Key Points
Question
Do current donation metrics adequately represent organ procurement organization (OPO) performance?
Findings
In this national study of 75 769 organ donors, OPO performance varied significantly over time and across donation metrics, which were not entirely described by population demographics. The performance of OPOs was positively associated with use of donors who were ineligible, which varied significantly among OPOs and demographic subgroups.
Meaning
Use of donors who are ineligible should be considered within OPO performance metrics to provide a more accurate assessment of performance, and this represents a target metric for improving organ availability.
This national cohort study uses 10 years of data to evaluate the performance of organ procurement organizations using combined mortality and donation data and population demographic data.
Abstract
Importance
Organ transplant is a life-saving procedure for patients with end-stage organ failure. In the US, organ procurement organizations (OPOs) are responsible for the evaluation and procurement of organs from donors who have died; however, there is controversy regarding what measures should be used to evaluate their performance.
Objective
To evaluate OPO performance metrics using combined mortality and donation data and quantify the associations of population demographics with donation metrics.
Design, Setting, and Participants
This national cohort study includes data from the US organ transplantation system from January 2008 through December 2017. All individuals who died within the US, as reported by the National Death index, were included.
Exposures
Death, organ donation, and donation eligibility.
Main Outcomes and Measures
Evaluation of the variation in donation metrics and the use of ineligible donors by OPO and demographic subgroup.
Results
This study included 17 501 742 deaths and 75 769 deceased organ donors (45 040 men [59.4%]; 51 908 White individuals [68.5%]). Of these donors, 15 857 (20.9%) were not eligible, as defined by the OPOs. The median donation metrics by OPO were 0.004 (range, 0.002-0.012) donors per death, 0.89 (range, 0.68-1.30) donors per eligible death, and 0.72 (range, 0.57-0.86) eligible donors per eligible death. The OPOs in the upper quartile of the overall eligible donors per eligible death metric were in the upper quartile of annual rankings on 90 of 140 occasions (64.3%). There was little overlap in top-performing OPOs between metrics; an OPO in the upper quartile for 1 metric was also in the upper quartile for the other metrics on 37 of 570 occasions (6.5% of the time). The median donor eligibility rate, defined as the number of eligible donors per donor, was 0.79 (range, 0.61-0.95) across OPOs. Age (eg, 65 to 84 years, coefficient, −0.55 [SE, 0.03]; P < .001; vs those aged 18 to 34 years), sex (male individuals, −0.09 [SE, 0.02]; P < .001; vs female individuals), race (eg, Black individuals, 0.35 [SE, 0.02]; P < .001; vs White individuals), cause of death (eg, central nervous system tumor, 0.48 [SE, 0.08]; P < .001; vs anoxia), year (eg, 2016-2017: −0.10 [SE, 0.03]; P < .001; vs 2008-2009), and OPO were associated with the use of ineligible donors; OPO was a significant factor associated with performance in all metrics (χ256, 500.5; P < .001; coefficient range across individual OPOs, −0.15 [SE, 0.09] to 0.75 [SE, 0.09]), even after accounting for population differences. Female and non-White individuals were significantly less likely to be used as ineligible donors.
Conclusions and Relevance
We demonstrate significant variability in OPO performance rankings, depending on which donation metric is used. There were significant differences in OPO performance, even after accounting for differences in potential donor populations. Our data suggest significant variation in use of ineligible donors among OPOs, a source for increased donors. The performance of OPOs should be evaluated using a range of donation metrics.
Introduction
Organ transplant is a life-saving intervention for patients with organ failure. The need for organs, however, far outweighs supply, resulting in considerable waiting-list morbidity and approximately 7500 deaths of those on waiting lists annually.1 Organ procurement organizations (OPOs) are federally designated nonprofit organizations that are part of the Organ Procurement and Transplantation Network (OPTN). The OPOs are tasked with coordinating procurement and allocation of all organs from deceased donors in the US. Recent evidence suggests marked variation in donation rates among OPOs, which are largely unexplained.1,2,3,4,5,6,7
A key metric of OPO performance is the number of eligible donors per eligible death.2,3,8 Eligible deaths are defined by OPTN-developed criteria and self-reported by each OPO.9 Eligible deaths include the people living within each OPO’s service area who died 75 years or younger with a body mass index (BMI; calculated as weight in kilograms divided by height in meters squared) less than 50, in addition to other organ-specific criteria. Although these criteria are driven by demographic characteristics, prior studies found that differences in organ recovery rates across OPOs are not fully explained by differences in population characteristics.7 There have been calls for a more comprehensive, objective assessment of the potential donor pool when reporting donation rates.2,3,10
Two factors limit the ability to accurately compare donation rates by OPOs. First, validated estimates of the number of eligible deaths, which is the denominator for donation rates, are needed for each OPO. Second, complicating comparisons further, many successful transplants arise from ineligible donor donations, according to OPTN criteria. A previous study found that the use of ineligible liver grafts ranged from 0% to as high as 19.6% in some OPOs.11 It remains unclear how ineligible donor use varies across organ types, and how this use affects donation rates.
