Author + information
- Received November 23, 2011
- Revision received February 1, 2012
- Accepted February 21, 2012
- Published online July 3, 2012.
- Todd Dardas, MD, MS⁎,⁎ (, )
- Nahush A. Mokadam, MD†,
- Francis Pagani, MD, PhD‡,
- Keith Aaronson, MD, MS§ and
- Wayne C. Levy, MD⁎
- ↵⁎Reprint requests and correspondence:
Dr. Todd Dardas, Division of Cardiology, University of Washington, 1959 N.E. Pacific Street, Box 356422, Suite AA522, Seattle, Washington 98195-6422
Objectives The goal of this research was to identify disparities in risk within heart transplant urgency designations.
Background Patients with left ventricular assist devices (LVADs) are given 30 days of elective status 1A time. This allowance may create competition for organs between stable LVAD-supported registrants and less stable registrants listed status 1A or 1B.
Methods The Scientific Registry of Transplant Recipients database was analyzed for all status 1A and 1B listings between 2005 and 2010. Cox models were used to estimate the relative and absolute risk of adverse events (death or delisting as too ill) during status 1A or 1B listing.
Results Status 1A registrants supported with dual inotropes and right heart monitoring had a higher risk of adverse events compared to those supported with implanted LVADs using elective 1A time (hazard ratio: 3.2; 95% confidence interval: 1.8 to 5.7). The 30-day risk of events was 1% (95% confidence interval: 0.1% to 3%) for implanted LVADs using elective 1A time and 6% (95% confidence interval: 4% to 8%) for dual inotrope support. Registrants listed with paracorporeal ventricular assist devices had a higher risk of adverse events (hazard ratio: 9.1; p < 0.0001) compared with registrants with implanted LVADs using elective 1A time. The odds of transplant were higher for implanted LVADs (odds ratio: 1.5; p < 0.0001) compared with dual-inotrope and intra-aortic balloon pump support.
Conclusions The historic allowance for 30 days of elective status 1A time for implanted LVADs creates disparities in risk among status 1A registrants. The allowance of 30 days of elective status 1A time should not be allocated to stable registrants with implanted LVADs. Registrants supported with paracorporeal ventricular assist devices should be listed status 1A indefinitely.
Early in the course of ventricular assist device (VAD) development, 30 days of Organ Procurement and Transplantation Network (OPTN) status 1A time was given to patients immediately after VAD placement to allow for recovery from VAD surgery before transplant listing. This policy was later changed in 2002 to allow 30 days of status 1A time to be used at the transplant center's discretion any time after VAD placement. A subsequent revision of OPTN policy no longer required hospitalization for VAD-supported registrants to use elective 1A time (1). Patients without a VAD complication or those who have used their 30 days of elective 1A time are listed status 1B.
Survival has steadily improved for patients supported with implanted left ventricular assist devices (LVADs) as experience with LVAD placement and long-term management increases and devices become more reliable (2,3). The 30-day allowance still exists but may allow competition between patients with very different risks of death. Stable registrants with implanted LVADs are likely to have a much lower risk of death than other patients listed status 1A (4). Because LVADs are used more commonly, there is a growing belief that placement of an LVAD is necessary for successful transplantation. Increasing numbers of stable, VAD-supported patients listed status 1A may prevent organ allocation to sicker patients, decreasing survival among those waiting for transplant and leading to an overall reduction in the efficacy of the transplant urgency system.
We assessed the equality of risk among registrants listed status 1A and 1B who were actively listed for transplant and demonstrate significant disparities in risk within and between OPTN status designations.
This study used data from the Scientific Registry of Transplant Recipients (SRTR). The SRTR data system includes data on all donor, wait-listed candidates, and transplant recipients in the United States submitted by a member of the OPTN; this system has been described elsewhere. The Health Resources and Services Administration, U.S. Department of Health and Human Services, provides oversight to the activities of the OPTN and SRTR contractors. Subjects with complete listing dates, listing status justification, initial listing and list removal dates, VAD type, VAD implantation dates, and, for transplant recipients, transplant date and date of death or graft failure were included. The current analysis focused on January 2005 to November 2010 to obtain accurate information on modern implanted and paracorporeal VADs.
