Liver Committee Minutes Plus

OPTN/UNOS Liver and Intestinal Organ Transplantation Committee
Meeting Minutes
November 2, 2018
Chicago, IL
Julie Heimbach, MD, Chair

James Trotter, MD, Vice Chair
Introduction
The OPTN/UNOS Liver and Intestinal Organ (Liver) Transplantation Committee met in Chicago,
Illinois on 11/2/18 to discuss the following agenda items:
1. Operations and Safety Interview Data
2. Review Regional and Society Comments
3. Public Comment Feedback on Broader-2-Circles (B2C) v. Acuity Circles (AC)
4. Public Comment Feedback on Sharing Thresholds and Circle Sizes
5. Public Comment Feedback on Hawaii and Puerto Rico Variances
6. Public Comment Feedback on Pediatric Allocation and Other Issues
7. Vote on Proposal to Send to Board of Directors (BOD)
8. Other Significant Items
The following is a summary of the Committee’s discussions.
1. Operations and Safety Interview Data
Data summary:
The OPTN/UNOS Operations and Safety (Ops and Safety) Committee conducted a
questionnaire of 54 organ procurement organizations (OPOs) and ten transplant hospitals that
asked about each organization’s travel practices. Each organization submitted one response to
the questionnaire.
Summary of discussion:
UNOS staff presented the relevant data from the questionnaire. To conduct the questionnaire,
members of the Ops and Safety Committee interviewed key personnel from OPOs and
transplant hospitals. There is potential bias in the data because the responses are reflective of
only one person’s perspective at each organization.
Some members of the committee questioned the accuracy of the responses. The Chair stated
that the data indicates that there is large variation in the perceptions of OPO and transplant
hospital travel practices.
OPO and transplant hospital interviewees reported a wide range in the distance (in miles and
minutes) that their organization would travel to procure a liver. The responses also showed a
wide range in the percent of livers transported by air across each region. A separate report
prepared by the Scientific Registry of Transplant Recipients (SRTR) showed large variation in
the percent of livers currently flown in each Donation Service Area (DSA). The SRTR report
indicated that the percent of livers currently flown in each region ranges from 30% to 60%.
Areas with no liver program could be flying for 100% of their livers. All of this data indicates that
travel practices vary widely across the nation. In the questionnaire, the median percent of livers
flown by OPOs was approximately 50%. A committee member reminded the group that there
are roughly ten OPOs with their own recovery teams so they can fly donors instead of
procurement teams, which is an important distinction for traveling purposes.
1
The questionnaire also asked about each organization’s experience with the new lung allocation
policy, which went into effect in December 2017 and involved broader distribution. The
respondents reported a variety of experiences with the new policy. Committee members
questioned the usefulness of the questionnaire data because there has been more objective
data collected on the lung allocation model. UNOS staff clarified that the OPTN does not collect
travel or cost data. They also stated that the Ops and Safety Committee quickly put together the
questionnaire to help answer some of the questions about fixed-distance allocation and noted
that there is a need to do more detailed data collection and analysis in the future.
The committee then discussed the data in the six month report for the new lung allocation
model. Although there are differences in the distribution models and respective data for lungs
and livers, the report could be used to justify the decisions made by the Liver Committee
because it shows that there are geographic limits to distribution.
One committee member stated that since implementing the new lung allocation model, his/her
transplant hospital has gone from 30% to 70% imported lungs; the median costs of organ
acquisition went from $35,000 to $70,000; the volume of transplants stayed the same; and the
patient population did not change. Another committee member noted that in New York, the
costs have decreased from $75,000 to $55,000.
The committee discussed the costs associated with importing and exporting livers. A committee
member noted that OPOs can charge different rates for imported organs versus exported
organs. Some committee members also mentioned that OPOs are charging more for imported
and exported organs. However, an OPO administrator on the committee noted that his/her OPO
does not charge more. UNOS staff stated that the increased charge for imported/exported
organs occurs prior to the updating of insurance policies, so it is not a long-term issue.
Next Steps:
The committee will consider the data presented and use it, when possible, as justification.
2. Review Regional and Society Comments
Data summary:
UNOS staff presented data on the public comments submitted regarding the “Liver and Intestine
Distribution Using Distance from Donor Hospital” proposal.
Summary of Discussion:
Most of the comments were submitted by transplant hospitals, patients, and members of the
general public. Many comments came from Texas, New York, California, and South Carolina,
while other states also submitted a significant number of comments. UNOS staff reminded the
committee that the policy making process is not a popular vote.
Approximately 40% of respondents opposed both B2C and AC, approximately 10% supported
both models, approximately 10% supported B2C and opposed AC, and approximately 35%
supported AC and opposed B2C.
UNOS staff presented data on public comment feedback at the state level. There was mixed
support for the two models when looking at framework preference by state, where each state
counts as a single vote. A committee member commented that the data showing the amount of
support in either direction is relevant, but it is more important to consider the substantive and
constructive comments. Nonetheless, it is important to see that there is not unanimous support
in any direction.
2
One committee member stated that he/she believed that people may have thought that B2C
entailed broader distribution than AC because of the use of the word “broader” in the
nomenclature.
UNOS staff informed the committee that public comment closed at 5:00 PM Eastern Time on
November 1, and the data presented was finalized around 10:00 PM on November 1. The short
public comment period and turnaround were due to the committee’s request for additional
modeling. Prior to the meeting, UNOS staff distributed a document with all submitted comments
to the committee members. UNOS staff also reminded the committee that not every respondent
answered every question.
There were some comments that were mistakenly taken off the OPTN website, but they will be
put up. One comment did not meet the OPTN standards. The individual that submitted the
comment was contacted and edited the comment.
UNOS staff presented data on state preferences where each state was weighted by the number
of transplants in the state. From this perspective, there was more support for AC. One
committee member noted that similar states opposed both models and supported B2C. Another
committee member stated that it is important to look at the data this way (where responses are
weighted by number of transplants) because it shows the number of patients that will be
affected.
The Chair presented the feedback from the regional webinars. A committee member noted that
his/her regional webinar did not seem to be well attended and participants were generally
confused by the proposal. The Chair noted that there were fewer comments during the regional
webinars than expected, but this may have been due to the awkwardness of speaking during a
webinar or because the webinars occurred during the work day. Although there has been
significant public feedback overall, there has been less regional feedback on this proposal.
Another committee member noted that regional members did not know that they would be
voting on the proposal during their regional webinar. Committee members were critical of the
regional feedback process for this proposal. In the feedback that was submitted, there was a
low vote count and no clear consensus in most regions.
The Chair then presented the comments from transplant societies.
The American Society of Transplantation (AST) did not support either model, but preferred B2C
with a Model for End-Stage Liver Disease (MELD)/ Pediatric End-Stage Liver Disease (PELD)
threshold of 35 and a pediatric MELD/PELD threshold of 32. They were concerned about flying,
higher costs, increased risk, and more discards. They supported the variance for Puerto Rico.
They wanted additional stratification of pediatrics in modeling and monitoring and were
concerned that the new policy would dis-incentivize splitting.
The American Society of Transplant Surgeons (ASTS) supported building a population-based
model that could be adjusted incrementally to minimize unintended consequences. They also
suggested studying the impact of distribution circles on listed patient survival, transplantation
rates, transplant MELD, as well as the impact of increased travel on success rates and
associated costs.
The Association of Organ Procurement Organizations (AOPO) was generally supportive of the
process and believed that the committee is looking to provide an incremental change that meets
the Final Rule and maximizes utilization. They were concerned that broader sharing will result in
an increased risk of out-of-sequence allocation and that recipient complications will make
sequential reallocation more challenging. The Chair noted that data for out-of-sequence
allocations exists for Share 35 and can be used as justification if needed. AOPO also wanted
3
improvements to DonorNet, some of which are already in place. They supported the pediatric
changes and extending the variance to Puerto Rico.
NATCO preferred B2C at a MELD/PELD threshold of 35 and 150/250/500 nautical mile (nm)
circles. They did not want anyone other than the transplant community to make allocation policy.
They would have preferred that the committee considered population-adjusted circles, which
might have alleviated some concerns about geographic inequities.
Studies of Pediatric Liver Transplantation (SPLIT) supported the pediatric changes.
The law firm representing the plaintiffs in Cruz et al v. U.S. Dept. of Health and Human
Services, (S.D.N.Y 18-CV-06371) submitted a comment stating their belief that B2C prioritizes
geography too much. They also supported AC, although the letter was not favorable in general.
The Chair reiterated that the most important thing the committee can do is to provide
justification for the decisions they make so that their reasoning can be referenced in any future
legal proceedings.
A committee member noted that the litigants used waitlist deaths as part of the rationale for their
opinion, but the SRTR used a different definition for waitlist mortality in their report than what
program directors typically use. The SRTR modeling did not account for candidates that were
removed from the waitlist for being too sick in their waitlist mortality calculation. By using this
definition, the modeling showed a reduction in transplants and a decrease in waitlist deaths in
some areas. The Chair told the committee that even though this definition is different than what
they are used to, SRTR also modeled the current allocation system using this definition so the
comparison between the current system and any of the proposed systems remains pertinent.
The SRTR cannot model the number of people that are removed from the waitlist because the
Liver Simulation Allocation Model (LSAM) does not provide the reason for why candidates are
removed from the waitlist. They do have this data for real patients, but not for simulated
patients. The SRTR could provide modeling on total pre-transplant deaths if the committee
would like, although this probably does not change waitlist outcomes. The SRTR has been
modeling waitlist deaths in this way for a number of years. The committee remained critical of
the SRTR’s definition of waitlist deaths and maintained that it is not the proper definition. In
response to the comment that a decrease in waitlist deaths cannot occur if there are fewer
transplants in a DSA, UNOS staff noted that it could also be the case that transplants are going
to sicker patients, so fewer transplants could translate to a decrease in waitlist deaths. A
committee member also noted that there are many factors that impact waitlist death (besides