Improved metrics to assess OPO performance could spur efforts within low-performing OPOs to improve organ procurement, thereby increasing available organs for transplant.3 We therefore combined national mortality and donation data and aimed to analyze the association of the definition of eligible death with OPO performance assessment.
Methods
Data Sources
Data sources for this study included the Centers for Disease Control (CDC) National Death Index, United Network for Organ Sharing (UNOS) Standard Transplant Analysis and Research file data, and National Health and Nutrition Examination Survey data from January 2008 through December 2017. We exclusively assessed donations from adult deceased donors and did not consider donations from living donors. We limited our analysis to solid-organ transplants: heart, kidney, liver, lung, and pancreas. Additional data source details are available in the eMethods in the Supplement.
The University of Michigan institutional review board provided a waiver for the conduct of this study, and no informed consent was obtained for the conduct of this study. This was because the study consisted of secondary use of deidentified data.
Donation Metrics
For this analysis, we studied 3 metrics of organ donation from deceased donors (Figure 1): (1) donors per death, (2) donors per eligible death, and (3) eligible donors per eligible death. The value of donors per death measures the percentage of the population who become donors of 1 or more organs when deceased. Donors per eligible death is an adjusted metric that accounts for the number of deaths meeting the predefined eligibility criteria. The metric of eligible donors per eligible death represents the fraction of eligible deaths that are successfully converted into donors. Comparing donors per death with eligibility-based metrics provides insight on how the eligibility definition affects OPO performance. Comparing donors per eligible death with eligible donors per eligible death provides insight into the magnitude of ineligible donor use among OPOs.
Figure 1. Organ Donation Metrics.
Visual representation of the association of donation metric components. Regions are not drawn to scale.
We defined donors as any OPTN-reported deceased donor in the Standard Transplant Analysis and Research data files. Deaths are defined as mortality events reported in the CDC National Death Index. Eligible deaths are any death that was reported by an OPO to UNOS as meeting all donor eligibility criteria. Eligible donors are the subset of eligible deaths that were also reported to have donated at least 1 organ. An individual who does not meet the eligibility criteria can still be an organ donor; therefore, the donors per eligible death metric can be greater than 1.0 if an OPO uses a high number of ineligible donors. The OPO with the median eligible donors per eligible death value was used as the baseline OPO for all metrics.
Donor Eligibility Rate
To assess ineligible donor use, we calculated a donor eligibility rate. This rate represents the fraction of donors that meet all eligibility criteria. We used this metric to quantify how ineligible donor use varied among OPOs and was associated with performance differences in donation metrics.
Statistical Analysis
Details of the statistical approach are in the eMethods in the Supplement. We performed the data analysis from July 2019 to June 2020. The threshold of significance for all tests was P < .05. We performed all statistical analyses using R version 3.6.3 (R Foundation for Statistical Computing) and developed map figures using Tableau version 2019.2 (Tableau Software).
Results
From 2008 through 2017, there were more than 17 million adult deaths in the US and 75 769 donors (45 040 men [59.4%]; 51 908 White individuals [68.5%]; Table 1). The most common primary cause of death (COD) was from cardiovascular or cerebrovascular events. The aggregate donation rate was 0.4% during this period, and 15 857 donors (20.9%) did not meet all of the donation eligibility criteria.
Table 1. Study Population Description From 2008 Through 2017a.
