Underutilization of Hepatitis C-Positive Kidneys for Hepatitis C-Positive Recipients


Corresponding author: Dorry Segev, dorry@jhmi.edu


Hepatitis C-positive (HCV(+)) candidates likely derive survival benefit from transplantation with HCV(+) kidneys, yet evidence remains inconclusive. We hypothesized that lack of good survival benefit data has led to wide practice variation. Our goal was to characterize national utilization of HCV(+) kidneys for HCV(+) recipients, and to quantify the risks/benefits of this practice. Of 93,825 deceased donors between 1995 and 2009, HCV(+) kidneys were 2.60-times more likely to be discarded (p < 0.001). However, of 6830 HCV(+) recipients, only 29% received HCV(+) kidneys. Patients over 60 relative rate (RR 0.86), women (RR 0.73) and highly sensitized patients (RR 0.42) were less likely to receive HCV(+) kidneys, while African Americans (RR 1.56), diabetics (RR 1.29) and those at centers with long waiting times (RR 1.19) were more likely to receive them. HCV(+) recipients of HCV(+) kidneys waited 310 days less than the average waiting time at their center, and 395 days less than their counterparts at the same center who waited for HCV(−) kidneys, likely offsetting the slightly higher patient (HR 1.29) and graft loss (HR 1.18) associated with HCV(+) kidneys. A better understanding of the risks and benefits of transplanting HCV(+) recipients with HCV(+) kidneys will hopefully improve utilization of these kidneys in an evidence-based manner.


The prevalence of Hepatitis C virus (HCV) is approximately 12% among end-stage renal disease (ESRD) patients (1), and HCV(+) patients have an increased risk of death on dialysis when compared with patients who are HCV(−) (1–4). Similarly, the prevalence of HCV is 4.2% among deceased donors (5–7). Kidney transplantation (KT) in HCV(+) recipients is associated with slightly worse outcomes than transplantation in HCV(−) recipients, including increased risk of death and graft loss, and increased incidence of posttransplant diabetes (8–13). However, this practice is considered a safe alternative to dialysis treatment, and several single-center studies have shown that HCV(+) recipients derive a survival benefit from KT when compared with remaining on dialysis (14–16).

The use of HCV(+) kidneys is controversial, and initial studies recommended excluding them from the organ supply given the near certainty of HCV transmission to the recipient (17,18). However, in 1994 a cost-benefit analysis suggested that a policy where HCV(+) kidneys were transplanted into HCV(+) recipients might provide better patient outcomes than a discard policy (19).

Evidence suggests that outcomes of HCV(+) recipients who receive kidneys from HCV(+) donors are slightly worse than outcomes of HCV(+) recipients who receive kidneys from similar HCV(−) donors (5,20). So if an HCV(+) kidney and a comparable HCV(−) kidney were both available for a given HCV(+) patient, the choice would be intuitive. However, a given patient never faces this decision; rather, the true clinical decision is whether to accept the HCV(+) organ offer currently at hand or to wait on dialysis for the next HCV(−) offer (21). Whether HCV(+) candidates derive a survival benefit from being transplanted with HCV(+) kidneys (versus waiting for HCV(−) kidneys) has been a difficult question to study because no national registry collects HCV status of all candidates on the waiting list; UNOS collects HCV status only when a patient receives a kidney, not when the patient is added to the waiting list (22). The obvious potential benefit for an HCV(+) patient to accept the currently available HCV(+) kidney, as opposed to waiting for the next available HCV(−) kidney, would be decreased waiting time and as such decreased dialysis mortality. At least one single-center study has observed this, with shorter waiting times for HCV(+) patients who accept HCV(+) kidneys (23).

We hypothesized that the inability to quantify the survival benefit of HCV(+) KT in HCV(+) candidates has caused high discard rates of HCV(+) kidneys and varied practice patterns among those using them. We further hypothesized that those HCV(+) recipients who did receive HCV(+) kidneys would have significantly shorter waiting times (and thus lower risks of death on the waiting list) than those who waited for HCV(−) kidneys. The goals of our study were to explore national practice patterns in discard and utilization of HCV(+) kidneys for HCV(+) recipients, and to quantify risks and benefits associated with receiving an HCV(+) kidney.