MELD) that vary across the nation, including rates of hypertension, diabetes, etc. Because of
this, some places have high rates of transplant but also have high waitlist mortality.
Indiana University Health, The University of Kansas Health System, Vanderbilt University
Medical Center, and Washington University in St. Louis/Barnes-Jewish Hospital Transplant
Center collectively submitted a letter to the committee as public comment. They suggested
using the Cumulative Community Risk Score (CCRS) to measure the impact of the models.
They did not support either model. They focused on the idea that the Final Rule has many other
components that the committee is not paying attention to, such as efficiency, organ wastage,
and decreasing disparity in access. They objected to using median MELD at transplant (MMaT)
in the DSA to measure disparity because DSA should not be used even in this way. In their
view, MELD does not reflect mortality risk.
The committee continued to discuss the issue of waitlist mortality. There are differences in
waitlist mortality for the same MELD/PELD score by region, although the reasons for these
differences is not clear. A committee member stated that the variance in MMaT for non-
exception patients across the country is almost 21 points, which is a large disparity. There was
4
disagreement among the committee members over how exception patients influence the MMaT,
although there was agreement that exception patients are not as impacted by geographic
disparity. There is large variation in non-standard exception scores across the country, which is
why the committee is in the process of implementing a National Liver Review Board (NLRB) as
approved by the OPTN/UNOS BOD in December 2017.
The Chair presented SRTR data on the MMaT by DSA for each OPO. The lowest MMaT for an
OPO is 24 and the highest is 33. There is also wide variation within the same DSA. One
committee member noted that MMaT is an imperfect metric because it is dependent on a
multitude of factors. The committee member suggested using other metrics, such as offer rate
or a candidate’s likelihood of still being on the waitlist after a certain period of time, to measure
program performance.
The Chair reiterated that the way SRTR calculated waitlist mortality was different than the way
waitlist mortality is calculated at the transplant programs. However, the data is still useful in
examining the change between current waitlist mortality and the proposed policies. SRTR sent
the committee members new data that modeled total pre-transplant deaths, as previously
discussed. A committee member also sent a document comparing LSAM waitlist mortality data
to UNOS waitlist mortality data. The LSAM results are not predictions of actual values, but
instead are important in showing the relative change between the different scenarios. The same
idea should be applied to the percent of livers being flown. The modeling should be used to
understand the magnitude of change between scenarios, as the point prediction might be
inaccurate. The Chair ended the discussion by stating that there are differences in how waitlist
mortality is calculated, and the committee must consider these differences going forward.
Next steps:
The committee will consider the data provided and public comments in the policy making
process.
3. Public Comment Feedback on Broader-2-Circles (B2C) v. Acuity Circles (AC)
UNOS staff identified the common themes throughout the public comment period. The Chair
reminded the committee to continue thinking about sources of objective evidence for each
issue.
One common theme was concern about an increase in discarded organs due to more travel.
The Chair stated that the committee might be able use the data from Share 35 to provide
evidence for increased travel causing more discards. Another theme identified in public
comment was concern about additional cold ischemic time (CIT), also due to more traveling.
The Chair stated that current data indicates that after eight hours of CIT, the quality of the liver
decreases. Other committee members disagreed with this statement and instead stated that the
quality of the donor pool is going down and the percent of livers transplanted prior to five hours
of CIT is increasing. A committee member suggested using the decreasing quality of the donor
pool as justification for the concern about increased CIT.
Other themes identified in public comment were concerns about higher costs and increased risk
for recovery teams.
A committee member shared the results of Share 35 after two and three years. The variance in
MMaT increased, the discard rate decreased, there was no change in CIT, post-transplant
survival was not affected, and the number of transplants went up (although this is more likely
due to an increase in donors).
5
A committee member clarified that CIT and transport time are not the same thing. There are
factors beyond transport time that affect CIT. Also, behaviors will change to reduce CIT when
faced with longer transport time. It is also important to recognize that some high quality livers
can have a longer CIT and flying to recover a liver is typically done only when the recovery team
knows they are getting a high quality liver.
A committee member noted that there is work being done to better quantify the increased risk of
flying on procurement teams. The committee member also mentioned that there are many “near
misses” that did not become accidents but should be recognized. The committee also discussed
that more travelling and flying will force programs to utilize more manpower, and the opportunity
cost of sending a procurement team a far distance for a marginal liver may not be worthwhile.
This issue is more important at small programs, who may only have two or three surgeons.
A committee member, who is an OPO administrator, noted that his/her OPO does all of the
recoveries in his/her region and the transplant programs seem to be satisfied with this
arrangement. Another committee member noted that in his/her region, they offer local recovery
for every liver and are almost always turned down. One of the major issues with local recovery
is that the skill level of the recovering surgeon is often unknown and transplant programs do not
like this unknown. The committee suggested using live video technology so that the transplant
program can observe and comment on the organ recovery process, but this is more of a long-
term solution.
The Chair presented common themes in public comment related to disparity. These themes
were:
 Desire to reduce variation in MMaT, especially in non-exception candidates
 Desire to address long wait time in high MMaT areas
 Concerns that disparity creates a system in which patients with the means to move or to
list in another state have a better chance of transplant
 Desire for the sickest patients to get transplanted first
 Belief that organs are a national resource although this belief is not held by everyone
 Desire to prioritize offers to populations that have higher waitlist mortality rate.
 Desire to prioritize offers to populations with a higher incidence of liver disease
 Desire to prioritize offers to populations that have less access to the waitlist
 Belief that MMaT is not appropriate way to measure disparity because MELD doesn’t
predict mortality risk
Other comments suggested ways to address disparity. These comments were:
 Increase transplant rates instead of changing allocation policy
 Belief that median MMaT is driven by listing and acceptance practices
 Belief that the solution to disparity is additional education in areas with lower donation
rates
 Create OPO performance goals
 Education for the general public
The Chair then presented common themes in public comment related to unintended
consequences. The Chair noted that large changes in allocation policy could have many
unintended consequences, a possibility that the committee could use as justification for
whichever model they choose. Common themes in public comment related to unintended
consequences were:
 Concern that additional costs would cause smaller programs to close
 Desire to minimize the amount of change
6
 Concern that donors will be reluctant to donate if their liver will not stay in the local
community
Committee members disagreed on the relevance of the reluctance of individuals to donate if
their liver will not stay local. Some committee members said they had not heard this concern
raised in their DSA, while others had heard it. UNOS staff noted that HRSA has conducted
national surveys related to this concern and, while some people do prefer their organs to stay
local, the surveys have consistently shown that donors and donor families want their organs to
go to the sickest patient first.
A committee member noted that some populations, especially minorities, feel a lack of access to
obtaining a transplant and are therefore less likely to donate. If certain populations do not feel
like they are part of the recipient pool or they mistrust the medical system, then they are less
likely to donate.
The committee discussed the issue of regional variance. The committee must be thoughtful in
how the new allocation model is communicated, especially to the donor community. For
example, it will not resonate with people if the proposal says that places like California are
disadvantaged relative to places in the Southeast, even if this is true in the case of MMaT. Low
donation rates exacerbate this issue. OPO performance is part of the reason for the variance in
donation rate, but the variance is also caused by differences in the cause of death. For
example, the donation rate is higher in the Southeast because more people die from stroke and
at a younger age. Another committee member noted that rural populations across the nation
have less access to the medical system overall. A committee member sympathized with places
with a high prevalence of acute liver disease and suggested focusing on making sure these
people get on the waitlist.
A committee member observed that many of these arguments were the same ones made during
the development of Share 35. However, none of the smaller programs have closed due to Share
35. Additionally, it is true that populations across the nation have different levels of access to the
healthcare system and different rates of certain diseases, but the fact remains that the single
largest determining factor in getting a liver transplant is where the candidate is listed. There
should be efforts to increase access, but the purpose of the committee is to allocate the livers
available to the people on the waitlist. However, the committee does need to consider
unintended consequences and should not disenfranchise any population.
A committee member stated that avoiding unintended consequences is particularly important to
this discussion because the implementation of the NLRB could also lead to unintended
consequences and there are no rescue devices for liver patients. It is also necessary to mitigate
travel, especially for smaller programs who may turn down a marginal liver if they need to fly to
recover it, although it will be difficult to find objective evidence to support this conclusion.
Additionally, the committee member noted that much of the evidence is based on experience,
not objective data. One committee member stated that private insurance reimburses transplant
programs at a higher rate so programs may target privately-insured patients for organs that they
must fly to procure, thus disadvantaging those on the waitlist with public insurance.
The Chair then presented the other common themes in public comment. The other themes
included a desire to decrease waitlist mortality, concern that the process is moving too quickly,
and questioning how there can be a decrease in transplant numbers and a simultaneous
decrease in waitlist mortality.
The committee then started discussing which model they prefer. One committee member asked
why the group had not considered a model using proximity points. If they had time to consider
7
this, they would, but the compressed timeframe has made this impossible. The Chair told the
committee that they can make tweaks to either model based on the data they have.
A committee member commented that there are ways to alter the AC model to help mitigate
travel and this model prioritizes the sickest patients. It is possible to predict what will happen if
the committee elects to add three MELD/PELD points to the inner acuity circle by extrapolating
the data they have been given. This may mitigate some flying.
