Objectives: Liver allograft shortage has necessitated greater use of donations after circulatory death. Limited data are available to compare recipients’ health care utilization for donation after circulatory death versus brain death.
Materials and Methods: Liver transplant data for our center from November 2016 until May 2019 were obtained (208 donations after brain death and 39 after circulatory death). We excluded patients <18 years old and multiorgan transplants; for cost data only, we also excluded retransplants. Primary outcome was recipients’ health care utilization in donation after circulatory death versus brain death and included index admission length of stay, readmissions, and charges from transplant to 6 months. Secondary outcomes were patient and graft survival.
Results: Donors from circulatory death were younger than donors from brain death (median age 32 vs 40 years; P < .01). Recipient body mass index (31.23 vs 29.38 kg/m2), Model for End-Stage Liver Disease score (17 vs 19), portal vein thrombosis (15.8% vs 18.0%), length of stay (7 vs 8 days), and 30-, 90-, and 180-day posttransplant index admissions were not significantly different. Charges for index admission were equivalent for donation after circulatory death ($370771) and brain death ($374272) (P = .01). Charges for readmissions at 30 and 180 days were not significantly different (P = .80 and P = .19, respectively). Rates for graft failure (10.3% vs 4.8%; P = .08) and recipient death (10.3% vs 3.8%; P = .17) at 6 months posttransplant were similar.
Conclusions: Donation after circulatory death versus brain death liver transplant recipients had similar lengths of stay and equivalent index admission charges. Graft and patient survival and charges from transplant to 6 months were similar. Donation after circulatory death liver allografts provide a safe, cost-equivalent donor pool expansion after careful donor-recipient selection.
Key words : Biliary complications, Charges, Liver transplant outcomes, Readmission
The number of patients on the wait list for liver transplant (LT) at any point in time significantly outweighs the number of donor organs available for transplant, which leads to high rates of mortality among patients awaiting transplants.1 One of various ways to increase the donor pool is donation after circulatory death (DCD), which has been in practice for many years. The first attempted LT with a DCD organ was performed in 1963.2 However, the legal acceptance of brain death in the United States led to a surge in organs donated after brain death (DBD), which was favored over DCD because of the ability to continually oxygenate organs throughout procurement.3 The number of available DBD organs was far lower than the number of patients awaiting transplant, which necessitated the persistent use of DCD allografts; however, DCD livers have widely been considered marginal compared with DBD. Many early studies reported significantly higher primary nonfunction and graft failure from ischemic cholangiopathy and consequently higher rates of retransplant and inferior patient survival in recipients of DCD allografts.4,5
Over the years, transplant centers have continued to increase their use of DCD livers,6 supported by recent data that have shown similar rates of graft and patient survival between DCD and DBD recipients along with fewer postoperative and biliary complications than initially reported.7,8 Findings by Taylor and colleagues, from a contemporary United Kingdom cohort, refuted earlier publications and concluded that overall patient survival was improved by DCD LT compared with waiting for a DBD graft, despite individual recipient characteristics.9
Studies are emerging on the costs of health care utilization related to DCD versus DBD, including hospital stay and readmission costs. Early studies showed higher cost associated with DCD compared with DBD LT, caused by a longer durations of stay and more frequent and severe postoperative complications that required intervention.10 Another study noted ischemic cholangiopathy, a complication predominantly encountered with DCD transplant, and retransplant as the main causes of increased charges in DCD recipients.11 These studies emphasized the importance of risk factor mitigation for biliary complications, including shorter warm ischemia times (WIT) during organ recovery.
Despite the higher frequency of use of DCD livers, this mode of LT only accounted for 6.9% of deceased donor transplants in the United States in 2018. Although the discard rate for DBD livers has continued to decline over the past decade, the rate of discard of DCD livers has remained greater than 25%,12 which shows that many centers remain hesitant to use these organs. Early data showed poor patient outcomes, and there are limited data to compare the cost of care of DCD recipients versus DBD recipients.7 Because of the lack of data on health care utilization of DCD LT, we conducted this study. We compared DCD LT with DBD LT for various outcomes at The Ohio State University Comprehensive Transplant Center to understand the roles of the 2 donor types in health care utilization and the potential to enlarge the donor pool.
