MELD Related Files Liver MELD-Plus

134
LIVER DISEASE
MELD scoring system is useful for predicting prognosis in
Gut: first published as 10.1136/gut.52.1.134 on 1 January 2003. Downloaded from http://gut.bmj.com/ on November 24, 2019 by guest. Protected by copyright.
patients with liver cirrhosis and is correlated with residual
liver function: a European study
F Botta, E Giannini, P Romagnoli, A Fasoli, F Malfatti, B Chiarbonello, E Testa, D Risso,
G Colla, R Testa
.............................................................................................................................
Gut 2003;52:134–139
Background: Indices for predicting survival are essential for assessing prognosis and assigning prior-
ity for liver transplantation in patients with liver cirrhosis. The model for end stage liver disease (MELD)
has been proposed as a tool to predict mortality risk in cirrhotic patients. However, this model has not
been validated beyond its original setting.
Aim: To evaluate the short and medium term survival prognosis of a European series of cirrhotic
patients by means of MELD compared with the Child-Pugh score. We also assessed correlations
between the MELD scoring system and the degree of impairment of liver function, as evaluated by the
monoethylglycinexylidide (MEGX) test.
Patients and methods: We retrospectively evaluated survival of a cohort of 129 cirrhotic patients
See end of article for with a follow up period of at least one year. The Child-Pugh score was calculated and the MELD score
authors’ affiliations was computed according to the original formula for each patient. All patients had undergone a MEGX
....................... test. Multivariate analysis was performed on all variables to identify the parameters independently
Correspondence to: associated with one year and six month survival. MELD values were correlated with both Child-Pugh
Professor R Testa, scores and MEGX test results.
Gastroenterology Unit, Results: Thirty one patients died within the first year of follow up. Child-Pugh and MELD scores, and
Department of Internal
Medicine, University of
MEGX serum levels were significantly different among patients who survived and those who died.
Genoa, Viale Benedetto Serum creatinine, international normalised ratio, and MEGX60 were independently associated with six
XV, No 6, 16132 Genoa, month mortality while the same variables and the presence of ascites were associated with one year
Italy; rtesta@unige.it mortality. MELD scores showed significant correlations with both MEGX values and Child-Pugh scores.
Accepted for publication Conclusions: In a European series of cirrhotic patients the MELD score is an excellent predictor of both
30 April 2002 short and medium term survival, and performs at least as well as the Child-Pugh score. An increase in
....................... MELD score is associated with a decrease in residual liver function.
T
here is an increasing discrepancy between the number of Quantitative liver function tests that evaluate the clearance
cirrhotic patients on waiting lists for orthotopic liver of exogenous substances have been studied and combined
transplantation (OLT) and the number of available donor with the Child-Pugh classification or with other scoring
livers.1 The correct timing for surgery has an important impact systems to create composite prognostic scores in an attempt to
on both mortality and morbidity of the patients.2 Indeed, late improve prognostic accuracy.5 10–13 The monoethylglycinexylid-
transplantation affects patient outcome while performing OLT ide (MEGX) test, which depends on liver blood flow and cyto-
in those with less urgent indications penalises patients who chrome P450 activity, is a quantitative liver function test
could benefit most from the procedure.3 Moreover, the related to the severity of liver disease,14 and shows good corre-
economics, social burden, and ethical considerations of main- lation with medium and short term survival prognosis of cir-
taining patients on waiting lists should also be taken into rhotic patients.6 15–20
consideration.4 Recently, the model for end stage liver disease (MELD) was
Over the years, many clinical and biochemical parameters introduced as a tool to predict mortality risk and to assess
have been suggested in order to more accurately predict the disease severity in patients with liver cirrhosis so as to
prognosis of cirrhotic patients and correctly assess their short determine organ allocation priorities.21 Although previoulsy
and medium term survival. The Child-Pugh score is still con- formulated as a prognostic index for cirrhotic patients
sidered the cornerstone in the prognostic evaluation of undergoing transjugular intrahepatic portosystemic stent
cirrhotic patients although it was formulated more than 30 shunt (TIPSS), it was validated by the same authors on a
years ago. Nevertheless, it has some drawbacks such as broad series of patients with liver disease of various aetiology
subjectivity of clinical parameters and limited discriminant and severity.22 23 Nevertheless, although the MELD score takes
ability.5–7 into consideration objective parameters (serum creatinine,
Child-Pugh class A patients usually show good medium
term survival without OLT unless other events (for example, .............................................................
hepatocellular carcinoma, uncontrolled bleeding due to portal
hypertension, etc) occur,1 8 while Child-Pugh class C patients Abbreviations: OLT, orthotopic liver transplantation; MEGX,
monoethylglycinexylidide; MELD, model for end stage liver disease;
are considered the conventional candidates for the procedure.
TIPSS, transjugular intrahepatic portosystemic stent shunt; INR,
Child-Pugh class B patients can be considered a heterogene- international normalised ratio; HCV, hepatitis C virus; HBV, hepatitis B
ous group as their clinical condition may remain stable for virus; HDV, hepatitis D virus; ROC, receiver operating characteristic; SS,
more than a year or rapidly deteriorate.9 sensitivity; SP, specificity; RR, relative risk.
www.gutjnl.com
MELD scoring system 135
Table 1 Clinical and biochemical characteristics of the 129 cirrhotic patients

subdivided according to one year survival
Surviving patients Deceased patients p Value
Sex (M/F) 70/28 25/6 NS

Age (y) 48 (22–75) 51 (31–71) NS
Encephalopathy (yes/no) 7/91 5/26 NS
Ascites (yes/no) 32/66 21/10 0.0007
Bilirubin (mg/dl) 2 (0.3–12) 2.4 (0.4–7) NS
Albumin (g/dl) 3.5 (2.4–5) 3.4 (2.3–5) NS
Prothrombin activity (%) 59 (31–100) 52 (30–85) 0.01
Child-Pugh score 7 (5–12) 9 (5–14) 0.001
Creatinine (mg/dl) 0.9 (0.6–1.3) 1.0 (0.7–1.2) NS
INR 1.6 (1.0–2.9) 1.8 (1.1–2.9) 0.01
MELD score 6 (−5–17) 9 (0–19) 0.003
MELDaetiology score 12 (−5–23) 15 (0–25) 0.003
MEGX15 (ng/ml) 16 (0–110) 8 (0–44) 0.003
MEGX30 (ng/ml) 21 (2–91) 10 (1–63) <0.001
MEGX60 (ng/ml) 27 (4–93) 18 (1–46) <0.01
Data are median (range).

INR, international normalised ratio; MELD, model for end stage liver disease; MELDaetiology, model for end
stage liver disease computed including aetiology of liver disease; MEGX, monoethylglycinexylidide.
the international normalised ratio (INR), bilirubin levels) All patients had undergone the MEGX test as follows: lido-
and is computed with statistically derived coefficients on a caine was injected into a forearm vein over approximately 2–3
continuous scale with no upper or lower limits, thus avoiding minutes at a dose of 1 mg/kg body weight, and blood samples
many drawbacks of the Child-Pugh score, it has generated were then taken 15, 30, and 60 minutes after lidocaine injec-
some criticism.24 25 tion (MEGX15, MEGX30, and MEGX60). MEGX serum concen-
In this study our aim was to evaluate the short and medium tration was measured at the three sampling times by means of
term prognosis of a European series of cirrhotic patients by the TDx fluorescence polarisation immunoassay system,27 and
means of the MELD score compared with the Child-Pugh was calculated as follows: MEGXt − MEGX0, where MEGXt is
score. Moreover, we assessed correlations between the MELD MEGX concentration at sampling time and MEGX0 is MEGX
score and both the Child-Pugh score and residual liver concentration at baseline.
function, as evaluated by the MEGX test. Lastly, we performed Both Child-Pugh and MELD scores were calculated on
these analyses focusing on Child-Pugh class B patients due to parameters obtained at referral. Biochemical evaluations were
the peculiar importance of the prognostic accuracy in this carried out by the same laboratory. Prothrombin expressed as
subgroup of patients. per cent activity was converted to prothrombin time INR using
internal standards of the laboratory and were assessed by a
single operator (GC).
PATIENTS AND METHODS Statistical analysis was first performed on the whole group
We retrospectively studied 129 cirrhotic patients (95 males, of 129 patients and then on the subgroup of 63 Child-Pugh
34 females; median age 50 years, range 22–75) consecutively class B patients, evaluating six month and one year survival.
referred to our unit who were eligible for hepatic functional Comparisons between groups were performed using Fisher’s
assessment, and with a follow up of at least one year. exact test and the Mann-Whitney U test. Results are
Patients with hepatorenal syndrome were not eligible, while expressed as median (range). The correlation between
none of the patients had spontaneous bacterial peritonitis. variables was evaluated by means of Spearman’s rank correla-
Patients with hepatocellular carcinoma were excluded from tion test (rs). Receiver operating characteristic (ROC) curves
the study. Liver cirrhosis was diagnosed on the basis of were used to determine the cut off values of MEGX serum
histological, clinical, biochemical, and instrumental results. concentrations, Child-Pugh score, and MELD, with the best
Aetiology of liver disease was hepatitis C virus (HCV) in 60 sensitivity (SS) and specificity (SP) in discriminating between
patients, hepatitis B virus (HBV) in 21 (among these six were patients who survived and those who died. The validity of the
positive for hepatitis D virus (HDV)), HCV and HBV in models was measured by means of the concordance (c) statis-
two, HCV and HBV with HDV in one patient, alcohol abuse tic (equivalent to the area under the ROC curve).28 A c value of
in 28, HCV and alcohol abuse in nine, and autoimmune in 0.8–0.9 indicates excellent diagnostic accuracy; a model with a
eight. c value >0.7 should be considered useful. Univariate survival
The Child-Pugh score was calculated26 and patients were curves were estimated using the Kaplan-Meier method. Step-
classified as follows: 43 class A (31 males and 12 females; wise survival analysis was performed by means of Cox
median age 52 years, range 22–75), 63 class B (45 males regression to determine the variables independently associ-
and 18 females; median age 48 years, range 28–71), and 23 ated with one year and six month mortality. For all analyses a
class C (19 males and four females; median age 48 years, p value <0.05 was considered statistically significant. Data
range 31–71). The MELD score was calculated according were analysed using the SPSS package for Windows (SPSS
to the original formula proposed by the Mayo Clinic Inc., Chicago Illinois, USA).
group:
RESULTS
3.8 × loge (bilirubin (mg/dl)) + 11.2 × loge (INR) + 9.6 × During the one year follow up, 31 patients died: five were
loge (creatinine (mg/dl)) + 6.4 × (aetiology: 0 if choles- Child-Pugh class A (12%), 16 were class B (25%), and 10 were
tatic or alcoholic, 1 otherwise).20 21 class C (44%). The causes of death were all related to liver dis-
ease. Ninety eight patients survived more than one year: 38
Moreover, we computed the MELD score with or without aeti- were Child-Pugh class A (88%), 47 were class B (75%), and 13
ology of liver disease. were class C (56%).
www.gutjnl.com
136 Botta, Giannini, Romagnoli, et al
Table 2 Clinical and biochemical characteristics of the 129 cirrhotic patients

subdivided according to six month survival
Survived patients Deceased patients p Value
Sex (M/F) 86/31 9/3 NS

Age (y) 49 (22–75) 55 (35–71) NS
Encephalopathy (yes/no) 10/107 1/11 NS
Ascites (yes/no) 44/73 9/3 0.03
Bilirubin (mg/dl) 1.7 (0.3–12) 3.1 (1.6–6.1) 0.002
Albumin (g/dl) 3.5 (2.4–5) 2.9 (2.3–4.3) 0.002
Prothrombin activity (%) 58 (31–100) 50 (30–85) 0.01
Child-Pugh score 7 (5–12) 10 (6–14) 0.0002
Creatinine (mg/dl) 0.9 (0.6–1.3) 1.0 (0.8–1.3) NS
INR 1.6 (1.0–2.9) 1.9 (1.1–2.9) 0.01
MELD score 7 (−5–17) 11 (6–19) 0.0007
MELDaetiology score 12 (−5–23) 16 (13–25) 0.0002
MEGX15 (ng/ml) 13.8 (0–110.8) 7.2 (0–29.7) 0.03
MEGX30 (ng/ml) 20.2 (2.2–90.9) 9 (0.58–39) 0.0007
MEGX60 (ng/ml) 26.1 (4.1–92.9) 14.3 (0.95–40) 0.001
Data are median (range).

INR, international normalised ratio; MELD, model for end stage liver disease; MELDaetiology, model for end
stage liver disease computed including aetiology of liver disease; MEGX, monoethylglycinexylidide.
Tables 1 and 2 show the clinical and biochemical Multivariate analysis was performed on variables
characteristics of the patients, their MELD scores, and contained in both the Child-Pugh and MELD scores, and
their MEGX serum levels 15, 30, and 60 minutes after on all MEGX sampling times to identify parameters
lidocaine administration, respectively, after both a one independently associated with six month and one
year and six month period of follow up. Median Child-Pugh year mortality. This analysis showed that INR (relative risk
as well as MELD scores and MEGX serum levels at (RR) 4.282 (range 1.438–12.751); p=0.009), creatinine serum
each sampling time were significantly different both at levels (RR 8.785 (0.213–36.186); p=0.03), and MEGX60 (RR
six and 12 months between patients who survived and 0.920 (0.860–0.983); p=0.01) were the variables significantly
those who died. Ascites was clinically detected more associated with six month mortality, while the presence of
frequently among deceased patients compared with those ascites (RR 2.146 (1.203–3.895); p=0.01), creatinine serum
who survived. levels (RR 1.473 (1.091–2.014); p=0.01), INR (RR 2.47
The cut off values with the best SS and SP in predicting six (1.056–5.863); p=0.04), and MEGX60 (RR 0.969 (0.939–1);
month and one year survival for Child-Pugh score, MELD p=0.05) were significantly associated with one year mor-
scores, and MEGX serum levels were calculated using ROC tality.
curves. Moreover, c index was calculated to evaluate the accu- MELD scores showed a highly significant correlation with
racy of the three models. Child-Pugh score, MEGX30, and both MEGX serum levels (MEGX15, rs=−0.545, p<0.0001;
MEGX60 showed the best SS in predicting six month survival MEGX30, rs=−0.617, p<0.0001; MEGX60, rs=−0.542, p<0.0001)
while MEGX15 had the best SP. Both Child-Pugh and MELD and Child-Pugh scores (rs=0.817, p<0.0001). Moreover,
scores had excellent prognostic accuracy (table 3). Moreover, Child-Pugh scores were correlated with MEGX serum levels
we observed no differences in prognostic accuracy among the (MEGX15, rs=−0.529, p<0.0001; MEGX30, rs=−0.598,
various prognostic parameters (comparison of ROC curves). p<0.0001; MEGX60, rs=−0.577, p<0.0001).
The MELD score showed the best SS in predicting one year
survival compared with both the Child-Pugh score and Prognostic analysis of Child-Pugh class B patients
MEGX test, although with low SP (table 4). All variables Among the various clinical, biochemical, and functional
evaluated showed good one year prognostic accuracy, without parameters of the 63 Child-Pugh class B patients, only MEGX30
significant differences among the various parameters. Figure (16.4 ng/ml (2.9–63.3) v 9.3 (0.58–13.3); p=0.02) and MEGX60
1 shows that the Child-Pugh and MELD scores as well as (24.6 ng/ml (6.7–55.6) v 13.9 (0.9–18.5); p=0.02) were
MEGX60 cut offs values identified by means of ROC curves significantly lower among patients who died in the short term
clearly differentiated between patients with different survival (six months), while MEGX30 (19 ng/ml (4–54) v 11 (1–63);
times. p=0.05) was lower among patients who died in the medium
term (one year).
Table 3 Six month survival sensitivity, specificity, and Table 4 One year survival sensitivity, specificity, and
c index of the Child-Pugh score, MELD scores, and c index of the Child-Pugh score, MELD scores, and
MEGX serum levels at each sampling time MEGX serum levels at each sampling time
Sensitivity Specificity Sensitivity Specificity
Cut off (%) (%) c index Cut off (%) (%) c index
Child-Pugh score 8 92 72 0.824 Child-Pugh score 8 61 75 0.691

