ZurichOpenRepositoryand
Archive
UniversityofZurich
UniversityLibrary
Strickhofstrasse39
CH-8057Zurich
www.zora.uzh.ch
Year:2018
DeningBenchmarksinLiverTransplantation:AMulticenterOutcome
AnalysisDeterminingBestAchievableResults
Muller,Xavier;Marcon,Francesca;Sapisochin,Gonzalo;Marquez,Max;Dondero,Federica;Rayar,
Michel;Doyle,MajellaMB;Callans,Lauren;Li,Jun;Clavien,Pierre-Alain;Puhan,Milo;etal
DOI:https://doi.org/10.1097/SLA.0000000000002477
PostedattheZurichOpenRepositoryandArchive,UniversityofZurich
ZORAURL:https://doi.org/10.5167/uzh-144568
JournalArticle
PublishedVersion
Originallypublishedat:
Muller, Xavier;Marcon, Francesca;Sapisochin, Gonzalo;Marquez, Max;Dondero, Federica;Rayar,
Michel;Doyle,MajellaMB;Callans,Lauren;Li,Jun;Clavien,Pierre-Alain;Puhan,Milo;etal(2018).
Dening Benchmarks in Liver Transplantation:A Multicenter Outcome AnalysisDetermining Best
AchievableResults.AnnalsofSurgery,267(3):419-425.
DOI:https://doi.org/10.1097/SLA.0000000000002477
Defining Benchmarks in Liver Transplantation
A Multicenter Outcome Analysis Determining Best Achievable Results
Xavier Muller, MD,
Francesca Marcon, MD,y Gonzalo Sapisochin, MD,z Max Marquez, MD,z
Federica Dondero, MD,§ Michel Rayar, MD,ô Majella M. B. Doyle, MD,jj Lauren Callans,
Jun Li, MD,yy Greg Nowak, MD,zz Marc-Antoine Allard, MD,§§ Ina Jochmans, MD, PhD,ôô
Kyle Jacskon, MD,jjjj Magali Chahdi Beltrame, MD,
Marjolein van Reeven, MD,yyy Samuele Iesari, MD,zzz
Alessandro Cucchetti, MD,§§§ Hemant Sharma, MD,ôôôjjjjjj Roxane D. Staiger, MD,
Dimitri A. Raptis, MD, PhD,
Henrik Petrowsky, MD,
Michelle de Oliveira, MD,
Roberto Hernandez-Alejandro, MD,ôôôjjjjjj Antonio D. Pinna, MD,§§§ Jan Lerut, MD, PhD,zzz
Wojciech G. Polak, MD, PhD,yyy Eduardo de Santiban
˜
es, MD,
Martı
´
n de Santiban
˜
es, MD,
Andrew M. Cameron, MD, PhD,jjjj Jacques Pirenne, MD, PhD,ôô Daniel Cherqui, MD,§§
Rene
´
A. Adam, MD, PhD,§§ Bo
¨
-Go
¨
ran Ericzon, MD, PhD,zz Bjoern Nashan, MD, PhD,yy
Kim Olthoff, MD,
Avi Shaked, MD,
William C. Chapman, MD,jj Karim Boudjema, MD,ô
Olivier Soubrane, MD, PhD,§ Catherine Paugam-Burtz, MD, PhD,§ Paul D. Greig, MD,z
David R. Grant, MD,z Amanda Carvalheiro, MD,y Paolo Muiesan, MD,y Philipp Dutkowski, MD,
Milo Puhan, MD, PhD, and Pierre-Alain Clavien, MD, PhD
Objective: To propose benchmark outcome values in liver transplantation,
serving as reference for assessing individual patients or any other patient
groups.
Background: Best achievable results in liver transplantation, that is, bench-
marks, are unknown. Consequently, outcome comparisons within or across
centers over time remain speculative.
Methods: Out of 7492 liver transplantation performed in 17 international
centers from 3 continents, we identified 2024 low risk adult cases with a
laboratory model for end-stage liver disease score 20 points, a balance of
risk score 9, and receiving a primary graft by donation after brain death. We
chose clinically relevant endpoints covering intra- and postoperative course,
with a focus on complications graded by severity including the complication
comprehensive index (CCI
1
). Respective benchmarks were derived from the
median value in each center, and the 75 percentile was considered the
benchmark cutoff.