A committee member noted that most of the transplant community does not understand the AC
model and it is worrisome that the committee would put out a model that most people do not
understand. Another committee member stated that people are influenced by where they are
from and how the new policy affects them. If the committee had more time, they could explain
the policy better.
Before voting, the Chair asked each committee member to express their opinion on the models.
These comments are summarized below:
 The AC model benefits the vast majority of patients. It is not perfect, but the issues of
increased cost and potential program closure can be mitigated. AC does the best job at
reducing MMaT variation. The charge of the committee is to do what is best for people
across the nation.
 B2C is more broadly supported by the transplant community. Increased flying will be a
logistical nightmare with many unintended consequences. The concerns about socio-
economic status are legitimate.
 Support B2C because there is increased risk when flying, as evidenced by the plane
crash that killed a number of this committee member’s colleagues.
 AC is best for patients and the risks can be mitigated.
 People understand B2C better.
 There is a precedent for B2C with lung allocation.
 Region 4 has over 1,500 candidates on the waitlist and more lives will be affected than
almost anywhere else. The 150 nm circle will isolate the people in Houston. The most
important thing is that geography cannot preclude a patient from getting an offer. B2C
will isolate a large number of patients in Texas. It is also necessary to look at the travel
of the donor as well as the procurement team. Modeling shows a 10% increase in
transplants in Houston, but mortality goes up.
 The job of the OPO is to maximize the gift of life that donors are giving, and AC does
this.
 It does not make sense to have fixed-distance circles with variable population densities.
They will need to do more in the future, so it makes sense to go with B2C which is
understandable and is an incremental change.
 Older livers should stay local.
 B2C would result in the least amount of change but it has the ability to be altered based
on results
 AC compensates for population density due to narrowness of the MELD/PELD bands
 National sharing should be at MELD/PELD of 18 instead of 15
A formal vote was taken regarding: which model do you prefer?
Results were as follows: B2C 11 in favor (55%); AC 9 in favor (45%)
8
B2C in favor AC in favor Abstain
Number of votes 11 9 0
Percentage 55% 45% 0%
Next Steps:
The committee will move forward with a proposal for B2C.
4. Public Comment Feedback on Sharing Thresholds and Circle Sizes
The Chair presented information on public comment feedback regarding the sharing threshold
and circle sizes. Most comments supported larger circle sizes with a sharing threshold of
MELD/PELD 29 or smaller circles with a sharing threshold of MELD/PELD 35.
UNOS staff reiterated the responsibility of the committee to show why each of their decisions is
necessary to make the allocation system more efficient as outlined in the Final Rule. The model
that the BOD passes will probably not change for two years so that there is enough time to
analyze the allocation model and appropriately adjust it, provided that there are not major
unintended consequences.
Some of the themes identified in public comment regarding circle size were:
 Concerns about including water
 Concerns about making areas smaller than the current DSA
 Requests for a population-based model
 Requests to treat Alaska differently
There were a number of other circle sizes suggested but none gained much traction. In terms of
justification, the committee recognized 150nm as the agreed upon cutoff for driving versus
flying.
The Chair presented the themes identified in public comment regarding the sharing threshold for
the 250nm circle. These themes were:
 Support for MELD/PELD 35 as this is the closest to the December 2017 proposal
 Support for MELD/PELD 29 because it appears to be an inflection point on the mortality
curve and gives access to more patients within 250nm earlier
 Requests for a threshold of MELD/PELD 15
 Requests for no sharing threshold
SRTR modeled B2C at both MELD/PELD 32 and 35, so it is possible to understand what
MELD/PELD 29 would look like by extrapolating the modeling for MELD/PELD 32 and 35. Some
committee members were resistant to proposing a policy that had not been modeled. The
committee discussed the sharing threshold extensively prior to putting the proposal out for
public comment and there was support for MELD/PELD 29, 32, and 35.
A committee member proposed a new idea that would include candidates with MELD/PELD 35
or more with the Status 1A and 1B candidates in the 500nm circle.
The Chair presented what each of the sharing threshold options would look like. A threshold of
MELD/PELD 35 would lessen distribution, while MELD/PELD 29 would increase distribution.
9
A committee member stated that the MELD/PELD threshold should have biological meaning. A
threshold of MELD/PELD 32 has no biological meaning and would be difficult to justify.
However, MELD/PELD 29 is an inflection point where risk of death starts getting higher. Offers
should go out further for individuals with a higher risk of death.
A committee member noted that another issue is that the NLRB is not yet in place. Once it is
implemented, there will be patients with different exception scores with the same condition. If
the sharing threshold is set at MELD/PELD 29, there will be many patients with unequal
exception scores. There needs to be a cap on exception scores that is lower than the sharing
threshold. This cannot be changed until the exception patients cycle through the NLRB, so there
will be a disparity for three months. A committee member mentioned that hepatocellular
carcinoma (HCC) patients will be transplanted at a higher rate with B2C and the NLRB, but the
Chair stated that this effect is not known.
HRSA clarified what they are requesting from the committee. HRSA does not intend to direct the
OPTN on what the policy should be because there are many factors that must be balanced.
HRSA directed the OPTN to work with the public to create an allocation policy that considers all
of the factors described in the Final Rule. When the committee is setting a geographic
boundary, the committee must provide evidence for why the boundary is required. No policy is
going to be perfect and stand in perpetuity, but the onus is on the OPTN to justify the decisions
the committee makes.
A committee member noted that the policy will be updated and improved over time based on the
data so they should start at MELD/PELD 32 because they have modeling for this threshold.
Another committee member discussed the MELD/PELD cliff that the threshold would create in
B2C. For example with a threshold of MELD/PELD 32, a candidate with a MELD/PELD of 31
who is 155 nm from the donor hospital will be skipped over by a candidate with a MELD/PELD
of 16 that is 145 nm from the donor hospital. This is a 40% difference in mortality risk. A
threshold of MELD/PELD 29 would lower this cliff from MELD/PELD 31 to MELD/PELD 28,
which is still significant but less dramatic.
The Chair presented data on public comment feedback about the sharing threshold. Most
commenters supported MELD/PELD 29, followed by MELD/PELD 35, and finally MELD/PELD
32. The Chair pointed out that more than half of the BOD are not physicians so they will have a
different perspective than the committee. Patients supported the lowest threshold. The
committee discussed the ability of MELD/PELD to predict mortality and there was agreement
that it was valid up to a certain point, but it is not perfect. The committee is tasked with
balancing the threshold with efficiency, especially regarding travel. A committee member noted
that MELD/PELD scores will likely lower with time, but the Chair stated that this is not
necessarily true.
A committee member reiterated his/her proposal to move candidates with MELD/PELD 35 and
above to the 500 nm circle after the Status 1A and 1B candidates. In this proposal, livers would
be allocated to Status 1A and 1B within 500 nm, then MELD/PELD 35 within 500 nm, then
MELD/PELD 32 or 29 within 250 nm and so on. The proposal was seconded. Allocation would
still go to Status 1A and 1B first and then would be ordered by MELD/PELD within the 500 nm
circle. Nothing would change for pediatric candidates. Some committee members were
concerned that this would change the model too much without formal SRTR modeling. Another
committee member suggested that this is similar to AC and they should just go back to AC. The
percent of candidates on the waitlist with MELD/PELD 35 or more is small, so the committee
does not know how much increased travel there will be. The committee member that suggested
the proposal stated that candidates with MELD/PELD 35 or above have similar mortality to
Status 1A and 1B, so it is not possible to justify allocating them lower on the list.
10
A formal vote was taken regarding: do you support this modification?
Results were as follows: 7 (37%) Yes; 12 (63%) No
Yes No Abstain
The committee started discussing the sharing threshold again. One committee member stated
that the justification for a threshold of MELD/PELD 32 is that it is an incremental change that
allows for the NLRB to go into effect and can be studied over time. The Chair informed the
committee that the vote on the threshold will be between a threshold of MELD/PELD 32 and
threshold of MELD/PELD 29. There was no support for MELD/PELD 35. Another committee
member reiterated that MELD/PELD 29 can be justified because it is the inflection point on the
mortality risk curve. Many of the societies are concerned about the 10% increase in flying
associated with a threshold of MELD/PELD 32.
A committee member disagreed with the notion that incremental change can be used as
justification. Incremental change cannot be replaced by the idea that the committee makes no
progress. Incremental change is not what is best for patients.
Another committee member asked if there is any way the AC model can be modified to mitigate
the issues of travel. The committee agreed that it was not possible to do this within the allotted
timeframe. A committee member noted that AC is preferable because it allocates locally first,
thus minimizing travel. Additionally, if the committee alters AC to have proximity points or
different band sizes, then travel could be reduced. The Chair stated that the committee already
chose B2C and they cannot re-vote on B2C versus AC.
UNOS staff clarified the justification that the committee has discussed for each of the
MELD/PELD thresholds. The justification offered for MELD/PELD 32 was that it is an
incremental change and allows for time for the NLRB to be implemented. UNOS staff said that
this justification does not relate back to the Final Rule, and will therefore be difficult to use. A
threshold of MELD/PELD 32 was also modeled to have a 10% increase in travel over the
current system. This can be brought back to the Final Rule, but this modeling has not been
done for MELD/PELD 29. A committee member stated that they had already proposed
MELD/PELD 32 to the community and changing to MELD/PELD 29 could violate the trust of the
community. However, there was a plurality of support for MELD/PELD 29 when looking at public
comment feedback by state weighted by transplant volume. Patients also preferred
MELD/PELD 29. A committee member asked if it would be possible to predict the increase in
travel for a threshold of MELD/PELD 29. The threshold of MELD/PELD 29 would add about 2%
of the waitlist to the 250 nm band. A committee member suggested that proposing something
that has been modeled (MELD/PELD 32) is sufficient justification. However, the Chair said that
there is precedent for bringing something to the board that has not been modeled. A committee
member reminded the group that HRSA is asking for justification on why travel is a relatively
important consideration compared to reducing waitlist mortality. Another committee member
suggested that the inflection point in the mortality risk curve is actually at MELD/PELD 25, which
would only add 5% more patients to the 250 nm circle. However, this is the percentage of
patients on the waitlist but not the percent that gets transplants. The Chair motioned to vote.
A formal vote was taken regarding: do you support a MELD/PELD sharing threshold of 32 or
29?
11
Results were as follows: 11 (55%) in favor of MELD/PELD 29; 9 (45%) in favor of MELD/PELD
32
In favor of In favor of Abstain