Materials and Methods
This retrospective study included patients who underwent LT at our center from November 2016 to May 2019. This study period was selected because DCD LT started at our center in 2016. During the study period, a total of 247 LT were performed. Of these, 39 patients received livers from DCD donors, and 208 from DBD donors. We excluded patients <18 years and living donor liver donation and multiorgan transplants; for charge data only, we also excluded patients who received retransplants. Demographics and clinical, intraoperative, and hospital variables of donors and recipients were collected.
Multiple independent and dependent variables were collected for this study. Independent variables collected from electronic health records were recipient and donor demographics, intraoperative and postoperative information, and various interventional, pharmaceutical, and laboratory test data of DCD and DBD recipients. Dependent variables included length of stay (LOS) during index admission, readmissions at 30, 90, and 180 days, and patient care charges (the amount billed for treatment) incurred during the index admission and at postoperative day (POD) 180. We also collected information on post-LT complications and patient and graft survival between the 2 study groups.
This study was approved by our institutional review board (IRB No. 2019H0190), and all protocols conformed to the ethical guidelines of the 1975 Helsinki Declaration. Donor and recipient data were deidentified for data analyses.
Donation after circulatory death liver procurement
The DCD livers in our study were obtained from controlled DCD (Maastricht type III) by the rapid technique for organ recovery and intravenous administration of heparin (30000 U). With a cruciate abdominal incision, we obtained rapid access and cannulated the aorta above the bifurcation, followed by cross-clamp of the thoracic aorta. The abdominal aorta was then infused with cold, pressurized (200 mm Hg) histidine-tryptophan-ketoglutarate (HTK) solution, which was vented through the inferior vena cava below the right atrium. After procurement, the liver was again flushed with HTK solution on the back table through the portal vein before package preparation and transport. For DCD liver recovery, our center limits the functional WIT to <30 minutes (starting from the agonal phase, ie, systolic blood pressure <80 mm Hg, or arterial oxygen saturation <80% until the start of the cold ischemia time [CIT]), which included the 2- to 5-minute mandatory wait time implemented by the Organ Procurement Organization. This protocol assures that no autoresuscitation has occurred, in accordance with recommendations from the Institute of Medicine.3 Although our center permits CIT up to 10 hours, efforts are made to select uncomplicated recipients for DCD liver allografts and expedite the start of surgery to shorten the CIT. The DCD allografts were not offered to recipients with known complete portal venous thrombosis, extensive abdominal surgeries, and requirement for retransplant.
Liver transplant technique in donation after circulatory death and donation after brain death
The implantation technique did not vary between DCD and DBD recipients and involved side-to-side cavacavastomy and standard reconstruction of the portal vein, followed by a blood flush. Hepatic artery reconstruction was preceded by an infusion of tissue plasminogen activator (2 mg) in the donor hepatic artery for DCD livers. The biliary tract was reconstructed in a standard fashion with a duct-to-duct or Roux-en-Y anastomosis if duct-to-duct anastomosis was not feasible. Aminocaproic acid was used liberally, particularly in DCD donors, to limit fibrinolysis and reduce blood product utilization.
Postoperative immunosuppression was the same in both DCD and DBD recipients and included standard dosage and levels of tacrolimus and mycophenolate mofetil. Corticosteroids, administered initially in intravenous form and then orally, were tapered over 2 weeks. Tacrolimus was delayed and initially substituted with basiliximab in patients with significant acute kidney injury and/or delay in recovery, from a neurological standpoint, after transplant.
Liver transplant outcomes
The primary outcome of this study was health care utilization including (1) LOS during the index admission, (2) readmissions at 30, 90, and 180 days, and (3) patient care charges starting from index admission for LT to 180 days. Secondary outcomes were post-LT biliary interventions and patient and graft survival between the 2 study groups. Liver graft function was assessed clinically and with daily liver function tests including serum lactate, serum glucose, total bilirubin, alkaline phosphatase, liver enzymes, international normalized ratio, and fibrinogen after LT. Doppler ultrasonography of the liver was done routinely within 6 hours of transplant and on POD 1, 3, and 5. Graft failure was defined as listing for retransplant or patient death.