MELD score 8 75 68 0.796 MELD score 6 87 42 0.675
MELDaetiology score 14 75 72 0.823 MELDaetiology score 12 74 58 0.680
MEGX15 (ng/ml) 4 50 91 0.689 MEGX15 (ng/ml) 11 74 63 0.676
MEGX30 (ng/ml) 15.2 92 63 0.799 MEGX30 (ng/ml) 18 81 64 0.703
MEGX60 (ng/ml) 22.8 92 61 0.782 MEGX60 (ng/ml) 26 77 54 0.657
MELD, model for end stage liver disease; MELDaetiology, model for end MELD, model for end-stage liver disease; MELDaetiology, model for end
stage liver disease computed including aetiology of liver disease; stage liver disease computed including aetiology of liver disease;
MEGX, monoethylglycinexylidide. MEGX, monoethylglycinexylidide.
www.gutjnl.com
A
application to patients with different stages of liver
1.0 cirrhosis in order to evaluate their short term survival
Survival probability
0.9
0.8
prognosis.22 23 Lastly, two independent studies performed in
0.7
0.6
North American cirrhotic patients showed that the MELD
0.5 Child-Pugh score ≤8 score performed at least as well as the Child-Pugh score in
0.4
0.3
predicting patient outcome following acute variceal
0.2
Child-Pugh score >8 bleeding32 and mortality in patients referred for OLT.33
0.1
0.0 Nevertheless, in Europe the MELD scoring system has been
0 500 1000 1500 2000 2500
evaluated only in the prognostic assessment of patients
Time (days)
undergoing TIPSS.34
B In this study our aim was to evaluate the short and medium
1.0
term prognostic ability of the MELD scoring system compared
0.9
0.8
0.7 MELD score ≤6
with the Child-Pugh score in a European series of cirrhotic
0.6 patients who were referred to a tertiary care medical centre
0.5
0.4
and who were followed up for at least one year. Moreover, we
0.3 assessed the correlations among these staging systems and
0.2 MELD score >6
0.1 the degree of liver functional impairment, as evaluated by
0.0
0 500 1000 1500 2000 2500
means of the MEGX test. Lastly, we performed these analyses
Time (days) focusing on Child-Pugh class B patients due to the peculiar
C
importance of the prognostic accuracy in this subgroup of
1.0 patients.
0.9
0.8
Firstly, we observed that Child-Pugh and MELD scores, and
MEGX ≤ 26 ng/ml
0.7 60
MEGX could discriminate between deceased and surviving
0.6
0.5 patients both in the short and medium term. Moreover, we
0.4
0.3
MEGX
60
> 26 ng/ml showed that these parameters had high diagnostic accuracy
0.2 for predicting six and 12 month mortality although none per-
0.1
0.0
formed better than the others. Secondly, we found that among
0 500 1000 1500 2000 2500
the clinical variables of the traditional score, the presence of
Time (days)
ascites seemed to be significantly related to both short and
Figure 1 Kaplan-Meier estimated survival curves by (A) Child-Pugh medium term mortality as it was clinically detected more fre-
score (p<0.0001), (B) model for end stage liver disease (MELD) quently among deceased patients. Lastly, multivariate analysis
score (p<0.0001), and (C) monoethylglycinexylidide concentration showed that signs of liver decompensation, such as the
at 60 minutes (MEGX60) (p<0.0001). presence of ascites, subclinical decrease in renal function, INR,
and impairment of liver metabolic activity (MEGX) were
Multivariate analysis showed that MEGX60 (RR 0.867 independently associated with both one year and six month
(0.767–0.980); p=0.02) was independently associated with six mortality, although the presence of ascites was not associated
with earlier death.
month mortality while this analysis was unable to identify any
For the first time, we found a significant correlation
variable associated with one year mortality.
between MELD score and the degree of metabolic liver
Lastly, correlations between MELD scores and MEGX serum
functional impairment, as well as confirmation of the
levels (MEGX15, rs=−0.356, p=0.005; MEGX30, rs=−0.442,
relationship between MELD and Child-Pugh score.22 23
p<0.0001; MEGX60, rs=−0.262, p=0.039) and Child-Pugh
Child-Pugh class B patients were then analysed separately
scores (rs=0.556, p<0.0001) were maintained even in this
as this subgroup of patients displays a prognosis that may be
smaller subgroup of patients.
difficult to evaluate. Indeed, we found that correlations
between MELD and degree of liver functional impairment
DISCUSSION were maintained even in this subgroup of patients, although
Prognostic evaluation of patients with liver cirrhosis is only MEGX30 values were significantly lower in deceased
an important topic often challenging the clinician. The patients than in survivors. While impaired liver
number of patients on waiting lists for OLT is becoming function (MEGX60) seemed to be the only parameter
increasingly higher compared with the number of available independently associated with six month mortality, multi-
donor livers. Correct timing of OLT can reduce the mortality variate analysis was unable to identify any variable associated
of patients on waiting lists and improve post-transplant with one year mortality, thus emphasising the complexity of
survival. Moreover, accurate prognosis of patients with medium term prognostic assessment of this group of
cirrhosis is important so as to plan their management as well patients.
as the choice of major procedures (for example, TIPSS, In this study, we confirmed the prognostic ability
surgical, or locoregional treatment for hepatocellular and accuracy of the MELD scoring system in a setting
carcinoma). which was different from that where the score was
The Child-Pugh score is an important component of formulated (for example, ethnic background, features of
the prognostic evaluation of cirrhotic patients and of the disease, etc). In fact, the MELD score showed discriminant
current organ allocation policy, although this traditional ability between patients who survived and those who
score has several shortcomings such as subjectivity of some died both at six and 12 months of follow up. Moreover, it
parameters and limited discriminant ability. In order to over- showed excellent diagnostic accuracy in predicting short and
come the limits of the Child-Pugh score, previous studies medium term survival, at least comparable with the
have evaluated a “combined score” with quantitative liver Child-Pugh score. Lastly, for the first time we showed that the
function tests,5 6 10 11 15 have created new scores, or have MELD scoring system, which is simple, reproducible, and
applied scores that were originally formulated to evaluate contains variables with important physiopathological
multiorgan insufficiency in critically ill patients to cirrhotic connotations, correlated significantly with the degree of liver
patients.29–31 functional impairment.
Recently, the study group at the Mayo Clinic introduced a Our findings are further strengthened by the results of
new scoring system (MELD) to evaluate the prognosis of multivariate analysis. Indeed, multivariate analysis showed
patients undergoing TIPSS.21 They then generalised its that parameters contained in the MELD score (INR and
www.gutjnl.com
138 Botta, Giannini, Romagnoli, et al
serum creatinine levels) and the MEGX test results (MEGX60) 2 Carithers JL Jr. Liver transplantation. American Association for the study
were independently associated with six month mortality, and of liver diseases. Liver Transpl 2000;6:122–35.
3 Freeman R, Edwards E. Liver transplant waiting time does not correlate
that signs of liver decompensation such as the presence of with waiting list mortality: implications for liver allocation policy. Liver
ascites was an important parameter predicting one year mor- Transpl 2000;6:543–52.
tality. These results seem to confirm the complementary role 4 Rufat P, Fourquet F, Conti F, et al. Costs and outcomes of liver
of clinical and functional assessment in the prognostic transplantation in adults. Transplantation 1999;68:76–83.
5 Oellerich M, Burdelski M, Lautz HU, et al. Assessment of pre-transplant
evaluation of cirrhotic patients. Moreover, they indicate prognosis in patients with cirrhosis. Transplantation 1991;51:801–6.
the importance of both Child-Pugh and MELD scoring 6 Testa R, Valente U, Risso D, et al. Can the MEGX test and serum bile
systems. Lastly, they underscore the importance of some acids improve the prognostic ability of Child-Pugh’s score in liver
parameters that are pathophysiologically inter-related, and cirrhosis? Eur J Gastroenterol Hepatol 1999;11:559–63.
7 Shrescha R, McKinley C, Showalter R, et al. Quantitative liver function
that contribute to worsening of the patient’s condition, such tests define the functional severity of liver disease in early-stage cirrhosis.
as the subclinical decrease in renal function and the presence Liver Transpl Surg 1997;3:166–73.
of ascites. Furthermore, these results are in keeping 8 Keeffe EB. Summary of guidelines on organ allocation and patient listing
with those of recent studies showing that the presence of for liver transplantation. Liver Transpl Surg 1998;4:S108–14.
9 Oellerich M, Burdelski M, Lautz HU, et al. Predictors of one-year
ascites as well as renal function impairment represents one of pretransplant survival in patients with cirrhosis. Hepatology
the most important prognostic variables in cirrhotic 1991;14:1029–34.
patients.35 36 10 Arrigoni A, Gindro T, Aimo G, et al. Monoethylglycinexlydide test: a
The correlation between the MELD scoring system and the prognostic indicator of survival in cirrhosis. Hepatology 1994;20:
383–7.
degree of liver functional impairment that we observed in the
11 Adler M, Verset D, Bouhdid H, et al. Prognostic evaluation of patients
whole group was confirmed even in the subgroup of with parenchymal cirrhosis. J Hepatol 1997;26:642–9.
Child-Pugh class B patients, although in this group the MELD 12 Jalan R, Hayes PC. Review article: quantitative tests of liver function.
score was unable to discriminate between deceased and Aliment Pharmacol Ther 1995;9:263–70.
surviving patients, either in the short or medium term. It is 13 Gindro T, Arrigoni A, Martinasso G, et al. Monoethyl glycine xylidide
(MEGX) test evaluation in primary biliary cirrhosis: comparison with
likely that in this subgroup of patients evaluation of dynamic Mayo score. Eur J Gastroenterol Hepatol 1997;12:1155–9.
changes in the MELD score would help identify patients at 14 Testa R, Caglieris S, Risso D, et al. Monoethylglycinexylidide formation
higher risk of death.37 Nevertheless, functional evaluation measurement as a hepatic function test to assess severity of chronic liver
seemed to be the best tool for predicting one year (MEGX30) disease. Am J Gastroenterol 1997;92:2268–73.
15 Shiffman ML, Luketic VA, Sanyal AJ, et al. Hepatic lidocaine metabolism
and six month (MEGX60) mortality in Child-Pugh class B and liver histology in patients with chronic hepatitis and cirrhosis.
patients. Hepatology 1994;19:933–40.
The fact that the various analyses identified different 16 Oellerich M, Hartmann H, Ringe B, et al. Assessment of prognosis
MEGX sampling times as variables associated with medium transplant candidates by use of the Pugh-MEGX score. Transpl Proc
1993;25:1116–19.
and short term prognosis has a sound physiopharmacological 17 Shiffman ML, Luketic VA, Sanyal AJ, et al. Use of hepatic lidocaine
basis. In fact, MEGX30 most likely represents comprehensive metabolism to monitor patients with chronic liver disease. Ther Drug
functional reserve in cirrhotic patients and reflects the com- Monit 1996;18:372–7.
bination of both impaired blood flow and metabolic 18 Shiffman ML, Fisher RA, Sanyal AI, et al. Hepatic lidocaine metabolism
and complications of cirrhosis. Implications for assessing patient priority
activity.38 MEGX15 is predominantly influenced by hepatic for hepatic transplantation. Transplantation 1993;55:830–5.
blood flow and probably represents a “load stress test” while 19 Ercolani G, Grazi G, Callivà R, et al. The lidocaine (MEGX) test as an
MEGX60 is more an expression of residual metabolic index of hepatic function: its clinical usefulness in liver surgery. Surgery
capacity.39 2000;127:464–71.
20 Fabbris L, Jemmolo RM, Toffolo G, et al. The monoethylglycinexylidide
In conclusion, in this study we have confirmed the test for grading of liver cirrhosis. Aliment Pharmacol Ther
prognostic ability of the MELD scoring system for both 1999;13:67–75.
short and medium term survival, even in a European 21 Malinchoc M, Kamath PS, Gordon FD, et al. A model to predict poor
series of cirrhotic patients. Moreover, for the first time we survival in patients undergoing transjugular intrahepatic portosystemic
shunts. Hepatology 2000;31:864–71.
have shown that the MELD scoring system is correlated with
22 Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict
the degree of liver functional impairment. As both the MELD survival in patients with end-stage liver disease. Hepatology
and Child-Pugh scores showed similar accuracy, we feel that 2001;33:464–70.
they should be considered complementary tools in evaluating 23 Forman LM, Lucey MR. Predicting the prognosis of chronic liver disease:
an evolution from Child to MELD. Hepatology 2001;33:473–5.
the progressive worsening of the clinical condition of
24 McCaughan GW, Strasser SI. To MELD or not to MELD? Hepatology
cirrhotic patients. Furthermore, particular attention should 2001;34:215–16.
be paid to the presence of ascites and creatinine serum levels 25 Wiesner RH, McDiarmid SV, Kamath PS, et al. MELD and PELD:
as clinical signs of derangement of more complex activities application of survival models to liver allocation. Liver Transpl
and markers of a higher risk of death. Lastly, among 2001;7:567–80.
26 Pugh RNH, Murray Lyon IM, Dawson JL, et al. Transection of the
Child-Pugh class B patients, evaluation of liver functional oesophagus for bleeding oesophageal varices. Br J Surg
reserve has an important role in assessing their short term 1973;60:646–64.
prognosis. 27 Oellerich M, Raude E, Burdelski M, et al. Monoethylglycinexylidide
formation kinetics: a novel approach to assessment of liver function. J
Clin Chem Biochem 1987;25:845–53.
..................... 28 Hanley JA, McNeil BJ. The meaning and use of the area under receiver
Authors’ affiliations operating characteristic (ROC) curve. Radiology 1982;143:29–36.
F Botta*, E Giannini*, P Romagnoli, A Fasoli, F Malfatti, 29 Singh N, Gayowski T, Wagener MM, et al. Outcome of patients with
B Chiarbonello, E Testa, R Testa, Gastroenterology Unit, Department cirrhosis requiring intensive care unit support: prospective assessment of
of Internal Medicine, University of Genoa, Italy predictors of mortality. J Gastroenterol 1998;33:73–9.
D Risso, Department of Health Science, University of Genoa, Genoa, 30 Muto P, Freeman RB, Haug CE, et al. Liver transplant candidate
Italy stratification systems. Implications for third-party payors and organ
G Colla, Department of Laboratory Medicine, S Martino Hospital, allocation. Transplantation 1994;57:306–8.
Genoa, Italy 31 Zauner C, Schneeweiss B, Schneider B, et al. Short-term prognosis in
critically ill patients with liver cirrhosis: an evaluation of a new scoring
*F Botta and E Giannini contributed equally to this work. system. Eur J Gastroenterol Hepatol 2000;12:517–22.
32 Chalasani N, Kahi CJ, Francois F, et al. Mayo clinic end-stage liver
disease model (MELD) for predicting patient outcomes following acute
REFERENCES variceal bleeding. Hepatology 2001;34:345A.
1 Lucey MR, Brown KA, Everson GT, et al. Minimal criteria for placement 33 Abouassi SG, Mihas AA, Williams LM, et al. MELD and CTP scores are
of adults on the liver transplant waiting list. Transplantation equivalent predictors of mortality in cirrhotic veterans referred for
1998;66:956–62. orthotopic liver transplantation (OLT). Hepatology 2001;34:207A.
www.gutjnl.com
34 Schepke M, Roth F, Brensing KA, et al. MELD-score compared to the 37 Yousfi MM, Douglas DD, Harrison E, et al. Model for end-stage liver
Child-Pugh and the Emory-score for the prediction of long-term survival in disease (MELD). Dynamic changes in MELD score is important in
patients undergoing transjugular intrahepatic portosystemic shunting predicting mortality for patients awaiting liver transplantation (LTX).
(TIPS). Hepatology 2001;34:183A. Hepatology 2001;34:254A.
35 Gonzalez E, Rimola A, Navasa M, et al. Liver transplantation in patients 38 Bargetzi MJ, Aoyama T, Gonzalez FJ, et al. Lidocaine metabolism in
with non-biliary cirrhosis: prognostic value of preoperative factors. J human liver microsomes by cytochrome P450IIIA4. Clin Pharmacol Ther
Hepatol 1998;28:320–8. 1989;46:521–7.
36 Fernàndez-Esparrach G, Sànchez-Fueyo A, Ginès P, et al. A 39 Testa R, Campo N, Caglieris S, et al. Lidocaine elimination and
prognostic model for predicting survival in cirrhosis with ascites. J monoethylglycinexylidide formation in patients with chronic hepatitis or
Hepatol 2001;34:46–52. cirrhosis. Hepatogastroenterol 1998;45:154–9.
www.gutjnl.com
Liver 2019.qxp_Liver Policy 6/26/19 4:11 PM Page 1
TA L K I N G A B O U T T R A N S P L A N TAT I O N
Questions and Answers for Transplant Candidates about

the Liver Allocation System
United Network for Organ Sharing (UNOS) is a non-profit, charitable
organization that serves as the Organ Procurement and Transplantation
Network (OPTN) under contract with the federal government. The OPTN
helps create and define organ allocation and distribution policies that make
the best use of donated organs. This process involves continuously evaluating
new advances and discoveries so policies can be adapted to best serve
patients waiting for transplants.
All transplant programs and organ procurement organizations throughout

the country are OPTN members and are obligated to follow the policies the
OPTN creates for allocating organs.
As part of this process, UNOS developed a system for prioritizing

candidates waiting for liver transplants based on statistical formulas that are
designed to predict who needs a liver transplant most urgently. The MELD
(Model for End- Stage Liver Disease) is used for candidates age 12 and
older and the PELD (Pediatric End-Stage Liver Disease Model) is used for
patients age 11 and younger.
This document explains the system and how it affects those needing a
transplant.
What is MELD? How is it used?