| Characteristic | Patients, No. (%) | |||
|---|---|---|---|---|
| Deaths per CDC | Donors per UNOS | Eligible deaths per UNOS | Ineligible donors per UNOS | |
| Total | 17 501 742 | 75 769 | 84 271 | 15 857 |
| Age group, y | ||||
| 18-34 | 738 099 (4.2) | 23 863 (31.5) | 25 929 (30.8) | 3806 (24.0) |
| 35-49 | 1 432 633 (8.2) | 22 324 (29.5) | 24 666 (29.3) | 4565 (28.8) |
| 50-64 | 4 456 606 (25.5) | 23 377 (30.9) | 27 737 (32.9) | 4875 (30.7) |
| 65-84 | 10 874 404 (62.1) | 6205 (8.2) | 5939 (7.1) | 2611 (16.5) |
| Sex | ||||
| Female | 7 617 297 (43.5) | 30 729 (40.6) | 34 898 (41.4) | 6196 (39.1) |
| Male | 9 884 445 (56.5) | 45 040 (59.4) | 49 373 (58.6) | 9661 (60.9) |
| Race | ||||
| White | 13 320 408 (76.1) | 51 908 (68.5) | 48 939 (58.1) | 12 589 (78.4) |
| Black | 2 370 271 (13.5) | 11 832 (15.6) | 18 676 (22.2) | 1602 (10.1) |
| Hispanic | 1 274 830 (7.3) | 9509 (12.6) | 12 362 (14.7) | 1268 (8.0) |
| Other | 536 233 (3.1) | 2520 (3.3) | 4294 (5.1) | 398 (2.5) |
| BMI | ||||
| <30 | 10 955 848 (62.6) | 50 962 (67.3) | 54.660 (64.9) | 10 156 (64.1) |
| ≥30 | 6 545 894 (37.4) | 24 807 (32.7) | 29.611 (35.1) | 5701 (36.0) |
| Cause of death | ||||
| Anoxia | 3 448 711 (19.7) | 23 646 (31.8) | 24 088 (28.6) | 6306 (39.8) |
| Cerebrovascular causes/stroke | 10 089 023 (57.7) | 27 816 (36.7) | 34 077 (40.4) | 5.087 (32.1) |
| Central nervous system tumor | 73 590 (0.4) | 261 (0.3) | 375 (0.4) | 28 (0.2) |
| Head trauma | 3 685 799 (21.1) | 22.694 (30.0) | 24 252 (28.8) | 3880 (24.5) |
| Other | 204 619 (1.2) | 1352 (1.8) | 1479 (1.8) | 556 (3.5) |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CDC, Centers for Disease Control and Prevention; UNOS, United Network for Organ Sharing.
Ineligible donors are donors who did not meet at least 1 of the Organ Procurement and Transplantation Network donation eligibility criteria. Patients from Puerto Rico were excluded from all analyses because the death data for this subgroup were not included in the available CDC mortality data.
Donation Metrics by OPO
eFigure 1 in the Supplement provides OPO-level maps for the 3 donation metrics. Aggregating across all years and demographics, the median number of donors per death was 0.004 (range, 0.002-0.012) among OPOs. In comparison, the median number of donors per eligible death was 0.89 (range: 0.68-1.30), and the median number of eligible donors per eligible death was 0.72 (range, 0.57-0.86). There was no significant correlation between the donors per death metric and either of the other donation metrics (eTable 1 in the Supplement). The OPO size was not significantly associated with any of the donation metrics.
Stratified Donation Metrics
The β regression models of donor demographics on donation metrics are summarized in Table 2. Positive coefficients represent a positive association with metrics for that subgroup after controlling for other variables.
Table 2. Selected β Regression Results for Donation Rate Metricsa.
| Characteristic | Donors per death | Donors per eligible death | Eligible donors per eligible death | |||
|---|---|---|---|---|---|---|
| Coefficient (SE) | P value | Coefficient (SE) | P value | Coefficient (SE) | P value | |
| Age group, y | ||||||
| 18-34 | Baseline | NA | NA | NA | NA | NA |
| 35-49 | −0.02 (0.01) | .09 | 0.12 (0.02) | <.001 | 0.05 (0.02) | .004 |
| 50-64 | −0.15 (0.01) | <.001 | 0.06 (0.02) | <.001 | −0.004 (0.02) | .78 |
| 65-84 | −0.41 (0.01) | <.001 | −0.45 (0.02) | <.001 | −0.65 (0.02) | <.001 |
| Sex | ||||||
| Female | Baseline | NA | NA | NA | NA | NA |
| Male | 0.04 (0.01) | <.001 | 0.10 (0.01) | <.001 | 0.07 (0.01) | <.001 |
| Race/ethnicity | ||||||
| White | Baseline | NA | NA | NA | NA | NA |
| Black | −0.39 (0.01) | <.001 | −0.90 (0.02) | <.001 | −0.78 (0.02) | <.001 |
| Hispanic | −0.48 (0.01) | <.001 | −0.90 (0.02) | <.001 | −0.78 (0.02) | <.001 |
| Other | −0.67 (0.01) | <.001 | −1.36 (0.02) | <.001 | −1.23 (0.03) | <.001 |
| BMI | ||||||
| <30 | Baseline | NA | NA | NA | NA | NA |
