Transplant Center Variability in Organ Offer Acceptance and Mortality Among US Patients on the Heart Transplant Waitlist | Cardiology | JAMA Cardiology | JAMA Network
[Skip to Navigation]
Figure 1.  Flow Diagram of Cohort Selection
Flow Diagram of Cohort Selection
Figure 2.  Per-Center Acceptance Rates and Waitlist Mortality
Per-Center Acceptance Rates and Waitlist Mortality

A, Distribution of unadjusted per-center offer acceptance rate. Centers were stratified by United Network of Organ Sharing (UNOS) region. The diameter of each circle corresponds to the number of offers received by the program during the study period. B, Association between unadjusted per-center acceptance rate and unadjusted per-center 1-year cumulative incidence of waitlist mortality among transplant candidates who received a first-rank offer. C, Association between unadjusted per-center acceptance rate and unadjusted per-center 1-year cumulative incidence of transplant among transplant candidates who received a first-rank offer. Each dot represents a transplant center. Locally weighted scatterplot smoothing curve is added to the plot to help visualize the association.

Figure 3.  Cumulative Incidence Function of Waitlist Mortality for Candidates Listed for Heart Transplant
Cumulative Incidence Function of Waitlist Mortality for Candidates Listed for Heart Transplant

Stratified by transplant center–adjusted acceptance rate.

Table 1.  Demographics by First-Rank Offer Acceptance
Demographics by First-Rank Offer Acceptance
Table 2.  Summary of Match Runs and Distribution of LVAD by Transplant Centers Stratified by Adjusted Center Acceptance Rates
Summary of Match Runs and Distribution of LVAD by Transplant Centers Stratified by Adjusted Center Acceptance Rates
1.
Colvin  M, Smith  JM, Hadley  N,  et al.  OPTN/SRTR 2017 annual data report: heart.   Am J Transplant. 2019;19(suppl 2):323-403. doi:10.1111/ajt.15278 PubMedGoogle ScholarCrossref
2.
Kasiske  BL, Wey  A, Salkowski  N,  et al.  Seeking new answers to old questions about public reporting of transplant program performance in the United States.   Am J Transplant. 2019;19(2):317-323. doi:10.1111/ajt.15051 PubMedGoogle ScholarCrossref
3.
Goldberg  DS, French  B, Lewis  JD,  et al.  Liver transplant center variability in accepting organ offers and its impact on patient survival.   J Hepatol. 2016;64(4):843-851. doi:10.1016/j.jhep.2015.11.015 PubMedGoogle ScholarCrossref
4.
Wey  A, Valapour  M, Skeans  MA,  et al.  Heart and lung organ offer acceptance practices of transplant programs are associated with waitlist mortality and organ yield.   Am J Transplant. 2018;18(8):2061-2067. doi:10.1111/ajt.14885 PubMedGoogle ScholarCrossref
5.
Wey  A, Salkowski  N, Kasiske  BL, Israni  AK, Snyder  JJ.  Influence of kidney offer acceptance behavior on metrics of allocation efficiency.   Clin Transplant. 2017;31(9). doi:10.1111/ctr.13057 PubMedGoogle Scholar
6.
Massie  AB, Kucirka  LM, Segev  DL.  Big data in organ transplantation: registries and administrative claims.   Am J Transplant. 2014;14(8):1723-1730. doi:10.1111/ajt.12777 PubMedGoogle ScholarCrossref
7.
Muller  CJ, MacLehose  RF.  Estimating predicted probabilities from logistic regression: different methods correspond to different target populations.   Int J Epidemiol. 2014;43(3):962-970. doi:10.1093/ije/dyu029 PubMedGoogle ScholarCrossref
8.
Kilic  A, Weiss  ES, Allen  JG,  et al.  Simple score to assess the risk of rejection after orthotopic heart transplantation.   Circulation. 2012;125(24):3013-3021. doi:10.1161/CIRCULATIONAHA.111.066431 PubMedGoogle ScholarCrossref
9.
Hong  KN, Iribarne  A, Worku  B,  et al.  Who is the high-risk recipient? predicting mortality after heart transplant using pretransplant donor and recipient risk factors.   Ann Thorac Surg. 2011;92(2):520-527. doi:10.1016/j.athoracsur.2011.02.086 PubMedGoogle ScholarCrossref
10.
Bergenfeldt  H, Stehlik  J, Höglund  P, Andersson  B, Nilsson  J.  Donor-recipient size matching and mortality in heart transplantation: influence of body mass index and gender.   J Heart Lung Transplant. 2017;36(9):940-947. doi:10.1016/j.healun.2017.02.002 PubMedGoogle ScholarCrossref
11.
Kilic  A, Emani  S, Sai-Sudhakar  CB, Higgins  RS, Whitson  BA.  Donor selection in heart transplantation.   J Thorac Dis. 2014;6(8):1097-1104.PubMedGoogle Scholar
12.
Madan  S, Saeed  O, Shin  J,  et al.  Donor troponin and survival after cardiac transplantation: an analysis of the United Network of Organ Sharing Registry.   Circ Heart Fail. 2016;9(6):e002909. doi:10.1161/CIRCHEARTFAILURE.115.002909 PubMedGoogle Scholar
13.
Jay  C, Schold  JD.  Measuring transplant center performance: the goals are not controversial but the methods and consequences can be.   Curr Transplant Rep. 2017;4(1):52-58. doi:10.1007/s40472-017-0138-9 PubMedGoogle ScholarCrossref
14.
Kirklin  JK, Pagani  FD, Kormos  RL,  et al.  Eighth annual INTERMACS report: special focus on framing the impact of adverse events.   J Heart Lung Transplant. 2017;36(10):1080-1086. doi:10.1016/j.healun.2017.07.005 PubMedGoogle ScholarCrossref
15.
Linblad K, Lehman RR. Four-month monitoring of heart allocation proposal to modify the heart allocation system. Accessed February 25, 2020. https://optn.transplant.hrsa.gov/media/2924/post-implementation-heart-policy-report_20190417_ready-to-post.pdf
16.
Klassen  DK, Edwards  LB, Stewart  DE, Glazier  AK, Orlowski  JP, Berg  CL.  The OPTN Deceased Donor Potential Study: implications for policy and practice.   Am J Transplant. 2016;16(6):1707-1714. doi:10.1111/ajt.13731 PubMedGoogle ScholarCrossref
Original Investigation
April 15, 2020

