Introduction

The lower risk of death associated with kidney transplantation relative to maintenance dialysis extends to transplantation with deceased donor kidneys for patients with varying demographics and comorbidities.1–7 Despite lower expected long-term graft survival, transplantation with kidneys from donors with increased risk factors, including high–Kidney Donor Profile Index (KDPI) donor kidneys, is also associated with significantly longer patient survival for most transplant candidates as compared with waiting longer for a higher-quality donor transplant.8,9 Waitlisted candidates must explicitly consent to receive offers for donors with KDPI >85%. While the mode of consenting patients for high-KDPI donor kidney offers may vary by transplant center or individual physician, the implicit benefit of consenting to receive these offers is greater likelihood to receive donor offers with potentially a shorter time to transplantation.

There are no specific guidelines regarding identifying candidates who would most likely benefit by consenting to receive high-KDPI donor offers. The willingness to accept high-KDPI donor offers may be based on personal preferences and patients' individual relative trade-offs between remaining on the waiting list and acceptance of donor kidneys with lower expected post-transplant graft survival.10,11 In a seminal study by Merion et al., on the basis of the expanded criteria donor (ECD) definition of kidneys with lower expected graft survival, findings illustrated the survival benefit of accepting ECD transplants relative to waiting for standard criteria donors.12 In addition, on the basis of candidates from 1995 to 2002, the study characterized patients who may accrue particular benefit from these transplants, including candidates older than 40 years, candidates of non-Hispanic ethnicity, nonsensitized candidates, and candidates with diabetes or hypertension.12 The potential benefit of consenting to receive high-KDPI transplants may also depend on center- and regional-level factors.12,13 A study by Massie et al. demonstrated that among patients older than 50 years, candidates with longer waiting time, and patients with diabetes, there was a survival benefit with almost all deceased donor kidneys relative to remaining on the waiting list. In addition, even among younger patients with either longer expected waiting times or with diabetes, there is substantially longer life expectancy with higher KDPI donor transplantation.8 However, there have been no contemporary national studies evaluating characteristics of patients who consent to receive high-KDPI donor kidneys nor quantifying the potential benefit of consent for high-KDPI donor kidneys on the basis of the likelihood of receiving a transplant.

In this study, we sought to characterize processes and outcomes associated with consenting for high-KDPI donors among kidney transplant candidates in the United States. First, our aim was to evaluate the proportion and characteristics of candidates who consented for high-KDPI donor transplantation at the time of waitlist placement. Second, we intended to describe center-level variation in the proportion of candidates with high-KDPI consent. Third, we sought to evaluate the incidence of receiving a deceased donor transplant on the basis of consent status. Finally, we intended to evaluate whether outcomes associated with consent for high-KDPI donor transplants were modified by patient and center characteristics.

Methods

This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donors, waitlisted candidates, and transplant recipients in the United States, submitted by the members of the Organ Procurement and Transplantation Network. The Health Resources and Services Administration, US Department of Health and Human Services, provides oversight to the activities of the Organ Procurement and Transplantation Network and SRTR contractors. The data reported here have been supplied by the Hennepin Healthcare Research Institute as the contractor for the SRTR. The interpretation and reporting of these data are the responsibility of the authors and in no way should be seen as an official policy of or interpretation by the SRTR or the US Government.

We used data from the SRTR and selected adult candidates (18 years and older) listed for kidney transplantation between January 1, 2015, and May 30, 2022. We excluded candidates who did not have a KDPI consent form documented on the listing date, candidates who received living donor transplantation on the day of listing or activation, and candidates listed as inactive who were not activated during our study period (Figure 1). In addition, for candidates with waitlist placements at multiple institutions, we only retained the initial waitlist period for the purpose of the analyses.

Figure 1:

Study flow chart. Duplicate listing refers to multilisting at a different institution. For multilisted candidates, their first listing fulfilling inclusion criteria is included in the study. KDPI, Kidney Donor Profile Index; LDTx, live donor transplant.