The Model for End-Stage Liver Disease (MELD) is a numerical scale, ranging
from 6 (less ill) to 40 (gravely ill), used for liver transplant candidates age
12 and older. It gives each person a ‘score’ (number) based on how urgently
he or she needs a liver transplant within the next three months. The
number is calculated by a formula using four routine lab test results:
1
• bilirubin, which measures how effectively the liver excretes bile

• INR (prothrombin time), which measures the liver’s ability to
make blood clotting factors
• creatinine, which measures kidney function (Impaired kidney
function is often associated with severe liver disease.)
• serum sodium, which measures the severity of conditions such as portal
hypertension.
The only priority exceptions to MELD are the categories known as Status
1A and 1B. Status 1A patients have acute (sudden and severe onset) liver
failure and a life expectancy of hours to a few days without a transplant.
Status 1B is reserved for very sick, chronically ill pediatric patients (age
less than 18). Less than one percent of liver transplant candidates are in
these categories at any one time. All other liver candidates age 12 and older
are prioritized by the MELD system.
A patient’s score may go up or down over time depending on the status of

his or her liver disease. Most candidates will have their MELD score assessed
a number of times while they are on the waiting list. This will help ensure
that donated livers go to the patients in greatest need at that moment.
MELD has been shown to rank patients on the waiting list reliably in terms
of their short-term risk of death. The MELD formulas are simple, objective
and verifiable, and yield consistent results whenever the score is calculated.
What is PELD? How does it differ from MELD?

Candidates age 11 and younger are placed in categories according to the
Pediatric End-Stage Liver Disease (PELD) scoring system. Again, a small
group of urgent patients may be listed as a Status 1A or 1B. All other
candidates in this age range receive priority through PELD.
2
PELD is similar to MELD but uses some different factors to recognize the
specific growth and development needs of children. PELD scores may also
range higher or lower than the range of MELD scores. The measures used
are as follows:
• bilirubin, which measures how effectively the liver excretes bile

• INR (prothrombin time), which measures the liver’s ability to
make blood clotting factors
• albumin, which measures the liver’s ability to maintain nutrition
• growth failure
• whether the child is less than one year old
As with MELD, a patient’s score may go up or down over time depending

on the degree of his or her disease severity. Most candidates will have
their PELD score assessed a number of times while they are on the waiting
list. This will help ensure that donated livers go to the patients in greatest
need at that moment.
How are livers allocated?

First, transplant candidates that are not compatible with the donor based
on a number of characteristics (blood type, height, weight, etc.) are
screened from the match run that determines the order a liver is offered.
The remaining candidates on this match run are prioritized based on the
following factors:
• the donor’s age

• their medical urgency
• their geographical proximity to the donor (local–defined by the
Organ Procurement Organization’s service area; regional – UNOS
has 11 allocation regions in the U.S.; national-all remaining
candidates in the nation)
3
Livers from adult donors are allocated first to the most urgent candidates
located in the same region as the donor; Status 1A candidates, followed by
Status 1B candidates. The allocation sequence provides broader access
to those most in need of a liver (those with scores higher than 35) and
those who would receive the most benefit (those with scores higher
than 15). Therefore, after regional Status 1A and 1B candidates, liver
offers are then made to
• candidates with MELD/PELD scores 35 and higher within the donor‘s

region, with offers first made locally, then regionally (i.e., local 40)
regional 40, local 39, regional 39, etc.)
• local candidates with scores greater than 15
• regional candidates with scores greater than 15
• national candidates in Status 1A or 1B
• national candidates with scores greater than 15
• candidates with scores less than 15 locally, regionally, then
nationally
If a combined liver-intestine is being offered, candidates waiting for a

liver-intestine anywhere in the country may be offered the combination
(based on their MELD/PELD score) after local candidates with MELD/PELD
scores of 29 or higher.
Partly because pediatric transplant candidates need smaller organs, they will
receive priority in the liver offer sequence if the donor is younger than 18.
Liver offer process for donors 0-10 years of age

1. Offers are first extended to all compatible pediatric Status 1A
candidates located in the same region as the donor.
2. Next, the liver is offered to the remaining Status 1A candidates across
the nation that are 0-11 years old.
4
3. If the liver has not been accepted yet, it is offered to local adult Status 1A
potential transplant recipients then to Status 1A adults in the same region.
4. Next, all pediatric Status 1B candidates in the region receive the liver
offer, followed by all candidates 0-11 years old in the region in order
of decreasing PELD score.
5. If no one has accepted the liver at this point, it is offered to adolescent
(12-17 years old) candidates that are local to the donor and have a
MELD score greater than or equal to 15, then to local adults that have
a MELD score greater than or equal to 15.
6. That same adolescent/adult MELD score greater than or equal to 15
sequence of offers would then be made to those potential transplant
recipients in the region.
7. Following these offers, candidates with a MELD score less than 15 are
offered the liver using the same adolescent/adult progression locally,
then regionally.
8. If not accepted for any of these patients, the liver is then offered to
potential recipients nationwide, with similar pediatric priority and those
most urgent patients being offered the liver first.
Is waiting time counted in the sysytem?

Various studies report that waiting time is a poor indicator of how
urgently a patient needs a liver transplant. This is because some
patients are listed for a transplant very early in their disease, while
others are listed only when they become much sicker.
Under the MELD/PELD system with a wide range of scores, waiting

time is not often used to break ties. Waiting time will only determine
who comes first when there are two or more patients in the same
allocation classification with the same MELD or PELD score.
5
Do MELD and PELD account for all conditions?

MELD/PELD scores reflect the medical need of most liver transplant
candidates. However, there may be special exceptions for patients with
medical conditions not covered by MELD and PELD. If your transplant
team believes your case qualifies for an exception, they may submit
information to their regional review board (RRB) and request a higher
score. The RRB will consider the medical facts and determine whether
or not to grant a higher score.
Is this system likely to change?

Liver allocation policy based on MELD and PELD has changed as transplant
professionals have applied and learned from the system, and future
changes will likely be required to better meet patients’ needs. In fact, this
system is designed to be flexible and allow improvements. In transplantation,
as in all scientific fields, new studies are taking place all the time to learn
how to save more lives and help people live longer and better.
For more information

Start with your doctor or the medical team at your transplant center. They
know the most about your specific medical condition and treatment. Don’t
be afraid to ask questions. It will help you to have a detailed understanding
of all your treatment options.
UNOS’ Patient Services phone line (888-894-6361) can provide information

about the OPTN allocation policy and other resources available to you.
Additional information is available online on the following websites:
http://www.transplantliving.org
http://optn.transplant.hrsa.gov
http://www.unos.org
http://www.srtr.org
6
Our mission is to unite and strengthen the donation

and transplant community to save lives.
www.unos.org/social
700 North 4th Street, Richmond, VA 23219

www.unos.org
Copyright © 2019 United Network for Organ Sharing 104 7.19

Public Comment Proposal
MELD Exception Scores during

NLRB Transition
OPTN/UNOS Executive Committee
Prepared by: Elizabeth Miller

UNOS Policy and Community Relations Department
Contents
Executive Summary 1
Is the sponsoring Committee requesting specific feedback or input about the proposal? 1
What problem will this proposal address? 2
Why should you support this proposal? 2
How was this proposal developed? 2
What is the Committee proposing? 5
MMaT/DSA during the transition period 5
Which populations are impacted by this proposal? 6
How does this proposal impact the OPTN Strategic Plan? 7
How does this proposal comply with the Final Rule? 7
How will the OPTN implement this proposal? 8
How will members implement this proposal? 9
Will this proposal require members to submit additional data? 9
How will members be evaluated for compliance with this proposal? 9
Policy or Bylaws Language 10
OPTN/UNOS Public Comment Proposal
MELD Exception Scores during

NLRB Transition
Affected Policies: 9.4.D Calculation of Median MELD or PELD at Transplant
Sponsoring Committee: Executive Committee
Public Comment Date: January 31-March 22, 2019
Executive Summary
Liver allocation currently uses donor service areas (DSAs) and OPTN regions as units of distribution. That
will change when recently approved allocation changes are implemented (targeted for April 30, 2019).
The National Liver Review Board (NLRB) was to be implemented on January 31, 2019, but that plan was
delayed because unintended consequences were identified in the plan to base the model for end stage
liver disease (MELD) exception scores on median MELD at transplant within 250 NM of the transplant
hospital (MMaT/250). Because the MMaT/250 for transplant programs within the same DSA can vary,
similarly situated candidates at different hospitals within a DSA can have different scores, and would
consistently appear lower or higher than one another on every match run while allocation is still based on
DSA and region. The OPTN is now soliciting comments on a policy change that would apply only with
respect to the calculation of exception scores to address this issue. This transition policy would be in
effect for a brief period after the public comment period (scheduled to end March 22, 2019) and before
implementation of the revised allocation policy (targeted for April 30, 2019).
In an attempt to address this situation, the OPTN Executive Committee approved the OPTN Liver
Committee’s proposed temporary solution: using the median MELD at transplant in the DSA (MMaT/DSA)
instead of MMaT/250 during the period between implementation of the NLRB and implementation of liver
allocation changes (hereafter, “transition period”). This transition period will be interim, and will expire
upon the implementation of liver allocation changes (targeted for April 30, 2019). The goal of this
approach is to better align the calculation of exception scores with the current allocation system during
the transition period.
Is the sponsoring Committee requesting specific

feedback or input about the proposal?
The Committee requests feedback on the proposed temporary solution of using the MMaT/DSA instead
of MMaT/250 during the transition period. In addition to feedback on the proposed solution, the
Committee welcomes comment on the other options considered by the committee as well as whether to
convert exception scores that were granted by Regional Review Boards (RRB) but will expire after the
implementation of the new allocation system (see pages 5-6). The Executive Committee may reconsider
these alternate solutions for this proposal.
Page 1
What problem will this proposal address?

Upon the implementation of NLRB, MMaT will be used to calculate exception scores for candidates. As
part of the Board-adopted proposal in December of 2018, MMaT is set to be calculated based on a 250
nautical mile radius from the transplant candidate’s hospital. This calculation was tied to a unit of
allocation in the new allocation system – 250 nautical miles from the donor hospital. However, preliminary
data recently reviewed by the Liver Committee showed assigning exception scores relative to MMaT/250
while DSAs are still in use for allocation results in certain similarly situated candidates at different
hospitals within a DSA receiving different exception scores, and consistently appearing lower or higher
than one another on every match run. Though this is a temporary problem, only existing during the
transition period, the Executive Committee seeks the feedback of the community on whether to address
this problem for the short transition period, and if so, how.
Why should you support this proposal?

This proposal will reduce the disparate impact on some exception candidates in the same DSA during the
transition period.
How was this proposal developed?

Development of underlying proposal
In 2017, two proposals from the OPTN/UNOS Liver and Intestinal Organ Transplantation Committee
(Liver Committee) were approved by the Board. The first proposal, originally approved in June 2017,
created the NLRB and changed exception scoring to be based on MMaT/DSA. 1 The second was
approved in December 2017 and changed allocation of livers to allocation by DSA and region, with some
consideration for distance between donor hospital and transplant hospital. 2
Before either of those proposals was implemented, the Board directed the Liver Committee to propose
revisions to policy that included providing OPTN Final Rule compliant replacements for the use of Region
and DSA in liver allocation and the use of DSA in the median MELD or pediatric end-stage liver disease
(PELD) at transplant scoring for exception patients.3
After reviewing the OPTN’s determination that the “disparate sizes, shapes, and populations of DSAs as
drawn today are not rationally determined in a manner that can be consistently applied equally for all
candidates,”4 the Health Resources and Services Administration (HRSA) Administrator (Administrator)
sent a letter to the OPTN on July 31, 2018, stating that “the OPTN Board is directed to adopt a liver
allocation policy that eliminates the use of DSAs and OPTN regions and that is compliant with the OPTN
Final Rule.”5 The letter contained a deadline for the Board to adopt a new liver allocation policy by its
December 2018 meeting. Additionally, the Administrator explained that “[t]he OPTN may also implement
transition patient protections. See 42 CFR § 121.8(d)(1) (providing that when the OPTN revises organ
allocation policies, it shall consider whether to adopt transition procedures that would treat people on the
waiting list and await transplantation prior to the adoption or effective date of the revised policies no less
favorably than they would have been treated under the previous policies).”6
1 Proposal to Establish a National Liver Review Board, OPTN/UNOS Liver and Intestinal Transplantation Committee,
June 2017, https://optn.transplant.hrsa.gov/media/2176/liver_boardreport_nlrb_201706.pdf (accessed January 20,
2019).
2 Enhancing Liver Distribution, OPTN/UNOS Liver and Intestinal Transplantation Committee, December 2017,
https://optn.transplant.hrsa.gov/media/2329/liver_boardreport_201712.pdf (accessed January 20, 2019).