For status 1A registrants, status justifications were recorded by OPTN as either VAD support using elective time, VAD support with a complication, ventilator dependence, inotropic support and/or intra-aortic balloon pump support (referred to hereafter as “medical support”), and exception status. We further separated registrants justified as “VAD support using elective time” into those with an implanted LVAD and those with a paracorporeal device to create 6 subgroups for analysis. Those with an implanted LVAD and simultaneous paracorporeal right-sided VAD were assigned to the paracorporeal support group. Total artificial hearts were excluded from the analysis. For status 1B registrants, status justifications included low-dose single inotrope, implanted LVADs, paracorporeal devices that were defined similarly to status 1A registrants, and all others not meeting those criteria, which include those listed status 1B with exception.
Demographic characteristics are reported according to transplant urgency status and status justification code present at the time of initial listing. Trends over time in status 1A and 1B listing justifications were analyzed by year using the Poisson regression or Cochran-Armitage test for trend. Time to listing after VAD placement surgery was calculated from the VAD placement date until the next status 1A or 1B listing. Median listing refers to the time from initial listing until removal from that listing status for any reason (i.e., transplant, death).
We assessed the frequency of status 1A listing of stable, implanted LVAD registrants simultaneous to other types of status 1A candidates within transplanting centers and within organ procurement organizations (OPO). Listing was only considered to be simultaneous when a registrant with an implanted LVAD using status 1A elective time was listed at the same time as another registrant listed status 1A with the same blood type.
The composite endpoint was defined as death while on the waiting list or delisting as too ill for transplant. Two separate types of Cox hazard model analysis were performed to measure risk on the basis of the following: 1) active status designation and status justification (in-status analysis), which only counts a registrant at-risk when listed in a specific status and justification (i.e., time at-risk is only measured while listed status 1A and justified by a VAD with a complication); and 2) risk after first status designation and justification (first-status analysis), which includes all time after initial listing regardless of subsequent status or justification after initial listing (i.e., time at-risk begins with the first listing date, status, and justification and ends at the time of an event or censoring regardless of subsequent status or justification changes). The in-status analysis is useful to compare risk among registrants in the same status, whereas the first-status analysis allows assessment of risk after the initial presentation and includes all time on the list regardless of events after the first listing. Models were stratified according to status justification to measure the cumulative hazard. Hazard ratios (HRs) were obtained using dummy variable coding and were adjusted for age, sex, ABO blood type, and heart failure etiology. The counting method was used to compile the time at-risk to adjust for discontinuous listing episodes and changes in status justification over time. Registrants were censored as free of the composite endpoint at the time of transplant or if the composite outcome had not occurred before November 10, 2010. We also measured the cumulative hazard and relative hazard of death 30 days before and after VAD implantation to separate acute VAD support from long-term support.
The odds of transplant were assessed with logistic regression models adjusted for age, etiology of heart failure, ABO blood type, sex, and status justification for status 1A and status 1B registrants.
Mortality after transplant was calculated with standard Cox models stratified by transplant status and status justification. Mortality was adjusted for age.
All analyses were conducted by using SAS version 9.2 (SAS Institute Inc., Cary, North Carolina).
The sample included 11,908 status 1A listings with 267 events for the in-status analysis and 7,064 individuals with 1,073 events for the first-status analysis (Table 1). There were 13,260 status 1B listings with 972 events for the in-status analysis and 8,368 individuals with 1,135 events for the first-status analysis. The type of VAD placed changed from pulsatile devices to continuous flow devices over the course of the study. The most common device placed in 2005 was the HeartMate XVE (Thoratec, Pleasanton, California) (>55%). By 2010, 75% of devices placed were the HeartMate II (Thoratec).
Adverse events while waiting for transplant
Compared with elective status 1A registrants listed with an implanted LVAD, registrants with a VAD and a complication had a risk of adverse events that was not significantly higher (HR: 1.8; 95% confidence interval [CI]: 0.90 to 3.5), whereas those with paracorporeal VADs (HR: 9.1; 95% CI: 5.2 to 16) and those with medical support (HR: 3.2; 95% CI: 1.8 to 5.7) had a significantly higher risk of events (Table 2, Fig. 1), as analyzed according to the in-status analysis. These associations did not change after adjusting for age, ABO blood type, sex, and etiology of heart failure (Table 2). The first-status analysis, allowing for subsequent status changes and periods of delisting or hold status, were not appreciably different with respect to the relative risk of adverse events between status 1A justifications, while tending to increase the rate of adverse events for all groups (Table 3, Fig. 2). The cumulative hazard within 30 days of status elective 1A listing for an implanted LVAD was 1.0% (95% CI: 0.05% to 3%), VAD with complication was 6% (95% CI: 3% to 14%), medical therapy was 6% (95% CI: 4% to 8%), and 1A paracorporeal VAD was 15% (95% CI: 12% to 20%) (Table 3). Biventricular support with either implanted or paracorporeal devices (HR: 2.1; 95% CI: 0.64 to 6.8) or right-sided VAD (HR: 2.3; 95% CI: 0.2 to 21) was associated with a higher risk of mortality that was not statistically different from implanted LVAD support, although insufficient numbers of patients had complete VAD support information to adequately assess the differences between implanted and paracorporeal support.
