Characterizing Within-Center Placement of Out-of-Sequence Deceased-Donor Kidney Transplants

Abstract

Introduction

Deceased-donor kidney (DDK) allocation is intended to follow a ranked match-run of potential recipients. Organ procurement organizations (OPOs) can bypass this sequence and allocate kidneys “out-of-sequence” (OOS) to transplant centers, which then decide which of their candidates receives the kidney. We examine OOS placement practices within individual transplant centers.

Methods

We included the 5 US transplant centers that performed the most OOS kidney transplants from 2021 to 2023, representing 1598 OOS transplants. Candidates were classified as “skipped” if their transplant center declined an OOS kidney on their behalf and allocated the kidney to a patient lower on the center’s waitlist. “Never-skipped” candidates received OOS kidney transplants without having been skipped. For each OOS transplant, we assessed how often the top-ranked candidate was skipped. We used mixed effects models to determine factors associated with being skipped.

Results

Among included centers, the top candidate was skipped in 98% of OOS placements. The median OOS recipient within-center match-run rank was 23. Skipped candidates were younger (56 vs. 63 years) and had more dialysis exposure than never-skipped patients. Fewer than 50% of skipped candidates received a DDK transplant later in the study period (through 2023), and those who did received kidneys with a median Kidney Donor Profile Index (KDPI) 13 points higher than their lowest-KDPI skipped offer.

Conclusion

Younger candidates and those with longer dialysis durations were more likely to be skipped. A minority of skipped patients later received a lower KDPI (lower risk of graft failure) kidney. More transparency in OOS placement decisions is needed.

Graphical abstract

There is a worsening shortage of kidneys available for transplant in the US, with over 49,000 waitlist additions but only approximately 28,000 kidney transplants performed in 2024.¹ In light of this scarcity, the national transplant system uses a predefined algorithm to determine each candidate’s ranked priority for every deceased donor kidney on the basis of factors including waiting time and allosensitization.^2,3 However, because donated kidneys have a wide array of expected longevity, transplant centers often decline offers for lower-quality kidneys offered to candidates with high waitlist priority in anticipation of better offers in the future.⁴ This process can lead to placement inefficiency for organs with undesirable characteristics, such as advanced donor age or unfavorable donor health history. These kidneys must be offered to many candidates before acceptance, leading to accumulation of excessive cold ischemia time that threatens short- and long-term organ function and therefore increases the risk of no center accepting the organ.5, 6, 7, 8

OPOs are responsible for donation authorization, organ recovery, and the logistics of organ placement.⁹ Because OPO performance is assessed in part by the transplantation rate of procured organs, OPOs have increasingly moved away from the ordered list of candidates on the match-run and are instead skipping over centers and allocating kidneys directly to a specific center.10, 11, 12 This practice of OOS allocation is intended to reduce kidney nonuse by channeling marginal kidneys to “aggressive” transplant centers, who in turn can transplant these kidneys into whichever candidates they determine may benefit from earlier transplantation with a suboptimal organ regardless of priority on the match-run.^13,14 Importantly, skipped centers are not aware that they are being skipped for a given OOS kidney. In the past several years, there has been a marked rise in this practice, with a recent analysis showing that nearly 1 in 5 transplanted DDKs are now allocated OOS and that these kidneys encompass a broad range of quality.^13,15

As an increasing proportion of kidneys are allocated OOS rather than on the basis of established allocation system prioritization, it is crucial to understand how transplant centers decide which of their own candidates receive these kidneys.¹⁶ Although prior national studies have shown inequitable access to OOS kidneys, interpreting these findings is complicated by differences in the demographics of waitlisted candidates across centers.^11,13 For example, older patients may be more likely to receive OOS kidneys if the transplant centers receiving the most OOS kidneys serve an older population compared with other institutions. Alternatively, centers may be selectively “list-diving” within their waitlisted patients—skipping their own younger, higher-ranked patients in favor of their own older, lower-priority patients.

To understand how transplant centers select recipients for OOS kidneys—and to avoid confounding from center-level differences—we examined OOS allocation patterns within the 5 transplant centers performing the most OOS transplants. We aimed to assess within-center allocation priority of OOS kidney recipients, characterize patients who are skipped by their transplant center compared with those who are not skipped, and describe outcomes among skipped patients. We hypothesized that transplant centers frequently skip patients when placing OOS kidneys, resulting in demographic differences between skipped and never-skipped patients, and that most skipped patients do not ultimately receive a kidney of better quality compared with the kidneys for which they were skipped.