3 Yolanda Becker, OPTN President, letter to the OPTN Liver and Intestinal Organ Transplant Committee, June 25,
2018.
4 Yolanda Becker, OPTN President, letter to Administrator Sigounas, June 25, 2018.
5 George Sigounas, letter to Sue Dunn, OPTN President, July 31, 2018.
6 Ibid.
Page 2
The Liver Committee proposed using distances of 150NM, 250NM, and 500NM between donor hospital
and transplant hospital instead of DSA and region for liver allocation. It also proposed using MMaT/250
from the transplant candidates’ hospital as the basis for exception scores. The Liver Committee originally
requested a delay of three months in between implementation of the scoring changes and implementation
of allocation changes. This would permit time for many exception candidates to transition to the new
scoring methodology for exception candidates according to the renewal schedules in policy. At the
December 2018 Board meeting, the Board approved both of these changes, as well as the time between
implementation.7 Following the Board meeting, the exception scoring changes were scheduled to be
implemented on January 31, 2019, and the allocation changes were scheduled for April 30, 2019.
After the Board’s December 2018 meeting, in preparation for implementation of scoring changes, the
Liver Committee reviewed a report of MMaT/250 for each liver transplant program.8 Within this cohort, the
report showed four DSAs where the liver transplant programs within the same DSA do not have the same
MMaT/250 (Table 1).9 In one DSA, there is a single transplant program that has a MMaT/250 one point
below all of the others in the DSA. In another, there is a single transplant program that has a MMaT/250
two points above all of the others in the DSA. In one DSA, there are two transplant programs two points
higher than the other two transplant programs. In the final DSA, there are four transplant programs with
the same MMaT/250, one that is a point higher, and another that is a point lower.
Table 1: DSAs in which MMaT within 250NM of Liver

Transplant Programs differs by Transplant Program10
DSA Code Transplant Hospital Code MMaT
MAOB CTYN 30
MABI 31
MACH 31
MALC 31
MAMG 31
MAUM 32
LAOP LACH 27
LAOF 27
LATU 27
LAWK 29
TXSB TXJS 30
TXCM 31
TXMC 31
TXSP 31
TXTX 31
MNOP MNMC 30
MNSM 30
MNUM 32
SDMK 32
If MMaT/250 is implemented during the transition period, exception candidates that extend an exception
or apply for a new exception at some of the affected transplant hospitals within these four DSAs will
7 OPTN Policy Notice Liver and Intestine Distribution Using Distance from Donor Hospital, Liver and Intestinal
Transplantation Committee, January 2019, https://optn.transplant.hrsa.gov/media/2788/liver_policynotice_201901.pdf
(Accessed January 19, 2019).
8 OPTN/UNOS Liver Committee meeting January 17, 2019.
9 Ibid.
10 OPTN/UNOS Descriptive Data Request. “NLRB: Median Transplant Scores to be Implemented.” Prepared for Liver
and Intestinal Transplantation Committee Conference Call, January 17, 2018.
Page 3
consistently appear in lower classifications on the match run than those candidates at another transplant
hospital in DSA-level allocation sequences.11
The Liver Committee proposed an adjustment to the implementation plan to align scoring during the
transition period. The Liver Committee’s proposal was to use the MMaT/DSA during the transition period.
The Executive Committee considered the Liver Committee’s proposal on January 25, 2019. At that time,
the Executive Committee was of the opinion that it could be enacted without soliciting public comment
prior to implementation. Therefore, the Executive Committee approved this proposal for implementation12.
On January 30, 2019, the Administrator informed the OPTN by letter that “HRSA views the transition
policy approved January 25, 2019, as a policy change that should be subject to additional public
comment before it is implemented.”13
In light of the letter from the Administrator, the OPTN reconsidered the options available to solve the
problem of exception scores that were not aligned with allocation during the transition period. On January
30, 2019, the Executive Committee repealed the January 25, 2019 approval in order to solicit public
comment on the proposal prior to a final determination.14 In light of the new information about the need for
public comment15, the OPTN proposes that exception scores be assigned relative to MMaT/DSA during
the transition period, and that the transition period be shortened to approximately a month to permit time
for public comment, which is outlined further below, under What is the Committee Proposing?. The
options the OPTN considered are discussed below.
Other options considered

The OPTN considered several different options to resolve this issue. The Executive Committee welcomes
comment on these options and may reconsider these options following public comment.
1) Leave MMaT/250 during the transition period. (2018 Board approved)

This option would implement the entire NLRB system immediately, including assigning exception
scores based on MMaT/250. It would have the advantage of being easy to explain and would not
require any policy change. It also would result in more similar scores among transplant hospitals in a
single region compared to MMaT/DSA. This option was already approved by the Board, has already
been out for public comment, and has been communicated to members.
2) National MMaT during the transition period.

The OPTN considered using the national MMaT during the transition period, since it would be
unaffected by the distribution unit. However, the purpose of using a localized MMaT was to factor in
the differences in the ability to access transplant across the country. A score based on a national
11 Beginning on the day the new exception scoring takes effect, candidates will be assigned exception scores based
on MMaT when they apply for a new exception or extend existing exceptions. Since exception scores are valid for a
period of up to 3 months, candidates who already have existing exception scores on the day NLRB takes effect will
not all change to MMaT-based exception scores at the same time. Some will request a new exception or extend an
existing exception in the first week, while others may not need to do so until months later. During this transition
period, there will be some candidates with exception scores based on the current exception scoring and others based
on MMaT. Early in the transition period, there will be more candidates with scores assigned under the current system,
and at the end of the transition period, most exception candidates will have scores based on MMaT. During this time,
even with this change, there will be similarly situated candidates who have different exception scores, potentially at
the same transplant hospital. This particular difference in scores is unavoidable during the transition, and is part of
the reason the Liver Committee previously requested that the scoring changes take effect at least three months
before the implementation of allocation changes. This proposal does not change the fact that there will be some
candidates with old scores and some candidates with new scores during the transition period.
12 OPTN/UNOS Executive Committee meeting, January 25, 2019.
13 George Sigounas, letter to Sue Dunn, January 30, 2019.
14 OPTN/UNOS Executive Committee meeting, January 30, 2019.
15 42 CFR 121.4(b)(1) provides that the OPTN must “[p]rovide opportunity for the OPTN membership and other
interested parties to comment on proposed policies and shall take into account the comments received in developing
and adopting policies for implementation by the OPTN.
Page 4
MMaT would not grant the same chance of transplant to candidates in areas where the local MMaT is
high that it would in areas with a lower local MMaT.
3) Manually adjust MMaT for the affected transplant hospitals.

The OPTN considered using a different score in place of MMaT/250 only for candidates at the 10
affected transplant hospitals in the four affected DSAs. This could either involve assigning all of the
candidates in a DSA the highest MMaT/250 within the DSA or the lowest MMaT/250 in the DSA. This
would solve the problem of assigning different exception scores to similar candidates during the
transition period, but would create a fractured system that could be difficult to explain to members and
candidates.
4) Delay implementation of NLRB so that there is no transition period.

The OPTN considered implementing NLRB and allocation changes simultaneously. This option would
address the specific problem of allocation using DSA while scores are based on a different
geographic area.
If the transition period was eliminated, candidates would transition from their current scores to MMaT-
based scores under the new allocation sequence; this would result in some candidates receiving
offers based on exception scores assigned under the old scoring system. The RRB system and
associated exception scores were created to work in an allocation system based on DSAs and OPTN
regions. They include regional agreements and RRBs, both designed to address geographic
variability in access to transplant and clinical practice. The use of MMaT/250 by the NLRB and
associated exception scoring system were designed to work with a system of allocation based on
circles around the donor hospital. When scores assigned under one system are used in the other
system, it creates the potential for unintended disadvantage to certain patients. The OPTN wanted to
avoid this result, so it considered options for converting exception scores if there was a simultaneous
implementation of NLRB and allocation changes.
While most of the existing exception scores could be relatively easily converted by assigning the
standard score listed in policy for the candidate’s diagnosis, approximately 1/3 of the exception
scores are not so easily converted. This other group is comprised of candidates with exception scores
that cannot be readily tied to a standard score. For these, the OPTN considered not converting these
scores, expiring the exception scores early, or creating a method of converting based on the regional
median transplant scores. The OPTN would like public comment on whether non-standard RRB-
generated exception scores should be converted, and by what method.
What is the Committee proposing?

MMaT/DSA during the transition period
The Executive Committee supports assigning exception scores relative to MMaT/DSA during the
transition period.16 This would provide for an interim transition period during which the exception scores
assigned under the RRB system would begin to phase out and new scores assigned relative to MMaT
under the NLRB system would phase in. Once approximately a third of the scores have transitioned, after
approximately one month, the allocation changes would take effect. Then, the OPTN computer system
would automatically replace MMaT/DSA with MMaT/250 and keep each exception candidate’s score that
was assigned during the transition the same relative to MMaT. For instance, if a candidate received a
score under NLRB transition scoring of MMaT/DSA minus 3, it would be automatically converted to
MMaT/250 minus 3 on the day allocation changes take effect. This would ensure aligned scores for this
portion of the exception candidates from the first day of the new allocation system.
16The Executive Committee initially supported this as proposed by the Liver Committee, with a three month transition
that would allow for conversion of most of the exception candidates. Including time for public comment has caused a
delay in implementation of the NLRB, and the transition period will now be approximately one month.
Page 5
The Executive Committee supports this transition plan because the MMaT based exception scoring would
align with the unit of allocation during the transition period. When it was proposed in the DSA and region
system in 2017, it was proposed as MMaT/DSA, and then when it was amended in 2018 for the nautical
mile circle based allocation system, the NLRB scoring was changed to MMaT/250. Reverting to
MMaT/DSA for the transition period (in which the NLRB exists in and DSAs and regionals are used as
units of distribution in the allocation system) would maintain this alignment because both exception
scores and allocation would use DSA during the short transition period. Although the transition period will
no longer be long enough to allow rollover of most of the existing exception scores, a transition period of
approximately one month still allows for transition of approximately a third of the existing exception
candidates prior to the new allocation system being implemented and allows for some time in between
implementation of NLRB and allocation so that there is time to allow the community to adjust to the
revised processes related to submission and review of exception forms.
In order for similarly situated liver transplant candidates within the same DSA to receive similar offers
during the transition period, the OPTN proposes using MMaT/DSA on initial implementation of the scoring
changes, and MMaT/250 only when the allocation changes are also implemented.
Which populations are impacted by this proposal?

There were a total of 1221 candidates on the liver waiting list on January 25, 2019 with a MELD
exception, 181 (15%) of which are in the four affected DSAs. 17 Of those candidates, approximately 1/3
would be expected to be due for an extension of their exception score during the transition period. 18 This
subset of these candidates requiring extensions during this transition period and any candidates that
might apply for a MELD exception in these DSAs will experience more equal access to transplant during
the transition period. This is the population directly targeted by this proposal.
Because MMaT/DSA would be used for all exception scores awarded during the transition period,
candidates at transplant hospitals outside of the four affected DSAs who are awarded exception scores
during the transition period may also be impacted because they may experience an additional change in
their exception score. Of the 145 active liver transplant programs in the country with MELD score
candidates:
 62 (43%) have the same MMaT, whether calculated as MMaT/DSA or MMaT/250.
 83 (57%) will have a different MMaT/250 than MMaT/DSA.
This means that, if this proposal is approved, individual candidates’ exception scores assigned during the
transition period are likely to change when the transition period ends. Because the MMaT/DSA scores will
only be used during the transition period, some of these candidates will see their scores change when
allocation changes are implemented and their scores are converted from MMaT/DSA to MMaT/250.
Most of the liver transplants (approximately 63%) under current allocation take place within the DSA.
Another approximately 31% of transplanted organs are accepted and transplanted within the region.19 In
some regions, there may be more variation in scores between DSAs during the transition period if this
proposal is adopted. Changing to MMaT/DSA scores in the same region for the transition period could
impact a portion of the 31% of organs allocated within the region, but outside of the DSA during the
transition period.
17 There were 14 MELD exception candidates in LAOP, 69 in MAOB, 63 in MNOP, and 35 in TXSB currently waiting
on January 25, 2019. This count represents an approximate number of exception candidates on the liver waitlist on a
given day. OPTN data.
18 Exception scores must be renewed every 3 months. If the transition period is approximately one month, it can be
assumed that approximately 1/3 of the existing exception scores will come due for renewal in that time.
19 In 2016, 63% of liver transplants were from donors in the DSA, 30.6% were from within the region, and 6.4% were
from within the nation. In 2017, 62.5% were from donors in the DSA, 32.3% were from within the region, and 5.4%
were from within the nation. In 2018, 63.2% were from donors in the DSA, 31% were from within the region, and 5.7%
were from within the nation. OPTN data.
Page 6
Because candidates under the age of 12 will be awarded exception pediatric model for end-stage liver
disease (PELD) scores based on the national median PELD at transplant (MPaT), they will not be directly
impacted by this proposal. There is a slight change to the calculation of MPaT proposed, to exclude
organs recovered outside of the region instead of outside 500NM during the transition period to align with
the allocation units and the intent of excluding nationally allocated livers.
How does this proposal impact the OPTN Strategic

Plan?
1. Increase the number of transplants: There is no impact to this goal.
2. Improve equity in access to transplants: This proposal will improve equity in access to transplants
for some liver candidates by ensuring that candidates with MELD exceptions granted at the same
time for the same diagnosis in the same DSA will have the same access to livers allocated at the
DSA level during the transition period. This proposal may also reduce equity in access to
transplant among similarly situated candidates in different DSAs within a region during the
transition period.
3. Improve waitlisted patient, living donor, and transplant recipient outcomes: There is no impact to
this goal.
4. Promote living donor and transplant recipient safety: There is no impact to this goal.
5. Promote the efficient management of the OPTN: There is no impact to this goal.
How does this proposal comply with the Final Rule?

The Final Rule requires that allocation policies “(1) Shall be based on sound medical judgment; (2) Shall
seek to achieve the best use of donated organs; (3) Shall preserve the ability of a transplant program to
decline an offer of an organ or not to use the organ for the potential recipient in accordance with
§121.7(b)(4)(d) and (e); (4) Shall be specific for each organ type or combination of organ types to be
transplanted into a transplant candidate; (5) Shall be designed to avoid wasting organs, to avoid futile
transplants, to promote patient access to transplantation, and to promote the efficient management of
organ placement;…(8) Shall not be based on the candidate's place of residence or place of listing, except
to the extent required by paragraphs (a)(1)-(5) of this section.”20
1. Sound Medical Judgment: The Committee proposes the transition period based on sound
medical judgment. The materials provided include descriptive data concerning the impact of the
proposed options.
2. Best Use of Donated Organs: The Committee believes that offering organs to the most medically
urgent candidates first is the best use of donated livers. This policy seeks to make the best use of
donated organs by granting similar access to exception candidates with the same diagnosis listed
at the same time in the same DSA by assigning them an equal MELD exception score during the
transition period, while livers are still allocated based on DSA.
3. Preserve Ability to Decline an Offer: This does not affect the ability of a transplant program to
decline an offer.
4. Specific to Organ Type: This is specific to an organ type. In this case, the proposed policy is
specific to the allocation of deceased donor livers.
5. Avoid Wasting Organs: Organs are wasted when a transplantable organ is not transplanted. This
policy would not result in any increase in organs that would not be transplanted.
Avoid Futile Transplants: A futile transplant may occur if a recipient is transplanted with an organ
that does not continue to function soon after transplantation. This proposed policy change does
not incentivize futile transplants.
Promote Patient Access to Transplantation: This proposal promotes liver candidate access to
transplant by ensuring that candidates requesting an exception score for the same degree of
illness at the same time in the same first unit of allocation have the same MELD exception score
20 21 C.F.R. §121.8(a).
Page 7
during the transition period. This achieves more equity in access to transplant for these liver
candidates during this transition period.
Promote Efficient Management of Organ Placement: A proposal that reduces logistical
complications associated with procuring an organ and transporting it from the donor to the
candidate promotes efficient management of organ placement. This proposal would have no
impact on the efficient management of organ placement. Candidates within the same DSA have
similar logistical complications, and changing the order in which they appear is unlikely to affect
efficient management of organ placement.
8. Geographic Considerations: A policy may be based in a candidate’s residence or place of listing
only to the extent required to achieve the considerations listed above. This proposal seeks to
reduce the geographic considerations in the placement of livers during the interim transition
period that is expected to end on April 30, 2019 with the implementation of allocation changes.
Using MMaT/DSA will tie exception scores to the first allocation unit while the first unit of
allocation is still DSA. Although geography will be considered, it is required in order to promote
access to transplant for similarly situated exception candidates, because it will treat candidates at
all of the hospitals in the DSA the same, instead of treating candidates differently based on
individual listing program. Doing so therefore also adheres strictly to the Final Rule, which
specifically cites the candidate’s “place of listing” as a geographic constraint to avoid. Using
MMaT/DSA as a temporary geographic limitation is also required to ensure the best use of the
organ, offering organs to the sickest candidates first. Additionally, the temporary use of
MMaT/DSA during the approximately one month transition period conforms to the requirement to
use geography “only to the extent required,” and shall expire upon the implementation of
allocation changes.
The Final Rule provides for transition procedures to “treat people on the waiting list and awaiting
transplantation prior to the adoption or effective date of the revised policies no less favorably than they
would have been treated under the previous policies.” 21,22 The OPTN is proposing a transition, aimed at
equalizing the impact of the transition on candidates within the same DSA. Currently, liver candidates are
awarded exception scores based on their disease severity and regional agreements. Therefore, similar
candidates with exception scores in the same DSA or the same region can expect to be awarded similar
exception scores. This proposal would create a transition provision that would ensure that similar
candidates in the same DSA are awarded similar exception scores.
The Final Rule requires that the Board “[p]rovide opportunity for the OPTN membership and other
interested parties to comment on proposed policies and shall take into account the comments received in
developing and adopting policies for implementation by the OPTN”.23 The letter from the Administrator
provided clarification that this is a “policy change that should be subject to additional public comment
before it is implemented.”24 Given that this change to exception scoring for the transition period requires a
change to policy, the OPTN proposes this solution for consideration during public comment.
How will the OPTN implement this proposal?

