• Deceased donor;
  • organ donation;
  • organ procurement;
  • organ utilization


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Understanding the additional life-years given to patients by deceased organ donors is necessary as substantial investments are being proposed to increase organ donation. Data were drawn from the Scientific Registry of Transplant Recipients. All patients placed on the wait-list as eligible to receive or receiving a deceased donor solid organ transplant between 1995 and 2002 were studied. The benefit of transplant was determined by the difference in the expected survival experiences of transplant recipients and candidates expecting transplant soon. An average organ donor provides 30.8 additional life-years distributed over an average 2.9 different solid organ transplant recipients, whereas utilization of all solid organs from a single donor provides 55.8 additional life-years spread over six organ transplant recipients. The relative contribution of the different organs to the overall life-year benefit is higher for liver, heart and kidney, and lowest for lung and pancreas. The life-year losses from unprocured and unused organs are comparable to suicide, congenital anomalies, homicide or perinatal conditions and half that of HIV. Approximately 250 000 additional life-years could be saved annually if consent for potential deceased donors could be increased to 100%. Therefore, increasing organ donation should be considered among our most important public health concerns.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Within the transplant community, the lack of suitable organs for transplantation is recognized as the most critical problem facing transplantation today. However, within the greater medical community at large, the relative role that the organ shortage plays in relation to other public health issues is uncertain. This is understandable for a number of reasons. First, transplantation, although with high visibility, is a low-volume medical endeavor; approximately 20 000 organ transplants are performed in the United States annually (1), and more than 80 000 individuals are currently waiting for transplants on various solid organ transplant lists. This contrasts with over 1 million coronary artery revascularization procedures performed annually (2). Second, there are no data quantifying the true impact of the organ shortage in terms of loss of life in a format that would allow ready comparisons to other public health problems such as the HIV epidemic, infectious diseases or heart disease.

Using data procured from the United States Scientific Registry of Transplant Recipients (SRTR), we developed a model capable of estimating the impact of an extra deceased organ donor in terms of additional life-years obtained by patients awaiting transplantation. Although the benefits of the existence of deceased donor kidney or pancreas transplantation have been well studied (3–6), the potential benefits of increasing donation rates overall have not been reported. Previous studies have compared the outcomes of transplant recipients to those of potential transplant recipients while they were waiting on the national lists to receive organ transplants, that is, while they were ‘wait-listed’. In these studies the outcomes of wait-listed patients had been considered from the time of listing until transplant or death and represented expected increased life expectancy from transplantation compared to a lifetime without transplant. Essentially, these studies compared transplantation to hypothetical situations where transplantation did not exist, thus estimating the average benefit available to recipients because transplant is available, a valid and important assessment.

However, in the present study, we considered the benefits to all patients awaiting transplantation provided by an extra deceased donor through increased donation. Although similar to previous studies (3–6), this perspective calls for a different analytical structure. A patient receiving an organ from this extra donor is transplanted and removed from the waiting-list before they would have been otherwise. Since this patient would have received a different organ, had the extra donor not been available, a second organ is made available, or ‘released’, to another patient. Viewing the wait-list as a line or queue with patients receiving organs as they approach the front of the line, it is easy to see that adding organs removes patients near the front of the line and shortens the wait of patients behind. In this study, we calculated the averted deaths expected from increased donation. Therefore, we compared survival expectations following transplantation with survival expectations among patients expecting transplant soon for the most common solid organ transplants: liver, heart, kidney, kidney-pancreas, lung and pancreas. Combined, we estimate the potential life-saving benefits and consider the public health implications of increasing deceased organ donation.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Data source and sample selection

The data were collected by the United Network for Organ Sharing (UNOS) and obtained from the Scientific Registry of Transplant Recipients (SRTR). We studied all US patients wait-listed for and/or transplanted with deceased donor liver, heart, kidney, kidney-pancreas, lung or pancreas transplant in the United States between January 1, 1995 and December 31, 2002. Wait-listed candidates were compared to all patients who received a deceased donor liver, heart, kidney, kidney-pancreas, lung or pancreas transplant. The SRTR determination of last follow-up date was used for each patient.