Transplant Center Variability in Organ Offer Acceptance and Mortality Among US Patients on the Heart Transplant Waitlist

Author Affiliations
  • 1Medical student, School of Medicine, Duke University, Durham, North Carolina
  • 2Division of Cardiothoracic Surgery, Duke University Medical Center, Durham, North Carolina
  • 3Department of Biostatistics and Bioinformatics, Duke University, Durham, North Carolina
  • 4Division of Cardiology, Duke University Medical Center, Durham, North Carolina
JAMA Cardiol. 2020;5(6):660-668. doi:10.1001/jamacardio.2020.0659
Key Points

Question  Is transplant center variability in offer acceptance associated with mortality among patients on the heart transplant waitlist?

Findings  In this cohort study of 9628 candidates for heart transplant over 10 years, every 10% increase in adjusted center acceptance rate of offers made to the highest-priority candidates was associated with a 27% reduction in the rate of mortality among patients on the waitlist, with no detriment in 5-year adjusted posttransplant patient survival or graft failure.

Meaning  The center-level decision to accept or decline an allograft may represent a modifiable behavior that is associated with mortality among patients on the heart transplant waitlist.

Abstract

Importance  Under the current Centers for Medicare & Medicaid Services guidelines, there is incentivization to optimize posttransplant outcomes regardless of mortality among patients on the waitlist and transplant rates; few data exist with regard to transplant center acceptance practices and survival to heart transplant.

Objectives  To evaluate the extent of variability in organ acceptance practices in the US and whether this center-level behavior is associated with heart transplant candidate survival.

Design, Setting, and Participants  In this retrospective cohort study, the US National Transplant Registry was queried for all match runs of adult candidates listed for isolated heart transplant between May 1, 2007, and March 31, 2017. Data analysis was conducted from October 30, 2018, to May 1, 2019. The final cohort included 93 transplant centers, 19 703 donors, and 9628 candidates.

Main Outcomes and Measures  Center acceptance rates for heart allografts offered to the highest-priority candidates, association between center acceptance rate and mortality among patients on the waitlist, and posttransplant outcomes between candidates who accepted their first-rank offers vs those who accepted previously declined offers.