This proposal will require calculation of the MMaT for each DSA during the transition period, in addition to
the recalculation of MMaT/250 to convert existing MMaT based exception scores upon implementation
and programming of the new allocation policy (including distribution based on nautical mile distances) in
UNetSM. It will require rapid communication to members to explain the change to the policy and the impact
21 21 C.F.R. 121.8(d).
22 The provision for creating transition protections was reiterated in the letter from Administrator Sigounas, which
stated, “The OPTN may also implement transition patient protections. See 42 CFR 121.8(d)(1) (providing that when
the OPTN revises organ allocation policies, it shall consider whether to adopt transition procedures that would treat
people on the waiting list and await (sic) transplantation prior to the adoption or effective date of the revised policies
no less favorably than they would have been treated under the previous policies). Of course, the OPTN will also have
opportunities to refine, modify, and improve any OPTN liver policy.” George Sigounas, letter to Sue Dunn, OPTN
President, July 31, 2018.
23 42 C.F.R. 121.4(b)(1).
24 George Sigounas, letter to Sue Dunn, January 30, 2019.
Page 8
on patients whose scores may be different. The expected timeline for implementation of the National Liver
Review Board has been adjusted to allow for public comment on this proposal. This proposal is not
expected to change the planned implementation date of April 30, 2019 for the remainder of Liver and
Intestine Distribution Using Distance from Donor Hospital.
How will members implement this proposal?

If this proposal is approved, transplant hospitals may need to review their MMaT/DSA and MMaT/250 and
evaluate whether they need to educate their patients about any changes in score they may experience.
Some exception candidates will experience two score changes within an approximately one month
period. Candidates who apply for a new exception or an extension of an exception score after
implementation of NLRB and prior to the implementation of allocation changes will be awarded a new
MMaT/DSA score during the transition period, and then that score will be converted to a MMaT/250
based score on the day that allocation changes are implemented (targeted April 30, 2019).
Will this proposal require members to submit additional data?

No, this proposal does not require additional data collection.
How will members be evaluated for compliance with

this proposal?
This will not change the current routine monitoring of members.
Page 9
Policy or Bylaws Language

Proposed new language is underlined (example) and language that is proposed for removal is struck
through (example).
9.4 MELD or PELD Score Exceptions
9.4.D Calculation of Median MELD or PELD at Transplant
Median MELD at transplant (MMaT) is calculated by using the median of the MELD scores at the time of
transplant of all recipients at least 12 years old who were transplanted at hospitals within 250 nautical
miles of the candidate’s listing hospital the DSA of the candidate’s transplant hospital in the last 365 days.
Median PELD at transplant (MPaT) is calculated by using the median of the PELD scores at the time of
transplant of all recipients less than 12 years old in the nation.
The MMaT and MPaT calculations exclude recipients who are either of the following:
1. Transplanted with livers from living donors, DCD donors, and donors from donor hospitals more than
500 nautical miles away from outside the region of the transplant hospital
2. Status 1A or 1B at the time of transplant.
The OPTN Contractor will recalculate the MMaT and MPaT every 180 days using the previous 365-day
cohort. If there have been fewer than 10 qualifying transplants within 250 nautical miles of a transplant
hospital the DSA of the candidate’s transplant hospital in the previous 365 days, the MMaT will be
calculated based on the previous 730 days.
Exceptions scores will be updated to reflect changes in MMaT or MPaT each time the MMaT or MPaT is
recalculated. The following exception scores are not awarded relative to MMaT or MPaT and will not be
updated:
1. Exception scores of 40 or higher awarded by the NLRB according to Policy 9.4.A: MELD or PELD
Score Exception Requests
2. Any exception awarded according to Policy 9.5.D: Requirements for Hepatic Artery Thrombosis
(HAT) MELD Score Exceptions
3. Exceptions awarded to candidates less than 18 years old at time of registration according to Policy
9.5.I: Requirements for Hepatocellular Carcinoma (HCC) MELD or PELD Score Exceptions
4. Initial and first exceptions awarded to candidates at least 18 at time of registration according to Policy
9.5.I: Requirements for Hepatocellular Carcinoma (HCC) MELD or PELD Score Exceptions
Effective pending notice to OPTN members and implementation of the National Liver Review Board, and
shall expire upon the implementation of the Liver and Intestine Distribution Using Distance from Donor
Hospital proposal, adopted by OPTN/UNOS Board of Directors December 3, 2018 and subsequent notice
to OPTN members.
Page 10
ORIGINAL ARTICLE CROOME ET AL.
Intraregional Model for End-Stage

Liver Disease Score Variation
in Liver Transplantation: Disparity
in Our Own Backyard
Kristopher P. Croome , David D. Lee, Justin M. Burns, Andrew P. Keaveny,
and C. Burcin Taner
Department of Transplant, Mayo Clinic Florida, Jacksonville, FL
Variation in average Model for End-Stage Liver Disease (MELD) score at liver transplantation (LT) by United Network
for Organ Sharing (UNOS) regions is well documented. The present study aimed to investigate MELD variation at the
interregional, intraregional, and intra–donation service area (DSA) levels. Patients undergoing LT between 2015 and 2016
were obtained from the UNOS standard analysis and research file. The distribution of allocation MELD score including
median, skew, and kurtosis was examined for all transplant programs. Intraregional median allocation MELD varied sig-
nificantly within all 11 UNOS regions. The largest variation between programs was seen in region 5 (MELD 24.0 versus
38.5) and region 3 (MELD 20.5 versus 32.0). Regions 1, 5, and 9 had the largest proportion of programs with a highly
negative skewed MELD score (50%, 57%, and 57%, respectively), whereas regions 3, 6, 10, and 11 did not have any pro-
grams with a highly negative skew. MELD score distribution was also examined in programs located in the same DSA,
where no barriers exist and theoretically no significant difference in allocation should be observed. The largest DSA varia-
tion in median allocation MELD score was seen in NYRT-OP1 LiveOnNY (MELD score variation 11), AZOB-OP1
Donor Network of Arizona (MELD score variation 11), MAOB-OP1 New England Organ Bank (MELD score variation
9), and TXGC-OP1 LifeGift Organ Donation Ctr (MELD score variation 9). In conclusion, the present study demon-
strates that this MELD disparity is not only present at the interregional level but can be seen within regions and even
within DSAs between programs located as close as several city blocks away. Although organ availability likely accounts for
a component of this disparity, the present study suggests that transplant center behavior may also play a significant role.
Liver Transplantation 24 488–496 2018 AASLD.

Received November 21, 2017; accepted January 11, 2018.
This variation can be demonstrated through a patient
SEE EDITORIAL ON PAGE 459 with a MELD score 38 facing a 90-day probability of
wait-list death ranging from 14% in some donation ser-
Since 2002, liver transplantation (LT) candidates have vice areas (DSAs) to 82% in others, with a 90-day proba-
been prioritized on the waiting list according to their bility of undergoing a LT ranging from 18% to 86%
Model for End-Stage Liver Disease (MELD) scores. based on geographic location.(3)
Previous studies have demonstrated variation in average The 11 UNOS regions are not monolithic, however.
MELD score at transplant and wait-list outcomes by Previous studies have demonstrated that MELD score
United Network for Organ Sharing (UNOS) region.(1-3) variation exists at both the interregional and intrare-
gional levels. In 1 study, MELD score at LT varied by
more than 10 points between DSAs and more than 7
Abbreviations: DRI, donor risk index; DSA, donation service area;
LT, liver transplantation; MELD, Model for End-Stage Liver Dis- points between regions.(4) In addition, significant vari-
ease; OPO, organ procurement organization; SC, single center; ability in MELD score can be seen between centers
UNOS, United Network for Organ Sharing. within the same DSA, where no geographic barriers in
Address reprint requests to C. Burcin Taner, M.D., Department of Trans- organ allocation exist. This variability alludes to differ-
plant, Mayo Clinic Florida, 4500 San Pablo Road, Jacksonville, FL ent transplant center practice patterns in addition to
32224. Telephone: 904-956-3261; E-mail: taner.burcin@mayo.edu
“organ availability” as an influence on median MELD
score at transplantation.
488 | ORIGINAL ARTICLE
LIVER TRANSPLANTATION, Vol. 24, No. 4, 2018 CROOME ET AL.

When comparing MELD score between programs,
using only means/averages ignores a great deal of infor-
mation about the practice pattern of that institution.
Although standard criteria organs are usually allocated
and accepted by the highest MELD score recipient on
the list, marginal or extended criteria organs may be
available for a broader number of recipients with lower
MELD scores if they have been declined by multiple
centers. Previous publications have suggested that stea-
totic, elderly, and donation after cardiac death livers
can have good results if used for appropriate recipients
with lower biological MELD scores.(5-8) With this in
mind, a transplant center using marginal organs for
their lower MELD recipients after they have been
declined by multiple other centers will have a lower
mean MELD score than a program who does not use FIG. 1. Categorization of skew and kurtosis.
these organs, despite a similar MELD score require-
ment for standard criteria organs. In this way, averag-
ing a program’s MELD score does not fully represent allocation MELD scores between transplant programs.
the availability of organs at any given transplant center In doing so, we sought to not only further our under-
and unfairly represents the utilization of any given standing of the variability between UNOS regions, but
organ by the transplant center. Understanding the dis- also the large variability in allocation MELD within
tribution of MELD score at transplant may provide regions and within DSAs.
more insight into the transplant center’s true availabil-
ity and utilization of organs. Skew and kurtosis are 2
ways in which a distribution can be described. Skew- Patients and Methods
ness provides a measure of the symmetry of data,
After approval from the Mayo Clinic institutional
whereas kurtosis is a measure of the “tailedness.” These
review board, data were obtained and extracted from
metrics can be used to assess the MELD score distri-
the UNOS standard analysis and research file. The
bution of patients transplanted at a particular center
study population included all adult LTs performed in
and can provide a better representation of any trans-
the United States from January 1, 2015, to December
plant center’s access to organs.
31, 2016.
The present study aimed to investigate intraregional
The distribution of allocation MELD score was
variability and to compare the distribution and skew of
examined for all transplant programs. If skewness was
less than 21 or greater than 11, the distribution was
categorized as highly skewed. If skewness was between
21 and 21=2 or between 11=2 and 11, the distribution
Kristopher P. Croome, David D. Lee, Andrew P. Keaveny, and C.
was categorized as moderately skewed and if skewness
Burcin Taner participated in research design. Kristopher P. Croome, was between 21=2 and 11=2, the distribution was cate-
David D. Lee, and C. Burcin Taner participated in data analysis. gorized as approximately symmetric (normal; Fig. 1).(9)
Kristopher P. Croome and David D. Lee participated in the perfor-
mance of the research. Kristopher P. Croome, David D. Lee, Justin M.
Kurtosis was defined in terms of the central peak in
Burns, Andrew P. Keaveny, and C. Burcin Taner participated in the comparison to a normal distribution (kurtosis 5 3 or
writing of the article. mesokurtic). A distribution with kurtosis <3 (excess
Copyright V
C 2018 by the American Association for the Study of Liver kurtosis <0) was called platykurtic. Compared with a
Diseases. normal distribution, its tails are shorter and thinner,
View this article online at wileyonlinelibrary.com. and often its central peak is lower and broader. A dis-
tribution with kurtosis >3 (excess kurtosis >0) was
DOI 10.1002/lt.25021
called leptokurtic. Compared with a normal distribu-
Potential conflict of interest: Nothing to report. tion, its tails are longer and fatter, and often its central
peak is higher and sharper (Fig. 1).(10)
ORIGINAL ARTICLE | 489

CROOME ET AL. LIVER TRANSPLANTATION, April 2018
Wait-list mortality was calculated as rate per 100

person-years. Deaths following removal from the wait-
Results
ing list for reasons other than transplant, transfer, or Between January 1, 2015, and December 31, 2016, a
recovery, and before any subsequent transplant, were total of 13,417 adult LTs were performed in the United
included. Donor risk index (DRI) was calculated using States. Median allocation MELD score at transplant
previously defined methods.(11) All statistical analyses varied significantly between the UNOS regions
were performed using STATA, version 12 (Stata (P < 0.001). The highest region was region 5 (MELD
Corp., College Station, TX). Differences between 5 34), and the lowest region was region 10 (MELD 5
groups were analyzed using the unpaired t test for 24; Fig. 2). Intraregional median allocation MELD also
continuous variables and by the v2 test or continuity varied significantly within all 11 UNOS regions. The
correction method for categorical variables. Wilcoxon largest variation between programs was seen in region 5
rank sum was used for variables that did not display a (MELD 24 versus 38.5) and region 3 (MELD 20.5
normal distribution. All statistical tests were 2-sided, versus 32). Biological median MELD scores at trans-
and differences were considered significant when plant had much less variability between the UNOS
P < 0.05. regions. The highest region was region 5 (MELD 5

FIG. 2. Allocation and calculated MELD scores in each respective UNOS region.



On linear regression larger program size was associ-
ated with lower median MELD score (P 5 0.02),
higher DRI (P 5 0.001), a lower median MELD score
than the regional median (P 5 0.03), and lower
median MELD score than the DSA median (P 5
0.001), respectively.
The distribution of allocation MELD scores for
each transplant program was evaluated for skew and
kurtosis. Overall nationally, allocation MELD score
distribution showed a moderate negative skew and was
platykurtic (Fig. 3). Regions 1, 5, and 9 had the largest
proportion of programs with a highly negative skewed
MELD score and leptokurtosis (50%, 57%, and 57%,
respectively), whereas regions 3, 6, 10, and 11 did not
have any programs with a highly negative skew (Table
1). In the highest MELD regions (5 and 9), a wide
FIG. 3. Aggregate distribution of allocation MELD score at
transplant for the United States (moderately negative skew and range of MELD score distributions were seen, with
platykurtic). some programs being highly negatively skewed and

leptokurtic, whereas others were normally distributed
and platykurtic (Fig. 4). Programs with a positively
skewed MELD score were only seen in regions 3, 10,
and 11. The majority of programs with a positive skew
25), whereas the lowest were regions 1 and 9 (MELD were single-center organ procurement organizations
5 18). The percent of MELD exception cases under- (OPOs; 77%; Fig. 5).
going transplantation also varied significantly by region, MELD score distribution was also examined in
with the highest being regions 6 and 9 (46.6% and programs located in the same DSA, where no barriers
46.2%) and the lowest being regions 3, 10, and 11 exist and theoretically no significant difference in allo-
(28.0%, 25.6%, and 28.5%, respectively). Median pro- cation should be observed (Table 2). The largest DSA
gram wait-list mortality was highest in region 7 (19.3 variation in median allocation MELD score was seen
per 100 patient years) and lowest in region 9 (8.8 per in NYRT-OP1 LiveOnNY (MELD score variation
100 patient years). There was no correlation between 11), AZOB-OP1 Donor Network of Arizona
wait-list mortality and median allocation MELD score (MELD score variation 11), MAOB-OP1 New
at transplant (P 5 0.21). Wait-list mortality was England Organ Bank (MELD score variation 9), and
inversely correlated with the number of exception cases TXGC-OP1 LifeGift Organ Donation Ctr (MELD
undergoing transplantation (P 5 0.048). score variation 9).
TABLE 1. Skewness of MELD in Transplant Programs in Each Respective UNOS Region

Skewness of MELD (Number of Programs)
Highly Moderate Moderate Highly
Region Negative Negative Normal Positive Positive
Region 1 3 1 2 0 0
Region 2 2 4 9 0 0
Region 3 0 0 9 5* 1*
Region 4 2 2 8 0 0
Region 5 8 2 4 0 0
Region 6 0 0 5 0 0
Region 7 2 2 6 0 0
Region 8 1 3 5 0 0
Region 9 4 1 2 0 0
Region 10 0 0 5 3 (2*) 0
Region 11 0 0 6 4 (2*) 0
*Single-center OPOs.


