassam G. Abu Jawdeh, Yu-Hui Chang, Lavanya Kodali, Amit K. Mathur, Girish Mour, Sumi S. Nair, Rachana Punukollu.
Funding
None.
Declarative Statements
The clinical and 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.
Data Availability Statements
Original data generated for the study will be made available upon reasonable request to the corresponding author. Data Type: Observational Data. Reason for Restricted Access: All data displayed in the tables and figures are contained within the main manuscript, and the supplemental table. Data sharing is restricted by Mayo institutional policy. Data may be available on reasonable request to the corresponding author.
Supplemental Material
This article contains the following supplemental material online at http://links.lww.com/KN9/B408.
Supplemental Table 1. Number of readmissions and post-transplant surgical interventions.
Supplemental Table 2. HCU between the high KDPI and low KDPI group excluding preemptive transplant.
Supplemental Table 3. Days alive and outside hospital.
Supplemental Table 4. Outpatient and ED utilization between the high KDPI and low KDPI group excluding preemptive transplants.
Supplemental Table 5. eGFR at various times points for elderly KT recipients by KDPI group.
References
- 1.United States Renal Data System. 2023 Annual Data Report, 2023. Accessed October 15, 2024. https://usrds-adr.niddk.nih.gov/2023/end-stage-renal-disease/7-transplantation [Google Scholar]
- 2.Organ Procurement and Transplantation Network. National Data Reports. Accessed March 4, 2025. https://optn.transplant.hrsa.gov/data/view-data-reports/national-data/# [Google Scholar]
- 3.Rao PS Schaubel DE Guidinger MK, et al. A comprehensive risk quantification score for deceased donor kidneys: the kidney donor risk index. Transplantation. 2009;88(2):231–236. doi: 10.1097/TP.0b013e3181ac620b [DOI] [PubMed] [Google Scholar]
- 4.Organ Procurement and Transplantation Network. A Guide to Calculating and Interpreting the Estimated Post-Transplant Survival (EPTS) Score Used in the Kidney Allocation System (KAS). Accessed March 4, 2025. https://optn.transplant.hrsa.gov/media/1511/guide_to_calculating_interpreting_epts.pdf [Google Scholar]
- 5.Jay CL, Washburn K, Dean PG, Helmick RA, Pugh JA, Stegall MD. Survival benefit in older patients associated with earlier transplant with high KDPI kidneys. Transplantation. 2017;101(4):867–872. doi: 10.1097/TP.0000000000001405 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Chopra B, Sureshkumar KK. Kidney transplantation in older recipients: preemptive high KDPI kidney vs lower KDPI kidney after varying dialysis vintage. World J Transplant. 2018;8(4):102–109. doi: 10.5500/wjt.v8.i4.102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bui K, Kilambi V, Mehrotra S. Functional status-based risk-benefit analyses of high-KDPI kidney transplant versus dialysis. Transpl Int. 2019;32(12):1297–1312. doi: 10.1111/tri.13483 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Massie AB, Luo X, Chow EK, Alejo JL, Desai NM, Segev DL. Survival benefit of primary deceased donor transplantation with high-KDPI kidneys. Am J Transplant. 2014;14(10):2310–2316. doi: 10.1111/ajt.12830 [DOI] [PubMed] [Google Scholar]
- 9.Axelrod DA Schnitzler MA Xiao H, et al. An economic assessment of contemporary kidney transplant practice. Am J Transplant. 2018;18(5):1168–1176. doi: 10.1111/ajt.14702 [DOI] [PubMed] [Google Scholar]
- 10.Mour GK Chang YH Calderon E, et al. Kidney donor profile index and post-transplant health care utilization: implications for value of transplant care delivery. Clin Transplant. 2022;36(6):e14618. doi: 10.1111/ctr.14618 [DOI] [PubMed] [Google Scholar]
- 11.Crannell WC, Perkins JD, Leca N, Kling CE. Deceased donor kidneys are discarded at higher rates when labeled as high kidney donor profile index. Am J Transplant. 2022;22(12):3087–3092. doi: 10.1111/ajt.17197 [DOI] [PubMed] [Google Scholar]
- 12.Stewart DE, Garcia VC, Aeder MI, Klassen DK. New insights into the alleged kidney donor profile index labeling effect on kidney utilization. Am J Transplant. 2017;17(10):2696–2704. doi: 10.1111/ajt.14379 [DOI] [PubMed] [Google Scholar]
- 13.Lentine KL Smith JM Lyden GR, et al. OPTN/SRTR 2022 annual data report: Kidney. Am J Transplant. 2024;24(2S1):S19–S118. doi: 10.1016/j.ajt.2024.01.012 [DOI] [PubMed] [Google Scholar]