Results  Among 19 703 unique organ offers, 6302 hearts (32.0%) were accepted for first-rank candidates. After adjustment for donor, candidate, and geographic covariates, transplant centers varied in acceptance rates (12.3%-61.5%) of offers made to first-rank candidates. Higher acceptance rates were associated with lower cumulative incidence of 1-year mortality among patients on the waitlist. For every 10% increase in adjusted center acceptance rate, the risk of mortality decreased by 27% (subdistribution hazard ratio, 0.73; 95% CI, 0.67-0.80). No statistically significant difference was observed in 5-year adjusted posttransplant patient survival (adjusted hazard ratio, 1.02; 95% CI, 0.94-1.11) and graft failure (subdistribution hazard ratio; 0.95; 95% CI, 0.83-1.09) between hearts accepted at the first-rank compared with lower-rank positions.

Conclusions and Relevance  Variability in heart allograft acceptance rates appears to exist among transplant centers, with candidates listed at lower acceptance rate centers being more likely to experience mortality while on the waitlist. Comparable posttransplant survival suggests that allografts that were declined as a first offer perform as well as those that were accepted at their first offer. These findings suggest that organ acceptance rate or time to transplant from being added to the waitlist may be an additional measure of heart transplant program performance.

Introduction

Health services research performed in urgency-driven heart transplant has traditionally assessed outcomes beginning at the time of transplant. Although these studies generate important findings related to the benefit of transplant in recipients, they fail to examine a larger denominator comprising candidate recipients who must survive to transplant before any measure of posttransplant survival can be calculated. At present, the mean mortality among adults on the waitlist (hereafter referred to as waitlist mortality) for a heart transplant in the US is 12 deaths per 100 waitlist years.1 When considering the sickest patients in the previous status 1A group, this rate exceeds 30 deaths per 100 waitlist years.1 To our knowledge, this information is not yet available for the revised allocation system.

Survival to transplant and practice patterns of transplant centers remain poorly understood.2-5 The association between transplant centers’ organ offer acceptance behavior and waitlist mortality was first described for kidney and liver transplant, demonstrating that candidates listed at centers with high acceptance rates of first-rank organ offers had comparable posttransplant survival but reduced risk of waitlist mortality.3,5 For thoracic transplant, Wey et al4 reported, for the first time in their 1-year outcomes assessment in the United Network of Organ Sharing database, the association between organ offer acceptance practices and survival among patients on the waitlist.

Because each center must elect to accept or decline an allograft before a candidate may proceed to transplant, center-level acceptance patterns represent a modifiable behavior that may substantially affect equitable organ allocation and overall mortality among candidates who are on the waitlist. We sought to examine (1) within and across-region variability in center acceptance patterns for heart allografts offered to the highest-priority candidates, (2) the association between acceptance behavior and candidate outcomes, and (3) the posttransplant outcomes in candidates who accepted the first-rank offer vs a previously declined offer.

Methods
Data Source

We performed a retrospective cohort analysis using the United Network of Organ Sharing Standard Analysis and Research data. These data were subsequently linked with information from the Potential Transplant Recipient file, which provides match-run information for every offer made for each donor heart allograft that was ultimately used for transplant. These data sets have been described previously.6 Full institutional review board approval was granted by Duke University before initiation of this study; informed consent is not separately pursued by the institution using the database for research. The data source used for this analysis included all candidates wait-listed for heart transplant from May 1, 2007, to March 31, 2017. Data analysis was conducted from October 30, 2018, to May 1, 2019. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.

Study Population and Cohort Determination

The Potential Transplant Recipient file was queried for all match runs for adult (age ≥18 years at the time of being wait-listed for transplant) candidates wait-listed for isolated heart transplant who received a genuine first-rank offer for a heart allograft. Only match runs that resulted in transplant were considered. Offers that were determined to be bypassed—an uncommon event in which an organ is offered first to a lower-rank candidate through means such as directed donation, natural disaster, or donor medical urgency—were not considered to be genuine and were excluded. Offers in which critical donor-specific or candidate-specific information (eMethods in the Supplement) was missing were excluded. In addition, transplant centers that received fewer than 10 first-rank offers in a year were excluded because acceptance patterns may be artificially variable owing to small sample size (649 first-rank offers made to 43 centers during the study period). Complete cohort selection is shown in Figure 1.

Statistical Analysis

For the model of first-rank offer acceptance, the unique match runs were the units of analysis. The unadjusted center acceptance rate was calculated by dividing the number of accepted first-rank offers by the total number of first-rank offers received at the center during the study period. To estimate the adjusted center acceptance rate, we modeled the outcome of acceptance of the first-rank offer on the transplant center using logistic regression. We used marginal standardization to estimate the adjusted acceptance rate for each center,7 which allows inference to the total first-rank heart allograft offers. We estimated the counterfactual acceptance probability for each offer as if it was listed at all 93 transplant centers and then calculated the mean estimated probability for each center.