Study population

We studied 6830 HCV(+) patients who received a deceased donor KT between January 1, 1995 and February 20, 2009 as reported to UNOS. We also studied 93 825 deceased donors during the same time period where at least one organ was recovered, of whom 93 120 had information about HCV status and 3321 were HCV(+).

Recipients of HCV(+) kidneys

To estimate the relative rate (RR) that HCV(+) recipients were transplanted with HCV(+) kidneys (as opposed to HCV(−) kidneys), we built a multivariate generalized linear model (GLM) with a Poisson family and log link as previously described (24), adjusted for age, gender, ethnicity, insurance, body mass index (BMI), diabetes, hypertension, previous transplant, peak panel reactive antibody (PRA), years on dialysis and center waiting time, accounting for center-level clustering.

Race and receipt of HCV(+) kidneys

Initially, we found that African-American (AA) HCV(+) recipients had almost twice the rate of receiving an HCV(+) kidney than their non-AA HCV(+) counterparts. Since this disparity might have occurred at the center-level, such that centers with higher proportions of AA recipients utilized HCV(+) kidneys more than other centers, we adjusted for the proportion of AA recipients at each center. To further explore the relationship between race and receipt of an HCV(+) kidney, we calculated the number of AAs that would have been expected to receive HCV(+) kidneys per center if distribution had been uniformly random, by multiplying the proportion of HCV(+) recipients who were AA by the number of HCV(+) kidneys transplanted at that center. We then compared the expected and observed number of HCV(+) kidneys allocated to AA recipients for each center.

Center-level distributions of HCV(+) recipients and HCV(+) donors

For each center, we calculated the total number of HCV(+) recipients and HCV(+) kidneys, and the percentage of HCV(+) recipients transplanted with HCV(+) kidneys. To better understand center clustering (i.e. what proportion of the national volume was performed by what number of centers), we compared the center-level cumulative distribution of HCV(+) recipients and HCV(+) kidneys. The more area under each cumulative distribution curve, the fewer centers that performed the bigger bulk of transplants in HCV(+) recipients. To test whether center-level variation correlated with center-level patient characteristics, we examined the distribution of the difference between the percentage of HCV (+) recipients expected to receive HCV (+) kidneys and that observed for each center. To determine the percent expected for each center, we used the GLM described above to calculate each patient's predicted probability of receiving an HCV(+) kidney based on national practice.

Waiting time and HCV(+) kidneys

For each center, we calculated the average waiting time among HCV(+) recipients at that center (regardless of donor HCV status). We then calculated the difference between an individual recipient's waiting time and the average waiting time at that recipient's center and compared the difference in waiting time by donor HCV status.

Discard of HCV(+) kidneys

From donors in our study population, each kidney available for possible transplantation was analyzed separately. We built a GLM as above (24) to estimate the RR of discard of HCV(+) kidneys, adjusted for donor race, gender, age, year of recovery, cause of death, donation after cardiac death (DCD) status, creatinine, blood type, BMI, hypertension, expanded criteria donor (ECD) status and hepatitis B status (core antibody and surface antigen). Fifteen HIV(+)donors were excluded from analysis. We repeated the analysis with only non-DCD donors ages 15–45 with creatinine <2. We also repeated the analysis to examine the role of classification as ‘high infectious risk’ by the Center for disease control (CDC); since this was only captured after January 1, 2004, our repeated analysis was limited to donors after that date.

Survival in HCV(+) recipients

We built a Cox proportional hazards model to examine associations between donor HCV status and (1) patient survival and (2) death-censored graft survival among HCV(+) recipients. These associations were studied within the entire cohort, and then repeated in subgroups to explore possible effect modification by race, gender, age, diabetes, PRA and BMI. All models were adjusted for donor age, ECD status, DCD status, cold ischemic time (CIT) and creatinine, and recipient race, gender, BMI, insurance, diabetes, hypertension, angina, previous malignancy, peak PRA and years on dialysis.


Recipients of HCV(+) kidneys

Of 6830 HCV(+) recipients, the majority (71%) received an HCV(−) kidney, with only 1998 receiving an HCV(+) kidney. This did not vary significantly by year (Figure 1A). A higher proportion of recipients of HCV(+) kidneys were AA (66.4% vs. 46.0%), diabetic (35.3% vs. 26.7%) and privately insured (24.6% vs. 20.0%), while a lower proportion were female (20.1% vs. 31.0%), Caucasian (22.2% vs. 37.8%), highly sensitized (3.6% vs. 13.8%) and retransplants (13.1% vs. 22.6%). HCV(+) donors were significantly less likely to be ECD, DCD or have a creatinine >1.5 compared to HCV(−) donors (Table 1).