MELD/PELD 29 MELD/PELD 32
Following the vote, a committee member put forth a motion that the model proposed by the
committee be implemented at least three months after implementation of the NLRB with a cap
of MELD/PELD 28 for standard exceptions. The motion was seconded. A committee member
noted that it will be necessary to communicate to patients that their MELD/PELD score could go
down due to the new allocation policy. The committee discussed grandfathering candidate’s
MELD scores over for the first three months or until there is time for their MELD scores to be
readjusted. In the motion put forth, the exception cap of MELD/PELD 28 would go into effect the
same day that B2C is implemented, and in the meantime the cap would be MELD/PELD 34. An
exception cap of MELD/PELD 34 is in currently approved but not yet implemented policy. The
committee discussed lowering the cap to MELD/PELD 28 once B2C is implemented.
Implementing B2C three months after the NLRB would allow for time for exception patients to
cycle through the NLRB. The cap at MELD/PELD 28 is necessary because with MMaT minus
three there will still be candidates higher than MELD/PELD 29 in some areas of the country.
UNOS staff clarified that new exception scores could not go into effect all at once because
moving to a model that is based on MMaT is a logistical challenge. The metadata that will be
collected for exception patients when the model is MMaT-based would allow the committee to
transition to a different cap. Policy states that time on the waitlist will be used as the tiebreaker
between patients with the same exception score. There was some confusion as to how the new
exception score would be implemented but the committee agreed to vote on the proposal to
implement B2C three months after the NLRB and have an exception cap at MELD/PELD 28
when B2C is implemented. The committee will figure out the logistics of the motion at a later
time. A committee member clarified that programs will still be allowed to apply for exceptions to
these rules. Another committee member noted that in San Francisco, the MMaT is MELD/PELD
33 so exception patients would get a score of MELD/PELD 30, but this proposal has a cap of
MELD/PELD 28.
A formal vote was taken regarding: do you support implementing B2C at least three months
after implementing the NLRB?
Results were as follows: 14 (74%) Yes; 2 (11%) No; 3 (16%) Abstain
Yes No Abstain
A formal vote was taken regarding: do you support a cap for standard MELD/PELD exceptions
at 28?
Results were as follows: A majority of the committee supported the proposal.
12
UNOS staff clarified that this exception would only be for adult standard exceptions. There
would not be a cap on pediatric exceptions.
A committee member suggested keeping organs from donors after cardiac death (DCD) that are
age 60 and older within 150 nm. This could decrease travel. A committee member stated that
most programs will not use DCD organs for candidates over MELD/PELD 30. This proposal
would give programs an incentive to use these organs locally. The committee decided not to
make this change.
UNOS staff asked about the impact of the MELD/PELD 29 threshold on blood type O donors.
Currently, blood type O donors are allocated in the same classification as blood types O and B
recipients who also have a MELD/PELD of 30 or higher. UNOS staff asked if the sharing
threshold for blood type B recipients should also be MELD/PELD 29. These thresholds are in
place for different reasons. This would be an extra line in the policy but does not change the
difficulty of programming. No vote was taken.
Next Steps:
The committee will include these modifications in the policy proposal.
5. Public Comment Feedback on Hawaii and Puerto Rico Variances
The committee discussed extending the existing blood type variance in Hawaii to Puerto Rico.
When the committee put the proposal out for public comment, they supported keeping the
variance in Hawaii and did not want to extend the variance to Puerto Rico. However, public
comment supported extending the variance to Puerto Rico. The committee did not originally
support extending the variance because Puerto Rico is not as geographically isolated as
Hawaii. There is also some sharing between Puerto Rico and Region 3. The Chair presented
data on the distribution of deceased donor liver transplants by recipient blood type in Puerto
Rico. The variance being discussed allocates blood group O livers first to blood group O and
blood group B candidates with MELD/PELD greater than 30.
A committee member stated that extending the variance to Puerto Rico means that a blood type
O liver recovered in Miami would not get offered in Miami until it went through all the different
blood types in Puerto Rico. The flight from Miami to Puerto Rico is two hours. Puerto Rico has
roughly 70 candidates on the waitlist. Puerto Rico has a relatively large donor pool because of
high consent rates and high homicide rates. However, many of their donated organs are
exported. A committee member stated that blood type O candidates in Miami should have
access to blood type O livers, but the exception does make sense for Puerto Rico. The blood
type O candidates in Miami would be the ones disadvantaged. A committee member expressed
empathy for the overall situation of the people of Puerto Rico and mentioned how they are a
disadvantaged population. Another committee member stated that although Puerto Rico is
closer to the contiguous United States than Hawaii, it is still isolated. Puerto Rico does export
some marginal organs to Miami. A representative from the Minority Affairs Committee noted that
this idea came from his/her committee.
A formal vote was taken regarding: do you support extending the Hawaii variance to Puerto
Rico?
Results were as follows: 15 (83%) Yes; 2 (11%) No; 1 (6%) Abstain
13
Yes No Abstain
Next Steps:
The committee will include this modification in the policy proposal.
6. Public Comment Feedback on Pediatric Allocation and Other Issues
The committee then discussed their plan for liver allocation in Alaska. Alaska does not have a
liver program and is geographically isolated. One option is to nationally share all of their livers,
but this is not efficient because the state is roughly 2500 miles from the closest transplant
program.
A committee member put forth the motion that for any donated liver from Alaska, the donor
circle starts from the northwest corner of the United States. The motion was seconded. This
makes practical sense because it allows for better distribution across the Northwest. UNOS staff
stated that if this proposal moves forward, the committee will need to select a specific location
from which the donor circle is drawn. The donor hospital location is used for two things. First, it
is used to determine where candidates show up on the match run when they are being
classified. And second, it is used to screen for the candidate-specific distance to travel. The
committee suggested that the Seattle-Tacoma (Sea-Tac) International Airport be used for both
of these purposes. UNOS staff stated that the Geography Committee looked at this same issue
and they are not recommending this course of action. The Geography Committee is
recommending that Alaska be treated where it is, like Hawaii. The committee agreed that if a
liver program opens in Alaska, this policy will be changed.
A formal vote was taken regarding: do you support the proposal to move Alaska’s donor hospital
location to Seattle-Tacoma International airport?
Results were as follows: 16 (100%) Yes; 0 (0%) No
Yes No Abstain
UNOS staff asked the committee if the change to a MELD/PELD threshold of 29 would have an
impact on simultaneous liver-kidney (SLK) eligibility policy. Current policy states that there is a
mandatory share above MELD/PELD 35, meaning that when a patient with a MELD/PELD of 35
or higher is getting a regional liver, they can also get the kidney. As part of the public comment
proposal, the committee proposed changing the threshold to MELD/PELD 32, but they have
since changed the sharing threshold to MELD/PELD 29. The Chair asked if the liver-kidney
proposal could be a separate policy so that the BOD does not reject the entire policy based on
just this part. A committee member disagreed with this idea and stated that the committee
should be consistent. If they separate one policy, then they should separate them all. Another
committee member stated that the members of the BOD who are from the kidney community
14
will not like this proposal. Regardless, the BOD ultimately are the decision makers who pass
policy. The new SLK policy, which went into effect in 2016, did not lead to an explosion of SLK
transplants or a reduction in SLK transplants. The Kidney Committee is also looking into how
often the safety net is being used.
A formal vote was taken regarding: do you support a MELD/PELD threshold of 29 for SLK
allocation?
The results were as follows: 16 (100%) Yes; 0 (0%) No
Yes No Abstain
UNOS staff clarified that this policy will be implemented in conjunction with the new allocation
model.
The Chair of the committee spoke to the group about some of the metrics that will be monitored
with the implementation of the new policy. The committee will monitor waitlist mortality, post-
transplant mortality, transplant rate, transplant count (across race, age, etc.), community risk
score, etc. The committee will also monitor the impact on exception scores. The Ad Hoc
Systems Performance Committee is working on a better way of recording decline codes and
reasons for declines. A committee member asked if it is possible for the OPTN or SRTR to
monitor the offer rate. Offer rate is less dependent on offer acceptance practices, so it could be
a good way of measuring disparity and the effect of the new policy. There would need to be a
definition of what constitutes an offer. The committee will need to monitor the exception patients
to make sure they are not being disadvantaged.
The committee clarified that the proposal to move allocation from Alaska to Sea-Tac includes
exception patients.
7. Vote on Proposal to Send to Board of Directors (BOD)
A formal vote was taken regarding: do you support sending the proposal to the BOD?
The results were as follows: 13 (93%) Yes; 1 (7%) No
Yes No Abstain
The Chair thanked the committee members for their effort in creating this proposal.
Next Steps:
The committee will include these modifications in the policy proposal. The proposal will be
presented to the BOD at their meeting on December 3-4.
8. Other Significant Items
No other significant items were discussed.
15
Upcoming Meeting
 November 15, 2018
16
RESEARCH ARTICLE
The MELD-Plus: A generalizable prediction risk