For candidates who were placed on the waiting list with active status, we evaluated whether patients consented to receive a high-KDPI donor, defined as KDPI >85% at the time of listing. For those patients who were initially placed on the waiting list with inactive status, we evaluated consent on the date of activation. We compared listing characteristics by consent for KDPI >85% versus KDPI <85% using two-sample t tests and chi-squared tests. We used multivariable logistic regression to evaluate factors associated with consent for KDPI >85%. The following factors were evaluated in the model: age group at listing/activation, sex, race/ethnicity, primary insurance, dialysis time at activation, educational attainment, body mass index (BMI), history of malignancy, peripheral vascular disease, candidate working for income, previous kidney transplant, blood group, and primary diagnosis (GN, diabetes, polycystic kidney disease, hypertension, other). To describe and evaluate center-level variation in the proportion of candidates waitlisted with high-KDPI consent, we limited the sample for these outcomes to candidates at centers with at least 30 candidates over the study period, to remove unstable estimates.

For all time-to-event analyses, the inception date was the date of waitlist placement for candidates who were listed as active and date of activation for candidates who were initially inactive. The cohort censoring date provided in the SRTR data was May 31, 2022. We used Fine and Gray competing risks models to graph estimated cumulative incidence of the following outcomes: deceased donor transplantation, living donor transplantation, waitlist death, and removal from the list (for any reason other than death or transplantation). We used Cox proportional hazards models to evaluate the association between consent for KDPI >85% with time to events censoring remaining outcomes. All models were adjusted for age group at listing/activation, sex, race/ethnicity, primary insurance, dialysis time, education, BMI, prior malignancy, peripheral vascular disease, working for income, previous transplant, blood group, and primary diagnosis. We also fit adjusted models for the deceased donor transplant model stratified for different characteristics (age groups, primary diagnosis, blood type, dialysis time, and candidates with diabetes of different ages) to test for effect modification.

We also evaluated the primary outcome of adjusted time to deceased donor transplantation stratified by center characteristics (limited to centers with a minimum of 30 candidates). Center characteristics were stratified by quartile on the basis of average deceased donor transplant volume during the study period, proportion of candidates consented for high-KDPI donors at the center, and proportion of high-KDPI transplants of all deceased donor transplants over the study period. We also performed several sensitivity analyses on our primary outcome of time to deceased donor transplantation. We fit adjusted Cox models on data censored at 1, 2, 3, and 4 years and fit an adjusted Cox model on listings activated in 2015, 2016, and 2017.

All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC). 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. The study was approved by the University of Colorado Institutional Review Board.

Results

The study population included 213,364 adult kidney transplant candidates placed on the waiting list between January 1, 2015, and May 30, 2022. Of those, 86,824 (41%) consented to receive a deceased donor kidney with KDPI >85%. Characteristics of candidates on the basis of high-KDPI consent status are presented in Table 1. Candidates who consented to high-KDPI donor kidneys were significantly older (57 versus 49 years, P < 0.001). The percentage of candidates providing consent to KDPI >85% was 63.7% for candidates older than 70 years as compared with 23.5% for candidates aged 18–49 years. Other notable characteristics associated with a higher proportion of high-KDPI consent included candidates with peripheral vascular disease (52%), candidates with diabetes as the primary diagnosis (49%), and candidates with shorter duration of dialysis before waitlist placement. There was a modest increase in the proportion of patients consented for high-KDPI donor organs over time. Center-level characteristics were also associated with differences in the proportion of candidates listed for high-KDPI donor organs (Table 1). Approximately 42% of candidates waitlisted for high-KDPI offers were at centers with the top quartile of high-KDPI consented candidates. Both the largest and smallest centers on the basis of transplant volume had a relatively lower proportion of candidates consented for high-KDPI donor organs. Centers that performed the highest quartile of high-KDPI transplants included 39% of all high-KDPI consented candidates.