FIG. 4. (A) Example of the MELD distribution of a transplant program in a high MELD region with a median DRI of 1.3 (highly
negatively skewed and leptokurtic). (B) Example of the MELD distribution of a transplant program in the same high MELD region
with a median DRI of 1.7 (normal distribution and platykurtic).

Discussion Regional differences in allocation MELD score are

likely multifactorial (organ availability, use of exception
Disparity in allocation MELD score across the United scores, wait-list practices, organ acceptance behavior)
States continues to be a major issue in the field of and have evolved over time. Interestingly, a higher pro-
LT.(12) Most discussions of MELD disparity focus on portion of MELD score exceptions were observed in
differences between the 11 UNOS regions viewed as the regions with the highest median MELD scores at
conglomerate averages. The present study demon- LT, whereas a lower proportion was observed in
strates that there was a large variability between trans- regions with the lowest median MELD scores at LT.
plant program median allocation MELD score at all Previous authors have introduced the term “MELD
levels of allocation during the study period. inflation” to describe the influence of MELD excep-
The present study supports prior publications that tions on median MELD score over time and have sug-
have demonstrated regional variation in allocation gested that the primary driver of MELD inflation is
MELD score.(1-3) Region 5 had the highest median the upgrade of MELD points given to exception
allocation MELD score (MELD, 34), whereas region candidates.(13,14) In the present study, there was no
10 had the lowest MELD score (MELD, 24). correlation between wait-list mortality and median


FIG. 5. Example of the MELD distribution of a transplant program in a single-center OPO (highly positively skewed and
leptokurtic).

allocation MELD score, although there was an inverse volume programs and were less likely to be seen in
correlation between wait-list mortality and the propor- single-center OPOs. The present study demonstrated
tion of MELD exceptions. that higher-volume programs were more likely to have
The present study also demonstrated a large amount a median allocation score lower than both the regional
of intraregional variation in median MELD scores. All and DSA median. The association between transplant
11 UNOS regions had programs that had median volume and aggressiveness is likely related to surgeon
MELD scores at transplant of 25 (normally distrib- attitude and experience or institutional experience and
uted and platykurtic), and all regions except 10 and 11 protocols for certain higher-risk organs. In addition,
had transplant programs with median MELD scores given the current regulatory climate, low-volume
of 30 (negatively skewed and leptokurtic). When centers may be more significantly affected by a single
individual program allocation MELD score distribu- bad outcome and as such be less inclined to use these
tions were investigated, regions 1, 5, and 9 had a larger organs.(16) Although the current study was not de-
proportion of programs with highly negatively skewed signed to investigate the influence of program out-
distributions toward higher MELD scores. Programs comes on median allocation MELD score, this
within these regions that displayed a normal MELD likely represents an area that should be studied further.
score distribution had significantly higher median The natural inclination of many programs suffering
donor DRI, likely accounting for their ability to trans- from inferior outcomes may be to become more
plant patients with lower MELD scores despite being “conservative” in organ acceptance. However, this can
in a high MELD score region. These “aggressive” pro- potentially increase expected outcomes and have a
grams appeared to be preferentially using higher DRI potentially counterproductive effect.
livers in their lower MELD score patients and as such Programs with a positive skew in their MELD score
were able to transplant patients across a range of distribution (primarily lower MELD score patients)
MELD scores despite the high median MELDs in were seen almost exclusively in single-center OPOs.
their region. Previous authors have evaluated individual Programs with a positive skew were only transplanting
transplant program organ acceptance behavior and a narrow range of patients around a lower median
assigned an aggressiveness score based on relative utili- MELD score. With the exception of organs that are
zation of higher-risk livers.(15) Although there is no offered for Share 35, these programs have the ability to
universal definition of a higher-risk liver donor, this transplant whatever patient they feel is suitable when
study used components of the widely accepted DRI as there is an organ offer within their DSA.
well as elevated LFTs and Public Health Service An interesting finding of the present study was
increased risk. Aggressive centers were generally high- that a large variability was seen between programs

TABLE 2. Variation in Median Allocation MELD Score Between Programs in Each Respective OPO
Region and OPO Median MELD Score Range
Region 1
MAOB-OP1 New England Organ Bank 9
CTOP-OP1 LifeChoice Donor Services SC
Region 2
PATF-OP1 Center for Organ Recovery and Educ. 8.5
PADV-OP1 Gift of Life Donor Program 7.5
MDPC-OP1 The Living Legacy Foundation of MD 1
NJTO-OP1 NJ Organ and Tissue Sharing Network 1
DCTC-OP1 Washington Reg Transplant Community SC
Region 3
FLUF-IO1 LifeQuest Organ Recovery Services 7
LAOP-OP1 Louisiana Organ Procurement Agency 1.5
GALL-OP1 LifeLink of Georgia 1
FLMP-OP1 Life Alliance Organ Recovery Agency 0
AROR-OP1 Arkansas Reg. Organ Recovery Agency SC
FLWC-OP1 LifeLink of Florida SC
FLFH-IO1 TransLife SC
MSOP-OP1 Mississippi Organ Recovery Agency SC
ALOB-OP1 Alabama Organ Center SC
PRLL-OP1 LifeLink of Puerto Rico SC
Region 4
TXGC-OP1 LifeGift Organ Donation Ctr 9
TXSB-OP1 Southwest Transplant Alliance 8
TXSA-OP1 Texas Organ Sharing Alliance 3
OKOP-OP1 LifeShare Transplant Donor Svcs of OK 3
Region 5
AZOB-OP1 Donor Network of Arizona 11
UTOP-OP1 Intermountain Donor Services 4
CASD-IO1 Lifesharing - A Donate Life Org. 3
CAOP-OP1 OneLegacy 2
CADN-OP1 Donor Network West 1
Region 6
WALC-OP1 LifeCenter Northwest 4
ORUO-IO1 Pacific NW Transplant Bank 1.5
HIOP-OP1 Legacy of Life Hawaii SC
Region 7
ILIP-OP1 Gift of Hope 6
WIDN-OP1 Wisconsin Donor Network 4
MNOP-OP1 LifeSource Upper Midwest OPO 2.5
WIUW-IO1 UW Health Organ and Tissue Donation SC
Region 8
CORS-OP1 Donor Alliance 3
MWOB-OP1 Midwest Transplant Network 3
MOMA-OP1 Mid-America Transplant Svcs 1
NEOR-OP1 Nebraska Organ Recovery System SC
IAOP-OP1 Iowa Donor Network SC
Region 9
NYRT-OP1 LiveOnNY 11*
NYFL-IO1 Finger Lakes Donor Recovery Network SC*
Region 10
OHLB-OP1 LifeBanc 5
MIOP-OP1 Gift of Life Michigan 3
OHLP-OP1 Lifeline of Ohio SC
OHOV-OP1 LifeCenter Organ Donor Network SC
INOP-OP1 Indiana Donor Network SC
Region 11
VATB-OP1 LifeNet Health 3
KYDA-OP1 KY Organ Donor Affiliates 0
NCNC-OP1 Carolina Donor Services 0
SCOP-OP1 LifePoint, Inc. SC
TNDS-OP1 Tennessee Donor Svcs SC
TNMS-OP1 Mid-South Transplant Foundation SC
NCCM-IO1 LifeShare of the Carolinas SC
*Regional Sharing Variance in Place.

SC 5 Single center.
located in the same DSA, where there are no barriers DSA and intraregional levels because such a program
to organ allocation and therefore no disparity should would have a lower median MELD score than a pro-
exist. The largest intra-DSA disparity was seen in gram who does not use these organs, despite a similar
NYRT-OP1 LiveOnNY (MELD score variation 11), MELD score requirement for standard criteria organs.
AZOB-OP1 Donor Network of Arizona (MELD There is no question that disparity in median
score variation 11), MAOB-OP1 New England MELD score at LT exists in the United States. The
Organ Bank (MELD score variation 9), and TXGC- present study demonstrates that this disparity was not
OP1 LifeGift Organ Donation Ctr (MELD score only present at the interregional level but was seen
variation 9). This disparity highlights the effects of within regions and even within DSAs between pro-
transplant program behavior on median MELD score grams located as close as several city blocks away.
at transplant. Again, in these DSAs with the largest Although organ availability likely accounts for a com-
disparity, some programs displayed a highly negative ponent of this disparity, the present study suggests that
skewed leptokurtic distribution, whereas others dis- transplant center behavior may have also play a signifi-
played normally distributed platykurtic distributions. cant role in the differences noted.
It should be noted that a large range in median
MELD was found in region 9, where a variance
exists so that all livers are shared at the regional and REFERENCES
not DSA level. In that region, all programs had a 1) Washburn K, Pomfret E, Roberts J. Liver allocation and distri-
median MELD score of 33-34 with the exception of bution: possible next steps. Liver Transpl 2011;17:1005-1012.
2 programs that had median MELD scores of 23 2) Gentry SE, Massie AB, Cheek SW, Lentine KL, Chow EH,
Wickliffe CE, et al. Addressing geographic disparities in liver
and 25, respectively. These lower MELD score pro- transplantation through redistricting. Am J Transplant 2013;13:
grams both displayed normal, platykurtic distributions 2052-2058.
suggesting that they were transplanting patients with 3) Massie AB, Caffo B, Gentry SE, Hall EC, Axelrod DA,
Lentine KL, et al. MELD exceptions and rates of waiting list
a wide range of MELD scores.
outcomes. Am J Transplant 2011;11:2362-2371.
One question arising from this study is as follows: 4) Yeh H, Smoot E, Schoenfeld DA, Markmann JF. Geographic
How much variability in transplant center practice and inequity in access to livers for transplantation. Transplantation
median MELD score is inevitable? Higher DRI organs 2011;91:479-486.
5) Laing RW, Scalera I, Isaac J, Mergental H, Mirza DF, Hodson
result in increased resource utilization and cost.(17,18)
J, et al. Liver transplantation using grafts from donors after circu-
Different transplant programs may have access to differ- latory death: a propensity score-matched study from a single cen-
ent resources that affect their ability to accept higher ter. Am J Transplant 2016;16:1795-1804.
DRI organs. Mandating programs to accept higher 6) Croome KP, Lee DD, Perry DK, Burns JM, Nguyen JH, Keaveny
AP, Taner CB. Comparison of longterm outcomes and quality of
DRI livers may be unsafe as well as threaten the avail- life in recipients of donation after cardiac death liver grafts with a
able resources for a given program. The tragedy of the propensity-matched cohort. Liver Transpl 2017;23:342-351.
national opioid epidemic has at least partially resulted in 7) Barbier L, Cesaretti M, Dondero F, Cauchy F, Khoy-Ear L,
record increases in the number of organ donors recently. Aoyagi T, et al. Liver transplantation with older donors: a com-
parison with younger donors in a context of organ shortage.
Between 2015 and 2016, there was an 11% increase in Transplantation 2016;100:2410-2415.
the number of LTs performed nationally. Because dif- 8) Bertuzzo VR, Cescon M, Odaldi F, Di Laudo M, Cucchetti A,
ferent regions of the country have been affected to vary- Ravaioli M, et al. Actual risk of using very aged donors for unse-
lected liver transplant candidates: a European single-center experi-
ing extents by this opioid crisis, the implications of this ence in the MELD era. Ann Surg 2017;265:388-396.
on MELD score disparity cannot be discounted. 9) Bulmer, MG. Principles of Statistics. Mineola, NY: Dover; 1979.
Analyzing MELD score distribution (skew and kur- 10) Westfall PH. Kurtosis as Peakedness, 1905-2014. R.I.P. Am
tosis), instead of using median MELD score when Stat 2014;68:191-195.
11) Feng S, Goodrich NP, Bragg-Gresham JL, Dykstra DM, Punch JD,
investigating and comparing transplant programs, is a
DebRoy MA, et al. Characteristics associated with liver graft failure:
novel approach with the ability to provide more robust the concept of a donor risk index. Am J Transplant 2006;6:783-790.
information about program behavior. By using these 12) Deshpande R, Hirose R, Mulligan D. Liver allocation and distri-
distribution metrics, the programs with lower median bution: time for a change. Curr Opin Organ Transplant 2017;
22:162-168.
score as a result of using marginal organs for their lower 13) Northup PG, Intagliata NM, Shah NL, Pelletier SJ, Berg CL,
MELD recipients after they have been declined by mul- Argo CK. Excess mortality on the liver transplant waiting list:
tiple other centers can be differentiated from programs unintended policy consequences and Model for End-Stage Liver
mainly transplanting lower MELD score recipients. It Disease (MELD) inflation. Hepatology 2015;61:285-291.
may also help to explain MELD score disparity at the

14) Biggins SW, Feng S. In a MELD-based economy, how can we centers on mortality after liver transplantation. N Engl J Med
fight off inflation? Liver Transpl 2007;13:2-4. 1999;341:2049-2053.
15) Garonzik-Wang JM, James NT, Arendonk KJV, Gupta N, Orandi 17) Jay CL, Lyuksemburg V, Kang R, Preczewski L, Stroupe K, Holl
BJ, Hall EC, et al. The aggressive phenotype revisited: utilization of JL, et al. The increased costs of donation after cardiac death liver
higher-risk liver allografts. Am J Transplant 2013;13:936-942. transplantation: caveat emptor. Ann Surg 2010;251:743-748.
18) Croome KP, Hernandez-Alejandro R, Chandok N. Early allo-
16) Edwards EB, Roberts JP, McBride MA, Schulak JA, Hunsicker
graft dysfunction is associated with excess resource utilization
LG. The effect of the volume of procedures at transplantation
after liver transplantation. Transplant Proc 2013;45:259-264.

TA L K I N G A B O U T T R A N S P L A N TA T I O N
Question s & Answers
Transplant
for
Candidates about
MELD and PELD
U N I T E D N E T WO R K F O R O RGA N S H A R I N G
The United Network for Organ Sharing (UNOS), a non-profit charitable
organization, operates the Organ Procurement and Transplantation Network
(OPTN) under federal contract. On an ongoing basis, the OPTN/UNOS
evaluates new advances and research and adapts these into new policies to best
serve patients waiting for transplants.
As part of this process, the OPTN/UNOS developed a system for prioritizing

candidates waiting for liver transplants based on statistical formulas that are
very accurate for predicting who needs a liver transplant most urgently.
The MELD (Model for End Stage Liver Disease) is used for candidates age 12
and older and the PELD (Pediatric End Stage Liver Disease Model) is used
for patients age 11 and younger.
This document will explain the system and how it affects those needing a transplant.
What is MELD? How will it be used?

The Model for End-Stage Liver Disease (MELD) is a numerical scale, ranging
from 6 (less ill) to 40 (gravely ill), used for liver transplant candidates age 12
and older. It gives each person a ‘score’ (number) based on how urgently he
or she needs a liver transplant within the next three months. The number is
calculated by a formula using three routine lab test results:
■ bilirubin, which measures how effectively the liver excretes bile;

■ INR (prothrombin time), which measures the liver’s ability to make bloo
clotting factors; and
■ creatinine, which measures kidney function. (Impaired kidney function is
often associated with severe liver disease.)
The only priority exception to MELD is a category known as Status 1. Status 1 patients
have acute (sudden and severe onset) liver failure and a life expectancy of
hours to a few days without a transplant. Less than one percent of liver
transplant candidates are in this category. All other liver candidates age 12 and
older are prioritized by the MELD system.
A patient’s score may go up or down over time depending on the status of his
or her liver disease. Most candidates will have their MELD score assessed a
number of times while they are on the waiting list. This will help ensure that
donated livers go to the patients in greatest need at that moment.
What is PELD? How does it differ from MELD?