Estimation of survival functions

The analysis was structured to assess the average benefit from an extra deceased organ donor in terms of the increased life-years expected for wait-listed patients, produced by increasing donation rates. To quantify life-year benefits, we estimated probability distribution functions separately for each organ describing the expected survival of transplant candidates near the time of transplant (i.e. those expecting transplant soon) and compared these functions with estimated and projected survival expectations following organ transplantation.

Waiting-list survival: ‘Candidates expecting transplant soon’:  A survivor function describes the probability of survival within a population over time. We estimated transplant-censored survivor functions separately for patients on each specific solid organ transplant wait-list from the time of listing. To generate survivor functions representing observed wait-list experience, we used Cox's proportional-hazards regression adjusted for each organ to average characteristics of transplanted patients by age, race, gender, cause of disease and year of waiting-list placement. Avoided deaths during intervals of shortened waiting time would have been expected to occur at the average rate of death faced by transplant candidates expecting transplant soon. Therefore, we weighted observed survival experiences according to organ-specific measures of allocation priority to model the survival of wait-listed patients expecting transplant soon.

Organs for transplant from deceased donors are distributed based on organ-specific prioritization criteria that include biological suitability, waiting time and urgency (7). To estimate the survival experience of patients with the highest allocation priority, we considered organ-specific allocation algorithms. For organs distributed by urgency, that is liver and heart, wait-list survival was first stratified by urgency at the time of listing (7). Urgency for heart distribution is determined by status, a description of the intensity of support required by the patient. Urgency for liver distribution is determined by laboratory results scaled by specific models predictive of death; MELD (Model for End-stage Liver Disease) for adults or PELD (Pediatric End-stage Liver Disease) for children. We calculated the death rates between 30 and 121 days after initial listing for each urgency stratum and determined the average death rate weighted by the distribution of urgency observed at the time of transplant. We evaluated mortality experiences starting at 1 month after listing because transplantation of urgency-distributed organs was less common within the first month of listing and early mortality is very high. This may increase the estimate of expected life-years while wait-listed, thus conservatively reducing the estimated life-year benefit of liver and heart transplant. For patients at high priority for transplantation with organs distributed by waiting time rather than by urgency, that is kidney, kidney pancreas, pancreas and lung, organ-specific expected death rates while waiting were calculated as the average of annualized mortality among transplant candidates in 1-year intervals after listing. These annualized mortality estimates were weighted by the observed distribution of time from listing to performed transplantations for recipients by organ.

The slope or derivative of a survival curve represents the rate of death in a short interval of time, also known as the hazard rate. The hazard (h) is related to the survivor function (S (t)) by the equation, S (t) = e −xβt. Using the weighted-average observed death rate as the hazard of death, we estimated survivor functions representing the average survival experience of transplant candidates expecting transplant soon. To enable comparison to post-transplant survival, for each organ, the survival function time was the average time transplant would be anticipated if an organ were available (specifically, at the average time from listing until transplant among organ recipients).

We assessed the proportionality of hazards over time for significant mortality predictors by testing interactions between each factor and a continuous linear function of time after listing. The earliest available date of wait listing was used for patients wait-listed for transplant at multiple centers. For patients listed for retransplantation, we considered each listing as a distinct observation. Following intention to treat, patients were not censored upon delisting. Dates of death, collected by UNOS, were augmented by Social Security-linked death records supplied by the SRTR, providing improved ascertainment of deaths following delisting and survivable transplant failure without retransplant for kidney, kidney-pancreas and pancreas transplant. Censoring for last available follow-up was structured following the recommendations of the SRTR (8). Confidence intervals were derived using Monte Carlo simulation techniques using standard errors of the estimates.

Post-transplant survival:  We modeled the risk of post-transplant death or retransplantation over time for each organ with patient survivor functions through 5-year post-transplant adjusted to the average characteristics of transplant recipients. Retransplantation was included as an endpoint in the analysis because the life-extending benefit given by an individual organ ends at retransplant. Time-specific rates of death or retransplantation beyond 5-year post-transplant were projected for each organ individually from average estimated rates between 3- and 5-year post-transplant assuming a constant hazard of death or retransplant.