We examined left ventricular assist device (LVAD) use patterns in candidates who received a first-rank offer. The proportion of such candidates was calculated within each adjusted acceptance rate group.

For the model of waitlist mortality, the unique candidates who received first-rank offers were the units of analysis. Competing risk analysis using a Fine-Gray subdistribution hazards model was performed to assess the association between the primary exposure—adjusted center acceptance rate—and the waitlist mortality outcome. Waitlist mortality was defined as removal from the waitlist owing to death or clinical deterioration, including death while on the waitlist or removal for being too sick to undergo a transplant. The unadjusted overall and center-specific cumulative incidence functions of waitlist mortality were estimated and assessed at 1 year after the patient was added to the waitlist. Because each center had a unique adjusted acceptance rate, this rate was included as a continuous covariate in the model for waitlist mortality. To improve interpretability, we also grouped adjusted center acceptance rates into approximate quartiles (<25.0%, 25.0%-29.9%, 30.0%-39.9%, and ≥40.0%) and estimated the cumulative incidence functions for each group. Adjusted cumulative incidences for adjusted center acceptance rate groups were estimated using the marginal standardization method as performed in the acceptance rate model.

The events of interest were patient survival and graft failure. Graft failure was defined as an event in which the recipient’s transplanted organ was removed, the recipient died, or the recipient began receiving long-term allograft support. The allografts were the units of analysis, and they were classified into 2 groups based on whether they were accepted by the highest-priority candidates. Unadjusted patient survival and cumulative incidence of graft failure were estimated yearly up to 5 years after transplant for allografts accepted by first-rank and lower-rank patient centers using the Kaplan-Meier and competing risks methods. Adjusted patient survival probabilities were estimated using the Cox proportional hazards regression model, and adjusted cumulative incidences of graft failure were estimated using the Fine-Gray subdistribution hazards model.

Covariates included in the adjustment of acceptance rate, waitlist mortality, and posttransplant outcomes were based on publicly available Scientific Registry of Transplant Recipients models (eMethods in the Supplement). All statistical analyses were performed with SAS, version 9.4 (SAS Institute Inc), and graphics were generated by the ggplot2 package in R, version 3.5.1 (R Core Team). A 2-sided significance level of P = .05 was used for all statistical tests.

Results
Study Population

A total of 9628 candidates across 93 transplant centers received first-rank offers from 19 703 unique donors after application of the inclusion criteria (Figure 1). The overall acceptance rate for the first-rank offer was 32.0% (n = 6302). The demographic information for candidates and donors is reported in Table 1. Candidates whose centers accepted the first-rank offer tended to be older, male, and had no prior cardiac surgery at the time of listing. Every candidate is listed for transplant. These first-rank candidates’ centers were more likely to accept organs from donors who were younger, male, and without public health service (PHS)–determined increased risk of infectious diseases (hepatitis B, hepatitis C, and HIV infection) or a history of cocaine use, intravenous drug use, and tobacco use. Of the 5606 candidates who declined the initial first-rank offer (total number of declined offers in Table 1 is higher because some candidates declined multiple first-rank offers), 1932 candidates (34.5%) did not receive a subsequent first-rank offer. At 1 year after declining the initial first-rank offers, 4230 candidates (75.5%) eventually underwent transplant and 455 candidates (8.1%) were removed from the waitlist owing to death or decompensation. A comparison of demographic information between missing and complete cases for all offers is provided eTable 1 in the Supplement, and comparison between high-volume and low-volume centers is provided in eTable 2 in the Supplement.

Variability of Per-Center Organ Offer Acceptance Rate

The rate at which centers accepted a first-rank offer for a candidate on their waitlist varied significantly from 12.3% to 61.5%, as depicted in Figure 2A. After adjusting for confounders, the offer acceptance rates were still significantly different across centers (test of center effect, Wald χ2, 1078.8292; P < .001). Table 2 summarizes the number of first-rank offers and unique candidates at centers grouped by adjusted per-center acceptance rates. There were 3409 candidates (35.4%) from centers with first-rank offer acceptance rates below 30.0%, but the same centers received approximately half of all first-rank offers (9004 [45.7%]). Candidates from centers with acceptance rates lower than 25.0% received a mean (SD) of 2.8 (4.6) first-rank offers, and those from centers with acceptance rates greater than 40.0% received a mean (SD) of 1.5 (1.5) first-rank offers.