Table 1 - Candidate characteristics by consent for Kidney Donor Profile Index >85% versus not (Kidney Donor Profile Index Candidate Characteristics Total (N=213,364) No Consent at Activation (N=126,540) Consent KDPI >85 at Activation (N=86,824) Age at activation, yr 52.7±13.1 49.4±13.3 57.4±11.2 Age at activation, yr, n (%)  18–49 79,449 (37) 60,796 (48) 18,653 (22)  50–59 58,655 (28) 32,780 (26) 25,875 (30)  60–64 32,078 (15) 15,430 (12) 16,648 (19)  65–69 27,893 (13) 11,988 (10) 15,905 (18)  70–99 15,289 (7) 5546 (4) 9743 (11) Patient sex, n (%)  Female 80,595 (38) 48,758 (38) 31,837 (37)  Male 132,769 (62) 77,782 (62) 54,987 (63) Race and ethnicity, n (%)  Black 62,667 (29) 36,022 (29) 26,645 (31)  Hispanic 41,238 (19) 25,248 (20) 15,990 (18)  Other 20,503 (10) 13,148 (10) 7355 (9)  White 88,956 (42) 52,122 (41) 36,834 (42) Insurance, n (%)  Veterans Administration and other 6356 (3) 3825 (3) 2531 (3)  Private 91,398 (43) 58,215 (46) 33,183 (38)  Medicaid/Children's Health Insurance Program 19,662 (9) 13,088 (10) 6574 (8)  Medicare 95,948 (45) 51,412 (41) 44,536 (51) Dialysis time at activation, mo, n (%)  Preemptive 67,443 (32) 41,658 (33) 25,785 (30)  >0–12 53,668 (25) 32,107 (25) 21,561 (25)  >12–24 37,045 (17) 20,639 (16) 16,406 (19)  >24–36 18,253 (9) 10,217 (8) 8036 (9)  >36–48 10,775 (5) 5965 (5) 4810 (6)  >48–60 7092 (3) 4061 (3) 3031 (4)  >60–72 5093 (2) 2969 (2) 2124 (2)  >72 13,995 (7) 8924 (7) 5071 (6) Candidate education, n (%)  Unknown 6027 (3) 3700 (3) 2327 (3)  High school or less 91,611 (43) 53,828 (43) 37,783 (44)  Some college 54,686 (26) 33,010 (26) 21,676 (25)  College or more 61,040 (29) 36,002 (29) 25,038 (29) BMI categories, kg/m 2 , n (%)  Missing 1015 (<1) 673 (1) 342 (<1)  13–20 8152 (4) 5564 (4) 2588 (3)  >20–25 46,661 (22) 29,138 (23) 17,523 (20)  >25–30 70,249 (33) 40,414 (32) 29,835 (34)  >30–35 55,730 (26) 32,487 (26) 23,243 (27)  35–50 31,557 (15) 18,264 (14) 13,293 (15) Diabetes type, n (%)  No 119,765 (56) 77,820 (62) 41,945 (48)  Type 1 7236 (3) 4687 (4) 2549 (3)  Type 2 84,463 (40) 43,124 (34) 41,339 (48)  Type other 1900 (1) 909 (1) 991 (1) Candidate malignancy, n (%) 19,734 (9) 9922 (8) 9812 (11) Candidate peripheral vascular disease, n (%) 23,387 (11) 11,304 (9) 12,083 (14) Working for income, n (%) 76,224 (36) 49,186 (39) 27,038 (31) Previous transplant, n (%) 25,250 (12) 15,575 (12) 9675 (11) Blood group, n (%)  O 103,786 (49) 61,553 (49) 42,233 (49)  B 31,697 (15) 18,528 (15) 13,169 (15)  A 69,469 (33) 41,433 (33) 28,036 (32)  AB 8412 (4) 5026 (4) 3386 (44) Diagnosis etiology group, n (%)  Other 39,513 (19) 25,101 (20) 14,412 (17)  GN 40,245 (19) 27,911 (22) 12,334 (14)  Diabetes 75,689 (36) 38,848 (31) 36,841 (42)  Polycystic kidney disease 15,344 (7) 9625 (8) 5719 (7)  Hypertension 42,573 (20) 25,055 (20) 17,518 (20) Year listed, n (%)  2015 30,390 (14) 19,427 (1) 10,963 (13)  2016 29,580 (14) 18,364 (15) 11,216 (13)  2017 28,446 (13) 17,107 (14) 11,339 (13)  2018 30,081 (14) 17,275 (14) 12,806 (15)  2019 31,013 (15) 17,243 (14) 13,770 (16)  2020 25,918 (12) 14,812 (12) 11,106 (13)  2021 27,472 (13) 16,161 (13) 11,311 (13)  2022 10,464 (5) 6151 (5) 4313 (5) Percentage of high-KDPI consent candidates at transplant center (range within quartile), n (%)  Quartile 1 (0–18.5) 55,457 (26) 50,941 (40) 4516 (5)  Quartile 2 (18.6–41.0) 51,822 (24) 35,738 (28) 16,084 (19)  Quartile 3 (41.1–65.8) 58,762 (28) 29,256 (23) 29,506 (34)  Quartile 4 (65.9–99.4) 46,753 (22) 10,244 (8) 36,509 (42) Average annual deceased donor transplant volume (range within quartile), n (%)  Quartile 1 (6–29) 14,842 (7) 9352 (7) 5490 (6)  Quartile 2 (30–63) 33,161 (16) 18,818 (15) 14,343 (17)  Quartile 3 (64–103) 53,820 (25) 29,296 (23) 24,524 (28)  Quartile 4 (104–321) 110,971 (52) 68,713 (55) 42,258 (49) Percentage of high-KDPI transplants among deceased donor transplant, n (%)  Quartile 1 (6–29) 37,139 (18) 27,082 (22) 10,057 (12)  Quartile 2 (30–63) 61,002 (29) 42,181 (33) 18,821 (22)  Quartile 3 (64–103) 57,594 (27) 33,707 (27) 23,887 (28)  Quartile 4 (104–321) 57,059 (27) 23,209 (18) 33,850 (39)