Quantifying the benefits of transplantation

Rationale for considering allocation priority: ‘The chain of benefits’:  An extra organ donor can generate up to seven additional solid organ transplants, a liver, heart, pancreas, two kidneys and two lungs, producing a domino effect through the organ specific waiting lists. A recipient of one of these extra organs would either have died prior to transplant without the organ or later received an organ now released for another wait-listed patient. This second wait-listed patient would either have died prior to transplant without the first patient's released organ or later have received an organ now released for a third wait-listed patient. This ‘chain of benefits’ continues until a patient who receives a transplant otherwise would have died had the extra donor not become available, thereby preventing a death before transplant without releasing another organ.

As a consequence of allocation priority, a patient receiving an organ from an extra donor, or receiving a released organ because of an extra donor, is positioned in the waiting list to receive another organ soon. Therefore, any single patient receiving an extra or released organ would benefit from avoiding death without a transplant for a relatively short interval of time. However, these intervals are linked together sequentially in time until a death is actually avoided.

Calculation of life-years saved:  The area under a survivor function quantifies the survival experience of a population in terms of average expected life-years per patient. To estimate the life-year expectations of the study samples, we integrated the post-transplant survivor functions and allocation-priority-weighted survivor functions for transplant candidates from the time of transplantation or anticipated transplantation if an organ were available, respectively, over 40 years. We calculated the additional life-years provided from an organ by subtracting the survival expectations among wait-listed candidates near the time of transplant from the area under the survivor function given by the estimated and projected rates of death or retransplant after transplantation. No benefits were included beyond 40-year post-transplant.

Sensitivity analysis

We assessed the robustness of our results to changes in model assumptions with sensitivity analyses. Specifically, we tested how estimated life-years of benefit were affected by: (i) employing multivariate-adjusted average transplant-censored wait-list survival experience, as used in previous analyses (3,4), instead of survival experience near the time of transplant;(ii) assuming a constant incidence of death (linear decline) instead of the more commonly applied constant hazard of death (exponential decline) when generating survivor functions from average death rates; (iii) altering the interval lengths for calculation of average death rates among patients awaiting urgency-distributed organs; (iv) altering the intervals for calculation of time-specific death rates among candidates for organs allocated according to waiting time and (5) estimating wait-list survivor functions without adjustment for the average characteristics of transplanted patients. Outcomes reported in the medical literature, such as life-years, are frequently discounted to present value at an annual rate of 3%, although this is relatively uncommon in the transplant literature. We thus also, (vi) calculated the life-year benefit provided by an organ donor at the time of transplant after discounting to present value.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Characteristics of deceased transplant donors

All 45 845 deceased organ donors procured in the United States between 1995 and 2002 were included. Donors procured in 2002, the most recent year studied, were used as the reference for procurement and utilization rates (Table 1). Donors most commonly provided kidneys and livers, followed by hearts, pancreata and lungs. Donor age had a strong impact on the use of organs. With the exception of the liver, organs from donors between 18 and 40 years were the most utilized, while organs from donors over 59 years were the least utilized. Organs from donors with head trauma as a cause of death were used the most frequently, especially lungs, hearts and pancreata. There was relatively little effect of race or gender on the utilization of organs. However, white donors were the most common and donors were more commonly male than female.

Table 1.  Characteristics of deceased transplant donors procured in 2002
 Number of donors*Percent of donors providing one or more:
Kidney Liver Lung HeartKidney- pancreas Pancreas
  1. *Demographic information was not available for the 279 deceased donors whose organs were not used for transplantation. These patients are included in the total count of donors but not in demographic breakdowns.

  2. **CVA, Cerebrovascular Accident.

All donors*617581.1%79.2%14.0%34.9%14.6%8.9%
Donor age
 0–17 years119768.3%64.7%11.5%36.9%15.3%11.0%
 18–40 years221492.1%86.9%21.1%53.8%26.8%15.4%
 41–59 years200385.9%77.2%12.1%25.0%6.4%3.8%
 >59 years76156.9%84.6%1.8%2.5%0.0%0.0%
Donor cause of death
 Head trauma256292.8%86.3%19.4%52.0%23.7%14.7%
Donor race
Donor gender

Survival experience of transplant candidates and recipients

Adjusted wait-listed patient survival curves are presented in Figure 1. Wait-listed mortality risk was least for pancreas and sequentially greater for kidney, kidney-pancreas, lung, heart and liver. Urgency strata were highly predictive of wait-listed survival for heart and liver. Post-transplant patient survival curves are presented in Figure 2. Post-transplant survival was similar for pancreas, kidney and kidney-pancreas, less and similar for heart and liver, and least for lung.