Lengths of intensive care unit (ICU) stay and total stay after transplant, post-LT care charges during the 6 months after transplant, and rates of graft failure and patient survival were compared between DCD and DBD recipients with the Mann-Whitney U test for distributional differences and the Fisher exact test. The Mann-Whitney U test for equivalence was performed to assess DBD and DCD charges during the index admission and from discharge to 6 months posttransplant. A tolerance limit of 0.15 was used to determine equivalence, on balance with strict (0.10) and liberal (0.20) standard recommendations for Mann-Whitney U test for equivalence.14,15 A separate Poisson analysis of variance model was fitted to each of the readmission count time ranges (number of readmissions from POD 1-30, POD 1-90, and POD 1-180) with an effect for donor criteria (DCD or DBD) to test for differences in counts between DCD and DBD. Time to graft failure and patient survival were compared with Kaplan-Meier curves and log-rank tests.
Donor and recipient characteristics
Donor and recipient characteristics are shown in Table 1 and Table 2. During the study period, 247 LT procedures were performed, including 39 from DCD and 208 DBD donors. The DCD donors were younger than DBD donors (32 vs 40 years; P < .01). There was no difference in the body mass index (BMI, calculated as weight in kilograms divided by height in meters squared) between the 2 donor groups(27 vs 27; P = .94). Median WIT in DCD donors was 19 minutes. Of total donors, 40% were positive for hepatitis C virus (HCV), ie, positive for either anti-HCV antibody or the nucleic acid test.
There were no statistical differences in the median age (57 vs 57 years; P = .80), BMI (31.23 vs 29.38; P = .46), CIT (236 vs 270 min; P = .24), or secondary WIT (27 vs 27 min; P = .36) between DCD and DBD recipients. The median Model for End-Stage Liver Disease (MELD) score for DCD recipients was slightly lower than for DBD recipients (17 vs 19; P = .09), although not to a significant degree.
Intraoperative parameters included recipient operative times, cell saver return, and use of blood products (cryoprecipitate, platelets, and plasma) and were not significantly different between the DCD group and the DBD group (Table 2).
Postoperative complications and interventions
There was no statistically significant difference between the rate of postoperative biliary interventions between DCD and DBD groups (15.4% vs 25%; P = .22).
Primary outcomes (health care utilization)
Median recipient ICU stay (39 vs 38 hours; P = .74) and posttransplant LOS (7 vs 8 days; P = .23) were not significantly different between DCD and DBD recipients. Alanine aminotransferase (850 vs 447 U/L; P < .01) and aspartate aminotransferase (840 vs 637 U/L; P < .01) levels indicated reperfusion injury and were significantly higher in DCD recipients at 72 hours posttransplant compared with DBD recipients (Table 2).
Graft and patient survival
Rates of graft and patient survival at 6 months were similar in DCD and DBD recipients. Rates of graft loss at 1, 3, and 6 months were 0%, 7.7%, and 10.3% in DCD recipients compared with 1.9%, 2.9%, and 4.8% in DBD recipients, respectively (P = .08). Patient mortality rates at 1, 3, and 6 months were 0%, 7.7%, and 10.3% among DCD recipients and 1%, 1.9%, and 3.8% among DBD recipients, respectively (P = .17) (Figure 1 and Figure 2).
Readmissions and charges
There were no significant differences in readmission counts from POD 1 to 30, from POD 1 to 90, or from POD 1 to 180 between the DBD and DCD groups (Figure 3). The charges of the index admission ($370771.56 vs $374272.42; P = .75), from discharge until 30 days ($2133.00 vs $4145.60; P = .80), and from 31 days after discharge until 180 days ($6398.80 vs $15663.61; P = .19) were lower in DCD recipients compared with DBD recipients, but not statistically significant. Further analysis of charges with a Mann-Whitney U equivalence test showed that charges for the index admission were equivalent (P = .01); however, the postdischarge charges were not equivalent between DCD and DBD recipients (P = .06 and P = .08, respectively) (Table 3).
Our center began the use of livers from DCD donors in 2016. During the study period, our DCD LT volume has increased from none to 15.8% (39/247). In this study, we have found that use of DCD liver allografts caused no additional burden on health care utilization (LOS, readmission, and cost of care) compared with DBD allografts. Furthermore, graft and patient survival rates were found to be similar between the 2 groups. Based on these findings, we are expecting a further increase in the utilization of DCD liver allografts to address the expansion of the donor pool.