score in cirrhosis
Uri Kartoun1,2¤, Kathleen E. Corey1,3, Tracey G. Simon1,3, Hui Zheng4, Rahul Aggarwal1,2,
Kenney Ng5, Stanley Y. Shaw1,2*
1 Harvard Medical School, Boston, Massachusetts, United States of America, 2 Center for Systems Biology;
Center for Assessment Technology & Continuous Health (CATCH), Massachusetts General Hospital, Boston,
Massachusetts, United States of America, 3 Gastrointestinal Unit, Massachusetts General Hospital, Boston,
Massachusetts, United States of America, 4 Center for Biostatistics, Massachusetts General Hospital,
Boston, Massachusetts, United States of America, 5 IBM Research, Cambridge, Massachusetts, United
States of America
a1111111111
a1111111111 ¤ Current address: IBM Research, Cambridge, Massachusetts, United States of America
a1111111111 * stanley_shaw@hms.harvard.edu
a1111111111
a1111111111
Abstract
OPEN ACCESS Background and aims

Citation: Kartoun U, Corey KE, Simon TG, Zheng H, Accurate assessment of the risk of mortality following a cirrhosis-related admission can
Aggarwal R, Ng K, et al. (2017) The MELD-Plus: A enable health-care providers to identify high-risk patients and modify treatment plans to
generalizable prediction risk score in cirrhosis.
decrease the risk of mortality.
PLoS ONE 12(10): e0186301. https://doi.org/
10.1371/journal.pone.0186301
Editor: Sheng-Nan Lu, Chang Gung Memorial Methods

Hospital Kaohsiung Branch, TAIWAN We developed a post-discharge mortality prediction model for patients with a cirrhosis-
Received: July 13, 2017 related admission using a population of 314,292 patients who received care either at Massa-
Accepted: September 28, 2017 chusetts General Hospital (MGH) or Brigham and Women’s Hospital (BWH) between 1992
and 2010. We extracted 68 variables from the electronic medical records (EMRs), including
Published: October 25, 2017
demographics, laboratory values, diagnosis codes, and medications. We then used a regu-
Copyright: © 2017 Kartoun et al. This is an open
larized logistic regression to select the most informative variables and created a risk score
access article distributed under the terms of the
Creative Commons Attribution License, which that comprises the selected variables. To evaluate the potential for generalizability of our
permits unrestricted use, distribution, and score, we applied it on all cirrhosis-related admissions between 2010 and 2015 at an inde-
reproduction in any medium, provided the original pendent EMR data source of more than 18 million patients, pooled from different health-
author and source are credited.
care systems with EMRs. We calculated the areas under the receiver operating characteris-
Data Availability Statement: The institutional tic curves (AUROCs) to assess prediction performance.
review board of Partners HealthCare and IBM
approved this study and all its methods, including
the EMR cohort assembly, data extraction, and Results
analyses. Data contain potentially identifying
information and may not be shared publicly.
We identified 4,781 cirrhosis-related admissions at MGH/BWH hospitals, of which 778
Deidentified data may be requested from The resulted in death within 90 days of discharge. Nine variables were the most effective predic-
Partners Human Research Committee, the tors for 90-day mortality, and these included all MELD-Na’s components, as well as albumin,
Institutional Review Board of Partners HealthCare
total cholesterol, white blood cell count, age, and length of stay. Applying our nine-variable
(Address: 399 Revolution Drive, Suite # 710,
Somerville MA, 02145, USA, Telephone: 857-282- risk score (denoted as “MELD-Plus”) resulted in an improvement over MELD and MELD-Na
1900; ecor@partners.org). scores in several prediction models. On the MGH/BWH 90-day model, MELD-Plus
PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 1 / 15

The MELD-Plus
Funding: The study was funded in part by grants improved the performance of MELD-Na by 11.4% (0.78 [95% CI, 0.75–0.81] versus 0.70
from the NIH K23 DK099422 (KEC), and NIH U54 [95% CI, 0.66–0.73]). In the MGH/BWH approximate 1-year model, MELD-Plus improved
LM008748 (SYS). NIH did not play a role in the
study design, data collection and analysis, decision
the performance of MELD-Na by 8.3% (0.78 [95% CI, 0.76–0.79] versus 0.72 [95% CI,
to publish, or preparation of the manuscript and 0.71–0.73]). Performance improvement was similar when the novel MELD-Plus risk score
only provided financial support in the form of was applied to an independent database; when considering 24,042 cirrhosis-related admis-
authors’ salaries and/or research materials. NIH
sions, MELD-Plus improved the performance of MELD-Na by 16.9% (0.69 [95% CI, 0.69–
provided support in the form of salaries for authors
UK, KEC, and SYS. IBM neither provided authors 0.70] versus 0.59 [95% CI, 0.58–0.60]).
[UK, KN] salaries related to the study nor played
any role in the study design, data collection and
analysis, decision to publish, or preparation of the
Conclusions
manuscript. We developed a new risk score, MELD-Plus that accurately stratifies the short-term mortal-
Competing interests: The authors have declared ity of patients with established cirrhosis, following a hospital admission. Our findings demon-
that no competing interests exist. UK and KN strate that using a small set of easily accessible structured variables can help identify novel
confirm that the commercial affiliation with IBM predictors of outcomes in cirrhosis patients and improve the performance of widely used tra-
does not alter their adherence to all PLOS ONE
policies on sharing data and materials.
ditional risk scores.
Abbreviations: AUROC, Area under the receiver

operating characteristic curve; CI, Confidence
interval; COPD, Chronic obstructive pulmonary
disease; CPT, Current procedural terminology;
EMR, Electronic medical record; GGT, Gamma Introduction
glutamyl transpeptidase; HE, Hepatic
encephalopathy; HIPAA, Health Insurance Cirrhosis-related complications account for 1.1% and 1.8% of all deaths in the United States
Portability and Accountability Act; OR, Odds ratio; and Europe, respectively [1, 2]. In addition to increased mortality, individuals with cirrhosis
ICD-9-CM, International Classification of Diseases, suffer from significantly worse health issues and greater disability compared to those without
Ninth Revision, Clinical Modification; LASSO, Least
cirrhosis [3].
absolute shrinkage and selection operator; MELD,
Model for end-stage liver disease; NAFLD,
Although risk-stratification tools for the prediction of cirrhosis-related mortality are avail-
Nonalcoholic fatty liver disease; Na, Sodium’s able [4–9], these models are based on small populations and use a limited number of prese-
chemical element symbo; NS, Not significant; SBP, lected traditional predictors. Improved mortality prediction scores may highlight the clinical
Spontaneous bacterial peritonitis; SE, Standard variables that contribute to mortality risk, including modifiable factors, and guide the alloca-
error; STD, Standard deviation; WBC, White blood tion of resources to improve cirrhosis care for high-risk patients.
cell count; eGFR, Estimated glomerular filtration
The recent availability of large cohorts of data from electronic medical records (EMRs)
rate.
allows for the development of improved mortality prediction scores through inclusion of a
broader set of clinically applicable, unbiased variables. Not only that, but developing such pre-
diction models allows clinicians to identify the clinical variables that contribute to mortality
risk, including modifiable factors. Improving current standard models like the model for end-
stage liver disease (MELD) and MELD-Na can guide clinicians in better targeting treatment to
improve cirrhosis care and outcomes for high-risk patients.
Cohorts assembled from EMRs represent a powerful resource to study disease complica-
tions at the population level. Recent studies have demonstrated the usefulness of EMR analysis
to discover or confirm outcome correlations, sub-categories of disease, and adverse drug
events [10–14]. The MELD score is based on three commonly used laboratory tests available in
the EMRs, and it is the most widely used tool to predict outcomes in patients with cirrhosis
[15, 16]. An extended version of MELD, one that incorporates serum sodium levels, the
MELD-Na score, has been recently adopted by The Liver and Intestine Transplantation Com-
mittee for liver transplant allocation [17]. Although the two scores are simple to calculate and
apply in a practical sense, the improved accessibility of a wide variety of variables from EMRs
raises the possibility that prediction models could benefit from the inclusion of a broader,
unbiased set of clinical variables. Identifying a combination of the most informative variables
may improve the prognostic utility beyond that of current risk scores.