Figure 1. Wait-listed patient survival censored at the time of transplantation but including deaths following delisting. Wait-listed patient survival functions for organs primarily distributed by waiting time adjusted for patient characteristics are shown in Panel A. The expected 5-year transplant censored wait-listed patient survivals were 80.3% for pancreas, 66.5% for kidney, 54.4% for kidney-pancreas and 37.7% for lung. The expected death rates were nearly constant over time for pancreas, increased through approximately 5 years following listing then were nearly constant subsequently for kidney and kidney-pancreas, and decreased over time for lung. Time-specific death rates were weighted by the distribution of time from listing to transplant for recipients of transplants to estimate the average rates of death for patients at high priority for transplantation, the rates of death avoided by extra organs. These estimates of the average rates of death near the time of transplantation were 4.1% for pancreas, 6.9% for kidney, 9.3% for kidney-pancreas and 22.1% for lung. Wait-listed patient survival for organs distributed by urgency are shown stratified by MELD or PELD quartiles for liver in Panel B and stratified by status for heart in Panel C. Urgency at listing has a large impact on the expected survival of patients waiting for livers or hearts. For patients listed for liver or heart transplant, the expected transplant censored wait-listed patient death rates stratified by urgency between 30 and 121 days following listing were adjusted to the average characteristics of patients receiving transplants. These death rates were weighted by the distribution of urgency observed at the time of transplant and used as estimates of the rates of death avoided by additional transplanted organs. For liver, stratified by MELD or PELD quartiles observed at transplant, annualized death rate estimates were 32.6% for MELD or PELD less than 15 years, 47.9% for MELD or PELD between 15 and 21 years, 78.6% for MELD or PELD between 22 and 29 years and 99.3% for MELD or PELD greater than 29 years. For heart stratified by status annualized death rate estimates were 20.9% for status 2, 44.1% for status 1B and 57.0% for status 1A.


Figure 2. Patient survival following transplantation ending at the time of death or retransplantation. Survival following transplantation is characterized by two distinct phases. There is an early post-transplant period of several months with considerable risk of death or need for retransplantation. This early high-risk period can be seen easily in the recipients of liver, heart and lung transplants, but is also apparent for the recipients of kidney-pancreas, kidney and pancreas transplants. The early high-risk period is followed by a pattern of approximately constant and lower risk. The estimated survival functions adjusted to the average characteristics of transplant recipients for each organ were used as estimates of the time-specific death or retransplant risk following transplantation through 5-year post-transplant. The long-term, greater than 5 years, rate of death or retransplantation was estimated by the average adjusted rate between 3- and 5-year post-transplant. These long-term annual estimates of the rates of death or retransplant were 2.9% for kidney pancreas, 3.8% for kidney, 4.3% for pancreas, 3.6% for liver, 4.4% for heart and 10.9% for lung.

Download figure to PowerPoint

Expected life-year gains from an extra deceased donor

Table 2 presents the expected gain in life-years from an extra deceased organ donor. The average deceased organ donor gives 30.8 additional life-years, accounting for utilization of organs. All organs, with the exception of the pancreas, were found to provide a significant life-extending benefit. Liver transplantation was found the most beneficial, providing an additional expected 16.9 life-years. Other benefits included: heart (14.5 life-years), kidney pancreas (12.9 life-years), kidney (7.2 life-years) and lung (2.1 life-years). Figure 3 presents an example of the determination of liver transplant benefits. An average donor who provided seven organs and generated a kidney, liver, heart, kidney-pancreas and two lung transplants provided an additional 55.8 life-years.

Table 2.  Average expected gain in life-years for wait-listed patients from an extra deceased organ donor
  Kidney Liver Lung HeartKidney- Pancreas PancreasTotal of all organs
  1. *The calculation of the average benefit if all organs are transplanted assumes the pancreas is put to its estimated best use, kidney-pancreas transplantation.