Offer and Donor Characteristics by Acceptance Rates

Candidate status at the time of match run, share type, and donor characteristics, including age and PHS-determined increased risk of infectious diseases, were stratified by centers’ first-rank offer acceptance rates (eTable 3 in the Supplement). At centers with the highest offer acceptance rates (≥40.0%), 1423 of 2878 first-rank offers (49.4%) for status 1A candidates were accepted, and 981 of 4860 first-rank offers (20.2%) for status 1A candidates were accepted at the lowest acceptance rate centers (<25.0%). Similarly, the highest acceptance centers accepted 442 of 1175 first-rank offers (37.6%) from donors aged 40 years or older and 286 of 644 PHS-determined donors (44.4%) with increased risk of infectious diseases, whereas the lowest acceptance centers accepted 205 of 1675 first-rank offers (12.2%) from donors aged 40 years or older and 148 of 1031 donors (14.4%) with increased risk of infectious diseases.

LVAD Use in Patients Wait-listed for Heart Transplant by Acceptance Rates

Use of LVAD by adjusted center acceptance rates is reported in Table 2. A total of 2027 LVADs were implanted in candidates awaiting heart transplant over the course of the study period. Of those, 808 LVADs (39.9%) were implanted in centers with acceptance rates below 30.0%, and approximately half of all first-rank offers (9004 [45.7%]) were made to the same centers. Centers with lower acceptance rates had higher percentages of LVADs implanted in candidates while on the waitlist and a greater number of first-rank offers made per candidate.

Per-Center Waitlist Mortality and Transplant

The 1-year unadjusted cumulative incidence of waitlist mortality from the time of receiving the initial first-rank offer was 4.8% (95% CI, 4.4%-5.2%) and for transplant was 86.7% (95% CI, 86%-87.4%). Across centers, unadjusted waitlist mortality ranged from 0% to 22.5%. Figure 2B and C demonstrate the association between unadjusted per-center acceptance rate and per-center, 1-year cumulative incidence of waitlist mortality and transplant, respectively. The unadjusted cumulative incidence functions of waitlist mortality stratified by adjusted acceptance rate groups are depicted in Figure 3. The 1-year cumulative incidence of waitlist mortality ranged from 2.8% (95% CI, 2.2%-3.5%) for candidates at centers with an acceptance rate greater than 40.0% to 7.3% (95% CI, 6.2%-8.4%) for candidates at centers with an acceptance rate lower than 25.0%. By incorporating adjusted center acceptance rate as a continuous variable in the Fine-Gray model of waitlist mortality, a 10% increase in acceptance rate was associated with a 27% lower rate of waitlist mortality (subdistribution hazard ratio [SHR], 0.73; 95% CI, 0.67-0.80), and a 10% increase was associated with a 22% higher rate of transplant (SHR, 1.22; 95% CI, 1.20-1.24).

Posttransplant Outcomes

Complete cases (19 297 of 19 703 allografts) were used to assess posttransplant outcomes. The patient survival probabilities were 90.6% (95% CI, 89.8%-91.3%) for 1 year, 85.2% (95% CI, 84.2%-86.1%) for 3 years, and 80.5% (95% CI, 79.3%-81.6%) for 5 years for allografts accepted at the first-rank position and 90.3% (95% CI, 89.7%-90.8%) for 1 year, 84.4% (95% CI, 83.7%-85.0%) for 3 years, and 78.7% (95% CI, 77.9%-79.5%) for 5 years for allografts accepted at a lower-rank position (eFigure in the Supplement). The adjusted HR for allografts accepted at the first-rank vs lower-rank position was 1.02 (95% CI, 0.94-1.11; P = .62). The cumulative incidences of graft failure were 3.3% (95% CI, 2.8%-3.7%) at 1 year, 5.3% (95% CI, 4.8%-6.0%) at 3 years, and 6.7% (95% CI, 6.0%-7.4%) at 5 years for allografts accepted at the first-rank position. These data were comparable to the adjusted HRs of cumulative incidences of graft failure at 1 year (3.1%; 95% CI, 2.8%-3.4%), 3 years (5.0%; 95% CI, 4.7%-5.5%), and 5 years (6.8%; 95% CI, 6.3%-7.3%) (eFigure in the Supplement). After adjusting for confounders, the rates of graft failure were also comparable (SHR, 0.95; 95% CI, 0.83-1.09; P = .49).