The MELD-Plus
The aim of the present study was to develop a risk score to predict mortality following a cir-
rhosis-related admission. We demonstrated that a score composed of a small set of easily
accessible clinical variables improves the prediction performance of both the MELD and
MELD-Na scores. We further demonstrated the generalizability of our model through inde-
pendent validation in a large EMR-based data source.
Methods
Study population
We analyzed a previously defined cohort of 314,292 patients at increased risk for metabolic
disease who were admitted to Massachusetts General Hospital (MGH) or Brigham and Wom-
en’s Hospital (BWH) between 1992 and 2010 [13]. We identified an admission as cirrhosis-
related when the keyword “cirrhosis” was present in the discharge summary of the admission
and we observed at least one ICD-9 code (571.2, 571.5, or 571.6 as in [18]) within the 30 days
preceding the discharge date, including during the admission. This identification method was
validated by a physician (Dr. Kathleen Corey) chart review.
We excluded elective admissions if they included at least one diagnosis or procedure code
for liver biopsy, radiofrequency ablation, transarterial chemoembolization, hepatic resection,
or liver transplant. We included only patients 18 years of age or older at the time of the admis-
sion, and we tracked the records of all patients for 90 days after their discharge. We deter-
mined mortality through linkage to the social security master death index.
Prediction modeling
To predict mortality within 90 days, we developed a model that included a large set of struc-
tured variables extracted from the EMRs. In addition to variables available during the period
of admission, we considered variables available for the period of 12 months preceding the dis-
charge date (see Table 1).
The variables included demographics (e.g., gender, ethnicity, marital status), laboratory
measurements (e.g., albumin, sodium), and medications (e.g., anticoagulants, lipid lowering
agents). For laboratory variables, we used the most recent values found during admission
(when no value was found during admission, we considered the preceding 12 months). Typi-
cally, common laboratory measurements were available during the admission (as seen in
Table 1). We determined comorbidities from the number of diagnosis codes within the 12
months prior to the discharge date, and we determined medication count by recording the
number of prescriptions within the 12 months preceding the discharge date.
Additional variables included body mass index, NAFLD fibrosis score (Eq 1), and the
MELD score (Eq 2). Missing values were imputed with the mean of the available data for each
variable. We randomly selected two thirds of the admissions to serve as a derivation set,
whereas the remaining one third served as a validation set. A complete list of the variables we
used is available in S1 Table, and all diagnoses and procedure definitions used in this study are
available in S2 Table.
NAFLD Fibrosis Score ¼ 1:675 þ 0:037 Age þ 0:094 BMI
þ1:13 IFG=Diabetes ðyes ¼ 1; no ¼ 0Þ ð1Þ
þ0:99 AST=ALT ratio 0:013 Platelet 0:66 Albumin
MELD Score ¼ þ6:43 þ 9:57 lnðCreatinineÞ

ð2Þ
þ3:78 lnðTotal BilirubinÞ þ 11:2 lnðINRÞ

The MELD-Plus
Table 1. Baseline characteristics. All values extracted during the 12 months preceding discharge date. For
laboratory variables, values are the most recent. Comorbidity calculations count the number of diagnosis
codes. Prevalence calculations consider admissions with at least one measurement for laboratories and at
least one diagnosis code for comorbidities.
Variable and category Cirrhosis-related admissions (n = 4,781)
Age (years); Mean (SD) 60.0 (13.7)
Gender (%)
Male 64.4
Female 35.6
Ethnicity (%)
Caucasian 77.4
African American 6.8
Other 2.3
Unknown 13.5
Insurance Type (could be 1 types per patient) (%)
Medicaid 5.6
Medicare 60.0
Other 98.6
BMI (kg/m2); Mean (SD) 28.7 (8.2)
Laboratory values; Mean (SD) / Prevalence (%)
Sodium (mmol/l) 136.5 (5.8) / 99.7
eGFR (ml/min/1.73m2) 59.6 (33.7) / 31.9
WBC (th/cumm) 6.7 (3.8) / 99.8
Platelets (th/cumm) 141.3 (100.0) / 99.8
Prothrombin time (INR) 1.5 (0.5) / 91.3
Albumin (g/dl) 2.9 (0.7) / 98.7
Total Bilirubin (mg/dl) 2.5 (4.3) / 98.7
Transaminase SGOT (u/l) 60.2 (76.6) / 98.8
Transaminase SGPT (u/l) 36.1 (38.2) / 96.7
GGT (u/l) 249.4 (346.6) / 6.8
MELD score 14.21 (6.1) / 84.8
NAFLD Fibrosis score 1.70 (2.1) / 13.7
Comorbidities; Mean (SD) / Prevalence (%)
Variceal hemorrhage / Gastrointestinal bleed 0.8 (2.1) / 24.2
Spontaneous bacterial peritonitis 0.1 (0.6) / 3.4
Hepatocellular carcinoma 0.7 (5.5) / 4.3
Hepatorenal syndrome 0.1 (0.5) / 3.2
Hepatic encephalopathy 1.4 (3.6) / 28.6
Ascites 2.4 (5.7) / 37.9
Renal failure 1.8 (8.4) / 13.1
Cerebrovascular disease 0.5 (2.4) / 10.8
Congestive heart failure 2.7 (7.9) / 36.7
Hypertension 3.1 (7.2) / 58.9
Acute myocardial infarction 0.5 (2.1) / 13.4
Ischemic heart disease 0.3 (1.2) / 12.4
Peripheral vascular disease 0.5 (2.2) / 13.0
Diabetes 5.9 (10.3) / 57.0
https://doi.org/10.1371/journal.pone.0186301.t001

The MELD-Plus
To select the most informative variables, we applied feature selection on the derivation set.
We used logistic regression with the adaptive least absolute shrinkage and selection operator
(LASSO) algorithm [19] because it is considered an efficient algorithm for parsimoniously
ranking variables in clinical predictive modeling [20, 21]. We considered all variables that
were statistically significantly different from a univariate analysis (P < 0.05) as in [22]. The
generalized linear model (GLM) equations used to calculate prediction risk at MGH/BWH
and at the independent EMR source are presented in Eqs 3 and 4.
L¼
þ11:794383
þ2:076192 Log10ð1 þ Total BilirubinÞ
þ2:494291 Log10ð1 þ CreatinineÞ
6:049540 Log10ð1 þ AlbuminÞ
þ2:525904 Log10ð1 þ INRÞ
ð3Þ
þ1:911856 Log10ð1 þ WBCÞ
þ0:015411 Length of stay
ðnumber of nights the patient spent in the hospital during the cirrhosis related admissionÞ
þ0:041047 Age ðyearsÞ
6:625270 Log10ð1 þ SodiumÞ
1:445666 Log10ð1 þ Total CholesterolÞ
expðLÞ
MELD Plus ¼ pð90 day mortalityÞ ¼ ð4Þ
1 þ expðLÞ
To calculate 95% confidence intervals, we applied the bootstrap procedure with 1,000
replicates. We calculated the area under the receiver operating characteristic curves
(AUROC) to measure the model’s accuracy in the validation set. Additionally, we evaluated
for overfitting by comparing the AUROC in the validation set to an average AUROC value
for 100 permutations of randomly selected derivation and validation sets (each including
two thirds and one third of the derivation set’s cirrhosis-related admissions, respectively).
We compared categorical variables using a chi-squared test, and we compared the differ-
ences in the means of continuous variables using a t-test or Wilcoxon rank sum test, as
appropriate. We further compared the differences in standard deviations by using an F-test.
All statistical tests were two-sided, with Bonferroni corrections for the 68 comparisons, and
the adjusted P value was 7.410−4 for each comparison. We performed all programming
using the R statistical language [23].
Independent validation
We were granted access to a data source of 18,345,793 individuals, pooled from multiple differ-
ent health-care systems with EMRs (“The IBM Explorys Network”) [24]. The data were stan-
dardized and normalized using common ontologies, searchable through a HIPAA-enabled,
de-identified cloud-computing platform. Patients were seen in multiple health-care systems
between 2010 and 2015, with a combination of data from clinical EMRs, outgoing health-care
system bills, and adjudicated payor claims.
We first identified all cirrhosis-related admissions in this database, and then we extracted
values for the selected variables of our MGH/BWH 90-day model. We further deployed the