Results: Benchmark cases represented 8% to 49% of cases per center. One-
year patient-survival was 91.6% with 3.5% retransplantations. Eighty-two
percent of patients developed at least 1 complication during 1-year follow-up.
Biliary complications occurred in one-fifth of the patients up to 6 months after
surgery. Benchmark cutoffs were 4 days for ICU stay, 18 days for hospital
stay, 59% for patients with severe complications ( Grade III) and 42.1
for 1-year CCI
1
. Comparisons with the next higher risk group (model for end
stage liver disease 21–30) disclosed an increase in morbidity but within
benchmark cutoffs for most, but not all indicators, while in patients receiving
a second graft from 1 center (n ¼ 50) outcome values were all outside of
benchmark values.
Conclusions: Despite excellent 1-year survival, morbidity in benchmark
cases remains high with half of patients developing severe complications
during 1-year follow-up. Benchmark cutoffs targeting morbidity parameters
offer a valid tool to assess higher risk groups.
Keywords: benchmark, complication, liver transplantation, morbidity,
outcome
(Ann Surg 2017;xx:xxx–xxx)
M
ore than 300.000 liver transplantations (LT) have been per-
formed worldwide, since its introduction in clinical practice in
the early 1980s.
1–3
With major improvements in surgical techniques,
immunosuppression therapy, and patient selection 1-year patient
survival rates have continuously improved to over 85%.
2,4– 6
LT is
From the
Department of Surgery and Transplantation, Swiss HPB Center Zurich,
University Hospital Zurich, Zurich, Switzerland; yDepartment of HBP Surgery
and Transplantation, Queen Elizabeth Hospital and Birmingham Children’s
Hospital, Birmingham, UK; zDepartment of Surgery, Abdominal Transplant
and HPB Surgical Oncology, Toronto General Hospital, University of Toronto,
Toronto, ON, Canada; §Department of Surgery and Transplantation, Ho
ˆ
pital
Beaujon, Clichy, France; ôDepartment of HBP Surgery and Transplantation,
University Hospital Rennes, Rennes, France; jjDivision of General Surgery
and Section of Abdominal Transplantation, Washington University School of
Medicine, St. Louis, MO;
Division of Transplantation, University of Penn-
sylvania, Philadelphia, PA; yyDepartment of Hepatobiliary Surgery and
Transplantation, University Medical Center Hamburg-Eppendorf, Hamburg,
Germany; zzDivision of Transplantation Surgery, CLINTEC Karolinska
Institute, Karolinska University Hospital Huddinge, Stockholm, Sweden;
§§Department of Surgery, The Hepatobilairy Center, Paul Brousse Hospital,
University Paris Sud, Villejuif, France; ôôAbdominal Transplant Surgery, KU
Leuven, University Hospitals Leuven, Leuven, Belgium; jjjjDepartment of
Surgery, The Johns Hopkins Hospital, Baltimore, MD;
Division of HBP
Surgery, Department of Surgery, Liver Transplant Unit, Hospital Italiano de
Buenos Aires, Buenos Aires, Argentina; yyyDepartment of Surgery and
Transplantation, Erasmus University Medical Center Rotterdam, Rotterdam,
The Netherlands; zzzDepartment of Abdominal and Transplant Surgery,
University Hospital St. Luc, Brussels, Belgium; §§§Department of Medical
and Surgical Sciences, University of Bologna, Bologna, Italy; ôôôDivision of
Transplantation, University of Rochester Medical Center, New York, NY;
jjjjjjDepartment of Surgery, Western University, London, ON, Canada; and
Department of Epidemiology, Epidemiology, Biostatistics and Prevention
Institute, University of Zurich, Zurich, Switzerland.
CCI
1
refers to registered trademark from the University of Zurich, Zurich,
Switzerland.
The authors report no conflicts of interest.
Supplemental digital content is available for this article. Direct URL citations
appear in the printed text and are provided in the HTML and PDF versions of
this article on the journal’s Web site (www.annalsofsurgery.com).
Reprints: Pierre-Alain Clavien, MD, PhD, Department of Surgery and Transplan-
tation, University Hospital Zurich, Raemistrasse 100, CH-8091 Zurich,
Switzerland. E-mail: clavien@access.uzh.ch.