Figure 1.

Temporal trends in (A) percentage of HCV(+) recipients transplanted using HCV(+) kidneys, and (B) percentage of HCV(+) kidneys discarded.

Table 1.  Characteristics of kidney transplant recipients who are HCV(+) and their donors, stratified by donor HCV status
 HCV(−) donorHCV(+) donorp-Value
Total 6830 HCV(+) recipientsN = 4832N = 1998 
Recipient characteristics
Age (%)
 60 and over15.714.4 
Female (%)31.020.1<0.001
Ethnicity (%)
Education (%)
 High school or less54.765.6<0.001
 College/technical school25.722.0 
 Bachelors or more14.212.2 
Medicaid (%)
Private insurance (%)20.024.6<0.001
BMI (mean)26.026.5<0.001
Diabetes (%)26.735.3<0.001
Angina (%)
Malignancy (%)
Peak PRA > 80 (%)13.83.6<0.001
Previous transplant (%)22.613.1<0.001
Years on dialysis (mean)4.93.5<0.001
Donor characteristics
Age (%)
 60 and over8.61.6 
Female (%)40.435.3<0.001
Expanded criteria donor (%)16.39.2<0.001
Donor after cardiac death (%)
BMI (mean)
Cold ischemic time (mean)19.720.9<0.001
Creatinine > 1.5 (%)15.49.7<0.001

In multivariate analysis of HCV(+) recipients, we found that patients over 60 received HCV(+) kidneys 14% less often than patients under 60 (RR = 0.86, 95% CI: 0.77–0.96, p = 0.005, Table 2), and women received them 27% less often than men (RR = 0.73, 95% CI: 0.66–0.80, p < 0.001). Patients with PRA >80 were 58% less likely to receive HCV(+) kidneys compared to patients with lower PRA (RR = 0.42, 95% CI: 0.32–0.56, p < 0.001), likely due to the fact that highly sensitized patients receive higher allocation priority. Diabetes was the only comorbid condition associated with increased receipt of HCV(+) kidneys (RR = 1.29, 95% CI: 1.18–1.40, p < 0.001), consistent with the well established significantly higher dialysis death rates for diabetic patients and the resulting urgency to transplant these patients. Along those lines, patients from centers with longer waiting times were significantly more likely to receive HCV(+) kidneys (RR = 1.19 per quartile of waiting time, p = 0.002).

Table 2.  Multivariate model, relative rate of a HCV(+) patient receiving an HCV(+) kidney versus receiving an HCV(−) kidney, accounting for center-level clustering
Recipient characteristicRR (95% CI)p-Value
  1. *By quartile.

Age > 600.86 (0.77–0.96)0.005
Female0.73 (0.66–0.80)<0.001
African-American1.56 (1.39–1.75)<0.001
Medicaid1.04 (0.86–1.25)0.7
Private insurance1.06 (0.94–1.18)0.3
BMI > 350.91 (0.79–1.05)0.2
Diabetes1.29 (1.18–1.40)<0.001
Previous transplant0.99 (0.86–1.14)0.9
PRA > 800.42 (0.32–0.56)<0.001
Years on dialysis0.93 (0.92–0.95)<0.001
Waiting time at recipient's center*1.19 (1.06–1.33)0.002
Proportion of African-American recipients at center*1.20 (1.05–1.37)0.007

Race and receipt of HCV(+) kidneys

In a preliminary patient-level multivariate model, we found that AA HCV(+) recipients had 1.81-times the rate of receipt of HCV(+) kidneys (data not shown) compared with non-AAs. When we added to the model two center-level variables, (1) proportion of AAs at the center and (2) average waiting time at the center, we found the rate of receipt of HCV(+) kidneys in AAs was somewhat attenuated but by no means entirely explained (RR = 1.56, 95% CI:, p < 0.001). We also confirmed that, for the majority of centers (71%), the observed number of AA recipients of HCV(+) kidneys was greater than the expected number based on racial distributions at the center where they were transplanted (p < 0.001, Figure 2).

Figure 2.