  2. **Following the estimates of the benefits of transplant, each organ was considered separately in utilization counts. Therefore, donors providing two kidney, two lung or two liver (split liver) transplants were included twice in frequency counts. The total benefit expected from an average donor was calculated first by multiplying the average frequency of each type of transplant in 2002 by that transplant's expected benefit. These values, reported as the average benefit given the average frequency of transplants in 2002, were summed to the average total benefit expected from a deceased donor.

Average benefit by type of transplant (life-years)−0.3 
 95% CI(6.4, 8.1)(15.6, 18.4)(1.2, 3.4)(13.0, 16.2)(10.5, 15.7)(−4.8, 5.8) 
Average benefit if all organs were transplanted (life-years)* 55.8
 95% CI(6.4, 8.1)(15.6, 18.4)(2.2, 6.7)(13.0, 16.2)(10.5, 15.7) (52.0, 60.2)
Average organ utilization per donor in 2002**
Average benefit given average frequency of transplants in 2002 (life-years)
 95% CI(8.8, 11.2)(12.5, 14.7)(0.2, 0.6)(4.5, 5.6)(1.5, 2.3)(−0.4, 0.5)(29.1, 32.7)

Figure 3. The benefit of liver transplantation. For liver transplant recipients the adjusted survival function for death or retransplantation through 5-year post-transplant (green) is plotted together with its projection through 40-year post-transplant (red). Also plotted is an average wait-list survival function based on the average rate of death expected for patients near the time of transplantation (blue). This average wait-list survival rate is drawn from the survival expectations of patients wait-listed for liver transplantation between 30 and 121 days after listing stratified by quartiles of MELD or PELD scores observed at transplantation (see Figure 1 panel B). The benefit of liver transplantation, 16.9 life-years, is given by the shaded area between the post-transplant survival function and the average wait-list survival function. The majority, 65%, of the benefit has been realized by 20-year post-transplant, and by 30-year post-transplant 86% of the benefit has been realized.

Download figure to PowerPoint

The expected benefit provided by various subgroups of deceased donors is presented in Table 3. It is important to note that benefits from different groups of donors are conditional in part on the average characteristics of the patients receiving the organs. Although every donor age group was found to provide life-extending benefits, donor age strongly influenced the size of the estimates, perhaps explaining in part the utilization patterns of organs by age shown in Table 1. The benefits when seven organs were transplanted were similar through age 40, but declined at older donor ages. Combining average utilization of organs with average benefits, donors between the ages of 18 and 40 provided 41.4 life-years, under age 18 provided 30.7 life-years, between the ages of 41 and 59 provided 22.9 life-years and over the age of 59 provided 14.2 life-years. Cause of donor death was also associated with considerable variation in benefit. Anoxia victims were found the most beneficial donors when all organs were used, providing 70.2 life-years. However, due to greater utilization of organs, head trauma victims provided the greatest utilization adjusted average benefit, 39.6 life-years. Victims of cerebrovascular accidents or strokes were found to provide smaller benefits, although, still considerable, 46.8 life-years if seven organs were transplanted and 23.3 life-years adjusting for average organ utilization. Relatively small differences were found in the size of benefits provided by race and gender.

Table 3.  Average expected gain in life-years for wait-listed patients from an extra deceased organ donor by donor characteristics estimated uniquely for each organ within each subgroup
 Percentage of all donors in 2002 Expected benefit if all organs were transplanted Utilization of organs in 2002Expected benefit adjusted to average utilization of organs in 2002****
YearsNo. of organs Years
  1. *The estimated benefits provided by donors over the age of 59 exclude potential benefits from lung, heart, kidney-pancreas or pancreas transplantation. These benefits were not estimated because it was expected that too few of these transplants had occurred to obtain reliable estimates (see Table 1), and the contribution to total benefit from these transplants is expected to be small due to their rarity.

  2. **Potential benefits were not estimated from deceased donors who were neither white nor black and those who had a cause of death other than anoxia, CVA/stroke or head trauma. These benefits were not estimated because it was expected that too few of these transplants had occurred to obtain reliable estimates (see Table 1).