Copyright
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0003-4932/16/XXXX-0001
DOI: 10.1097/SLA.0000000000002477
Annals of Surger y Volume XX, Number XX, Month 2017 www.annalsofsurgery.com | 1
ORIGINAL ARTICLE
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
currently the standard of care for many patients with end-stage liver
disease and liver tumors; a success that has created a dramatic
imbalance between organ need and availability.
7,8
To face the
resultant organ shortage, higher-risk grafts are implanted in sick
recipients, potentially jeopardizing the postoperative course. One of
the main issues in evaluating the results of LT is the lack of reliable
comparison to reference groups.
With this in mind, we embarked on the concept of benchmark-
ing, which is widely known as quality assessment and improvement in
the field of banking or industrial manufacturing. Benchmarking has
been used in medicine as a tool for evaluating single-center outcomes
by risk-adjusted comparisons to national data.
9–11
We introduced the
idea of identifying best achievable results by selecting low risk cases in
experienced centers and presented a large multicenter study at the
American Surgical Association annual meeting in 2016, proposing
benchmark values for major hepatectomies. The ideal low risk cases
were those performed in healthy living-donor using a comprehensive
analysis of relevant endpoints.
12
To define benchmark values in LT, we targeted high-volume
programs, holding an audited prospective database from areas, where
cadaveric LT is common practice. Low risk cases were defined by
pretransplant characteristics, using the model for end-stage liver
disease (MELD) score, which predicts mortality on the waiting list,
and the balance of risk (BAR) score, which takes into account
recipient and donor factors. Other well-established risk factors were
considered exclusion criteria to define the benchmark cohort.
13– 16
We selected 13 endpoints – to compute benchmark cutoffs based on
their clinical relevance and reproducibility. These benchmark values
may serve as reference for future studies and help assess worldwide
results and trends in the field of LT.
METHODS
Center Selection and Data Sources
We screened countries on every continent for centers perform-
ing at least 50 cadaveric LT per year during the 5-year study period
(2010–2015) and holding a prospectively collected and audited
database. The highest volume centers in the respective countries
were selected and the final collaborative consortium included 17
centers: 11 from Europe, 5 from North America, and 1 from South
America. No Asian center could be included due to the small number
of available cadaveric grafts.
Study Population and Variables of Interest
Benchmark cases were identified in every center database. To
define ‘‘ideal’’ LT cases, we considered the mortality risk of the patients
on the waiting list as well as the donor/recipient match, and thus selected
patients based on the MELD score and the BAR score.
17,18
The MELD score is a scoring system widely used for prioriti-
zation of patients awaiting LT – based on serum levels of bilirubin
and creatinine, and the international normalized ratio for prothrombin
time – predicting the risk of mortality on the waiting list in patients
with chronic liver diseases.
18–20
We selected only patients with a
MELD score 20 at the time of transplantation, as patients with higher
figures have been shown to display a increasing mortality risk.
20
The BAR score predicts post-LT mortality by including donor
age, recipient MELD score, recipient age, retransplant status, the
need for mechanical ventilation, and cold ischemia time.
17
Patients
with a BAR score 9 present an ideal donor–recipient match with
very low mortality risk, and this cutoff was thus included as a
selection criteria for benchmark cases.
7,17,21
To further narrow the selection of the ‘‘best cases,’’ we
excluded patients with acute liver failure, patients on mechanical
ventilation at the time of surgery, and patients receiving a graft from
donors after circulatory death (DCD).
13,16,22– 24
In addition, special
situations and technical difficulties impacting outcomes, such as
recipient portal vein thrombosis, previous major abdominal surgery
(hepato-biliary surgery and extensive colorectal surgery), partial
graft implantation or retransplantation were excluded.
14,15,25
In summary, patients with a MELD score 20, a BAR score
9 and receiving a standard LT from a donor after brain death were
selected as benchmark cases.
Comparative Cohorts
To compare outcomes from the benchmark cohort, we identi-
fied 2 higher risk groups. First, we considered a slightly higher risk
group from the same centers and study period using identical
selection criteria except for the next higher MELD score ranging
from 21 to 30 points. Second, to test the applicability of the
established benchmark criteria for the assessment of a single-center
cohort, we selected one of the highest risk groups from one of the
benchmark centers – that is, those receiving a second graft due to
failure of the original graft – over the same study period.