Distribution of difference between observed and expected number of HCV(+) African-American recipients of HCV(+) kidneys at a center.

Center-level distributions of HCV(+) recipients and HCV(+) donors

There was substantial variation between centers in the proportion of HCV(+) recipients who received HCV(+) kidneys (Figure 3A). For example, 81 centers (representing 35% of centers who transplanted HCV(+) recipients) did not use any HCV(+) kidneys for their HCV(+) patients, while 31 (13%) transplanted over half of their HCV(+) recipients with HCV(+) kidneys. The average waiting time among centers that used no HCV(+) kidneys was 567 days (range: 44–1074 days), while the average waiting time was 767 days among centers that used any of these kidneys (range: 216–1609 days). Transplantation of HCV(+) kidneys was much more clustered at a small subset of centers than transplantation of HCV(+) recipients (Figure 3B), implying wider dissemination of comfort with HCV(+) recipients than comfort with HCV(+) kidneys. Center-level variation in utilization of HCV(+) kidneys did not appear to be explained by differences in composition of centers’ patients; when we calculated the percent of HCV(+) recipients expected to receive HCV (+) kidneys, based on the characteristics of these patients and national practice patterns in patients with similar characteristics, we found wide variation in expected versus observed percentages, ranging from 56% less than expected to 69% more than expected (Figure 3C).

Figure 3.

Center-level (A) distribution of the percentage of HCV(+) recipients transplanted with HCV(+) donors, (B) cumulative distribution of HCV(+) recipients and HCV(+) donors and (C) difference between observed and expected number of HCV(+) kidneys transplanted, by center.

Discard rate of HCV(+) kidneys

The proportion HCV(+) recipients who received HCV(+) kidneys increased from 20.1% in 1995 to 38.3% in 2008. However, there was little temporal variation in the proportion of HCV(+) kidneys discarded during the study period (Figure 1A). In general, HCV(+) kidneys were discarded at 2.90-times the rate of HCV(−) kidneys, even after adjusting for other factors associated with discard (95% CI: 2.52–2.68, p < 0.001) (Table 3). When we limited the analysis to kidneys recovered during 2004 and adjusted for CDC high-risk donor status, we found a similar association between HCV and discard (RR = 2.57, 95% CI: 2.47–2.68, p < 0.001, data not shown). During the study period, 53.6% of HCV(+) kidneys were discarded (a total of 3562) compared with only 22.4% of HCV(−) kidneys discarded. When the analysis was restricted to kidneys from non-DCD donors ages 15–45 with creatinine levels <2, HCV(+) kidneys were discarded at 4.72-times the rate of HCV(−) kidneys, adjusting for donor age, BMI, blood type and hypertension (95% CI: 4.44–5.03, p < 0.001). In this restricted donor pool, still 38.5% of HCV(+) kidneys were discarded (a total of 1127) compared with only 6.1% of HCV(−) kidneys discarded.

Table 3.  Relative rate of discard of HCV(+) kidneys, multivariate model considering (a) all donors and (b) only non-DCD donors under the age of 45 with creatinine < 2; *= BMI categorized as <25, 25–30, 30–35, 35–40 and >40
(A) Includes all donors
Donor factorRR of discardp-Value
Hepatitis C+2.60 (2.52–2.68)<0.001
HBV surface antigen+1.92 (1.67–2.20)<0.001
HBV core antibody+1.22 (1.18–1.27)<0.001
African-American1.04 (1.01–1.06)<0.001
Female1.07 (1.06–1.09)<0.001
 <180.92 (0.92–0.92)<0.001
 35–501.03 (1.03–1.04)<0.001
 50–651.04 (1.04–1.04)<0.001
 >651.05 (1.04–1.05)<0.001
Year of organ recovery1.00 (1.00–1.00)0.4
Cause of death = stroke1.08 (1.06–1.10)<0.001
Donor after cardiac death1.38 (1.33–1.43)<0.001
Donor creatinine >23.24 (3.19–3.29)<0.001
Biopsy >20% sclerosis1.88 (1.83–1.93)<0.001
AB blood type1.22 (1.18–1.27)<0.001
BMI*1.02 (1.02–1.03)<0.001
Hypertension1.50 (1.47–1.53)<0.001
Expanded criteria donor1.06 (1.03–1.09)<0.001
(B) Restricted to non-DCD donors < 45 with creatinine <2
Donor factorRR of discardp-Value
Hepatitis C+4.72 (4.44–5.03)<0.001
HBV surface antigen+3.18 (2.45–4.14)<0.001
HBV core antibody+1.69 (1.57–1.82)<0.001
African-American1.05 (0.98–1.13)0.001
Female1.19 (1.13–1.26)<0.001
Year of organ recovery0.99 (0.98–0.99)<0.001
Cause of death = Stroke1.08 (1.02–1.15)0.01
Biopsy >20% sclerosis4.47 (3.97–5.05)<0.001
Blood type AB1.32 (1.17–1.49)<0.001
BMI*1.04 (1.01–1.06)0.002
Hypertension1.70 (1.59–1.81)<0.001