The MELD-Plus
GLM equations on the independent source (Eqs 3 and 4). Missing values were imputed based
on the mean values of the MGH/BWH 4,781 cirrhosis-related admissions: total bilirubin
(2.486604493), creatinine (1.375274633), albumin (2.888940902), INR (1.499619615), WBC
(6.673836966), sodium (136.5401552), and total cholesterol (133.6246914).
Because death dates were not available for patients in the IBM Explorys Network, we used
the year of death to determine the outcome. We were able then only to use a 1-year estimated
death for this population (e.g., for a patient discharged on October 13, 2010, we could only
determine if the patient died either in 2010 or 2011, or survived after that). To compare the
performance of the IBM Explorys approximate one-year prediction model, we used the origi-
nal 314,292-patient population at MGH/BWH and applied the same approximate one-year
mortality outcome identification method. We calculated AUROCs for MELD, MELD-Na (Eq
5), and for our risk score (Eqs 3 and 4).
MELD Na ¼ MELD þ 1:59 ð135 NaÞ ð5Þ
The institutional review board of Partners HealthCare and IBM approved this study and all
its methods, including the EMR cohort assembly, data extraction, and analyses.
Results
Univariate analysis
We identified a total of 4,781 admissions as cirrhosis-related, of which 778 resulted in death
within 90 days of the discharge date (16.3%). In a sample of 50 randomly selected patients,
64% were admitted primarily for cirrhosis, for instance, due to the presence of ascites or spon-
taneous bacterial peritonitis, and the rest had the comorbidity of cirrhosis but were admitted
primarily for different reasons such as heart failure or chronic obstructive pulmonary disease
(COPD). Individuals who died within the 90-day period after discharge were older in compari-
son with those who survived (64.1 years versus 59.2 years, P = 4.2210−17); however, the two
populations did not differ by gender (65.0% male versus 64.0% female, P = 1.0) or ethnicity
(77.0% Caucasian for both, P = 1.0).
We calculated event ratios by dividing the values ascertained for the two populations (e.g.,
the mean MELD scores were 18.5 and 13.3 for admissions that resulted in death and survival,
respectively, yielding a ratio of 1.4, P = 4.4510−65). Individuals who died within the 90-day
period after discharge had higher ratios of liver-related comorbidities than those who survived,
and these comorbidities included hepatorenal syndrome (ratio = 5.1, P = 4.6010−21), hepatocel-
lular carcinoma (ratio = 5.0, P = 1.4010−16), and ascites (ratio = 2.1, P = 3.5510−24). Laboratory
measurements also significantly differentiated the two populations. For instance, albumin was
lower in those who died within the 90-day period (2.60 g/dl versus 2.95 g/dl, P = 5.6410−35),
and the total bilirubin (4.87 mg/dl versus 2.02 mg/dl, P = 1.8210−41), INR (1.70 versus 1.46,
P = 5.6310−30), and creatinine (1.80 mg/dl versus 1.29 mg/dl, P = 1.2410−36) were higher in
those who died within the 90-day period.
No difference was found in the prevalence of COPD, cerebrovascular disease, diabetes, cor-
onary artery disease, peripheral vascular disease, pneumonia, or sleep apnea between the popu-
lations. The complete list of variables, indicating the differences between the surviving and the
deceased populations, is presented in S3 Table.
Logistic regression model

The AUROCs of 0.78 were identical for all three models composed of multiple variables (Fig
1). With generalizability in mind and the potential ease of extraction of commonly available

The MELD-Plus
Fig 1. AUROCs using differing variable combinations in a 90-day mortality prediction model at MGH/
BWH.
https://doi.org/10.1371/journal.pone.0186301.g001
laboratory values and other trivial variables (e.g., age, length of stay), we decided to follow the
model that comprised the 9 readily available clinical variables. To evaluate the contribution of
the MELD score to the 90-day mortality prediction, we evaluated the performance of MELD
and MELD-Na scores alone. Considering the 4,781 admissions, using the MELD score alone
to predict the 90-day mortality resulted in an AUROC value of 0.69. An additional model
using the MELD-Na score alone yielded an AUROC value of 0.70.
Each of the MELD-Na components were associated with an increased mortality, including
INR (OR, 1.58; 95% CI, 1.30–1.96), creatinine (OR, 1.25; 95% CI, 1.16–1.34), total bilirubin
(OR, 1.11; 95% CI, 1.08–1.14), and sodium (OR, 0.97; 95% CI, 0.95–0.99). Other laboratory
measurements associated with mortality included WBC (OR, 1.10; 95% CI, 1.07–1.13), total
cholesterol (OR, 0.996; 95% CI, 0.993–0.999), and albumin (OR, 0.45; 95% CI, 0.37–0.52).
Additional predictors included age at time of the admission (OR, 1.04; 95% CI, 1.03–1.05) and
length of stay (OR, 1.02; 95% CI, 1.005–1.03).
Because total cholesterol and hospital length of stay are typically not uniform factors
across different hospitals and may vary in different countries, we evaluated an additional
model that included only 7 of the 9 variables. This yielded an AUROC of 0.77 and
resulted in the following associations with increased mortality: INR (OR, 1.66; 95% CI,
1.38–2.05), creatinine (OR, 1.25; 95% CI, 1.17–1.35), total bilirubin (OR, 1.11; 95% CI,
1.08–1.14), sodium (OR, 0.97; 95% CI, 0.95–0.98), WBC (OR, 1.10; 95% CI, 1.07–1.13),
albumin (OR, 0.43; 95% CI, 0.36–0.51), and age (OR, 1.04; 95% CI, 1.03–1.05). We pres-
ent the GLM equations used to calculate prediction performance at MGH/BWH in Eqs 6

The MELD-Plus
and 7.
L¼
þ8:53499496
þ2:06503238 Log10ð1 þ Total BilirubinÞ
þ2:59679650 Log10ð1 þ CreatinineÞ
6:34990436 Log10ð1 þ AlbuminÞ ð6Þ
þ2:99724802 Log10ð1 þ INRÞ
þ1:92811726 Log10ð1 þ WBCÞ
þ0:04070442 Age ðyearsÞ
6:47834101 Log10ð1 þ SodiumÞ
MELD Plus ðexcluding length of stay and total cholesterolÞ ¼

expðLÞ ð7Þ
pð90 day mortalityÞ ¼
1 þ expðLÞ
Prediction of 90-day mortality after a cirrhosis-related admission

Using our 9-variable risk score, we divided our population into quintiles and compared the
average predicted 90-day mortality with the observed mortality within each quintile. The pre-
dicted 90-day mortality derived from a logistic regression model for each admission and indi-
cated the probability that a patient who survived the admission would die within 90 days post
discharge. As shown in Fig 2A–2C, the predicted 90-day mortality was strongly correlated
with the observed mortality rate throughout the range of risk in both derivation and validation
sets (Kendall’s τ = 1.0; P = 0.027; Pearson correlation r = 0.995 for the correlation between the
average calculated and observed mortality). We provide the logistic regression equations used
to calculate the predicted 90-day mortality probabilities in Eqs 3 and 4. The complete list of
variables that indicate the differences between the highest-risk quantile and the lowest-risk
quantile populations are presented in S4 Table.
Generalization evaluation
Applying our 9-variable risk score (the MELD-Plus score) demonstrated an improvement
over MELD and MELD-Na scores in all prediction models, as shown in Fig 3. On the MGH/
BWH 90-day model, MELD-Plus improved the performance of MELD-Na by 11.4% (0.78
[95% CI, 0.75–0.81] versus 0.70 [95% CI, 0.66–0.73]). On the MGH/BWH approximate 1-year
model, MELD-Plus improved the performance of MELD-Na by 8.3% (0.78 [95% CI, 0.76–
0.79] versus 0.72 [95% CI, 0.71–0.73]). On the IBM Explorys Network model used for external
validation, MELD-Plus improved the performance of MELD-Na by 16.9% (0.69 [95% CI,
0.69–0.70] versus 0.59 [95% CI, 0.58–0.60]).
It is notable that the performance of MELD-Plus on the IBM Explorys data was lower in
comparison with both MGH/BWH models (0.69 versus 0.78). Consistent with MELD-Plus,
the performance of MELD and MELD-Na were also much lower on the IBM Explorys data in
comparison with MGH/BWH. A potential reason for this is that the IBM Explorys Network
population was relatively healthier. Patients in the IBM Explorys network had lower severity of
liver disease in comparison with the corresponding MGH/BWH 1-year prediction model

The MELD-Plus
Fig 2. Predicted versus observed 90-day mortality within each risk quintile. (A) Entire cohort of 4,781
cirrhosis-related admissions. (B) Derivation set of 3,187 cirrhosis-related admissions. (C) Validation set of
1,594 cirrhosis-related admissions.
(mean MELD: 9.4 versus 16.8; P < 0.0001, mean MELD-Na: 11.4 versus 18.1; P < 0.0001).
There may be other differences in the data or populations in the independent systems; the
Partners HealthCare Research Patient Data Registry collected the MGH/BWH data, whereas
dozens of distinct data aggregation mechanisms collected the data for the IBM Explorys Net-
work. Furthermore, the variability in the levels of prediction performance might be influenced
by the variability in the data; prediction performance might be higher when there is more vari-
ability in the data source (i.e., the population comprising patients with a broad spectrum of

The MELD-Plus
Fig 3. Prediction performance across different cirrhosis populations. (A) MGH/BWH 90-day mortality
(4,781 cirrhosis-related admissions). (B) The IBM Explorys Network approximate 1-year mortality (24,042
cirrhosis-related admissions). (C) MGH/BWH approximate 1-year mortality (4,680 cirrhosis-related
admissions).