Donation after circulatory death, by virtue of its organ recovery protocol, has greater risk of warm ischemic injury compared with donation after brain death. Therefore, it is not surprising that liver enzymes peak higher in DCD liver recipients than in DBD liver recipients after transplant.7,8,16 This unavoidable ischemic insult must be considered for appropriate selection of recipient-donor combinations. The donor liver should be sufficiently healthy to tolerate the additional ischemia, and the recipient should be able to tolerate any ischemia-reperfusion injury and/or graft dysfunction that may result from additional ischemia.
The recent improvement of DCD liver recipient outcomes may be attributed to various practices implemented since DCD transplant began, foremost of which is rigorous selection of donors and recipients. Donor variables associated with good outcomes in DCD are younger age, low BMI, and short WIT and CIT. Donation after circulatory death livers from donors of advanced age pose a higher risk for postoperative complications.6,17,18 However, a study by Ramirez and colleagues19 showed similar mortality and biliary and vascular complications from transplant of DCD livers from donors >70 years old compared with DBD livers of the same age. A debate persists regarding the role of the DCD donor age,20 although it is generally accepted that younger DCD livers21 have improved tolerance to ischemia.
Organ procurement teams have attempted to reduce WIT to less than 30 minutes in most cases.10 A WIT of 35 minutes is associated with significantly higher graft failure rates (hazard ratio 1.84; P = .002).21 A procurement strategy employed by all centers is to reduce CIT, as each additional hour is associated with a 6% higher rate of graft failure (hazard ratio, 1.06; P < .001),21 and prolonged ischemia time leads to biliary strictures.22
Higher donor BMI has also been shown to be associated with adverse outcomes.20,21 This may be associated with concomitant hepatic steatosis, which additionally increases ischemia-reperfusion injury and biliary complications.23 Some studies have attempted to define recipient risk factors associated with reduced graft survival and concluded that DCD liver recipient age >60 years old, MELD score >25, and hepatocellular carcinoma may all contribute.7,24 Care is also taken to match DCD donors to recipients without previous extensive upper abdominal surgery, known complete portal vein thrombosis, or previous LT, as these factors may increase surgery time or complexity, thereby delaying implantation of the organ.16 Hence, a preferred recipient for a DCD liver is someone without multiple previous abdominal surgeries and for whom no additional intraoperative procedure is anticipated, to allow for shorter surgical time and more immediate reperfusion of the liver.
Although these are a valuable addition to the already strained donor pool, the utility of DCD livers has been limited by higher rates of postoperative complications, especially biliary complications, compared with DBD livers. A recent comparison of DCD and DBD LTs from Spain25 has shown that the odds ratio for developing ischemic cholangiopathy is 10 times higher in DCD donors. Similar findings were reported by other authors26-28 for biliary complications in DCD recipients. Interestingly, in our cohort, more DBD liver recipients required biliary intervention than DCD recipients in the first 6 months posttransplant, although the difference was not statistically different. This lower intervention rate might be explained by our careful selection of DCD donor and recipient pairs, quick procurement by experienced transplant attending physicians and fellows, pressurized flush with HTK solution before removal of the liver, strict adherence to the 30-min WIT limit, efforts to reduce the CIT, and the use of tissue plasminogen activator infusion in the hepatic artery after portal perfusion of the liver, which has been shown to reduce the incidence of biliary strictures.29
Our results did not show a significant difference in patient or graft survival between DCD and DBD recipients. These survival results are consistent with previous studies that have shown survival data for up to 5 years.10,25,26,30 A few studies have shown similar survival data for up to 10 years.31,32 Livers from donors who tested positive for HCV by nucleic acid test were used with equal frequency in both groups in our study. These recipients at our center are treated with antiviral therapy only after their polymerase chain reaction blood tests become positive. Recipients who receive HCV-positive organs require close follow-up postoperatively as they are at higher risk for rejection33 and antiviral drug interaction with immunosuppression34 and also at risk for accelerated fibrosis.35,36 Recent meta-analyses have shown no difference in the HCV recurrence in recipients in both the groups.26