| ≥30 | −0.13 (0.01) | <.001 | −0.34 (0.01) | <.001 | −0.38 (0.01) | <.001 |
| Cause of death | ||||||
| Anoxia | Baseline | NA | NA | NA | NA | NA |
| Cerebrovascular causes/stroke | −0.03 (0.01) | .01 | 0.09 (0.02) | <.001 | 0.16 (0.02) | <.001 |
| Central nervous system tumor | −0.64 (0.02) | <.001 | −1.17 (0.06) | <.001 | −1.00 (0.06) | <.001 |
| Head trauma | −0.08 (0.01) | <.001 | −0.04 (0.02) | .02 | 0.03 (0.02) | .06 |
| Other | −0.50 (0.01) | <.001 | −1.00 (0.03) | <.001 | −0.99 (0.03) | <.001 |
| Year | ||||||
| 2008-2009 | Baseline | NA | NA | NA | NA | NA |
| 2010-2011 | 0.01 (0.01) | .26 | 0.07 (0.02) | <.001 | 0.05 (0.02) | .02 |
| 2012-2013 | 0.03 (0.01) | .03 | 0.09 (0.02) | <.001 | 0.05 (0.02) | .02 |
| 2014-2015 | 0.05 (0.01) | <.001 | 0.12 (0.02) | <.001 | 0.06 (0.02) | .003 |
| 2016-2017 | 0.08 (0.01) | <.001 | 0.15 (0.02) | <.001 | 0.11 (0.02) | <.001 |
| Intercept | −4.16 (0.03) | <.001 | −1.68 (0.05) | <.001 | −1.28 (0.05) | <.001 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable.
Coefficients are not shown for individual organ procurement organization, although organ procurement organization was included as a covariate. The analysis used a parameter significance level of α set to 0.05.
Age was significantly associated with most donation metrics, with older age groups generally corresponding with lower likelihood of donation. One exception in the donors per eligible death metric, for which donors aged 35 to 64 years had higher donation rates (coefficient, 0.12 [SE, 0.02]) than patients 18 to 34 years old (P < .001). Patients 65 to 84 years old had the lowest coefficients for all metrics, ranging from −0.41 (SE, 0.01) for donors per death to −0.65 (SE, 0.02) for eligible donors per eligible death (P < .001), indicating that older patients are less likely to donate, even after accounting for eligibility criteria.
Sex, race/ethnicity, BMI, and COD were significantly associated with each donation metric at the P < .001 level. Male donor subgroups, compared with female donors, had higher donation metrics, with coefficients ranging from 0.04 to 0.10 (donors per death, 0.04 [SE, 0.01]; donors per eligible death, 0.10 [0.01]; eligible donors per eligible death, 0.07 [0.01]; P < .001). After controlling for other variables, White individuals, compared with other racial/ethnic subgroups, had the highest likelihood of donation (eg, coefficients for Black individuals: donors per death, −0.39 [SE, 0.01]; donors per eligible death, −0.90 [0.02]; eligible donors per eligible death, −0.78 [0.02]; P < .001). Coefficients for non-White groups had the largest negative magnitudes of all variables, reaching as low as −1.36 (SE, 0.02) for donors per eligible death in individuals of other races/ethnicities. Body mass index greater than 30, compared with BMI less than 30, was significantly inversely associated with all donation metrics (donors per death, −0.13 [SE, 0.01]; donors per eligible death, −0.34 [0.01]; eligible donors per eligible death, −0.38 [0.01]; P < .001). Anoxia-associated deaths were associated with the highest rates of donors per death (reference; other coefficients ranged from −0.03 [SE, 0.01]; P = .01 for deaths from cerebrovascular causes/stroke to −0.64 [SE, 0.02]; P < .001 for central nervous system tumor), while deaths associated with cerebrovascular causes or stroke were associated with the highest donation rates for eligibility-based metrics (donors per eligible death, 0.09 [0.02]; eligible donors per eligible death, 0.16 [0.02]; P < .001).
Organ-procurement organization was significantly associated with all 3 donation metrics. Even after accounting for population demographics and COD, there were significant differences in the donation and use of ineligible organs among OPOs. The OPO coefficients ranged from −0.32 to 0.19 for donors per death, from −0.71 to 0.24 for donors per eligible death, and from −0.71 to 0.32 for eligible donors per eligible death (all P < .001).