Methods

We performed a retrospective cohort study using US transplant registry data. This study used data from the Scientific Registry of Transplant Recipients. The Scientific Registry of Transplant Recipients data system includes data on all donors, waitlisted candidates, and transplant recipients in the US, 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 Scientific Registry of Transplant Recipients contractors. This study was approved by the Columbia University Irving Medical Center institutional review board. The requirement for informed consent was waived, as this study uses deidentified data from a national transplant registry. This manuscript adheres to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines. All research activities were consistent with the principles of the Declaration of Istanbul.

Identifying All Out-of-Sequence DDK Transplants

We used the 2021–2023 Potential Transplant Recipient kidney data set, which includes approximately 193 million offers for DDK, to identify all single-organ deceased-donor kidney transplants (DDKT) with a match-run date between January 1, 2021, and December 31, 2023 (n = 57,306). We excluded transplants without a corresponding transplant recipient registration form (n = 1551). We identified 158 duplicate match-runs containing identical donor or recipient pairs. For these duplicates, we kept only the first match-run that led to a transplant. We excluded kidneys originating from Puerto Rico (n = 256) and those without an OPO code (n = 10).

We identified transplanted kidneys that were placed OOS, where transplants were classified as OOS if the following conditions were met:

• 1.The match-run contained a conventional OOS code (861, 862, or 863)

OR

• 2.The match-run contained an “other, specify” code (799, 898) with one of the following OOS-equivalent terms in the free-text field: “out of sequence,” “open offer,” “expedited,” “aggressive”

AND

• 3.The OOS or OOS-equivalent code appeared earlier in the match-run than the organ acceptance code

This classification approach was adapted from previously published analyses of OOS allocation and was designed to capture both explicit and free-text documented OOS placements.¹⁵

Categorizing Transplant Centers by Volume of OOS Transplants Performed

We classified transplant centers by the number of OOS kidney transplants performed during the study period. Our primary analysis included only those transplant centers with ≥ 200 OOS kidney transplants. We performed a sensitivity analysis that included all transplant centers (Supplementary Methods).

Identifying Skipped and Never-Skipped Candidates

For each OOS transplant, we defined “skipped” candidates as those who were listed at the same transplant center that received the OOS kidney and who appeared higher in the match-run than the eventual recipient. Skipped candidates were therefore skipped over by their own transplant center for a given OOS kidney. We identified “never-skipped” candidates as those who received a transplant with a kidney placed OOS with no previous instances of being skipped over for an OOS placement by their own transplant center during the study period. If an offer was refused on behalf of a patient or if that patient was bypassed in the setting of standard allocation, this was not counted as a “skip” (Supplementary Tables S3 and S4).

Variable Handling and Missing Data

Candidate characteristics were included as coded in the data set. The first event was defined by either first skip (for skipped candidates) or receipt of an OOS transplant (for never-skipped candidates). Dialysis time was defined by the number of months between dialysis initiation and first event, excluding candidates who were not on dialysis at the time of their first event (i.e., those listed preemptively). Because of low cell frequencies, we categorized candidate race as “Other or Multiracial” if they identified with the following groups: Native American, Pacific Islander, or multiracial. Education level was categorized as 'high school or less” versus “college or more,” and primary insurance payer was classified as either government-sponsored (e.g., Medicare) or private. Details on missing data handling are provided in the Supplementary Methods.

Outcomes were categorized on the basis of candidate removal codes between the date of first skip and the end of the study period (December 31, 2023) using the cand_kipa file. Candidates with discordant outcomes in the Potential Transplant Recipient file versus the cand_kipa file (e.g., the candidate removal code indicated that the patient received a DDKT, but no record of the DDKT existed in the Potential Transplant Recipient file) were excluded (n = 199), as were candidates who died before first skip (n = 4) (Supplementary Methods).

Analysis

We used descriptive statistics to compare characteristics of skipped candidates with those of never-skipped candidates, using chi-square tests and Kruskal-Wallis tests to compare categorical variables and continuous variables, respectively.