- 14.Axelrod DA Schnitzler MA Xiao H, et al. The changing financial landscape of renal transplant practice: a national cohort analysis. Am J Transplant. 2017;17(2):377–389. doi: 10.1111/ajt.14018 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Ellimoottil C Ye Z Chakrabarti AK, et al. Understanding inpatient cost variation in kidney transplantation: implications for payment reforms. Urology. 2016;87:88–94. doi: 10.1016/j.urology.2015.05.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Englesbe MJ Ads Y Cohn JA, et al. The effects of donor and recipient practices on transplant center finances. Am J Transplant. 2008;8(3):586–592. doi: 10.1111/j.1600-6143.2007.02098.x [DOI] [PubMed] [Google Scholar]
- 17.Englesbe MJ, Dimick JB, Fan Z, Baser O, Birkmeyer JD. Case mix, quality and high-cost kidney transplant patients. Am J Transplant. 2009;9(5):1108–1114. doi: 10.1111/j.1600-6143.2009.02592.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Taber DJ DuBay D McGillicuddy JW, et al. Impact of the new kidney allocation system on perioperative outcomes and costs in kidney transplantation. J Am Coll Surg. 2017;224(4):585–592. doi: 10.1016/j.jamcollsurg.2016.12.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Weeda ER Su Z Taber DJ, et al. Hospital admissions and emergency department visits among kidney transplant recipients. Clin Transplant. 2019;33(5):e13522. doi: 10.1111/ctr.13522 [DOI] [PubMed] [Google Scholar]
- 20.National Kidney Foundation. CKD-EPI Creatinine Equation (2021), 2021. Accessed October 15, 2024. https://www.kidney.org/professionals/kdoqi/gfr_calculator/formula [Google Scholar]
- 21.Naylor KL Knoll GA Allen B, et al. Trends in early hospital readmission after kidney transplantation, 2002 to 2014: a population-based multicenter cohort study. Transplantation. 2018;102(4):e171–e179. doi: 10.1097/TP.0000000000002036 [DOI] [PubMed] [Google Scholar]
- 22.Naylor KL Knoll GA Slater J, et al. Risk factors and outcomes of early hospital readmission in Canadian kidney transplant recipients: a population-based multi-center cohort study. Can J Kidney Health Dis. 2021;8:20543581211060926. doi: 10.1177/20543581211060926 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Zens T Danobeitia J Leverson G, et al. The impact of kidney donor profile index on delayed graft function and transplant outcomes: a single‐center analysis. Clin Transplant. 2018;32(3):e13190. doi: 10.1111/ctr.13190 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Merion RM Ashby VB Wolfe RA, et al. Deceased-donor characteristics and the survival benefit of kidney transplantation. JAMA. 2005;294(21):2726–2733. doi: 10.1001/jama.294.21.2726 [DOI] [PubMed] [Google Scholar]
- 25.Hydrick TC Zhang C Ruch B, et al. Declining medicare reimbursement in abdominal transplantation from 2000 to 2021. Surgery. 2023;173(6):1484–1490. doi: 10.1016/j.surg.2023.01.015 [DOI] [PubMed] [Google Scholar]
- 26.Hippen BE, Maddux FW. Integrating kidney transplantation into value-based care for people with renal failure. Am J Transplant. 2018;18(1):43–52. doi: 10.1111/ajt.14454 [DOI] [PubMed] [Google Scholar]
- 27.Paulus AB, Kumar D, Pontinha VM. The case for value-based care in kidney transplantation: insights into geography, growth, and financial models. Curr Opin Organ Transplant. 2025;30(2):87–95. doi: 10.1097/MOT.0000000000001204 [DOI] [PubMed] [Google Scholar]
- 28.Jain G, Kshirsagar AV. Can future value-based care models in nephrology promote kidney transplantation? Kidney Int Rep. 2024;9(6):1565–1567. doi: 10.1016/j.ekir.2024.04.067 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Scientific Registry of Transplant Recipients. OPTN/SRTR 2023 Annual Data Report: Kidney, 2025. Accessed September 4, 2025. https://srtr.transplant.hrsa.gov/ADR/Chapter?name=Kidney&year=2023 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
Original data generated for the study will be made available upon reasonable request to the corresponding author. Data Type: Observational Data. Reason for Restricted Access: All data displayed in the tables and figures are contained within the main manuscript, and the supplemental table. Data sharing is restricted by Mayo institutional policy. Data may be available on reasonable request to the corresponding author.