Discussion

This study represents use of a large, national cohort to examine the degree of variability in heart allograft acceptance patterns in the US before the revised allocation system. We observed wide variability in first-rank offer acceptance rates at US heart transplant centers, with lower organ offer acceptance rates being associated with an increased risk of waitlist mortality. After our evaluation of posttransplant outcomes, we noted comparable patient and graft survival in those who accepted their first-rank offer vs those who accepted a lower-rank offer, suggesting that allografts that were previously declined appear to function as well as those that were accepted at their first offer. Such acceptance behaviors may continue to influence the revised allocation system unless centers modify their acceptance behaviors to optimize survival among transplant candidates on the waitlist.

Acceptance patterns ranged from 12% to 62%. Several considerations should be made when evaluating the suitability of an allograft for transplant. In heart transplant, these considerations include candidate, donor, and geographic factors, such as donor ejection fraction, donor-recipient size match, history of cocaine or tobacco use, and cold ischemic time. Although these factors have been shown to be associated with posttransplant survival,8-11 it is rare that the perfect donor is available, and every center has developed its own guideline to balance these factors. As such, the lack of standardization of practice and acceptable range of offer acceptance rates appear to permit such variability. In addition, reasons for declining organ offers vary across centers, but resource limitation and risk aversion may help explain why centers elect to decline extended-criteria donor heart allografts.12 We noted that, compared with lower acceptance centers, higher acceptance centers accepted a larger proportion of first-rank offers from donors with increased risk, including those aged 40 years or older and those with PHS-determined increased risk of infectious diseases.

In 2013, public reporting of transplant outcomes began and became closely tied to funding from the Centers for Medicare & Medicaid Services,2 the largest transplant payer in the US. In the early stages of implementing the public reporting system, large-volume centers encountered a 3-fold higher rate of low-performance evaluations despite efforts to adjust for recipient and donor characteristics.13 Although these performance evaluations consider waitlist mortality, transplant rates, and posttransplant survival, only posttransplant survival outcomes are subject to loss of funding when standards are not met.13 Since the change, commercial payers have adopted the same principles when reviewing first-year patient and graft survival for determination of centers of excellence and funding.13 Candidates are at considerable disadvantage with their chances of receiving a transplant if they lack access to a large-volume center equipped with additional resources (eg, ex vivo perfusion) that can afford to take risks with extended-criteria donor hearts.

Use of continuous-flow LVADs as a bridge to transplant has grown over the past decade,14 yet its association with centers’ organ acceptance behavior and ramifications on the revised allocation system are not well described to date. In this study, the indication for LVAD was assumed to be a bridge to transplant rather than a destination therapy because the candidates received an LVAD after they had been wait-listed for transplant; however, this distinction is not noted in the database. We found that LVAD use was higher in centers with lower acceptance rates—a finding that generates several hypotheses. One possible explanation is that an LVAD was implanted to stabilize patients while awaiting other organs that may be offered at a later date. Alternatively, when centers did not take their first-rank offers, the patient’s condition deteriorated, requiring LVAD support while awaiting the next offer. Although we cannot determine a causal relationship or timing of implantation from our data (whether an LVAD was implanted before or after the first offer was declined), our findings suggest that centers with lower acceptance rates had more LVADs implanted in candidates while they were on the waitlist but also had a higher number of allograft offers and, thus, more opportunities for their candidates to receive transplants. It is also possible that centers with lower acceptance rates accumulate more time on the waitlist and have more exposure time to receive allograft offers. Nevertheless, improved LVAD technology and postoperative management have allowed heart transplant candidates additional time to wait for an alternative offer and allocate allografts to lower-rank patients, which may be associated with inefficiencies within the allocation system that was designed to serve the sickest patients first. The new medical urgency statuses that replaced status 1A, 1B, and 2 in late 2018 provide a clearer distinction between unstable, nondischargeable patients with an LVAD (status 1 and 2) and stable, dischargeable patients with an LVAD (status 3).15 It would be of interest to the transplant community whether this policy change demonstrates a measurable alteration in centers’ organ acceptance behavior and waitlist mortality.

Our study findings generated from 10 years of data suggest that lower acceptance rates of first-rank offers were associated with higher risk of waitlist mortality. The decision to remain on the waitlist does not guarantee another opportunity to receive an allograft offer, nor does LVAD therapy guarantee freedom from complications that may preclude the patient from another operation. For heart transplant, all patients remaining on the waitlist went on to receive at least 1 subsequent offer at any position after declining their initial first-rank offer. However, 455 patients (8.1%) were removed from the waitlist owing to death or decompensation at 1-year follow-up. This finding suggests that every center’s decision to accept or decline an organ offer has an association with mortality among candidates on the waitlist, particularly because patient and graft survival were comparable between the 2 recipient groups, suggesting that previously declined allografts were noninferior.