Figure 2 shows the magnitude of the coefficient range for each variable. A larger range of coefficients for a variable suggests that variable has a stronger association with donation rates, even after controlling for the other features. Donor race/ethnicity had the largest association with donation among all variables (magnitude of coefficient ranges: donors per death, 0.67; donors per eligible death, 1.36; eligible donors per eligible death, 1.23). We found that the individual OPO association was larger than the association with year, sex, and BMI in all metrics (Figure 2).
Figure 2. Covariate β Regression Coefficients by Donation Metric.
A larger range suggests a stronger association between a variable and the donation metric, even after accounting for the other covariates. Race/ethnicity had the largest range across all donation metrics, suggesting that a patient’s race/ethnicity has the largest association with donation among the variables studied. The organ procurement organization (OPO) also had a significant range for all metrics, suggesting that performance differences among OPOs cannot be entirely described by population demographics. BMI indicates body mass index (calculated as weight in kilograms divided by height in meters squared); COD, cause of death.
OPO Performance Over Time
We assessed relative OPO performance annually to characterize performance consistency. eFigure 2 in the Supplement shows annual metric variability for individual OPOs, and eFigure 3 in the Supplement shows the variability in OPO annual relative ranking. Rankings were done objectively in descending order according to the fractional value of each metric. Figures 1 and 2 show that OPOs had year-on-year and relative variability in performance. The OPOs with the largest annual variability in the eligible donors per eligible death metric were Upstate New York Transplant Services Inc (SD, 0.128), LifeShare Transplant Donor Services of Oklahoma (SD, 0.108), and LifeCenter Organ Donor Network (SD, 0.105). The OPOs with the smallest variability in the eligible donors per eligible death metric were LifeSource Upper Midwest Organ Procurement Organization (SD, 0.023), Gift of Life Donor Program (SD, 0.024), and New Jersey Organ and Tissue Sharing Network OPO (SD, 0.025).
eTable 2 in the Supplement shows the percentage of OPOs that remained in the upper or lower quartile of each performance metric year on year. Generally, high-performing and low-performing OPOs were most consistent in the donors per death metric. The OPOs in the upper quartile of this metric overall were in the upper quartile of annual rankings on 116 of 140 occasions (82.9%). In contrast, the eligible donors per eligible death metric saw wider variation, with OPOs remaining in the upper quartile of annual rankings on 90 of 140 occasions (64.3%). High-performing and low-performing OPOs varied widely depending on the choice of metric. In a given year, an OPO that was in the upper quartile for 1 metric was also in the upper quartile for the other metrics only on 37 of 570 occasions (6.5% of the time). An OPO in the lowest quartile for 1 metric was in the lowest quartile for other metrics 29 of 570 occasions (5.1%) of the time.
Donor Eligibility
The median donor eligibility rate (ie, number of eligible donors per donor) was 0.80 (range, 0.61-0.95) across OPOs (eFigure 4 in the Supplement). eFigure 5 in the Supplement shows OPO variability in the donor eligibility rate by organ, with kidney and heart donations having the highest and lowest variability (SDs: kidney, 0.077; heart, 0.012), respectively.
eFigure 6 in the Supplement shows annual eligibility rate variability for each OPO. We found a significant negative correlation with the donor eligibility rate and donors per eligible death (ρ = −0.87 [95% CI, −0.92 to −0.78]; P < .001), indicating ineligible donor use is strongly associated with an increased donor per eligible death metric. We found no significant correlation between the donor eligibility rate and the other metrics (eTable 1 in the Supplement). The OPO size was also not significantly associated with the donor eligibility rate, which indicates that the volume of deaths in a service area is not associated with either higher or lower use of ineligible donors.
Table 3 shows β regression coefficients for age, sex, race, BMI, COD, and year of death influencing eligible donors per donor. Here, negative coefficients suggest a higher likelihood of being an ineligible donor. Ineligible donors were significantly more likely to be in the group aged 65 to 84 years (−0.55 [SE, 0.03]; P < .001) and male (−0.09 [SE, 0.02]; P < .001). Ineligible donors were significantly more likely to be White than non-White, with coefficients for non-White subgroups ranging from 0.31 (SE, 0.04; for individuals of other races/ethnicities) to 0.35 (SE, 0.02; for Black individuals) (all P < .001); ineligible donors were also more likely to have died of anoxia than other causes, including central nervous system tumor (0.48 [SE, 0.08]; P < .001). We did not find significant differences in eligibility rates between BMI groups. Year was a significant parameter, with recent years associated with a higher use of ineligible donors (eg, 2016-2017: −0.10 [SE, 0.03]; vs 2010-2011: 0.003 [SE, 0.03]; P < .001).