Waiting time among recipients with HCV

On average, HCV(+) patients who received HCV(−) kidneys waited 856 days, while those who received HCV(+) kidneys waited 469 days. Looking at this from a center-level, making comparisons among recipients within the same centers, we found that HCV(+) recipients of HCV(−) kidneys waited 85 days longer than the average waiting time at their centers, while HCV(+) recipients of HCV(+) kidneys waited 310 days less than the average waiting time at their centers. In other words, recipients of HCV(+) kidneys waited on average 395 days less than those recipients who waited for HCV(−) kidneys at the same center.

Patient survival

Consistent with other studies, we found that, among HCV(+) patients, receipt of an HCV(+) kidney was associated with 1.29-times the hazard of death in adjusted analyses (HR = 1.29, 95% CI: 1.15–1.45, p < 0.001) (Table 4A). However, this hazard ratio only translates to a difference of 1% in 1-year survival (94% for HCV(−) kidneys vs. 93% for HCV(+) kidneys, per unadjusted Kaplan–Meier estimates) and a difference of 2% in 3-year survival (85% vs. 83%).

Table 4.  Hazard of (a) death and (b) graft loss associated with transplantation of kidneys from HCV(+) donors, compared with transplantation of kidneys from HCV(−) donors, in multivariate models, stratified and nonstratified analyses
SubgroupHR HCV(+) kidneyp-ValueSubgroupHR HCV(+) kidneyp-Value
  1. *All models adjusted for recipient factors including: age, gender, BMI, insurance status, race/ethnicity, diabetes, hypertension, PRA, previous transplant, years on dialysis and donor factors including: age, race/ethnicity, BMI, hypertension, diabetes, ECD status, DCD status, creatinine and cold ischemic time.

All HCV(+) recipients1.29 (1.15–1.45)<0.001   
Not African-American1.60 (1.35–1.90)<0.001African-American1.08 (0.91–1.29)0.4
Male1.21 (1.04–1.42)0.01Female1.53 (1.18–1.99)0.001
Age < 601.28 (1.11–1.47)0.001Age > 601.36 (0.96–1.89)0.1
No diabetes1.38 (1.19–1.60)<0.001Diabetes1.17 (0.93–1.49)0.2
PRA < 801.32 (1.17–1.49)<0.001PRA > 801.03 (0.53–2.01)0.9
BMI < 351.26 (1.11–1.42)<0.001BMI > 351.94 (1.14–3.31)0.01
SubgroupHR HCV(+) kidneyp-ValueSubgroupHR HCV(+) kidneyp-Value
All recipients1.18 (1.04–1.32)0.007   
Not African-American1.31 (1.03–1.67)0.03African-American1.09 (0.93–1.27)0.3
Male1.13 (0.98–1.30)0.1Female1.26 (0.98–1.62)0.1
Age < 601.15 (1.01–1.31)0.04Age > 601.43 (0.93–2.21)0.1
No diabetes1.20 (1.04–1.38)0.01Diabetes1.19 (0.93–1.52)0.2
PRA < 801.16 (1.02–1.32)0.02PRA > 801.22 (0.68–2.21)0.5
BMI < 351.19 (1.05–1.35)0.006BMI > 350.97 (0.52–1.80)0.9

Furthermore, when we repeated the analysis stratified by various factors, we found that some HCV(+) subgroups did not experience any difference in survival when transplanted with HCV(+) kidneys versus HCV(−) ones. For example, non-AAs had a significantly increased hazard of death associated with receipt of an HCV(+) kidney (HR = 1.60, 95% CI: 1.35–1.90, p < 0.001) (Table 4A, left column), while this increase was not seen in African Americans (HR = 1.08, p = 0.4) (Table 4A, right column). Similarly, patients over 60, diabetics, and those with PRA >80, did not have a statistically significantly increased hazard of death associated with receipt of an HCV(+) kidney, while patients under 60 had 1.28-times the hazard of death when transplanted with HCV(+) kidneys, diabetics had 1.38-times the hazard, and those with PRA <80 had 1.32-times the hazard (p < 0.001 for all estimates).