Collected Data, Outcomes, and Follow-up
Local in vestigators collected center-specific data, using a secure
online data entry management system, including the following recipient
and donor characteristics: age, MELD before LT , underlying liver
disease and comorbidities, cold ischemia time, and a variety of postop-
erati ve ev ents at 4 specific time points (discharge, 3, 6, and 12 mo).
The primary end-point was morbidity and mortality. To assess
LT morbidity, we selected peri- and postoperative parameters with
proven impact on outcomes and costs, namely duration of surgery,
intraoperative blood transfusion, hospital and intensive care unit
(ICU) stay, as well as post-LT complications.
26– 28
A systematic and comprehensive classification of all postop-
erative complications was performed using the Clavien–Dindo
grading system. This system ranks complications by severity accord-
ing to their therapeutic consequences.
29,30
Briefly, grade I and II
complications are events requiring only bedside procedures or a need
for pharmacologic treatments. Grade III complications require sur-
gical, radiological, or endoscopic treatment. Grade IV complications
are life-threatening complications requiring ICU care, while grade V
complications correspond to death. Patients requiring a retransplan-
tation were graded IVa. To get additional information on morbidity,
we then measured the cumulative morbidity using the comprehensive
complication index (CCI
1
).
31,32
The CCI
1
expresses morbidity on a
continuous numeric scale from 0 (no complications) to 100 (death)
by weighing all postoperative complications according to the
Clavien–Dindo classification for their respective severity.
31
Since
in LT, complications related to the biliary tract are the most common
complications and responsible for major costs and morbidity, they
were additionally analyzed separately.
33,34
Statistical Analysis
Benchmark Metrics
Benchmark cutoffs were determined for 1-year mortality,
retransplantations, and peri- and postoperative morbidity indicators
including length of surgery (skin-to-skin time), intraoperative blood
transfusions, renal replacement therapy after transplantation until
discharge, length of ICU, and hospital stay (from day of transplan-
tation until hospital discharge), patients with mild (grade II) and
severe (grade III) complications, biliary complications, hepatic
artery thrombosis (HAT) and the CCI
1
at discharge, 3 months,
6 months, and at 1 year. Every outcome indicator listed above was
defined in the study protocol to secure standardization in data
collection among all centers. Of note, biliary complications were
Muller et al Annals of Surger y Volume XX, Number XX, Month 2017
2 | www.annalsofsurgery.com
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
defined as every post-transplant event related to the biliary system
such as leaks, bilioma, strictures or infection, and graded according
to the Clavien–Dindo classification, and subsequently integrated in
the CCI
1
formula. The median value for every indicator in each
center was determined and the 75th percentile of all center-specific
median values for a given indicator was considered the benchmark
cutoff, as previously reported.
12
Thus, we opted for an equal count of
every center in this study, regardless of differences in caseloads.
We conducted descriptive statistics for various intra- and
postoperative parameters to characterize the patient population.
Survival curves were compared by using the log rank test and linear
regression analysis was performed for correlations. A value of P <
0.05 was considered statistically significant. Statistical analysis was
performed using Stat 10.1 and SPSS, version 19.
RESULTS
Benchmark Cohort
Seventeen centers provided 7492 cases of LT over the 5-year
study period. Approximately one quarter (27%, n ¼ 2024) were
benchmark cases with a large variation among centers (range: 8% to
49%) (Fig. 1).
Characteristics and outcomes in the benchmark cohort:
Patients in the benchmark cohort displayed a median donor and
recipient age of 55 (IQR: 43–67) and 57 (IQR: 50–62) years,
respectively. Median MELD and BAR scores were 12 (IQR: 9–
16) and 4 (IQR: 2–6), respectively. Underlying liver diseases in-
cluded hepatocellular carcinoma (47.4%), hepatitis C (78.5%), alco-
holic cirrhosis (26.5%), and hepatitis B (2%). The median graft
preservation time was short with 7 hours cold storage (IQR: 6–9).