Candidates age 11 and younger are placed in categories according to the Pediatric
End-stage Liver Disease (PELD) scoring system. Again there is a Status 1 category
for highly urgent patients, representing about one percent of those listed.
1
All other candidates in this age range receive priority through PELD.
PELD is similar to MELD but uses some different factors to recognize the specific
growth and development needs of children. PELD scores may also range higher
or lower than the range of MELD scores. The measures used are as follows:
■ bilirubin, which measures how effectively the liver excretes bile;

■ INR (prothrombin time), which measures the liver’s ability to make blood
clotting factors;
■ albumin, which measures the liver’s ability to maintain nutrition;
■ growth failure; and
■ whether the child is less than one year old.
As with MELD, a patient’s score may go up or down over time depending on the
status of his or her disease. Most candidates will have their PELD score as-
sessed a number of times while they are on the waiting list. This will help en-
sure that donated livers go to the patients in greatest need at that moment.
What Led To the MELD/PELD System?

Until 2002, patients needing liver transplants were grouped into four medical
urgency categories. The categories were based on a scoring system that in-
cluded some laboratory test results and some symptoms of liver disease.
One concern with using symptoms in scoring was that different doctors might
interpret the severity of those symptoms in different ways. In addition, this
scoring system could not easily identify which patients had more severe liver
disease and were in greater need of a transplant.
Research showed that MELD and PELD accurately predict most liver patients’
short-term risk of death without a transplant. The MELD and PELD formulas
are simple, objective and verifiable, and yield consistent results whenever the
score is calculated.
OPTN/UNOS committees developed the liver policy based on MELD and PELD,
with key support from transplant patient/family advocates. It was approved by
the OPTN/UNOS Board of Directors in November 2001 and went into effect in
February 2002.
How are livers offered through MELD and PELD?

Livers are offered first to urgent and compatible patients in the donor’s local
area (often defined as a state or large metropolitan area), then to a larger region
of the country (the OPTN/UNOS has 11 allocation regions in the U.S.), then
nationwide. Because Status 1 candidates are most medically urgent, each liver
is first offered to local Status 1 candidates, then regional Status 1 candidates.
The sequence of offers after Status 1 patients depends on the donor’s age.
2
If the donor is younger than 18, after any Status 1s are considered the liver
would next be offered to candidates in the region age 11 or younger. The
organ would then be considered for local and then regional candidates with
a MELD of 15 or higher. Any patients age 12 to 17 would be considered
ahead of adult patients.
If the donor is 18 or older, the liver would be offered first to local and re-
gional Status 1 candidates. If not accepted for any of these patients, the liver
is then offered to candidates with a MELD/PELD score of 15 or higher, first
locally and then regionally.
If the liver is not matched to any candidates with a MELD/PELD of 15 or

higher, it may then be considered for local, then regional candidates with a
MELD/PELD of 14 or less. Finally the liver would be offered for any compatible
candidates nationwide, beginning with Status 1 candidates and then to those
with the highest MELD/PELD scores.
How is waiting time counted in the system?

Various studies report that waiting time is a poor indicator of how urgently a
patient needs a liver transplant. This is because some patients are listed for a
transplant very early in their disease, while others are listed only when they
become much sicker.
Under the MELD/PELD system with a wide range of scores, waiting time is
not often used to break ties. Waiting time will only determine who comes first
when there are two or more patients with the same blood type and the same
MELD or PELD score.
If a patient’s MELD or PELD score increases over time, only the waiting time
at the higher level will count. (For example, if you have waited 40 days with
a score of 12, and 5 days with a score of 15, you would only get credit for 5
days of waiting time at the score of 15.) However, if the patient’s MELD or
PELD score decreases again, he or she would keep the waiting time gained at
the higher score. (Using the earlier example, if your score goes from 12 to 15
and back to 12, you would have 45 days of waiting time at the score of 12.)
Patients initially listed as a Status 1 also keep their waiting time if their con-
dition improves and they later receive a MELD/PELD score.
Patients with higher MELD/PELD scores will always be considered before

those with lower scores, even if some patients with lower scores have waited
longer. (For example, a patient waiting for one day with a score of 30 will
come ahead of a patient with a score of 29, even if the patient with a 29 has
waited longer. This is because the patient with a score of 30 has a higher risk
of dying without a transplant.)
3
What if I had been on the waiting list when the system changed?
For the first year of the MELD and PELD system, a transition plan was in place
for liver candidates awaiting a transplant at the time of the change to maintain
their priority gained under the previous policy. After the transition period
ended, those candidates still awaiting a transplant were prioritized only by
their MELD or PELD score.
If you were listed for transplant at the time the policy became effective (Febru-
ary 27, 2002), your transplant team would best be able to address any ques-
tions about the transition and/or your current medical status.
All patients newly listed for a liver transplant since February 2002 receive
priority only by their MELD or PELD score, based on their current lab results.
Do MELD and PELD account for all conditions?

MELD/PELD scores reflect the medical need of most liver transplant candidates.
However, there may be special exceptions for patients with medical conditions
not covered by MELD and PELD. If your transplant team believes your case
qualifies for an exception, they may submit information to a review board
and request a higher score. The review board will consider the medical facts
and determine whether or not to grant a higher score.
Is this system likely to change?

Liver allocation policy based on MELD and PELD has already been changed
as transplant professionals have applied and learned from the system, and fu-
ture changes will likely be required to better meet patients’ needs. In fact, this
system is designed to be flexible and allow improvements. In transplantation, as in
all scientific fields, new studies are taking place all the time to learn how to
save more lives and help people live longer and better.
What if I have more questions?

If you have any further questions or concerns, you should contact your
transplant team for further information. Additional details about the OPTN,
UNOS, allocation policy and patient informational resources are available on
the following websites:
www.optn.org • www.unos.org • www.transplantliving.org
4
The UNOS mission is to advance organ availability and transplantation
by uniting and supporting its communities for the benefit of patients
through education, technology and policy development.
P.O. Box 2484, Richmond, VA 23218