To assess factors associated with an increased likelihood of being skipped for an OOS kidney, we constructed mixed effects models with transplant center as a random effect to account for center-level variation in waitlist length and allocation practices. The fixed effects were age, gender, race, ethnicity, education, payor, preemptive listing, and diabetes. Dialysis duration was not included in the regression model because it was intrinsically linked with preemptive listing status. However, we performed a sensitivity analysis including dialysis duration instead of preemptive listing status, in which patients listed preemptively were treated as having 0 dialysis time (Supplementary Table S5).

For patients in each outcome category (e.g., death, waitlist removal), we described demographic and clinical characteristics and calculated the median number of skipped offers and the median time-to-event, defined as the number of days between the first skipped offer and the terminal event (transplant, death, waitlist removal, or end of the study period).

For each DDK, we calculated the KDPI from the 10-variable kidney donor risk index using the scaling factor for the year in which the match run occurred. KDPI is a percentile measure of DDK quality, with lower values indicating higher-quality kidneys and better expected graft survival. Because KDPI is recalibrated annually relative to the national donor pool, comparisons of KDPI across calendar years should be interpreted as relative rather than absolute measures of graft-failure risk. For skipped patients, we calculated the median KDPI of their first skipped offer, their final skipped (or accepted) offer, their lowest-KDPI offer, and their highest-KDPI offer.

For skipped patients who later underwent DDKT, we examined the relationship between time from the best skipped offer to eventual transplantation and the difference in KDPI between these kidneys using a Spearman correlation coefficient to assess whether longer waiting after being skipped was associated with receipt of a higher-quality kidney. As a sensitivity analysis, we repeated key outcome analyses after restricting the cohort to candidates whose first skip or OOS transplant event occurred before January 1, 2023, ensuring at least 1 year of follow-up.

We used box and whisker plots to show how many patients were being skipped within each transplant center as an absolute value and as a proportion of the within-center list. We also examined center-reported refusal codes for skipped candidates.

To examine intercenter variability, we graphed the total number of OOS kidney transplants performed at 181 transplant centers over the study period. We grouped transplant centers by the number of OOS kidney transplants performed (< 50, 50–99, 100–199, or 200+) over the study period. Within each group, we calculated the position of the median OOS kidney transplant recipient on the within-center match-run and the percentage of OOS kidneys being allocated to the top priority within-center candidate, which was defined as the highest-ranked candidate on the transplant center’s match run for a given offer.

We defined statistical significance as a 2-sided alpha of < 0.05. All statistical analyses were performed using Stata/MP 18.0 (StataCorp, College Station, Texas).

Result

Cohort Characteristics

We included 1598 OOS kidneys across 5 transplant centers and 56 OPOs (Supplementary Figure S1). Among candidates receiving at least 1 OOS offer, 522 (9.3%) were never skipped (waitlisted patients who received an OOS transplant without ever being skipped) and 5082 (90.7%) were skipped at least once by their transplant center.

Never-skipped candidates were older than the skipped cohort (63 vs. 56 years, P < 0.01), had a higher prevalence of public insurance (61% vs. 56%, P < 0.01) (Table 1), were more likely to identify as White or Asian (P < 0.01), and were less likely to be Black (21% vs. 28%, P < 0.01). Never-skipped candidates were also more likely to be preemptive transplant candidates (40% vs. 23%, P < 0.01) and had shorter median dialysis duration at the time of first skip or transplant (17 vs. 24 months, P < 0.01).

Table 1.

Baseline demographic and clinical characteristics of never-skipped vs. skipped patients listed for deceased-donor kidney transplant at the 5 transplant centers performing 200+ OOS kidney transplants in 2021–2023. A mixed effects model with transplant center as the random effect was used to generate adjusted odds of being skipped