The median duration of LT was 5 hours (IQR: 290–480) with low
transfusion requirements [2U RBC (unit of red blood cell), IQR: 0–
5]. Postoperative renal replacement therapy was performed in 6.6%
of cases. Consistently, median length of ICU and hospital stay were
short with 2 (1–5) and 13 (9–20) days, respectively. Overall 1- and 5-
year patient survival-rates were 91.6% (actual survival) and 78.2%
(actuarial survival), respectively. Post-LT complications, however,
occurred frequently during the 1-year follow-up; the majority of
patients (82%) presented at least 1 complication and 58% developed
more than 1 complication. Overall, two-thirds (68.7%) of the patients
developed mild complications ( grade II), while almost half
(48.9%) experienced a severe complication requiring intervention
( grade III). Biliary complications occurred in 18.8% of the
patients; the majority (57.7%) after hospital discharge. Accordingly,
the overall median post-transplant morbidity index (CCI
1
) increased
over time from 20.9 at discharge (IQR: 0–34.6) to 26.6 at 3 months
(IQR: 8.7–42.4), 29.6 at 6 months (IQR: 12.2–44.9) and 33.5 at
12 months (IQR: 20.9–49.5). Omitting grade I complications for the
CCI
1
calculation did not affect the overall median CCI
1
value at the
4 different time points.
Center-specific and Volume Effect on Outcomes
We recorded a significant variability in perioperative morbid-
ity indicators among centers, namely median length of surgery (IQR:
4.9–8.1 h), median ICU (IQR: 1–8 d), and hospital stay (IQR: 6–24
d). The same applied to the CCI
1
from discharge (IQR: 0–51.7) to
1 year (IQR: 20.9–62.9). These results, however, did not correlate
with the respective center volume. Interestingly, we noted that
centers performing less benchmark cases; that is, higher proportion
of more difficult cases, had less biliary complications in the bench-
mark cohort (Pearson R ¼ 0.68, P ¼ 0.002, Fig. 2).
Benchmark Values
The benchmark cutoffs, calculated as the 75th percentile of
the medians of each center, served to indicate the best achievable
results for each post-LT parameter (Tables 1 and 2). The benchmark
for 1-year mortality and graft-loss are 9% and 11%, respectively.
FIGURE 1. Distribution of LT benchmark
cases among centers.
FIGURE 2. More biliary complication occur in centers with a
higher proportion of benchmark cases.
Annals of Surger y Volume XX, Number XX, Month 2017 Benchmarks for Liver Transplantation
ß 2017 Wolters Kluwer Health, Inc. All rights reserved. www.annalsofsurgery.com | 3
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
Benchmark cutoffs – set for specific perioperative parameters –
are 6 hours surgical time, need for 3 red blood units, and 8%
of patients requiring postoperative renal replacement therapy.
Benchmarks for ICU and hospital stays are 4and18 days,
respectively. The cutoff for severe ( grade II I) compl ications i s
59%, for overall biliary complications 28% and for HAT 4.4%,
while the benchmark values for cumulative morbidity, expressed by
the CCI
1
at discharge, 3, 6, and 12 months, are 29.6, 34.5, 37.2, and
42.1, res pectively.
Higher MELD Cohort
In a next step, we compared outcomes in the benchmark
cohort with sicker transplant recipients, displaying a MELD
score r anging from 21 to 30 (median MELD 24, IQR: 22–27),
but otherwise same selection cri teria (Table 2). In this cohort (n ¼
699), 1-year survival was similar to the benchmark cohort (88.7%
vs 91.6%, P ¼ 0.154). Although morbidity increased, most out-
come endpoints remained inside the benchmark cutoffs. Only
intraoperative transfusion rates (5 U RBC vs 3 U RBC), patient s
with renal replacement therapy after LT (13.7% vs 3%), grade IV
complications (24% vs 20%) and retransplantations (7.1% vs
4%) were outside the benchmark cutoffs. Cumulative post-LT
morbidity, expressed by the CCI
1
at discharge and after 1 year,
remained within the benchmark threshol d (29.6 vs 29.6, and 39.7
vs 42.1).