Limitations

There are several limitations of the study. Retrospective studies using large, national databases have the inherent limitation of unmeasured confounders that cannot be accounted for in the analysis. Such confounders include surgeon experience, time elapsed between when the candidate was added to the waitlist and when the patient received their initial first-rank offer, sensitization status, and organ procurement organizations’ (OPOs’) varying policies in organ acceptance. The association of OPO behavior with transplant candidate outcomes, in particular, is notable because there is regional variation in OPO organ acceptance criteria based on how local transplant centers view organ quality and their available resources.16 The interdependent association between OPO and transplant center is made more complex by the current Centers for Medicare & Medicaid Services Conditions of Participation for OPO performance, which disincentivize OPOs from procuring less-than-ideal donors.16 The ability of centers to screen out a set of prespecified donor characteristics from their match runs poses another limitation and is not captured in the database. As a result, a center’s acceptance rate may be inflated because the denominator excludes offers that were screened out. In addition, we did not examine acceptance rates or posttransplant outcomes stratified by sequence number at which allografts were accepted.

Wey and colleagues4 found that allografts that were offered more than 10 times had substantially lower acceptance rates (3%) compared with those accepted at first offer (28%), but donor ejection fraction stayed constant early and late in the match run. Ejection fraction is 1 marker of donor allograft quality, but their finding suggests that, although acceptance rates vary across an offer sequence, allograft quality may not, which is corroborated in this study. In addition, this study does not reflect the updated allocation system, in which the 3 medical urgency statuses increased to 6 as of October 18, 2018. It is likely too early to measure the outcomes of this updated policy. We anticipate that acceptance behaviors will change as criteria for status assignments have become highly specific, leaving less room for variability in the determination of candidates’ medical urgency.

Conclusions

This study found wide variability in heart acceptance rates among transplant centers, with candidates listed at low first-rank offer acceptance rate centers being more likely to die while on the waitlist. Center-level decision to accept or decline an allograft may represent a modifiable behavior that is associated with equitable organ allocation and a candidate’s risk of waitlist mortality. As such, the findings suggest that organ acceptance rate or time to transplant from being wait-listed may be an additional measure of heart transplant program performance, contributing information beyond waitlist mortality and posttransplant survival.

Back to top
Article Information

Accepted for Publication: February 7, 2020.

Corresponding Author: Ashley Y. Choi, BA, School of Medicine, Duke University Medical Center, Box 3863, Durham, NC 27710 (choi.ashley@duke.edu).

Published Online: April 15, 2020. doi:10.1001/jamacardio.2020.0659

Author Contributions: Ms Choi and Dr Lee had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Choi, Mulvihill, Lee, Kuchibhatla, Schroder, Granger, Hartwig.

Acquisition, analysis, or interpretation of data: Choi, Mulvihill, Lee, Zhao, Kuchibhatla, Patel.

Drafting of the manuscript: Choi, Mulvihill, Lee, Kuchibhatla, Patel, Hartwig.

Critical revision of the manuscript for important intellectual content: Choi, Mulvihill, Lee, Zhao, Kuchibhatla, Schroder, Granger, Hartwig.

Statistical analysis: Choi, Mulvihill, Lee, Zhao, Kuchibhatla, Hartwig.

Obtained funding: Choi, Mulvihill, Hartwig.

Supervision: Schroder, Patel, Granger, Hartwig.

Conflict of Interest Disclosures: Ms Choi reported receiving a grant from the National Institutes of Clinical and Translational Science of the National Institutes of Health during the study. Dr Mulvihill reported receiving a grant from the National Institutes of Health during the study. No other disclosures were reported.

Funding/Support: This study was funded by the Bollinger Scholarship Committee within the Department of Surgery at the Duke University Medical Center, grant TL1TR002555 from the National Institutes of Clinical and Translational Science (Ms Choi), and grant UL1TR002553 from the National Center for Advancing Translational Sciences of the National Institutes of Health (Dr Mulvihill).