Readmission after LT is a major cause of concern for transplant centers because of financial implications, consumption of additional hospital resources, and negative effects on patient survival.37 In an analysis of 11937 LT recipients, 90-day readmissions accounted for $43785 of added charges compared with patients who were not readmitted.38 However, Russo and colleagues39 were able to decrease the 30-day readmission rate from 40% to 20% (P = ?.02) by earlier and more extensive patient education, more frequent follow-up appointments, and expansion of outpatient services to include urgent same-day visits and infusion services. Our readmission rate for this study was similar between both the DCD and DBD groups, and similar readmission rates have been cited by other authors.37,39
Earlier studies on transplant and postoperative charges have shown higher costs for DCD versus DBD recipients. However, the costs for DCD transplants at our center were equivalent to those for DBD transplants during the index admission. Interestingly, charges assessed after discharge from the index admission trended lower for DCD recipients than for DBD recipients. Similarly, Jay and colleagues18 did not find any difference in the charges between DCD and DBD LT from index admission to discharge (P = .53), although they reported higher mean 1-year posttransplant charges for DCD recipients in both the adjusted and unadjusted models. They attributed these higher posttransplant charges to costs associated with ischemic cholangiopathy and retransplant, which were not observed in our study. The European perspective was reported by van der Hilst and colleagues,17 with higher 1-year charges for DCD transplants; they cited reinterventions as the main drivers of cost. In a Markov model analysis for different MELD quintiles, the costs for DCD LT were more costly than for DBD transplant.8 Singhal and colleagues40 also showed higher charges with DCD LTs. In our cohort, there were no significant differences in the readmission rate or postoperative interventions between the 2 groups, and this could explain the similar charges in our study. Additional follow-up may be needed to account for typical biliary complications encountered with DCD, which did not manifest in our study.
The findings of our study are limited by the retrospective design and follow-up, which was limited to 6 months. Although extended follow-up may increase the incidence of ischemic cholangiopathy, the purpose of this study was to define early outcomes of DCD transplant, and there was not a difference in biliary complications between DCD and DBD recipients in our study. Our small number of DCD transplants may limit the strength of our conclusions; however, DCD transplants constitute almost 16% of the transplants at our center, which is comparable to other centers. Additionally, retransplants that occurred during our study were not included because we were unable to accurately separate charges from the first and second transplants.
Our data showed that equivalent postoperative outcomes are possible with DCD LT as shown by the DCD recipients in our cohort who had outcomes and charges similar to those of DBD recipients. The DCD liver allografts should be considered comparable to DBD allografts when used in a systematic fashion. Our contemporary cohorts provide evidence that DCD liver allografts are a safe, cost-equivalent alternative to DBD transplant that expands the donor pool. Donor-recipient pairs must be carefully selected to achieve optimum outcomes from DCD LT. The charges involved in DCD transplant may be lowered with careful planning in the postoperative period, more frequent follow-up appointments, and greater expansion of outpatient services.
Volume : 19
Issue : 8
Pages : 771 - 778
DOI : 10.6002/ect.2021.0013
From the 1Division of Transplantation, Department of Surgery, The Ohio State University Wexner Medical Center; the 2Medical Student Research Program, College of Medicine, The Ohio State University; and the 3Division of Hepatology, Department of Internal Medicine, The Ohio State University Wexner Medical Center, Columbus, Ohio, USA
Acknowledgements: This work was funded in part by The Ohio State University College of Medicine Bennett research scholarship (KH). Other than described, the authors have not received any funding or grants in support of the presented research or for the preparation of this work and have no declarations of potential conflicts of interest.
Corresponding author:Navdeep Singh, 395 W. 12th Ave. Columbus, OH, USA
Table 1. Summary of Donor Demographic Variables
Table 2. Summary of Recipient Demographic and Intraoperative and Postoperative Variables
Figure 1. Kaplan-Meier Graft Survival Curve by Postoperative Day for Donor Type
Figure 2. Kaplan-Meier Patient Survival Curve by Postoperative Day for Donor
Table 3. Charge/Cost Data with Mann-Whitney U Probability Tests, Significant Differences, and Equivalence
Figure 3. Readmission Counts by Postoperative Day