Candidate characteristic	Never skipped candidates	Skipped candidates	aOR of being skippeda
	522	5082
Age at first event	63 (55–70)	56 (45–65)	0.96 (0.95–0.96)
Gender
Male	357 (68%)	3282 (65%)	Ref
Female	165 (32%)	1800 (35%)	1.17 (0.96–1.43)
Race/ethnicity
White	225 (43%)	1936 (38%)	Ref
Asian	69 (13%)	426 (8%)	0.60 (0.44–0.81)
Black	110 (21%)	1442 (28%)	1.13 (0.87–1.47)
Hispanic	116 (22%)	1148 (23%)	0.83 (0.64–1.10)
Other or multiracial	2 (0%)	130 (3%)	4.01 (0.97–16.52)
Education
College or more	330 (63%)	3257 (64%)	Ref
High school or less	187 (36%)	1777 (35%)	0.88 (0.72–1.08)
Primary payor
Government	318 (61%)	2852 (56%)	Ref
Private	203 (39%)	2201 (43%)	1.00 (0.82–1.22)
Preemptive transplant
No	313 (60%)	3892 (77%)	Ref
Yes	209 (40%)	1190 (23%)	0.55 (0.45–0.68)
Dialysis time (mo)	17 (10–32)	24 (14–41)	----
Diabetes
No	289 (55%)	2903 (57%)	Ref
Yes	233 (45%)	2179 (43%)	1.09 (0.90–1.32)

aOR, adjusted odds ratio.

^aAdjusted for age, gender, race, education, insurance type, preemptive status, and diabetes.

Odds of Being Skipped

In mixed effects models with transplant center as a random effect, older age was associated with lower odds of a candidate ever being skipped (adjusted odds ratio, 0.96 per year; 95% confidence interval, 0.95–0.96). Asian race (adjusted odds ratio, 0.60; 95% confidence interval, 0.44–0.81; ref: White) and preemptive transplant listing (adjusted odds ratio, 0.55; 95% confidence interval 0.45–0.68; ref: listed after dialysis initiation) were also linked to lower odds of a candidate ever being skipped (Table 1).

Outcomes Among Skipped Patients

Among skipped patients, 3% later died on the waitlist, 11% were removed from the waitlist, 44% received a single-kidney DDKT, 1% later received a dual-kidney DDKT, 9% received a living-donor kidney transplant, and 31% remained on the waitlist at the end of the study period (Table 2). Patients who died or were removed from the waitlist were skipped a median of 5 and 4 times, respectively. The median time from first skip to death or removal was 212 and 276 days, respectively (Table 2).

Table 2.

Outcomes among candidates ever skipped by their transplant center for an OOS kidney transplant within the 5 transplant centers performing 200+ OOS kidney transplants in 2021–2023

Candidate/offer characteristic	Died on WL	Transplanted (DD)	Transplanted (LD)	Removed	Remaining on WL
	N = 155 (3%)	N = 2311 (45%)	N = 467 (9%)	N = 572 (11%)	N = 1,577 (31%)
Median (IQR)
# of skips before event	5 (1–10)	3 (1–8)	2 (1–5)	4 (2–10)	5 (2–14)
Days between first skip and event	212 (94–415)	83 (25–233)	100 (37–208)	276 (117.5–481)	248 (63–562)
KDPI of first skipped offer	61 (45–79)	60 (38–74)	49 (28–68)	60 (38–75)	59 (36–74)
KDPI of last skipped or accepted DDK offer	54 (36–77)	49 (30–69)	48 (28–68)	54.5 (34–73)	47 (20–70)
Lowest KDPI skipped	28 (18–52)	32 (18–52)	28 (12–47)	29 (14–48.5)	20 (12–38)
Highest KDPI skipped	80 (67–91)	77 (61–87)	68 (51–83)	80 (62–91)	78 (67–90)

DDK, deceased donor kidney; KDPI, Kidney Donor Profile Index; DD, deceased donor; LD, living donor; WL, wait list.

Among skipped patients who eventually underwent a DDKT, the median wait time from first skip to transplant was 83 days. The median patient received a kidney with a KDPI 13 points higher than their best skipped offer (Figure 1a) but 6 points lower than their first skipped offer (Table 2). Longer time spent waiting after the lowest-KDPI skipped offer was associated with receipt of a higher KDPI transplant (Figure 1b). Sensitivity analyses restricted to candidates with at least 1 year of follow-up showed similar findings, including a comparable distribution of KDPI differences between skipped and subsequently accepted deceased-donor kidneys (Supplementary Table S6; Supplementary Figure S2).

Figure 1.

Describing the quality of lowest-KDPI skipped offer vs accepted offer among patients who were initially skipped for an OOS kidney transplant by their transplant center and later received a deceased-donor kidney transplant. (a) Distribution of the difference between the KDPI of the lowest-KDPI skipped offer and the KDPI of the accepted offer. (b) Time elapsed between lowest-KDPI skipped offer and accepted offer vs KDPI difference between lowest-KDPI skipped offer and accepted offer.