We assessed the relative hazard of adverse events within groups of patients supported with a VAD relative to the time from VAD surgery to their first status 1A designation. Adverse events occurred more commonly among those with paracorporeal devices when the registrant was listed ≤30 days after VAD implantation (HR: 8.6; 95% CI HR: 2.1 to 36) and was 4.9 times higher (95% CI HR: 1.7 to 14) for those with implanted devices.
Status 1B registrants with medical support (HR: 2.2; 95% CI: 1.7 to 2.7), paracorporeal VADs (HR: 4.2; 95% CI: 2.9 to 6.0), and those listed 1B with exception (HR: 2.5; 95% CI: 1.8 to 3.4) all had a higher risk of adverse events compared with status 1B registrants listed with implanted VADs according to the in-status analysis (Table 2, Fig. 3). These trends did not change after adjusting for age, ABO blood type, sex, and etiology of heart failure; they also did not change in the first-status analysis (Table 3, Fig. 4). The cumulative hazard at 30 days after initial listing for implanted VADs was 3% (95% CI: 2% to 4%), single inotrope support was 9% (95% CI: 8% to 10%), and status 1B paracorporeal support was 26% (95% CI: 18% to 39%) (Table 3).
For status 1B registrants with VADs, the risk of adverse events among those with an implanted VAD (HR: 1.6; 95% CI: 0.98 to 2.5) or a paracorporeal VAD (HR: 2.0; 95% CI: 3.7 to 1.02) was higher when the registrant was listed within 30 days of implantation compared with those listed for transplant 30 days after implantation.
Status 1A listing trends
Of all status 1A listings between 2005 and 2010, candidates listed status 1A with an implanted LVAD without complication increased from 11% to 26% (p < 0.0001), VADs with complications increased from 11% to 13% (p < 0.0001), paracorporeal devices decreased from 25% to 14% (p = 0.03), and medical support decreased from 44% to 39% (p < 0.0001) (Fig. 5).
The median length of a single listing episode for implanted LVADs without complication was 17 days (interquartile range [IQR]: 6 to 30 days), VADs with complication was 19 days (IQR: 9 to 42 days), paracorporeal VADs was 6 days (IQR: 2 to 17 days), and medical support was 6 days (IQR: 3 to 12 days).
Patients with implanted VADs using elective status 1A time simultaneously with patients in other status justifications occurred 105 times at the center level and 119 times at the OPO level between 2005 and 2010. There were 2 centers with ≥5 simultaneous listings and 33 OPOs with ≥10 simultaneous listings.
Status 1B listing trends
The sample included 10,279 status 1B registrants with 17,295 listing episodes. Of all listings from 2005 to 2010, support with a single low-dose inotrope decreased from 82% to 70% (p < 0.0001), implanted VADs increased from 7% to 21% (p < 0.0001), and paracorporeal VADs decreased from 4% to 2% (p < 0.0001).
The median length of a single listing episode for implanted LVADs was 41 days (IQR: 11 to 114 days), paracorporeal VADs was 17.5 days (IQR: 5 to 42 days), and low-dose inotrope was 14 days (IQR: 3 to 49 days).
Factors influencing likelihood of transplant
The likelihood of transplantation for a patient supported with an implanted LVAD using status 1A elective time increased over the course of observation from 11% to 27% of all status 1A transplants performed. The likelihood of transplantation decreased for those with medical support (38% to 35%) and for those with paracorporeal support (27% to 14%). These trends in allocation are consistent with the frequency of listing for each subgroup (Fig. 5).
Among status 1A registrants, implanted LVADs without complications (odds ratio [OR]: 1.5; 95% CI: 1.3 to 1.6), paracorporeal VADs (OR: 1.3; 95% CI: 1.2 to 1.5), and VADs with a complication (OR: 1.9; 95% CI: 1.7 to 2.2) were all more likely to undergo transplantation than medically supported registrants. Women were less likely to be transplanted than men (OR: 0.88; 95% CI: 0.81 to 0.97). Blood type A (OR: 1.6; 95% CI: 1.5 to 1.7), type B (OR: 1.5; 95% CI: 1.4 to 1.7), and type AB (OR: 2.5; 95% CI: 2.0 to 3.2) registrants had an advantage over type O registrants for receipt of transplant.