  3. ***CVA, cerebrovascular accident.

  4. ****The adjustment for average utilization of organs was calculated for each transplant and summed as in Table 2.

All donors94.755.82.930.8
Donor age
 0–17 years19.461.22.530.7
 18–40 years35.960.13.641.4
 41–59 years32.444.52.822.9
 >59 years*12.319.01.814.2
Cause of death**
 Head trauma41.559.63.539.6

Sensitivity analysis

Use of multivariate adjusted transplant censored wait-list outcomes, using methods described by Wolfe et al. (6), had little effect on the estimated total benefit of a donor, increasing the utilization adjusted estimate by 7% to 33.8 life-years and reducing by 8% the estimated benefits if all organs are used to 51.3 life-years. However, this had considerable effect on relative organ benefits leaving liver the most beneficial, followed by kidney-pancreas, kidney, heart, lung and pancreas resulting in a benefit of 16.7, 15.2, 10.4, 8.1, 0.5 and −0.5 life-years, respectively. Assuming a constant incidence of death instead of a constant hazard of death reduced the overall estimates by 12% to 27.0 life-years for the average benefit and 13% to 48.5 life-years if all organs were used. Again, individual organ benefits were reordered with kidney-pancreas providing the most benefit followed by liver, heart, kidney, lung and pancreas resulting in a benefit of 13.2, 12.9, 10.3, 7.8, 2.2 and −0.8 life-years, respectively. Extending or shortening the period of the wait-list estimate for heart or liver had little effect on the results. In addition, extending or shortening the interval length for the waiting-list estimates for kidney, kidney pancreas, pancreas and lung had little effect on the results. Adjusting waiting-list death rates to reflect the average characteristics of patients receiving transplants compared to unadjusted estimates had almost no effect on the results. Finally, after discounting average benefit at the time of transplant by 3% annually to present value, the overall estimates were 19.2 life-years for the average benefit and 34.9 life-years if all organs are used.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Although the desirability of organ donation as a societal goal has not been questioned, until now, no clear estimate of the benefit given by a deceased organ donor has been put forth. In a previous model, we attempted to estimate the threshold for investment in organ donation by looking at cost savings and quality gains related to renal transplantation (4), yet clearly the true value of organ donation extends far beyond simply avoiding dialysis for end-stage renal disease patients. Here we have expanded our focus to estimate the life-years gained not only from renal transplantation, but also from the resulting liver, heart, lung and pancreas transplants that can be performed as a result of the act of donation. The results are staggering. A single donor gives on average over 30 additional life-years to patients awaiting transplantation. A fully utilized donor, providing seven organs, gives over 55 life-years. In addition, transplantation of every organ type has been associated with significant quality of life benefits (9). These findings give a new meaning to the phrase ‘gift of life’.

It has been estimated that only 42% of eligible donors, medically suitable patients who have met brain-death criteria, actually undergo donation for a variety of reasons, including systems problems, lack of identification and lack of consent (10). Our estimates translate this to over 250 000 life-years lost annually in the United States alone because of the inability to convert eligible organ donors into actual organ donors. This is similar to estimates of life-years lost annually in the United States from stroke or liver disease (11). Further, combining lost potential donors and unused or unprocured organs, our estimates suggest an annual loss of over 600 000 life-years, similar to the life-years lost as a result individually of suicide, congenital anomalies, homicide or perinatal conditions and approximately one half to one fourth the life-years lost from HIV infection, heart disease, cancer or accidents (11).

As unlikely as it is that all deaths due to heart disease will be prevented, we cannot expect all potential donors to be procured and all organs utilized. Although consent rates for donation approaching 100% would seem unlikely, several interventions may remove current barriers and improve consent rates (12,13). In addition, improved utilization of organs already procured may be obtained through a variety of mechanisms including altered allocation algorithms (14) and improved methods of organ preservation (15).