Transplant centers placed OOS DDKs with their top-ranked candidate in only 2% of cases and skipped a median of 22 of their own candidates (10% of the within-center match-run) when selecting a recipient (Table 3, Figure 2). Among the kidneys with best predicted longevity (KDPI ≤ 20), only 3% went to the top priority within-center candidate. Transplant centers performing fewer than 200 OOS transplants (not the primary study population) allocated the OOS kidney to the top-priority within-center candidate in 7% of cases and skipped between 8%–13% of their within-center match-run (Table 3).

Table 3.

Transplant center list-diving behavior according to quantity of OOS kidney transplants performed over 3 years (2021–2023)

Transplant center category (# of OOS kidney transplants)	Transplant centers (n)	# OOS transplants	Recipient within-center position (percentile)			Median % of within-center list skipped	% allocated to top-priority within-center candidate
			Median	25th	75th		All OOS kidneys	OOS kidneys with KDPI ≤ 20
< 50	141	1924	7	8	18	8	7	11
50–99	28	1947	9	8	25	8	7	14
100–199	7	966	8.5	13	22	13	7	14
200+	5	1598	23	11	49	11	2	3
Overall	181	6435	11	10	29	10	6	11

KDPI, Kidney Donor Profile Index; OOS, out-of-sequence.

Figure 2.

How high-volume OOS transplant centers placed OOS kidneys, 2021-2023. (a) Position of the recipient on the within-center match-run. (b) Proportion of the within-center match-run skipped.

A sensitivity analysis performed using all 181 transplant centers in the Scientific Registry of Transplant Recipients data set yielded similar findings, with younger candidates and those with longer dialysis duration being more likely to be skipped and a minority of skipped candidates later receiving a DDKT (Supplementary Tables S1 and S2).

Candidate-related factors, including temporary medical ineligibility and candidate refusal, and others, accounted for 12.6% of all center-reported refusal codes (Supplementary Tables S3 and S4).

Discussion

In this retrospective cohort study of US national transplant registry data, we found that transplant centers receiving large numbers of OOS kidneys heterogeneously allocated these kidneys, placing OOS kidneys with the highest-priority patient within their transplant center in only 2% of cases. Among skipped candidates, fewer than half later received a DDKT within the study period, and those who did often receive a higher-KDPI (i.e., higher risk of graft failure) kidney than that of the lowest-KDPI kidney for which they were skipped over. As OOS kidney transplants now make up nearly a fifth of all US deceased donor kidney transplants, it is critical to understand how centers are placing these kidneys.

Allocation of deceased donor kidneys in the US is primarily based on waiting time, with some exceptions for special circumstances, such as highly sensitized patients, former living donors, and pediatric candidates.³ Given the wide range of expected longevity of donated kidneys, it is often the case that a lower-quality organ may not be appropriate for the highest-priority waitlisted candidate, as that candidate will likely receive a higher-quality offer in the future. In OOS allocation, centers have the flexibility to place a kidney with whichever candidate on their waitlist they feel is most appropriate, regardless of allocation priority, and it is therefore not surprising that OOS kidneys were usually not given to the top-ranked candidate. However, we found that the majority of skipped patients had outcomes suggesting that they may have fared better by receiving one of the OOS organs for which they were passed over in favor of a lower-priority candidate—14% died on the waitlist or were removed without transplant, and 31% remained on the waitlist at the end of the study period, an undesirable outcome given that longer duration of dialysis exposure is associated with greater risk of mortality and runs counter to patient preferences.^17,18 Among the 45% who eventually received a DDKT, the median patient waited an additional 83 days. Longer time spent waiting after the best skipped offer was associated with receipt of a higher KDPI transplant, suggesting that prolonged waiting after being skipped did not typically result in access to a better-quality kidney.

Our analysis of skipped versus never-skipped patients shows that transplant centers are more likely to skip over younger patients and those with greater dialysis exposure. This finding is consistent with a prior analysis comparing recipients of OOS versus in-sequence kidneys at the national level, which found that the median recipient of an OOS kidney was 5 years older and had spent 1.6 fewer years on dialysis than the median recipient of an in-sequence kidney.¹³ Given that some OOS kidneys are at higher risk for graft failure, it is not surprising that centers bypass younger candidates who require longer graft survival compared with older candidates. However, the median KDPI of these OOS kidneys was 56%, suggesting they included a broad range of kidneys, and even the lowest-KDPI (≤ 20%) kidneys only went to the top-ranked within-center candidate 3% of the time. The finding that candidate-specific refusal codes accounted for 12.6% of all refusal codes suggests that candidate medical ineligibility, refusals, and other candidate-specific factors did not usually explain the decision to skip a candidate.