Role of the Funder/Sponsor: The funding organizations had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Colvin  M, Smith  JM, Hadley  N,  et al.  OPTN/SRTR 2017 annual data report: heart.   Am J Transplant. 2019;19(suppl 2):323-403. doi:10.1111/ajt.15278 PubMedGoogle ScholarCrossref
2.
Kasiske  BL, Wey  A, Salkowski  N,  et al.  Seeking new answers to old questions about public reporting of transplant program performance in the United States.   Am J Transplant. 2019;19(2):317-323. doi:10.1111/ajt.15051 PubMedGoogle ScholarCrossref
3.
Goldberg  DS, French  B, Lewis  JD,  et al.  Liver transplant center variability in accepting organ offers and its impact on patient survival.   J Hepatol. 2016;64(4):843-851. doi:10.1016/j.jhep.2015.11.015 PubMedGoogle ScholarCrossref
4.
Wey  A, Valapour  M, Skeans  MA,  et al.  Heart and lung organ offer acceptance practices of transplant programs are associated with waitlist mortality and organ yield.   Am J Transplant. 2018;18(8):2061-2067. doi:10.1111/ajt.14885 PubMedGoogle ScholarCrossref
5.
Wey  A, Salkowski  N, Kasiske  BL, Israni  AK, Snyder  JJ.  Influence of kidney offer acceptance behavior on metrics of allocation efficiency.   Clin Transplant. 2017;31(9). doi:10.1111/ctr.13057 PubMedGoogle Scholar
6.
Massie  AB, Kucirka  LM, Segev  DL.  Big data in organ transplantation: registries and administrative claims.   Am J Transplant. 2014;14(8):1723-1730. doi:10.1111/ajt.12777 PubMedGoogle ScholarCrossref
7.
Muller  CJ, MacLehose  RF.  Estimating predicted probabilities from logistic regression: different methods correspond to different target populations.   Int J Epidemiol. 2014;43(3):962-970. doi:10.1093/ije/dyu029 PubMedGoogle ScholarCrossref
8.
Kilic  A, Weiss  ES, Allen  JG,  et al.  Simple score to assess the risk of rejection after orthotopic heart transplantation.   Circulation. 2012;125(24):3013-3021. doi:10.1161/CIRCULATIONAHA.111.066431 PubMedGoogle ScholarCrossref
9.
Hong  KN, Iribarne  A, Worku  B,  et al.  Who is the high-risk recipient? predicting mortality after heart transplant using pretransplant donor and recipient risk factors.   Ann Thorac Surg. 2011;92(2):520-527. doi:10.1016/j.athoracsur.2011.02.086 PubMedGoogle ScholarCrossref
10.
Bergenfeldt  H, Stehlik  J, Höglund  P, Andersson  B, Nilsson  J.  Donor-recipient size matching and mortality in heart transplantation: influence of body mass index and gender.   J Heart Lung Transplant. 2017;36(9):940-947. doi:10.1016/j.healun.2017.02.002 PubMedGoogle ScholarCrossref
11.
Kilic  A, Emani  S, Sai-Sudhakar  CB, Higgins  RS, Whitson  BA.  Donor selection in heart transplantation.   J Thorac Dis. 2014;6(8):1097-1104.PubMedGoogle Scholar
12.
Madan  S, Saeed  O, Shin  J,  et al.  Donor troponin and survival after cardiac transplantation: an analysis of the United Network of Organ Sharing Registry.   Circ Heart Fail. 2016;9(6):e002909. doi:10.1161/CIRCHEARTFAILURE.115.002909 PubMedGoogle Scholar
13.
Jay  C, Schold  JD.  Measuring transplant center performance: the goals are not controversial but the methods and consequences can be.   Curr Transplant Rep. 2017;4(1):52-58. doi:10.1007/s40472-017-0138-9 PubMedGoogle ScholarCrossref
14.
Kirklin  JK, Pagani  FD, Kormos  RL,  et al.  Eighth annual INTERMACS report: special focus on framing the impact of adverse events.   J Heart Lung Transplant. 2017;36(10):1080-1086. doi:10.1016/j.healun.2017.07.005 PubMedGoogle ScholarCrossref
15.
Linblad K, Lehman RR. Four-month monitoring of heart allocation proposal to modify the heart allocation system. Accessed February 25, 2020. https://optn.transplant.hrsa.gov/media/2924/post-implementation-heart-policy-report_20190417_ready-to-post.pdf
16.
Klassen  DK, Edwards  LB, Stewart  DE, Glazier  AK, Orlowski  JP, Berg  CL.  The OPTN Deceased Donor Potential Study: implications for policy and practice.   Am J Transplant. 2016;16(6):1707-1714. doi:10.1111/ajt.13731 PubMedGoogle ScholarCrossref
×