Graft survival

Overall, HCV(+) patients who received HCV(+) kidneys had 1.18-times the hazard of graft loss compared to those who received HCV(−) kidneys (95% CI: 1.04–1.32, p = 0.007) (Table 4B). As with patient survival, this hazard ratio translated to minimal differences in actual graft survival, with no difference at 1-year survival (91% for both HCV(−) and HCV(+) kidneys) and only 3% difference at 3 years (80% vs. 77%). In stratified models, patients under 60, patients without diabetes, patients with PRA <80, and patients with BMI <35 all had an increased hazard of graft loss associated with receipt of an HCV(+) kidney (Table 4B, left column). While the hazard of graft loss also appeared to be increased for patients over 60, diabetics, patients with PRA >80, and patients with BMI >35, these increases were not statistically significant (Table 4B, right column).


Since 1995, approximately half of HCV(+) kidneys have been discarded, while 71% of HCV(+) recipients have waited on average a year longer to receive an HCV(−) kidney. The biggest center-level predictor of HCV(+) kidney utilization was waiting time; centers with longer waiting times were significantly more likely to transplant HCV(+) kidneys into HCV(+) recipients. The risk of having received an HCV(+) kidney translated to a 1% lower survival at 1 year and a 2% lower survival at 3 years, while the benefits to patients were potentially significant, as HCV(+) recipients who were transplanted with HCV(+) kidneys spent over a year less time on the waiting list than those HCV(+) recipients who waited for HCV(−) kidneys at the same transplant center.

The important clinical question is whether HCV(+) recipients derive a survival benefit from receiving an HCV(+) kidney as opposed to waiting longer for an HCV(−) kidney. Unfortunately, this clinical question has never been successfully answered, let alone in a nationally representative cohort. While several single-center studies have demonstrated that HCV(+) patients derive a survival benefit from receiving a kidney transplant (from any donor) as opposed to remaining on dialysis, no study has successfully examined whether this effect is modified by the HCV status of the donor kidney (14–16). Unfortunately, similar to previous studies on this topic (19,22), we were unable to answer this directly because HCV status is only captured at the time of transplant, not at the time of listing. We were, however, able to quantify the ‘risk’ side of the risk/benefit equation directly, showing that transplantation with an HCV(+) kidney resulted in 1% lower survival at 1 year and 2% lower survival at 3 years, and the ‘benefit’ side indirectly, showing that waiting for an HCV(−) kidney resulted in an extra year of dialysis. Considering that the death rate for dialysis patients on the waitlist is 7.5 per 100 person-years on average (25), and that the rate is increased by 25% among dialysis patients with HCV (1), opting for the currently available HCV(+) kidney rather than waiting for the next available HCV(−) kidney might be justified for the right patients. And with over 1000 HCV(+) kidneys from non-DCD donors under 45 with creatinines < 2.0 discarded during our study period, it seems that lack of organ availability did not likely drive the choice for an HCV(+) patient to wait for an HCV(−) kidney.

We found wide variation in utilization of HCV(+) kidneys across centers, with 35% of centers never transplanting an HCV(+) organ into their HCV(+) recipients over the entire 13-year study period. While this was partially explained by waiting time, there were centers with the longest 10% of waiting times that indeed transplanted HCV(+) recipients but never with HCV(+) kidneys. Furthermore, the discard rate of HCV(+) kidneys was two to six times higher than for HCV(−) kidneys, and four to seven times higher when restricting the comparison to ‘ideal’ donors. These findings suggest that HCV(+) kidneys are underutilized nationally, and that increasing utilization might (a) provide significant benefit for HCV(+) patients by decreasing waiting times and (b) expand the organ supply for all patients by increasing overall organ utilization.