www.unos.org 03.08
Medicine, Health Care and Philosophy
https://doi.org/10.1007/s11019-018-9845-y
SHORT COMMUNICATION
Toward an accelerated adoption of data-driven findings in medicine
Research, skepticism, and the need to speed up public visibility of data-driven findings
Uri Kartoun1
© Springer Nature B.V. 2018
Abstract
To accelerate the adoption of a new method with a high potential to replace or extend an existing, presumably less accurate,
medical scoring system, evaluation should begin days after the new concept is presented publicly, not years or even decades
later. Metaphorically speaking, as chameleons capable of quickly changing colors to help their bodies adjust to changes in
temperature or light, health-care decision makers should be capable of more quickly evaluating new data-driven insights and
tools and should integrate the highest performing ones into national and international care systems. Doing so is essential,
because it will truly save the lives of many individuals.
Keywords Clinical informatics · Prediction modeling · Electronic medical records · Machine-learning · Data-mining ·
Cirrhosis · Liver transplantation
Throughout history, skepticism has played an important role NAFLD is actually a clinical condition. An NAFLD diag-
in evaluating a variety of phenomena. In medicine, some nosis has important health and clinical implications because
scientists have occasionally been dismissed as irrational only it is a risk factor for the development of diseases such as
to be proven right many years later. For instance, Galen, a type 2 diabetes mellitus and an independent risk factor for
second-century philosopher and physician, believed that the cardiovascular-related mortality and all-cause mortality
liver was the source of all veins and the principle organ for (Musso et al. 2011; Byrne and Targher 2015). Nonalcoholic
blood production (ElMaghawry et al. 2014). Though most steatohepatitis, the progressive form of NAFLD, can result
of Galen’s writings were incorrect, people still held strong in cirrhosis and hepatocellular carcinoma and is estimated
to his beliefs even 1500 years later. Dr. William Harvey was to become the leading indication for liver transplant in the
the first to describe blood circulation to the heart, brain, and United States by 2020 (Charlton 2008).
body in detail. In 1628, in his book, De Motu Cordis (On the Recent remarkable advancements in computer hardware
Motion of the Heart and Blood), describing the structure of and software and the growing accessibility of electronic
the heart and arteries, he posited for the first time that blood medical records (EMRs) have accelerated research on pre-
passed through the heart, not the liver as previously believed. dicting patient outcomes. Such advances have allowed the
Harvey’s findings were ridiculed, and many doctors in the rapid development of massive-scale predictive models—
seventeenth century noted that they would “rather err with powerful resources to study disease complications at the
Galen than proclaim the truth with Harvey.” (Bushak 2015). population level. Such models have proved highly useful
Another example of skepticism in medicine concerns non- to discovering or confirming disease correlations, sub-cat-
alcoholic fatty liver disease (NAFLD). Until a few decades egories of diseases, and adverse drug events. The model of
ago, the scientific community was undecided about whether the end-stage liver disease (MELD) risk score, for instance,
is one of the most important and widely used risk predic-
tion scores in medicine. Unlike in the case of other scores,
* Uri Kartoun a patient’s MELD score may indicate the likelihood of a
uri.kartoun@ibm.com major clinical event for the patient. MELD determines the
1 patient’s rank on the organ allocation waiting list; notably,
Center for Computational Health, IBM Research,
Cambridge, MA, USA since 2002, MELD has played a crucial role in determining
13
Vol.:(0123456789)
U. Kartoun
which patient on a waiting list will be the next to receive a as they are today. Future risk scores will likely be composed
liver transplant (Kamath and Kim 2007). of tens of thousands of patient characteristics and be calcu-
Combining the ability to store and rapidly process the lated automatically as an integrated component of an EMR
records of millions of individuals by accessing the reposi- system to provide real-time decision support to monitor a
tories of Massachusetts General Hospital (MGH), Brigham disease or to prioritize organ transplant candidacy.
and Women’s Hospital (BWH), and the IBM Explorys Plat- Finally, we faced criticism that several of the variables
form using machine-learning algorithms has helped us cre- that we used (all selected by a feature-selection algorithm)
ate a new and highly accurate score to predict short-term were associated with cardiovascular risk rather than liver-
mortality in cirrhosis patients (Kartoun et al. 2017). We related mortality. Strikingly, researchers from the Cleveland
took an unbiased approach to the discovery of biomarkers. Clinic validated another of our liver-related studies in which
In this approach, we filtered a large collection of medical we strengthened the existing knowledge and discovered new
records through a feature-selection algorithm and identi- biomarkers regarding the interplay between cardiovascular
fied a small set of variables that could serve as the most risk and liver disease. Both studies were published in The
efficient predictors for a given medical outcome. We used American Journal of Gastroenterology (Corey et al. 2016;
the traditional supervised-learning paradigm to assess accu- Mehta et al. 2016). Medical publications that describe unbi-
racy and applied standard statistical methods to assess the ased approaches to feature selection for developing new
validity of our approach. We realized that combining the scores or to classifying diseases more accurately are rare. A
components of MELD with several easily accessible vari- few, however, have been published, including, for instance,
ables would enable us to construct a new score that would a prediction model for 30-day readmission for heart fail-
be approximately 10% more accurate. We named our new ure patients (Kartoun et al. 2015) and models to classify
score MELD-Plus. MELD-Plus is an attempt to create a new rheumatoid arthritis (Liao et al. 2010), Crohn’s disease, and
mortality prediction risk score in cirrhosis. Our unbiased ulcerative colitis (Ananthakrishnan et al. 2013). Although
data-driven approach, which involves the use of an algo- we were not criticized explicitly for favoring a data-driven
rithm to select predicting variables as well as the large and unbiased approach rather than relying on domain expertise,
independent databases used for validation, makes our score it could have been our use of an approach not yet broadly
a useful tool that could truly save lives. Furthermore, the accepted that have raised further criticism.
fact that MELD-Plus’s variables are available for any patient Furthermore, our approach also relied on a new text-
(including total bilirubin, creatinine, albumin, INR, WBC, processing method that we developed to accurately extract
sodium, total cholesterol, length of stay, and age) makes it concepts from clinical narrative notes. The method, text
easy to calculate the patient’s mortality risk using Excel or nailing (TN), raised skepticism in reviewers of medical
to deploy on any digital health repository. informatics journals who claimed that TN “relies on simple
Our preceding manuscript drafts, in which we outlined a tricks to simplify the text,” and “leans heavily on human
better scoring system than MELD, raised significant skepti- annotation.” TN indeed may seem just like a trick of the
cism from reviewers and editors. Although we were invited light at first glance, but it is actually a fairly sophisticated
to present our earlier findings at a medical informatics method that finally caught the attention of more adventurous
conference (Kartoun et al. 2016), leading medical journals reviewers and editors who ultimately accepted it for publica-
repeatedly criticized our work. The criticism always had a tion (Kartoun 2017a, b). We found TN to be highly accurate,
reasonable rationale, but our findings and the proposition for outperforming traditional machine-learning algorithms in
an alternative score did not change throughout our resubmis- multiple scenarios, such as extracting family history of coro-
sions and were, therefore, kept out of the public eye. Eventu- nary artery disease (Corey et al. 2016), classifying patients
ally we successfully published our study in October 2017 in with sleep disorders (Beam et al. 2017; Kartoun et al. 2018),
PLoS ONE, a peer-reviewed journal. and improving the accuracy of the Framingham risk score
The main criticism of our initial manuscript was valid: for patients with NAFLD (Simon et al. 2017).
until 2016 we had access to only one source of data (MGH/ As for the historical trajectory of adopting liver alloca-
BWH), and our claim of generalizability was indeed weak. tion scores, in 1998, the Committee on Organ Procurement
Another criticism was that the interest in such scores might and Transplantation Policy of the Institute of Medicine (cur-
be limited to individual clinicians who are making decisions. rently called “The National Academy of Medicine”) pub-
This claim, however, may rule out the usefulness of any lished “The Final Rule,” calling for “…standardized medi-
other type of risk score as well. Another concern was that cal criteria to be used to determine the status of a person’s
our predictive model contained too many variables, reducing illness and when that person can be placed on a waiting
its practicability in day-to-day use. Such criticism was valid list” and further stating “…Minimum listing criteria for
if powerful computers capable of instantaneously processing including transplant candidates on the national list shall
tremendous collections of EMRs were not in such broad use, be standardized and, to the extent possible, shall contain
13
explicit thresholds for listing a patient and be expressed broader usefulness beyond mortality prediction (Tudoroiu
through objective and measurable medical criteria” (Insti- et al. 2018).
tute of Medicine 1999). Independently, scientists reported The adoption of MELD-Na would had been faster if the
in 2000 on a well-validated model and on the creation of a scientific community had been able to publish convinc-
new equation to calculate survival probabilities for patients ing studies earlier to assess the contribution of sodium to
following a transjugular intrahepatic portosystemic shunt MELD. Organizations such as The American Medical Infor-
placement (Malinchoc et al. 2000). In February 27, 2002, mation Association (AMIA) have encouraged universities as
this equation was selected to serve as the basis for the new well as commercial companies to form “Challenges,” such
allocation policy (Freeman et al. 2002). The equation, form- as the de-identification and the smoking status challenges
ing MELD, has become the standard by which priorities (Uzuner et al. 2007, 2008). Such challenges have resulted
are determined in donor liver allocation, and as expected, in a variety of high-impact papers that have significantly
implementation of MELD led to an immediate reduction in enhanced the medical informatics subdomain, as well as the
liver transplant waiting list registrations for the first time in entire health-care domain. If UNOS had worked more col-
the history of liver transplantation (with a 12% decrease in laboratively with AMIA, as well as with The Institute of
2002) (Kamath and Kim 2007). In subsequent years, multi- Electrical and Electronics Engineers (IEEE)’s Engineering
ple studies proposed that the incorporation of sodium into in Medicine and Biology Society, new challenges could have
the original MELD equation could significantly improve been formed, with titles such as “The MELD-Plus Chal-
prediction accuracy for liver disease. For instance, a study lenge” or “The Liver Disease Challenge,” inviting investiga-
published in the New England Journal of Medicine in 2008 tors from all around the globe to assess current scores and
estimated that using an extended version of MELD, one that propose new scores that might even outperform MELD-Plus.
incorporated serum sodium levels, would save 90 lives in the Additional associations, such as the Association for Comput-
period from 2005 to 2006 (Kim et al. 2008). Additional stud- ing Machinery (ACM), might be encouraged to be involved
ies supported the usefulness of sodium to improve prediction in such efforts focused on computational assessments of
performance for liver disease (Ruf et al. 2005; Londoño et al. health. Such initiatives could help accelerate the adoption of
2007; Luca et al. 2007). The MELD-Na score, an equation health-related data-driven findings, as these challenges are
that incorporates sodium into MELD, was finally adopted in expected to produce scientific papers faster and thus support
2014 (Mulligan and Hirose 2014). or rule out the usefulness of the newest findings.
Why did it take many years to adopt MELD-Na, a score In a desirable future scenario, UNOS may decide to
that was created by using a data-driven approach, instead replace MELD (or its subsequent score, MELD-Na) with
of starting to use it, say, in 2008, right after multiple stud- MELD-Plus or even with more advanced futuristic scores
ies demonstrated the advantage of using sodium to improve that may be developed by other researchers that incorpo-
the prediction accuracy of MELD? The lives of hundreds rate, for instance, additional behavioral and genetic aspects.
would have been saved if MELD-Na was in use starting in Hypothetically, we can imagine a patient a decade from now
2008 rather than in 2014. The reason for the delay was most who is in need of a liver replacement. That patient might feel
likely to let the scientific community assess and discuss encouraged if MELD-Plus was in use, determining more
further the combination’s potential usefulness as well as its accurately his or her rank on the waiting list. MELD-Plus
drawbacks, a consideration undertaken by a large number will not cure that patient, of course, but its ability to assess
of independent investigators and through the use of patient the severity of a condition more precisely could mean that
data captured at multiple health systems. Only after broad the patient might wait 2 months less for a new liver than if
scientific evidence had been accumulated, was the United the original MELD was in use. MELD-Plus, therefore, could
Network for Organ Sharing (UNOS) convinced to extend save the patient’s life.
MELD to MELD-Na. Furthermore, UNOS estimated that On the one hand, skeptics are often proven wrong as
MELD-Na was expected to save between 50 and 60 lives per science advances. For instance, it took years for the main-
year (Mulligan and Hirose 2014), and relevant to MELD- stream scientific community to accept Harvey’s contri-
Plus, our experiments demonstrate that while MELD-Na butions over Galen’s. On the other hand, skepticism in
performed slightly better than MELD, MELD-Plus per- medicine is essential, especially regarding questionable
formed significantly better than MELD (> 10% better) (Kar- treatments and methods and the potential effects of using
toun et al. 2017). Thus, MELD-Plus, if incorporated into new medicines. Advances in medicine that have raised
hospital systems, could save hundreds of patients every year significant skepticism include, for example, a human
in the United States alone. Furthermore, as an encourag- head transplant operation proposed by neurosurgeon Dr.
ing first step toward adoption, a very recent study reported Sergio Canavero (former director of the Turin Advanced
that MELD-Plus plays a predictive role in the occurrence of Neuromodulation Group, Italy) or a new approach to slow
post-liver transplantation acute kidney injury, proposing a the progression of Alzheimer’s disease proposed by Dr.
13
U. Kartoun
Dale Bredesen (University of California, Los Angeles). IBM does not alter his adherence to all Medicine, Health Care and
The development of new risk scores, by contrast, and Philosophy policies.
especially those that are based on components of similar
widely used scores, such as MELD, should not be inter-
preted as questionable and thus should be expected to References
raise only minor levels of skepticism. Regardless of any
Ananthakrishnan, A. N., T. Cai, G. Savova, S. C. Cheng, P. Chen, R.
specific disease, when a new score is introduced publicly
G. Perez, V. S. Gainer, S. N. Murphy, P. Szolovits, Z. Xia, S.
in a peer-reviewed scientific journal, the scientific com- Shaw, S. Churchill, E. W. Karlson, I. Kohane, R. M. Plenge, and
munity would benefit from the availability of mechanisms K. P. Liao. 2013. Improving case definition of Crohn’s disease
that could evaluate the scores more rapidly, considering and ulcerative colitis in electronic medical records using natural
language processing: A novel informatics approach. Inflammatory
data derived from multiple health-care systems. Espe-
Bowel Diseases 19 (7): 1411–1420.
cially, official organizations that rely on the scores (e.g., Beam, A. L., U. Kartoun, J. K. Pai, A. K. Chatterjee, T. P. Fitzgerald,
the American Diabetes Association, the American Heart S. Y. Shaw, and I. S. Kohane. 2017. Predictive modeling of phy-
Association, UNOS) would benefit from such mechanisms. sician-patient dynamics that influence sleep medication prescrip-
tions and clinical decision-making. Scientific Reports 7: 42282.
Combining the most advanced data-driven algorithms with
Bushak, L. 2015. Mad Scientist: 6 scientists who were dismissed as
human expertise is expected to achieve a desirable increase crazy, only to be proven right years later. Medical Daily, Nov
in knowledge, and this could potentially affect decisions to 18, 2015.
either replace or extend current scoring methods. Incorpo- Byrne, C. D., and G. Targher. 2015. NAFLD: A multisystem disease.
Journal of Hepatology 62: S47–S64.
rating data-driven scores adjusted by human expertise is
Charlton, M. 2008. Cirrhosis and liver failure in nonalcoholic fatty
essential, especially within the context of medical ethics, liver disease: Molehill or mountain? Hepatology 47: 1431–1433.
and will help in deciding which research findings may be Corey, K. E., U. Kartoun, H. Zheng, R. T. Chung, and S. Y. Shaw.
worth accelerating and what safeguards need to be pro- 2016. Using an electronic medical records database to identify
nontraditional cardiovascular risk factors in nonalcoholic fatty
vided. For instance, a better scoring system might result
liver disease. The American Journal of Gastroenterology 111 (5):
in some patients being pushed up the queue for transplan- 671–676.
tation, but it would result in other patients being pushed ElMaghawry, M., A. Zanatta, and F. Zampieri. 2014. The discovery of
down. Additional components coming into play regarding pulmonary circulation: From Imhotep to William Harvey. Global
Cardiology Science and Practice 2014(2):103–116.
medical scoring systems must consider the ability to assess
Freeman, R. B. Jr, R. H. Wiesner, A. Harper, S. V. McDiarmid, J. Lake,
the true risk of learning algorithms (Kartoun 2018) and E. Edwards, R. Merion, R. Wolfe, J. Turcotte, and L. Teperman.
the potential for sociological biases at the individual level, 2002. UNOS/OPTN Liver Disease Severity Score, UNOS/OPTN
as well as at the social level. An algorithm favoring a can- Liver and Intestine, and UNOS/OPTN Pediatric Transplantation
Committees. The new liver allocation system: Moving toward
didate for transplant based on his or her political views,
evidence-based transplantation policy. Liver Transplantation 8
family status, or sexual preference are just a few examples (9): 851–858.
for such potential biases. Institute of Medicine. 1999. Analysis of waiting times. In Organ pro-
To accelerate the adoption of a new method with a high curement and transplantation: Assessing current policies and the
potential impact of the DHHS final rule, ed. Committee on Organ
potential to replace or extend an existing, presumably less
Procurement, and Transplantation Policy, 57–58. Washington,
accurate, medical scoring system, evaluation should begin D.C.: National Academy Press.
days after the new concept is presented publicly, not years Kamath, P. S., and W. R. Kim. 2007. The model for end-stage liver
or even decades later. Metaphorically speaking, as cha- disease (MELD). Hepatology 45 (3): 797–805.
Kartoun, U. 2017a. Text nailing: An efficient human-in-the-loop text-
meleons capable of quickly changing colors to help their
processing method 2017. ACM Interactions 24 (6): 44–49.
bodies adjust to changes in temperature or light, health- Kartoun, U. 2017b. Beyond brute force. Communications of the ACM
care decision makers should be capable of more quickly 60 (10): 8–9.
evaluating new data-driven insights and tools and should Kartoun, U. 2018. Toward an equation that anticipates AI risks. Com-
munications of the ACM 61 (2): 8–9.
integrate the highest performing ones into national and
Kartoun, U., V. Kumar, A. Brettman, S. Yu, K. Liao, E. Karlson, A.
international care systems. Doing so is essential, because Ananthakrishnan, Z. Xia, V. Gainer, A. Cagan, S. Murphy, S.
it will truly save the lives of many individuals. Churchill, I. Kohane, P. Szolovits, T. Cai, and S. Y. Shaw. 2015. A
risk model for 30-day heart failure re-admission using electronic
Funding The author received honoraria and travel funding from The medical records. 2015 Joint Summits on Translational Science,
American Association for the Study of Liver Diseases (October 2017). San Francisco, CA.
Kartoun, U., K. Corey, H. Zheng, and S. Shaw. 2016. A prediction
model to assess mortality risk in cirrhosis. 2016 Joint Summits
Compliance with ethical standards on Translational Science, San Francisco, CA.
Kartoun, U., K. Corey, T. Simon, H. Zheng, R. Aggarwal, K. Ng, and
Conflict of interest The author has declared that no competing inter- S. Shaw. 2017. The MELD-Plus: A generalizable prediction risk
ests exist. The author confirms that the commercial affiliation with score in cirrhosis. PLoS ONE 12 (10): e0186301. https://doi.
org/10.1371/journal.pone.0186301. (eCollection 2017).
13
Kartoun, U., R. Aggarwal, A. Beam, J. Pai, A. Chatterjee, T. Fitzgerald, correlates with fibrosis and hepatocyte ballooning. The American
I. Kohane, and S. Shaw. 2018. Development of an algorithm to Journal of Gastroenterology 111 (12): 1853–1854.
identify patients with physician-documented insomnia. Scientific Mulligan, D. C., and R. Hirose. 2014. OPTN/UNOS, Liver and Intes-
Reports 8: 7862. tinal Organ Transplantation Committee. Report to the Board of
Kim, W. R., S. W. Biggins, W. K. Kremers, R. H. Wiesner, P. S. Directors. June 23–24, 2014, Richmond, Virginia.
Kamath, J. T. Benson, E. Edwards, and T. M. Therneau. 2008. Musso, G., R. Gambino, M. Cassader, and G. Pagano. 2011. Meta-
Hyponatremia and mortality among patients on the liver-trans- analysis: Natural history of non-alcoholic fatty liver disease
plant waiting list. New England Journal of Medicine 359 (10): (NAFLD) and diagnostic accuracy of noninvasive tests for liver
1018–1026. disease severity. Annals of Medicine 43: 617–649.
Liao, K. P., T. Cai, V. Gainer, S. Goryachev, Q. Zeng-treitler, S. Ray- Ruf, A. E., W. K. Kremers, L. L. Chavez, V. I. Descalzi, L. G. Podesta,
chaudhuri, P. Szolovits, S. Churchill, S. Murphy, I. Kohane, E. and F. G. Villamil. 2005. Addition of serum sodium into the
W. Karlson, and R. M. Plenge. 2010. Electronic medical records MELD score predicts waiting list mortality better than MELD
for discovery research in rheumatoid arthritis. Arthritis Care & alone. Liver Transplantation 11 (3): 336–343.
Research 62 (8): 1120–1127. Simon, T. G., U. Kartoun, H. Zheng, A. T. Chan, R. T. Chung, S. Shaw,
Londoño, M. C., A. Cárdenas, M. Guevara, L. Quintó, D. de Las Heras, and K. E. Corey. 2017. MELD-Na score predicts incident major
M. Navasa, A. Rimola, J. C. Garcia-Valdecasas, V. Arroyo, and P. cardiovascular events, in patients with nonalcoholic fatty liver
Ginès. 2007. MELD score and serum sodium in the prediction of disease. Hepatology Communications 1 (5): 429–438.
survival of patients with cirrhosis awaiting liver transplantation. Tudoroiu, M. I., G. Constantin, L. Pâslaru, S. Iacob, C. Gheorghe, I.
Gut 56 (9): 1283–1290. Popescu, D. Tomescu, and L. S. Gheorghe. 2018. The combina-
Luca, A., B. Angermayr, G. Bertolini, F. Koenig, G. Vizzini, M. Ploner, tion of serum cystatin C, urinary kidney injury molecule-1 and
M. Peck-Radosavljevic, B. Gridelli, and J. Bosch. 2007. An inte- MELD-Plus score predicts early acute kidney injury after liver
grated MELD model including serum sodium and age improves transplantation. Surgery, Gastroenterology and Oncology 23 (2):
the prediction of early mortality in patients with cirrhosis. Liver 121–126.
Transplantation 13 (8): 1174–1180. Uzuner, O., Y. Luo, and P. Szolovits. 2007. Evaluating the state-of-
Malinchoc, M., P. S. Kamath, F. D. Gordon, C. J. Peine, J. Rank, and the-art in automatic de-identification. Journal of the American
P. C. ter Borg. 2000. A model to predict poor survival in patients Medical Informatics Association 14 (5): 550–563.
undergoing transjugular intrahepatic portosystemic shunts. Hepa- Uzuner, O., I. Goldstein, Y. Luo, and I. Kohane. 2008. Identifying
tology 31 (4): 864–871. patient smoking status from medical discharge records. Journal
Mehta, N., T. Singh, R. Lopez, and N. Alkhouri. 2016. The heart age of the American Medical Informatics Association 15 (1): 14–24.
is increased in patients with nonalcoholic fatty liver disease and
13

MELD Related Files Liver MELD-Plus

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

MELD Related Files Liver MELD-Plus

Uploaded by

Copyright:

Available Formats

134

MELD scoring system is useful for predicting prognosis in

Table 1 Clinical and biochemical characteristics of the 129 cirrhotic patients

Sex (M/F) 70/28 25/6 NS

Data are median (range).

Table 2 Clinical and biochemical characteristics of the 129 cirrhotic patients

Sex (M/F) 86/31 9/3 NS

Data are median (range).

Child-Pugh score 8 92 72 0.824 Child-Pugh score 8 61 75 0.691

Questions and Answers for Transplant Candidates about

All transplant programs and organ procurement organizations throughout

As part of this process, UNOS developed a system for prioritizing

What is MELD? How is it used?

• bilirubin, which measures how effectively the liver excretes bile

A patient’s score may go up or down over time depending on the status of

What is PELD? How does it differ from MELD?

• bilirubin, which measures how effectively the liver excretes bile

As with MELD, a patient’s score may go up or down over time depending

How are livers allocated?

• the donor’s age

• candidates with MELD/PELD scores 35 and higher within the donor‘s

If a combined liver-intestine is being offered, candidates waiting for a

Liver offer process for donors 0-10 years of age

Is waiting time counted in the sysytem?

Under the MELD/PELD system with a wide range of scores, waiting

Do MELD and PELD account for all conditions?

Is this system likely to change?

For more information

UNOS’ Patient Services phone line (888-894-6361) can provide information

Our mission is to unite and strengthen the donation

700 North 4th Street, Richmond, VA 23219

Copyright © 2019 United Network for Organ Sharing 104 7.19

MELD Exception Scores during

Prepared by: Elizabeth Miller

MELD Exception Scores during

Is the sponsoring Committee requesting specific

What problem will this proposal address?

Why should you support this proposal?

How was this proposal developed?

https://optn.transplant.hrsa.gov/media/2329/liver_boardreport_201712.pdf (accessed January 20, 2019).

Table 1: DSAs in which MMaT within 250NM of Liver

and Intestinal Transplantation Committee Conference Call, January 17, 2018.

Other options considered

1) Leave MMaT/250 during the transition period. (2018 Board approved)

2) National MMaT during the transition period.

3) Manually adjust MMaT for the affected transplant hospitals.

4) Delay implementation of NLRB so that there is no transition period.

What is the Committee proposing?

Which populations are impacted by this proposal?

How does this proposal impact the OPTN Strategic

How does this proposal comply with the Final Rule?

How will the OPTN implement this proposal?

How will members implement this proposal?

Will this proposal require members to submit additional data?

How will members be evaluated for compliance with

Policy or Bylaws Language

Intraregional Model for End-Stage

Liver Transplantation 24 488–496 2018 AASLD.

ORIGINAL ARTICLE | 489

Wait-list mortality was calculated as rate per 100

490 | ORIGINAL ARTICLE

TABLE 1. Skewness of MELD in Transplant Programs in Each Respective UNOS Region

ORIGINAL ARTICLE | 491

Discussion Regional differences in allocation MELD score are