BMI, body mass index; KDPI, Kidney Donor Profile Index.

The multivariable model for evaluating independent factors associated with high-KDPI consent identified numerous statistically significant factors (Table 2). Older age was associated with a higher adjusted likelihood of high-KDPI consent in a dose-response manner and candidates aged 70+ years had more than five-fold adjusted likelihood relative to candidates aged 18–49 years (adjusted odds ratio [AOR]=5.43; 95% confidence interval [CI], 5.22 to 5.65). Black (AOR=1.17; 95% CI, 1.14 to 1.20) and Hispanic (AOR=1.05; 95% CI, 1.02 to 1.08) candidates had higher adjusted likelihood of high-KDPI consent relative to non-Hispanic White candidates. Candidates with Medicare as the primary payer had a higher adjusted likelihood of high-KDPI consent (AOR=1.13; 95% CI, 1.10 to 1.16), as well as patients with shorter durations of dialysis before waitlist placement. Candidates with higher BMI, peripheral vascular disease, and prior transplants also had higher likelihood of high-KDPI consent. Relative to patients with GN, candidates with diabetes as the primary diagnosis had a 35% higher adjusted likelihood of high-KDPI consent (AOR=1.35; 95% CI, 1.31 to 1.39). There was wide variation in the proportion of candidates with high-KDPI consent by individual transplant center (Figure 2). The median center-level proportion of high-KDPI consent was 41%, with 25th and 75th percentiles equal to 19% and 66%, respectively.

Table 2 - Multivariable logistic regression model of factors associated with consent for high–Kidney Donor Profile Index (>85%) donor offers at candidate listing Candidate Characteristics AOR (95% CI) Age at activation, yr  18–49 Reference  50–59 2.43 (2.38 to 2.49)  60–64 3.34 (3.24 to 3.44)  65–69 4.04 (3.91 to 4.17)  70–99 5.43 (5.22 to 5.65) Female versus male 0.99 (0.97 to 1.01) Race and ethnicity  Black 1.17 (1.14 to 1.20)  Hispanic 1.05 (1.02 to 1.08)  Other 0.87 (0.84 to 0.90)  White Reference Primary insurance  Private Reference  Medicaid/Children's Health insurance Program 1.01 (0.98 to 1.05)  Medicare 1.13 (1.10 to 1.16)  Veteran's Administration and other 0.94 (0.89 to 0.99) Dialysis time at activation, mo  Preemptive 0.86 (0.84 to 0.88)  >0–12 Reference  >12–24 1.06 (1.03 to 1.09)  >24–36 1.03 (0.995 to 1.07)  >36–48 1.06 (1.01 to 1.10)  >48–60 1.02 (0.96 to 1.07)  >60–72 0.98 (0.92 to 1.04)  >72 0.83 (0.80 to 0.87) Educationa  High school or less Reference  Some college 0.97 (0.95 to 0.99)  College or more 0.99 (0.96 to 1.01) BMI groupa  13–20 0.98 (0.93 to 1.03)  20–25 Reference  >25–30 1.17 (1.13 to 1.20)  >30–35 1.07 (1.04 to 1.09)  >35–50 1.03 (1.003 to 1.06) Candidate malignancy versus not 1.03 (0.998 to 1.06) Candidate peripheral vascular disease versus not 1.26 (1.23 to 1.30) Working for income versus not 1.02 (1.002 to 1.05) Previous transplant versus not 1.14 (1.11 to 1.18) Blood group  O Reference  A 0.97 (0.95 to 0.99)  AB 0.96 (0.92 to 1.01)  B 1.02 (0.99 to 1.05) Diagnosis etiology group  Diabetes 1.35 (1.31 to 1.39)  GN Reference  Hypertension 1.15 (1.12 to 1.19)  Polycystic kidney disease 1.11 (1.07 to 1.16)  Other 1.06 (1.02 to 1.09)

AOR, adjusted odds ratio; BMI, body mass index; CI, confidence interval.

aNote that a missing level category was included but not displayed for BMI and education.