Table 3. Selected β Regression Results for the Rate of Eligible Donors per Donora.
| Characteristic | Eligible donors per donor | |
|---|---|---|
| Coefficient (SE) | P value | |
| Age group, y | ||
| 18-34 | Baseline | NA |
| 35-49 | −0.13 (0.02) | <.001 |
| 50-64 | −0.16 (0.02) | <.001 |
| 65-84 | −0.55 (0.03) | <.001 |
| Sex | ||
| Female | Baseline | NA |
| Male | −0.09 (0.02) | <.001 |
| Race/ethnicity | ||
| White | Baseline | NA |
| Black | 0.35 (0.02) | <.001 |
| Hispanic | 0.32 (0.03) | <.001 |
| Other | 0.31 (0.04) | <.001 |
| BMI | ||
| <30 | Baseline | NA |
| ≥30 | 0.02 (0.02) | .36 |
| Cause of death | ||
| Anoxia | Baseline | NA |
| Cerebrovascular causes/stroke | 0.21 (0.02) | <.001 |
| Central nervous system tumor | 0.48 (0.08) | <.001 |
| Head trauma | 0.14 (0.02) | <.001 |
| Other | −0.33 (0.05) | <.001 |
| Year | ||
| 2008-2009 | Baseline | NA |
| 2010-2011 | 0.003 (0.03) | .91 |
| 2012-2013 | −0.08 (0.03) | .003 |
| 2014-2015 | −0.11 (0.03) | <.001 |
| 2016-2017 | −0.10 (0.03) | <.001 |
| Intercept | 0.54 (0.07) | <.001 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); NA, not applicable.
Coefficients are not shown for individual organ procurement organizations, although each organ procurement organization was included as a covariate. This analysis used a parameter significance level of α set to .05.
Organ procurement organization was significantly associated with each donation metric (χ256, 500.5; P < .001). Individual OPO coefficients ranged from −0.15 (SE, 0.09) to 0.75 (SE, 0.09), indicating that OPOs had significant differences in ineligible donor use even after accounting for population demographics.
eFigure 7 through 9 in the Supplement shows the association over time of race, sex, and age on each donation rate, respectively. White donors had the lowest donors per death rate across all years but the highest donors per eligible death rate across all years, suggesting a higher likelihood of an ineligible White donor being accepted for donation compared with an ineligible non-White donor. A similar phenomenon was found with donors aged 65 to 84 years, who also had the lowest donors per death metric and were more likely to be accepted for donation despite not meeting eligibility criteria.
Discussion
The single best tool to improve the survival of patients waiting for an organ transplant is to increase the supply of organs by optimizing organ use. However, there is substantial regional variation in both the rate of eligible deaths and sources of organ donation. Efforts to optimize organ donation rates across the country, including targeted interventions for low-performing organizations and spreading best practices from high-performing organizations, depend on accurate, generalizable estimates of eligible donations and deaths. In this study, we have shown that OPO performance varies widely depending on the definition of eligibility. We found that the use of ineligible donors differs significantly among OPOs and this factor contributes significantly to OPO donation rate variation.
OPO Performance Metrics
To improve understanding of OPO performance, we recommend adopting the following 3 metrics: (1) donors per death, (2) eligible donors per eligible death, and (3) eligible donors per donor (or the donor eligibility rate). As long as the definition of donor eligibility remains constant, our results suggest that the combination of metrics presented here is important for accurate representation of OPO performance. As an example, we can consider Washington Regional Transplant Community (WRTC). Over the study period, WRTC ranked 54th among OPOs using the eligible donors per eligible death metric (0.64 donors per death). However, WRTC had the sixth-highest donors per death (0.0067) over the same time frame. Without more context, this discrepancy might be attributed to a larger eligible donor pool within WRTC’s service area. Using all 3 metrics, we see that WRTC also had the seventh-highest use of ineligible donors (donor eligibility rate: 0.69), suggesting that a significant component of the high donors per death metric comes from the use of ineligible donors. Combining these 3 metrics, we can consider OPOs such as the University of Wisconsin Health Organ and Tissue Donation and Midwest Transplant Network (0.86 and 0.78, respectively) to be examples of high-performing OPOs. They have high values of eligible donors per eligible death (0.70 and 0.73, respectively), which demonstrate success procuring eligible organ donations, and they also have low donor eligibility rates (0.70 and 0.73, respectively), which suggests that they are also successfully procuring extended-criteria donations to expand the donor pool. Consequently, they both have among the highest metrics on overall donors per death (0.0071 and 0.0075, respectively).