These findings contribute important context to the current discourse surrounding the rise of OOS allocation, which was intended only as a salvage allocation strategy but is increasingly being used to place kidneys across the quality spectrum.^11,13,15 Scrutiny of OOS allocation has been increasing among both the media and the federal government; in February 2025, the Health Resources and Services Administration issued a letter directing the Organ Procurement and Transplantation Network to develop a compliance plan that curtails OOS offers, noting that current OOS practices do not comply with Section 121.7(f) of the Organ Procurement and Transplantation Network Final Rule (Identification of Organ Recipient – Wastage) and that organs are to be allocated to patients and not to transplant centers.¹⁹ Much of the increase in OOS has been attributed to an increasingly complex allocation system that has led to a marked growth in organ offers, alongside increasingly regulatory pressures on OPOs to improve organ use rates.^4,7,8 However, our findings suggest that in the absence of structured and transparent decision-making about OOS placement, the increased use of this allocation strategy threatens the fairness and objectivity of the transplant system.

Strengths of this analysis include the within-center design, which isolates transplant-center decision-making from OPO behavior and local demographics. Although our primary analysis focused on the 5 transplant centers performing the most OOS transplants, our sensitivity analysis performed on the entire national data set yielded similar results. Limitations of this analysis include the absence of detailed, case-by-case information surrounding how placement decisions were made. Future studies could incorporate qualitative components, such as interviews of transplant center staff and surgeons—to understand their decision-making. The follow-up period was also relatively short, and many patients who were still awaiting transplant at the end of our study period may have later received transplants. Finally, although KDPI is a useful and well-validated marker of organ longevity, it is not the only indication of organ quality and does not consider all organ and patient-specific factors.

Conclusion

In conclusion, we found that transplant centers skip a median of 22 candidates (10% of the within-center match run) when allocating an OOS kidney for transplant, and most skipped candidates are either removed from the waitlist without a transplant, die on the waitlist, or are later transplanted with a kidney with a higher estimated risk of graft failure than the kidneys for which they were skipped. These findings highlight the importance of transparency and structure in out-of-sequence kidney placement and should inform ongoing efforts to design expedited pathways for kidneys that are at risk of discard.

Disclosure

All the authors have no competing interests to disclose.

Supplementary Material

Supplementary File (PDF)

Supplementary Methods. Figure S1. Flow diagram of study population. Figure S2. Distribution of the difference in kidney donor quality, measured by KDPI, between the best skipped OOS kidney offer and the subsequently accepted deceased-donor kidney transplant among candidates initially skipped for OOS transplantation. This analysis was restricted to patients who were first skipped before Jan 1, 2023 to allow at least 1 year of follow-up. Table S1. Baseline demographic and clinical characteristics of never-skipped vs. skipped patients listed for deceased-donor kidney transplant at all US transplant centers, 2021–2023. A mixed effects model with transplant center as the random effect was used to generate adjusted odds of being skipped. Table S2. Outcomes among candidates ever skipped by their transplant center for an OOS kidney transplant within the 5 transplant centers performing 200+ OOS kidney transplants in 2021–2023. Table S3. Prevalence of center-reported refusal codes by category, excluding bypass codes. Table S4. Categorization of specific center-reported refusal codes. Table S5. Baseline demographic and clinical characteristics of never-skipped vs. skipped patients listed for deceased-donor kidney transplant at the 5 transplant centers performing 200+ OOS kidney transplants in 2021–2023. A mixed effects model with transplant center as the random effect was used to generate adjusted odds of being skipped. Unlike Table 2, dialysis time was used instead of preemptive dialysis status in the mixed effects model. Table S6. Outcomes among candidates ever skipped for a locally allocated OOS kidney transplant within the 5 transplant centers performing 200+ OOS kidney transplants in 2021–2023. This analysis was restricted to patients who were first skipped before Jan 1, 2023 to allow at least 1 year of follow-up.