Figure 2:

Center-level variation of candidates listed for high-KDPI kidney donor offers. *Two hundred and three adult kidney transplant centers with at least 30 candidates listed during the study period.

The median follow-up time in the study was 1.2 years (quartile 1=0.4 years and quartile 3=2.6 years) and ranged from 0 to 7.4 years. In total, there were 68,500 deceased donor transplants, 28,355 live donor transplants, 15,336 deaths on the waitlist, and 29,005 listing removals (for reasons other than transplant or death). Figure 3 displays the cumulative incidence of deceased donor transplantation on the basis of high-KDPI consent status stratified by age groups. At 5 years, the cumulative incidence of deceased donor transplantation among those who consented to KDPI >85% donor kidneys versus those who did not was 44% versus 41% for candidates aged 18–49 years and 40% versus 32% for candidates aged 70+ years. Figure 4 displays the cumulative incidence function of deceased donor transplantation, living donor transplantation, death on the waitlist, and waitlist removal on the basis of consent status overall as well as stratified by age group. Notably, among both candidates who consented to receive high-KDPI donor kidneys and those who did not, rates of living donor transplantation were lower with age, while rates of waitlist removal were higher with age. However, the differences in death, waitlist removal, and living donor transplantation were relatively similar between consent groups.

Figure 3:

Cumulative incidence of deceased donor transplantation by high-KDPI consent status by activation age (panels by age group)*. *Panels stratified by candidate age at activation on the waiting list; percentages shown in plot are cumulative incidences of deceased donor transplantation at 2 and 5 years.

Figure 4:

Cumulative incidence of competing events by high-KDPI consent status by activation age (panels by age group)*. *Panels stratified by candidate age at activation on the waiting list and presence or absence of consent for high-KDPI donor kidney transplants; percentages shown in plot are cumulative incidences of each event at 5 years.

In the multivariable Cox model, candidates who consented to KDPI >85% donor kidneys had a 15% higher rate of deceased donor transplantation compared with those who did not consent (adjusted hazard ratio [AHR]=1.15; 95% CI, 1.13 to 1.17). Candidates who consented to KDPI >85% donor kidneys also had a higher rate of deceased donor transplantation from donors with KDPI <85% and higher hazard of waitlist removal, while rates of living donor transplantation and waitlist mortality were not statistically significantly different between consent groups (Table 3). The adjusted association of consent for high-KDPI donor kidneys and higher rates of deceased donor transplantation was evident by different patient characteristics, but the magnitude of the effect varied by subgroup (Figure 4). In particular, the adjusted rate of deceased donor transplantation was higher among candidates older than 65 years, candidates with diabetes as the primary diagnosis, candidates with type B blood, and candidates with duration of prelisting dialysis up to 4 years. Notably, candidates with both diabetes and older age had particularly elevated rates of deceased donor transplantation, including more than 30% greater adjusted hazard for deceased donor transplantation among candidates aged 65+ years with diabetes relative to similar candidates without consent (Figure 5).

Table 3 - Adjusted associations of consent for high–Kidney Donor Profile Index organ versus not with postlisting outcomes Outcome Consent for High-KDPI Offer (KDPI >85) versus Not,a HR (95% CI) Transplant from a deceased donor 1.15 (1.13 to 1.17) Transplant from a deceased donor with KDPI <85 only 1.05 (1.03 to 1.07) Transplant from a live donor 0.98 (0.95 to 1.002) Death on waitlist 1.00 (0.97 to 1.03) Removal from the waitlist 1.06 (1.03 to 1.08)

BMI, body mass index; CI, confidence interval; HR, hazard ratio; KDPI, Kidney Donor Profile Index.

aAdjusted for age at activation (grouped), sex, race, insurance, dialysis months at activation (grouped), education, BMI category, malignancy, peripheral vascular disease, working for income, previous transplant, blood group, and etiology diagnosis group (GN, diabetes, polycystic kidney disease, hypertension, other).