These results suggest that interventions with a reasonable chance of increasing organ donation are likely to be extremely beneficial to wait-listed patients and society. Nonetheless, there are limited health care resources and a variety of proposals to increase organ donation including, but not limited to, implementing best practices in organ donation across all organ procurement organizations (16), establishing donor registries (17), increasing public education (18), extending financial and other rewards for donation (19) and changing the cultural and legal process of consent (18,20). Our work does not provide information regarding which, if any, of these interventions might be most beneficial, but it does provide information useful for the evaluation and comparison of these and other interventions. For example, the recent HHS initiative to increase donation consent rates to 75% in hospitals with the most potential donors would save 60 000 life-years annually if successful according to our estimates (21)

The relative contributions of the various organs to the overall life-years gained are noteworthy. Traditionally, renal transplantation has been considered the least ‘life saving’ of all solid organ transplants. In fact, the relative contribution in terms of life-years gained from a single kidney transplant is substantially more than a lung transplant and roughly half of the life-years gained as the result of a liver or heart transplant. This gain is magnified by the ability to transplant two kidney recipients from a single donor routinely. In some cases, such as combined kidney-pancreas transplant, although the relative contribution of a transplant to added longevity is significant, the relatively small number of transplants produced from an average donor leads to a relatively lower contribution to added life-years. This effect is also important in lowering the contribution of a heart transplant to the average life-years gained, because only 34.9% of donors provide a heart transplant. In the case of lung transplant, the combination of very low procurement rates and a relatively small amount of incremental longevity from transplantation makes it the lowest contributor to the value of a donor. Further, while there is evidence of quality of life benefits of pancreas transplantation (9), our estimates suggest solitary pancreas transplantation provides no life-extending benefit, as has been reported recently (5). Therefore, the expected relative benefit given by a deceased donor is influenced importantly by the organs that are procured and how they are allocated.

It should be noted that in order to derive accurate values for life-years gained by an additional donor, survival for wait-listed patients need to reflect survival for patients expecting transplant soon, and not the average survival of the entire waiting-list. This is especially important for patients awaiting liver and heart transplantation and also has some importance for lung transplantation. While the long-term survival of patients wait-listed for liver, heart and lung transplantation may be considerably better than early survival, the large majority of these transplants are allocated to patients with relatively short waits, higher degrees of urgency and higher risks of death. Adjusting our model to accurately reflect death rates for the next available recipient to be transplanted substantially increased its complexity. However, because our focus was on the benefits of additional donors, we feel it is important, and especially so in examining the different contributions of different organs to the overall life-years gained. For example, our methods produce similar but marginally smaller estimates of the benefits of kidney transplantation compared to long-term wait-listed survival (6,3) as has been used in previous studies. However, the estimates of the total benefit from a deceased donor were similar when using our method or methods used in previous studies (3–6). Therefore, there is evidence of considerable life-year benefits provided by a deceased organ donor regardless of the chosen estimation methods.

Certainly there are limitations of any attempt to predict future outcomes. The future is unknown and advances in medicine, or the lack of them, could impact survival expectations considerably. It is difficult to predict with certainty what the future holds for transplant with the development of new medications, new sources of organs and new allocation mechanisms. For example, while this remains to be seen, we might expect that recent changes in lung allocation with more emphasis on urgency may increase the benefits of lung transplantation. On the other hand, new diseases may develop adversely impacting long-term transplant outcomes, for example, a virus, or other infectious agent, causing particular difficulties in the immune compromised, could adversely impact long-term transplant outcomes. As shown in the sensitivity analysis, using alternative methods and assumptions produced some variation in the estimates of the relative ordering of specific organs by expected benefit. Some of this observed variation is driven by recipient characteristics and further examination of the relative benefits of transplant to recipients with differing characteristics, such as the frequency of HLA types, insurance status, education, comorbidity and region, is clearly warranted. However, organs estimated to provide large benefits using our method were estimated to provide large benefits with other methods as well. Similarly, organs with relatively small benefits estimated by our methods have small benefit estimates with these alternative methods. Ultimately, under a variety of assumptions, the estimated total benefit provided by a deceased donor is important and large.

In conclusion, the benefit given by a deceased organ donor is substantial and justifies strongly current initiatives designed to increase organ donation. The magnitude of life lost because potential donors are not procured and organs are unused ranks organ donation with some of the most important public health concerns. More extensive investments in efforts to increase organ donation are warranted.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference

Supported in part by a grant from the National Institute of Diabetes, Digestive, and Kidney, Diseases K25-DK-02916–03, Mark A. Schnitzler, Ph. D., P. I. Data reported here from the Scientific Registry of Transplant Recipients (SRTR) have been supplied by the United Network for Organ Sharing and URREA under contract with HHS. The authors alone are responsible for reporting and interpreting of these data.