Among status 1B registrants, non-VAD supported registrants were less likely to undergo transplantation than those with implanted LVADs (OR: 0.52; 95% CI: 0.47 to 0.57) or those with paracorporeal VADs (OR: 0.79; 95% CI: 0.64 to 0.97). Women were less likely to be transplanted than men (OR: 0.82; 95% CI: 0.8 to 0.9). Blood type A (OR: 1.7; 95% CI: 1.5 to 1.8), type AB (OR: 2.4; 95% CI: 2.0 to 2.9), and type B (OR: 1.5; 95% CI: 1.4 to 1.6) registrants had an advantage relative to type O registrants for receipt of transplant.
Post-transplant mortality according to status justification
After status 1A transplantation, ventilator-dependent (HR: 2.6; 95% CI: 1.7 to 4.1) registrants had a significantly higher risk of death post-transplant compared with those with medical support, whereas those with an implanted LVAD (HR: 1.2; 95% CI: 0.9 to 1.3), those listed as exception (HR: 1.2; 95% CI: 0.9 to 1.6), paracorporeal VADs (HR: 1.3; 95% CI: 1.0 to 1.5), and VADs with complications (HR: 1.2; 95% CI: 1.0 to 1.4) all had similar outcomes after transplant (Table 4, Fig. 6).
After 1B transplant, those with paracorporeal VADs (HR: 1.3; 95% CI: 1.0 to 1.8), implanted LVADs (HR: 1.1; 95% CI: 0.9 to 1.3), and those listed with exception (HR: 1.2; 95% CI: 1.0 to 1.5) had no worse prognosis than those supported with low-dose inotropes before transplant (Table 4).
The OPTN heart transplant urgency system allows 30 days of elective status 1A time for LVAD patients. Historically, this option allowed patients to recover from LVAD placement before using status 1A time. On the basis of our analysis, allowing elective status 1A time for implanted LVAD patients creates a disparity in what should be equal risk among status 1A registrants.
The use of implanted LVADs as a bridge to transplant is a well-established strategy that allows patients to be stabilized in the setting of advanced heart failure and optimized for transplant surgery or long-term destination therapy (3,5–7). However, compared with registrants with dual-inotrope and intra-aortic balloon pump support, exception status, or VADs with complications, those using elective time with an implanted LVAD have a much lower risk of adverse events. Our study demonstrates that the stability achieved with implanted LVADs is much greater than can be achieved among other subgroups of patients listed status 1A. Implanted LVAD registrants have the advantage of waiting in status 1B for longer periods of time in a relatively stable clinical condition and the option of status 1A listing for 30 days. Implanted LVAD registrants can be escalated to status 1A for device complications, allosensitization, or other exception criteria. Their clinical stability and unique listing options result in a higher likelihood of transplant that is not proportional to their low risk of adverse events relative to other registrants listed status 1A.
Registrants with paracorporeal support have the highest risk of adverse events while awaiting transplant. Both patients with implanted LVADs and those with paracorporeal devices share the ability to receive 30 days of elective status 1A time, while clearly not sharing the same risk of adverse events. The absolute difference between paracorporeal VAD support and implanted LVADs was 14% at 30 days and 22% at 90 days. Registrants with paracorporeal devices who have used their 30 days of elective time remain high-risk when listed status 1B. The difference between implanted LVAD support and paracorporeal support, regardless of 1A or 1B status, is certainly enough to warrant a careful examination of OPTN policy. Registrants with paracorporeal device support may warrant indefinite 1A status.
Post-transplant survival after LVAD support is not superior to medical therapy or any other status 1A justification. As such, there is no compelling reason from the standpoint of organ utility to favor the transplant of VAD-supported registrants. Only registrants justified with ventilator support had markedly worse post-transplant outcomes. This group also had markedly higher risk of events while awaiting transplant.
There are significant differences in risk between status 1B registrants with implanted LVADs, paracorporeal VADs, and those supported with single, low-dose inotropes. The difference between implanted LVADs and medical support was 6% at 30 days and 11% at 90 days. The difference between inotrope-supported status 1B registrants and paracorporeal VADs listed status 1B was 15% at 30 days and 19% at 90 days, suggesting that patients supported with paracorporeal VADs who remain transplant candidates should remain listed status 1A indefinitely. The very low risk of adverse events among both status 1A and 1B subjects with implanted LVADs suggests that these registrants may warrant an urgency listing between status 1B and status 2.