Influence of Usage of Ineligible Donors
Previous studies suggest that certain populations may have a lower consent rate for organ donation, which may drive some of this difference.12 We found, however, that OPO is a significant factor, even after accounting for differences in causes of death and population demographics. Instead, we found that donation metrics were driven by additional, OPO-specific factors. Key among these is use of ineligible donors. There are likely several reasons for the wide variation in use of ineligible donors among OPOs, including reporting differences among OPOs, because eligible deaths are self-reported and can significantly shift the denominator of eligibility-based metrics. From our regression models, we found OPOs had the widest variability in performance under the metric of eligible donors per donor, indicating OPOs are relatively consistent in converting eligible deaths into donors but have more variability in ineligible donor use.
Influence of Donor Race/Ethnicity
We showed population demographics play an important role in organ donation, with significantly lower donation metrics for minority groups. Accounting for factors such as COD and OPO, we found that White race was associated with higher values of donors per death and donors per eligible death metrics. Importantly, the relative association of race/ethnicity in our β regression was larger after including the eligibility definition, suggesting that racial/ethnic disparities are present even in ideal donors. This association with race/ethnicity was also seen in ineligible donor use, where we found that White donors were significantly more likely than non-White donors to be outside eligibility criteria. The root cause of this difference is unclear but could in part be a result of ineffective organ authorization practices with non-White donors12 or implicit or explicit bias in the acceptance of extended criteria donors. While some have called for adjustments by race/ethnicity to proposed donation metrics, research has shown that such adjustments can in fact hide bias and disparities.13 Instead, metrics that capture OPO-specific differences are important to understand where OPOs are successfully obtaining donation authorization from members of minority groups and where they are not.
How to Empirically Validate the Choice of Metrics
Eligibility should be expanded to include the largest population from which organ donations produce equivalent posttransplant outcomes. A study focused on liver transplants found similar variation in use of ineligible donors, with no significant difference in survival outcomes between recipients of eligible and ineligible organs.11 Our results imply that this variability exists across all solid-organ transplants. We also found increasing use of ineligible donors over time, suggesting that definitions of eligibility may need to change. To improve donation in lower-performing OPOs, we can look for best practices in extended criteria donor usage. Specifically, OPOs with low donor eligibility rates may have identified a subset of potential donors who are ineligible yet consistently provide positive outcomes for patients waiting for organs. Systematically expanding the current eligibility criteria based on current patterns of use of ineligible donors could incentivize OPOs to use similar donors, increase the overall number of donations, and reduce variability in donors per eligible death. The definition of eligibility may need to change to incorporate currently putatively ineligible populations.
Strengths
A notable strength of this work is our ability to combine CDC mortality and OPTN donation data to perform a detailed analysis of multiple data rates. Beyond demonstrating significant variation in donation metrics, we were able to use statistical modeling to help identify potential sources of this variation.
Limitations
Our study had some limitations. First, privacy requirements did not allow us to directly match CDC mortality deaths to OPTN donations, preventing us from conducting patient-level regression analyses, such as logistic regression, which are arguably more traditional. Second, as with most elements of the organ donation process, ineligible donor use is also driven by individual clinicians, donor recipients, and other stakeholders and is not fully dependent on OPOs. However, given the limitations in the available data, our analysis was restricted to the OPO level. Third, CDC mortality data did not include granular donor variables (eg, comorbidities), which may have influenced donor use. We also note that there may be heterogeneity in eligibility reporting standards across OPOs, which can obscure the donor eligibility rate.
Conclusions
In conclusion, we showed that there are significant differences among donation metrics, even after accounting for differences in donor populations. We have shown wide variation in OPO performance depending on the metric used. Variation in use of ineligible donors accounts for variation in OPO performance in metrics, such as donors per eligible death. We also found significant differences in eligibility rates among certain demographics, which may be the result of variable consent rates among groups and/or bias in the acceptance of donors by OPO and transplant providers. The use of an expanded set of metrics, such as the ones in our analysis, could help identify and monitor disparities between OPOs, to improve organ availability in the US.
eMethods.
eTable 1. Spearman’s rank correlation coefficients between donation metrics.
eTable 2. Consistency of top overall performers (2008-2017) in annual rankings by donation metric.
eFigure 1. Overall donation metrics from 2008-2017 by OPO.
eFigure 2. Year-to year variation in donation rates by OPO.
eFigure 3. Variation in yearly national donation rate ranking by OPO.
eFigure 4. Overall donor eligibility rates from 2008-2017 by OPO.
eFigure 5. Donor eligibility rate by organ.
eFigure 6. Year-to year variation in donor eligibility by OPO.
eFigure 7. Donation metrics by race.
eFigure 8. Donation metrics by sex.
eFigure 9. Donation metrics by age.