  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Reference
  • 1
    URREA, UNOS. 2002 Annual Report of the US Organ Procurement and Transplantation Network and the Scientific Registry of Transplant Recipients and Transplant Data 1992–2001 [Internet]. 2003. Rockville , Maryland , HHS/HRSA/OSP .
  • 2
    National hospital discharge survey: 2000 annual summary with detailed diagnosis and procedure codes. Washington DC , Department of Health and Human Services, Center for Disease Control, National Center For Health Statistics 2001 . (Series 13, number 153, 10.)
  • 3
    Schnitzler MA, Whiting JF, Brennan DC et al. The expanded criteria donor dilemma in cadaveric renal transplantation. Transplantation 2003; 75: 19401945.
  • 4
    Whiting JF, Kiberd B, Kalo Z, Keown P, Roels L, Kjerulf M. The cost-effectiveness of organ donation: evaluating investment into donor action and other donor initiatives. Am J Transplant 2004; 4: 569573.
  • 5
    Venstrom JM, McBride M, Rother KI, Hirshberg B, Orchard TJ, Harlan DM. Survival after pancreas transplantation in patients with diabetes and preserved kidney function. JAMA 2003; 290: 28612863.
  • 6
    Wolfe RA, Ashby VB, Milford EL et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med 1999; 341: 17251730.
  • 7
    UNOS. UNOS Policy 3.0 Organ Distribution. UNOS Policies and Procedures, 3–1–3–56. 2003. Rockville , MD , HHS/HRSA/OSP/DOT .
  • 8
    Dickinson DM, Ellison MD, Webb RL. Data Sources and Structure. OPTN/SRTR 2002 Annual Report: 926.
  • 9
    Dew MA, Switzer GE, Goycoolea JM et al. Does transplantation produce quality of life benefits? Transplantation 1997; 64: 12611269.
  • 10
    Sheehy E, Conrad SL, Brigham LE et al. Estimating the number of potential organ donors in the United States. N Engl J Med 2003; 349: 667674.
  • 11
    Selik RM, Chu SY. Years of potential life lost due to HIV infection in the United States. AIDS 1997; 11: 16351639.
  • 12
    Boulware LE, Ratner LE, Sosa JA, Cooper LA, LaVeist TA, Powe NR. Determinants of willingness to donate living related and cadaveric organs: identifying opportunities for intervention. Transplantation 2002; 73: 16831691.
  • 13
    Terell F, Mosley KL, Terrell AS, Nickerson KJ. The relationship between motivation to volunteer, gender, cultural mistrust, and willingness to donate organs among blacks. J Natl Med Assoc 2004; 96: 5360.
  • 14
    Rosengard BR, Feng S, Alfrey E et al. Report of the Crystal city meeting to maximize the use of organs recovered from the cadaveric donor. Am J Transplant 2002; 2: 701711.
  • 15
    McLaren AJ, Friend PJ. Trends in organ preservation. Transpl Int 2003; 16: 701708.
  • 16
    Cohen B, Wight C. A European perspective on organ procurement: breaking down the barriers to organ donation. Transplantation 1999; 68: 985990.
  • 17
    Merion R, Vinokur AD, Couper MP et al. Internet-based intervention to promote organ donor registry participation and family notification. Transplantation 2003; 75: 11751179.
  • 18
    Oz MC, Kherani AR, Rowe A et al. How to improve organ donation: results of the ISHLT/FACT poll. J Heart Lung Transplant 2003; 22: 389410.
  • 19
    Wigmore SJ, Forsythe JLR. Incentives to promote organ donation. Transplantation 2004; 77: 159161.
  • 20
    Chouhan P, Draper H. Modified mandated choice for organ procurement. J Med Ethics 2003; 29: 157162.
  • 21
    HHS Initiative to increase organ donation; Organ Donor Breakthrough Collaborative. 2003. (Accessed March 11, 2003, at