Our study corroborates previous findings that, even for recipients who are already HCV(+), receipt of an HCV(+) kidney is associated with a small increase in hazard of death and graft loss compared to receipt of an HCV(−) kidney (5,20). However, we also found that for certain subgroups such as older patients, diabetics or those with high PRA, the increased hazard is not observed, suggesting that the donor HCV status might not be an issue at all for these patients. Previous studies have shown an increased risk of posttransplant diabetes in HCV(+) recipients (9,11,12), which may contribute to the increased risk of death and graft loss, and might also explain why this risk appears to be attenuated for those who had diabetes prior to transplant. Nevertheless, there may be multiple reasons that HCV(+) recipients choose to wait for HCV(−) kidneys, including patient preference, previous HCV treatment with good response, concerns about genotype coinfection, and other concerns of increased harm. CDC high-risk donor behavioral factors might also play a role in clinical decision making, although in our analysis, these factors did not attenuate the independent effect of HCV status on organ discard.

We speculate on another potential factor involved in the discard of HCV(+) kidneys, that of regulatory disincentive. Indeed, current risk-adjustment models used by the SRTR and CMS to evaluate center-specific outcomes do not account for donor HCV status (26), and the fear of potential legal and regulatory consequences of using an organ for which the risk would not be properly adjusted has been shown to influence practice patterns (27). The underutilization of HCV(+) kidneys might be explained by provider fears of regulatory consequences from using these organs without risks properly adjusted in these models. As such, adding donor HCV status to these models might attenuate these fears and increase national utilization of these organs. We acknowledge that this reasoning is purely speculative and not examined by our study.

AA HCV(+) recipients had high rates of receipt of HCV(+) organs compared to other races. While this was partially explained by center-level variation in utilization, we found these organs were disproportionately utilized in AAs even within many centers. More studies are needed to determine whether these decisions are occurring at the level of the provider (providers are more likely to offer HCV(+) kidneys to AA recipients), or the patient (AA recipients are more likely to accept HCV(+) kidney offers). Interestingly, this practice pattern might be reasonable in the current environment, as HCV(+) AAs transplanted with HCV(+) kidneys did not have any increased risk of death compared to HCV(+) AAs transplanted with HCV(−) kidneys, while HCV(+) kidney receipt was associated with a significant increase in mortality for patients of other races. That said, the current disparity in waiting times for AAs might be playing a role in this effect modification, and correcting these disparities might change these inferences.

Our study had the following limitations. First, national data do not exist on the HCV status of candidates on the waiting list; we could only identify HCV(+) patients among those who actually received kidneys. As such, we were unable to directly study survival benefit derived from receipt of HCV(+) kidneys compared to waiting on the list, although we provide for comparison death rates on dialysis versus increases in posttransplant death rates attributable to the HCV(+) status of the transplanted kidney. Second, HCV-RNA levels are not captured in UNOS; as such all our analyses were based on antibody status only. On the donor side, it is possible that the increase in death and graft loss may be more significant in recipients of kidneys from donors who are HCV viremic. On the recipient side, a recent review recommended restricting the use of HCV(+) kidneys to recipients with active HCV viremia (28). Further studies are needed to better understand the relationship between HCV viremia and outcomes. Combined pegylated-interferon alpha/ribavirin therapy has shown some promise in achieving sustained virologic response in some HCV(+) ESRD patients (29) but tolerance to these regimens has been limited in some studies (30). Third, we also did not have data on HCV genotype, so we were unable to examine whether genotype mismatch between donor and recipient contributed to the increased hazards of death in patients transplanted with HCV(+) kidneys. A study by the New England Organ Bank showed no increased risk of death or graft loss associated with HCV genotype mismatch in transplant (31), but this has not been replicated using national data.

Our study suggests that HCV(+) kidneys are underutilized, and that tremendous variation exists in national practice patterns independent of measurable center-level characteristics. In this article, we have quantified the risks associated with transplanting HCV(+) kidneys into various subgroups of HCV(+) patients, so that these risks can be incorporated into the risk/benefit decisions made when an organ offer is considered for a given patient. We also encourage consideration of the increased risk associated with HCV(+) kidneys for risk-prediction models used to determine center-specific outcomes, as the lack of adjustment for donor HCV status might create a disincentive to the use of these organs and might contribute to the high discard rates.


As a study of the United Network for Organ Sharing database, this work was supported in part by Health Resources and Services Administration contract 234–2005-370011C. The content is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services.