Second Graft (Retransplantation) Cohort
To test the applicabili ty of the benchmark thresholds in a
high-risk group, we looked at 50 patients receiving a second
transplantation due to graft failure at 1 center. The median MELD
score in that group was 30 ( IQR: 24–36) and the median BAR was
17 (IQR: 11–20). Overall, 28% of retransplantations were per-
formed in the first 2 weeks following primary LT, 34% between
day 15 and the first year, and 38% thereafter. The main etiologies
for a retransplantation were primary nonfunction in 26%, non-
anastomotic biliary stenosis in 20%, hepatitis C recurrence in 16%,
and hepatic artery thrombosis in 10% of cases. Except for t he
use of 1 DCD graft, all retransplantations were performed with
whole grafts from donors after brain death (n ¼ 49). We observed
a 30% reduction in 1st-year patient survival, when compared
with the benchmark cutoffs (58% vs 88%), and all outcome
parameters in the re-LT groups were unambiguously inferior to
the benchmark cutoffs (Table 2). For example, ICU- and hospital
stays were twice as high (8 d vs 4dand28dvs18 d,
respectively), and s evere complications ( Grade III) (96% vs
59%) as well as biliary complications (42% vs 28%), all largely
exceeded benchmark values.
DISCUSSION
This is the first attempt to quantify best possible outcomes
after LT using an international cohort of well-defined benchmark
cases, representing approximately 1 quarter of overall LTs. This
TABLE 1. Benchmark Cutoffs in Liver Transplantation
Perioperative Course
OP duration 6 hours
Intraoperative blood transfusions 3U RBC
Renal replacement therapy 8%
ICU stay 4 days
Hospital stay 18 days
Morbidity and Mortality Discharge 3 months 6 months 1 year
Any complication 80% 90% 90% 94%
Grade II complication 69% 81% 83% 83%
Grade III complication 42% 54% 58% 59%
Biliary complications 12% 18% 20% 28%
Hepatic artery thrombosis (HAT)
4.1% 4.4% 4.4% 4.4%
CCI
1
points 29.6 34.5 37.2 42.1
Graft-loss 4% 6% 9% 11%
Mortality 2% 4% 7% 9%
HAT are commonly divided into two distinct entities based on the time of
occurrence after LT. The benchmark cutoffs for early HAT occurring within the first
30 days and for late HAT occurring thereafter are 4.1% and 1%, respectively.
TABLE 2. First-year Outcomes After Liver Transplantation in Two Higher Risk Groups Compared With First-year Benchmark
Cutoffs
MELD 21–30 (n ¼ 699) Retransplantations (n ¼ 50)
y
Benchmark Cutoffs (at 12 mo)
OP duration, h (IQR) 6 (5–7.5) 6.1 (5–7.6) 6
Intraoperative blood transfusions, number (IQR) 5 (2–8)
6 (4–10) 3
Hospital stay (d), median (IQR) 16 (10–26) 28 (16–51) 18
ICU stay (d), median (IQR) 4 (2–9) 8 (4–22) 4
Renal replacement therapy, patient number (%) 96 (13.7)
21 (42) 8%
Any complication, patient number (%) 616 (88.1) 50 (100) 94%
Grade II complications, patient number (%) 460 (65.8) 39 (78) 76%
Grade IIIa complications, patient number (%) 188 (26.8) 34 (68) 41%
Grade IIIb complications, patient number (%) 180 (25.8) 20 (40) 28%
Grade IV complications, patient number (%) 173 (24)
35 (70) 20%
Grade II complication, patient number (%) 516 (73.8) 42 (84) 82%
Grade III complication, patient number (%) 406 (58.1) 48 (96) 59%
Biliary complications, patient number (%) 109 (15.5) 21 (42) 28%
Retransplantations, patient number (%) 50 (7.1)
5 (10) 4%
CCI
1
discharge, median (IQR) 29.6 (8.7–43) 62.1 (42.2–98.7) 29.6
CCI
1
3 mo, median (IQR) 33.5 (20.9–47.5) 72.7 (54.6–100) 34.5
CCI
1
6 mo, median (IQR) 36.1 (20.9–51.2) 81 (60–100) 37.2
CCI
1
at 1 yr, median (IQR) 39.7 (22.6–55.8) 85.8 (64.2–100) 42.1
1-year mortality, patients (%) 62 (8.8) 18 (36) 9%
Value is outside the benchmark cutoff.
yAll values in the retransplantation cohort are outside the benchmark cutoffs.
Muller et al Annals of Surger y Volume XX, Number XX, Month 2017
4 | www.annalsofsurgery.com
ß 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. Unauthorized reproduction of this article is prohibited.