References
- 1.Israni AK, Zaun D, Rosendale JD, Schaffhausen C, Snyder JJ, Kasiske BL. OPTN/SRTR 2017 annual data report: deceased organ donation. Am J Transplant. 2019;19(suppl 2):485-516. doi: 10.1111/ajt.15280 [DOI] [PubMed] [Google Scholar]
- 2.Goldberg D, Karp S, Shah MB, Dubay D, Lynch R. Importance of incorporating standardized, verifiable, objective metrics of organ procurement organization performance into discussions about organ allocation. Am J Transplant. 2019;19(11):2973-2978. doi: 10.1111/ajt.15492 [DOI] [PubMed] [Google Scholar]
- 3.Goldberg D, Kallan MJ, Fu L, et al. Changing metrics of organ procurement organization performance in order to increase organ donation rates in the United States. Am J Transplant. 2017;17(12):3183-3192. doi: 10.1111/ajt.14391 [DOI] [PubMed] [Google Scholar]
- 4.Cannon RM, Jones CM, Davis EG, Franklin GA, Gupta M, Shah MB. Patterns of geographic variability in mortality and eligible deaths between organ procurement organizations. Am J Transplant. 2019;19(10):2756-2763. doi: 10.1111/ajt.15390 [DOI] [PubMed] [Google Scholar]
- 5.Siminoff LA, Gardiner HM, Wilson-Genderson M, Shafer TJ. How inaccurate metrics hide the true potential for organ donation in the United States. Prog Transplant. 2018;28(1):12-18. doi: 10.1177/1526924818757939 [DOI] [PubMed] [Google Scholar]
- 6.Gentry SE, Chow EKH, Massie A, et al. Liver sharing and organ procurement organization performance. Liver Transpl. 2015;21(3):293-299. doi: 10.1002/lt.24074 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Goldberg DS, French B, Abt PL, Gilroy RK. Increasing the number of organ transplants in the United States by optimizing donor authorization rates. Am J Transplant. 2015;15(8):2117-2125. doi: 10.1111/ajt.13362 [DOI] [PubMed] [Google Scholar]
- 8.Luskin R, Nathan H. Eligible death statistic: not a true measure of OPO performance nor the potential to increase transplantation. Am J Transplant. 2015;15(8):2019-2020. doi: 10.1111/ajt.13361 [DOI] [PubMed] [Google Scholar]
- 9.Organ Procurement and Transplantation Network OPTN policies contents. Published 2019. Accessed October 22, 2020. https://optn.transplant.hrsa.gov/media/1200/optn_policies.pdf
- 10.Federal Register Executive Order 13879 of July 10, 2019: advancing American kidney health. Published July 10, 2019. Accessed October 19, 2020. https://www.federalregister.gov/documents/2019/07/15/2019-15159/advancing-american-kidney-health
- 11.Croome KP, Lee DD, Keaveny AP, Taner CB. Noneligible donors as a strategy to decrease the organ shortage. Am J Transplant. 2017;17(6):1649-1655. doi: 10.1111/ajt.14163 [DOI] [PubMed] [Google Scholar]
- 12.Wilder JM, Oloruntoba OO, Muir AJ, Moylan CA. Role of patient factors, preferences, and distrust in health care and access to liver transplantation and organ donation. Liver Transpl. 2016;22(7):895-905. doi: 10.1002/lt.24452 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Goldberg D, Doby B, Siminoff L, Shah M, Lynch R. Rejecting bias: the case against race adjustment for OPO performance in communities of color. Am J Transplant. 2020;20(9):2337-2342. doi: 10.1111/ajt.15865 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eTable 1. Spearman’s rank correlation coefficients between donation metrics.
eTable 2. Consistency of top overall performers (2008-2017) in annual rankings by donation metric.
eFigure 1. Overall donation metrics from 2008-2017 by OPO.
eFigure 2. Year-to year variation in donation rates by OPO.
eFigure 3. Variation in yearly national donation rate ranking by OPO.
eFigure 4. Overall donor eligibility rates from 2008-2017 by OPO.
eFigure 5. Donor eligibility rate by organ.
eFigure 6. Year-to year variation in donor eligibility by OPO.
eFigure 7. Donation metrics by race.
eFigure 8. Donation metrics by sex.
eFigure 9. Donation metrics by age.