Transplant listing and organ allocation
Registrants with an implanted LVAD have significantly higher rates of transplantation (OR: 1.5; p < 0.0001) compared with those with dual-inotrope and intra-aortic balloon pump support registrants, even after adjusting for factors commonly associated with the likelihood of organ allocation. We identified few instances in which a listing center created competition between those with an implanted LVAD using elective time and other groups of status 1A registrants. Few additional instances of apparent mismatch were seen when expanding the level of analysis to the OPO level. These findings suggest that transplant centers rarely create internal competition between stable LVAD-supported registrants and more high-risk registrants. Between-center competition is also rarely generated among groups of disparate risk, as measured by conflicts at the OPO level. However, there may be important exceptions within certain regions where the transplant center density is high and between regions where boundaries create competition for organ allocation.
Allowing continued elective use of status 1A time for implanted LVAD registrants, although unlikely to result in a meaningful number of major disparities, is associated with an inequity in organ provision. Registrants not supported with LVADs are at higher risk of adverse events. In the absence of changes to OPTN policy concerning LVADs, those supported with LVADs will continue to have an advantage in organ allocation that is not proportional to the risk of adverse outcomes while listed for transplant. As the number of registrants waiting with LVADs increases and LVAD survival improves (survival with the HeartMate II device is reported as 85% at 12 months) (8), the overall efficacy of the transplant system, defined as maximizing survival by provision of transplant according to risk while waiting and following transplant, may decrease (8,9).
Although the differences in risk we report point toward significant inequities on the transplant list, the timing of listing and reasons for assigned urgency must be carefully considered. We were not able to adequately measure the differential risks between left-sided paracorporeal support and right-sided or biventricular support due to insufficient data. From the limited sample of registrants in whom this information is available, left-sided paracorporeal support may be lower risk than either biventricular (regardless of whether one device is an implanted LVAD) or right-sided paracorporeal support, although no statistical difference was evident (p = 0.46). Competition within OPOs between LVAD-supported and other types of registrants is rare on the basis of our analysis, although we cannot account for body size restrictions or allosensitization patterns, which would only reduce the number of registrants appearing to compete for the same organ.
The historic allowance of 30 days of elective status 1A time for implanted LVADs without complication no longer promotes equality within the OPTN 1A urgency status. Registrants with paracorporeal devices remain at high risk for adverse events, and allowing indefinite status 1A time for these registrants while reducing the allowance for stable implanted LVADs would promote equalization of risk among status 1A registrants. The popular idea that LVADs are not a bridge but a necessary gateway to transplant should be treated with caution to avoid over-application of LVAD therapy and policies that create inequality on the transplant list.
The data reported here were supplied by the Minneapolis Medical Research Foundation as the contractor for the Scientific Registry of Transplant Recipients (SRTR). The interpretation and reporting of these data are the responsibility of the author(s) and in no way should be seen as an official policy of or interpretation by the SRTR or the U.S. government. This work was funded by the University of Washington House Officers Association. Dr. Dardas has reported that he has no relationships relevant to the contents of this paper to disclose. Dr. Pagani has an uncompensated consulting relationship with HeartWare and National Heart, Lung, and Blood Institute (NHLBI) funding for the REVIVE-IT pilot trial. Dr. Aaronson reports a consulting relationship with HeartWare and NHLBI funding for the REVIVE-IT pilot trial. Dr. Levy has received research grants from HeartWare, General Electric, NHLBI; has licensing with Epocrates; has received honoraria from GlaxoSmithKline, Boehringer-Ingelheim; is on the Steering Committee for Amgen and Scios and Clinical Endpoint Committee for CardioMems; has consulting/stock options with Cardiac Dimensions; and he is on the American College of Cardiology–Heart Failure Writings Guidelines Committee. Dr. Mokadam is a consultant for Thoratec, Inc.
- Abbreviations and Acronyms
- confidence interval
- hazard ratio
- interquartile range
- left ventricular assist device
- organ procurement organization
- Organ Procurement and Transplantation Network
- odds ratio
- Scientific Registry of Transplant Recipients
- ventricular assist device
- Received November 23, 2011.
- Revision received February 1, 2012.
- Accepted February 21, 2012.
- American College of Cardiology Foundation
- ↵Organ Procurement and Transplant Network: Policies. http://optn.transplant.hrsa.gov/policiesAndBylaws/policies.asp. Accessed October 15, 2011.
- Pagani F.D.,
- Miller L.W.,
- Russell S.D.,
- et al.,
- HeartMate II Investigators
- Starling R.C.,
- Naka Y.,
- Boyle A.J.,
- et al.