The MELD-Plus
levels of cirrhosis severity). In the other direction, when the data is more uniform (e.g., most
patients have just been diagnosed with cirrhosis for the first time, and only a minority suffers
from an advanced cirrhosis), then prediction accuracy is lower. This hypothesis was confirmed
because the IBM Explorys network had a statistically significant lower standard deviation of
severity of liver disease in comparison with the MGH/BWH 1-year population (STD MELD:
1.8 versus 8.2; P < 0.0001, STD MELD-Na: 3.6 versus 8.2; P < 0.0001).
Discussion
In this study, we used accessible EMR variables to develop a highly accurate, predictive model
of 90-day post-discharge mortality in individuals with cirrhosis. We identified 9 variables that
accurately predicted 90-day mortality with an AUROC of 0.78. Our risk score improved the
performance of MELD and MELD-Na scores in multiple, independent patient populations,
and this also held true in a large external validation patient cohort. Furthermore, our model’s
calculated 90-day mortality risk was highly correlated with the observed mortality rate across
all five risk quintiles. In particular, the model’s performance on the highest-risk quintile (the
calculated and observed 90-day mortality was 31.6% and 31.2%, respectively) suggests that
high-risk patients can be accurately identified. An additional model that included only 7 of the
9 variables and excluded length of stay and total cholesterol yielded an AUROC of 0.77 [95%
CI, 0.74–0.80]. Although the 7-variable model demonstrated improved identification ability
compared to MELD or MELD-Na, the improved prediction performance achieved by includ-
ing total cholesterol in MELD-Plus suggests that it may be beneficial for cholesterol labs to be
routinely collected in cirrhosis admission order sets.
The MELD score has been used extensively to predict patient outcomes, mortality, and
readmission rates in individuals with cirrhosis [25, 26, 4]. Furthermore, although MELD-Na
[17] was superior to MELD, the MELD-Plus score yielded improved levels of discrimination
consistently in all prediction models, with AUROCs that significantly outperformed the tradi-
tional scores. These findings suggest that new types of cirrhosis-related risk indexes utilizing
novel risk indicators may improve prognostication in this high-risk population.
MELD-Plus includes all MELD-Na’s components, as well as additional variables (albumin,
total cholesterol, WBC, age, and length of stay). It is logical that a predication model that has all
the MELD-Na model variables and additional ones would perform better, as was observed by
MELD-Plus. Not only that, but many of the variables have physiological plausibility for inclu-
sion in a prediction model. Decreased albumin correlated with worse outcomes in our model,
which may be the result of decreased albumin marking decreased liver function in cirrhosis
patients [25, 5]. Increasing age and length of hospital stay helped predict worse outcomes as well
as could be expected. Along with that, higher WBC was correlated with a worse prognosis,
potentially indicating poorer patient status (e.g., infection) at time of score calculation. Although
patients may have multiple WBC measurements during admission, our model is both internally
and externally valid because it uses the most recent WBC lab value. We chose the most recent
WBC during model development because the last available set of labs is more reflective of the
current health of patients than older measurements. Surprisingly, increased total cholesterol pre-
dicted a more favorable prognosis. Although unintuitive at first, this aligns with previous reports
that claim cholesterol levels become less of a risk factor or even an inverse risk factor for mortal-
ity because serious diseases may lower cholesterol soon before death occurs [27].
Although our study describes analyses of retrospective medical databases, the proposed
score could be used to identify patients that are at a high-risk of mortality in real time and thus
may inform risk-stratification and therapeutic decision-making. In a desirable scenario, our
score could be calculated automatically as an integrated component of an EMR system; the

The MELD-Plus
clinician would see a risk score (probability) or a risk quantile (highest, lowest, or in between)
associated with the discharged patient, and this could be used to guide outpatient monitoring
strategies. With further validation, the MELD-Plus score could also be used longitudinally in
outpatients to monitor disease progression and/or responses to therapy.
Our study has limitations. First, it is a retrospective analysis limited to two academic, ter-
tiary-care hospitals. Even though we validated our model on a large external patient cohort,
subsequent studies must further assess the validity of our model in the external population and
consider different age ranges, coding systems, and data-collection methods. Second, the cir-
rhosis populations may vary at different centers—for example, alcohol use might significantly
vary between patients residing in the Boston area versus patients residing in other states [28].
Furthermore, although MGH and BWH are urban care facilities, the high prevalence of rural
populations at the IBM Explorys Network might affect prediction performance. Third,
although mortality was recorded, either through linkage to the social security master death
index as in the MGH/BWH models or through using EMR or billing/claims in the IBM Exp-
lorys model, such death indications may under-represent the true mortality rates. To minimize
this potential under-representation, we considered only patients who survived the study fol-
low-up. All patients had EMR data entries (such as laboratory measurements) after the study
follow-up, indicating survival, or had a recorded indication of death during the study follow-
up, with no EMR data entries found afterward.
Another limitation of MELD-Plus is that it did not specifically consider which procedures
patients underwent during the cirrhosis-related admissions. Furthermore, all the patients con-
sidered in our models survived the admission, but neither MGH/BWH’s nor IBM’s databases
contained information on post-discharge cause of death. To further assess MELD-Plus’s appli-
cability in clinical practice, future analyses should consider subgroups of patients to determine
linkages between invasive inpatient procedures and causes of mortality. Regardless of this limi-
tation, however, our MELD-Plus displayed validity in predicting overall mortality, which is
clinically applicable, because it provides clinicians with information on populations of patients
who need more intense or closer care.
Although we excluded elective admissions for liver biopsy, radiofrequency ablation, trans-
arterial chemoembolization, hepatic resection, or liver transplant, these criteria might exclude
patients with early and intermediate hepatocellular carcinoma (HCC), but not patients with
advanced HCC who underwent medical treatments only. Liver cancer can lead to early mortal-
ity, even in patients with mild liver cirrhosis, and, as such, our exclusion criteria may reduce
the applicability of our risk score when applying it to patients with more advanced HCC. Fur-
thermore, because we excluded admissions associated with a liver transplant, mortality risk
may decrease after a cirrhosis-related admission if patients successfully underwent a transplant
in a preceding admission.
Another limitation of our study is algorithmic. The adaptive LASSO method identified 9
predictors and left out variables that may also be correlated with predicting death. Feature
selection algorithms are known to be blind to the clinical importance of variables, and when
highly correlated predictors are identified, the algorithm randomly selects one. On the one
hand, important variables such as ascites, hepatocellular carcinoma, and diuretic medications
were not selected as predictors. On the other hand, the feature selection algorithm assures that
a minimal set of covariates produce a high level of prediction accuracy. Furthermore, we con-
ducted our model performance evaluation on a held-out data set not used for training.
Although a prediction model’s error usually decreases when more variables are included, this
is not always the case. This is true when performance is evaluated on the training set (due to
overfitting) but not the case when performance is evaluated on a held-out test dataset, as was
used across all our models.

The MELD-Plus
In conclusion, we describe an unbiased and well-validated score to estimate 90-day mortal-

ity after a cirrhosis-related admission. This score, comprising a small set of easily available clin-
ical variables extracted from EMRs, improved the MELD and MELD-Na scores in predicting
90-day mortality and approximate 1-year mortality. In addition, we identified high-risk
patients with great accuracy. MELD-Plus’s strong performance demonstrates potential for it to
replace current standard models, allowing for greater accuracy in the identification of high-
risk cirrhosis patients.
Supporting information
S1 Table. Summary of variables.
(DOCX)
S2 Table. Billing codes used to define conditions.
(DOCX)
S3 Table. Comparison of variables in patients who died vs. survived 90-days after dis-
charge.
(DOCX)
S4 Table. Comparison between the highest-risk (1st) and the lowest-risk (5th) quintiles.
(DOCX)
Acknowledgments
We acknowledge Chin Hur MD MPH (Director, GI Health Outcomes Research, GI Unit, Mas-
sachusetts General Hospital / Harvard Medical School) for his critical review of this manuscript.
Author Contributions
Conceptualization: Uri Kartoun, Kathleen E. Corey, Stanley Y. Shaw.
Data curation: Uri Kartoun, Rahul Aggarwal.
Formal analysis: Uri Kartoun, Rahul Aggarwal, Kenney Ng, Stanley Y. Shaw.
Funding acquisition: Kathleen E. Corey, Stanley Y. Shaw.
Investigation: Stanley Y. Shaw.
Methodology: Uri Kartoun, Kathleen E. Corey, Kenney Ng, Stanley Y. Shaw.
Resources: Stanley Y. Shaw.
Software: Uri Kartoun, Kenney Ng.
Supervision: Kathleen E. Corey, Stanley Y. Shaw.
Validation: Uri Kartoun, Rahul Aggarwal, Kenney Ng.
Writing – original draft: Uri Kartoun.
Writing – review & editing: Kathleen E. Corey, Tracey G. Simon, Hui Zheng, Rahul Aggar-
wal, Kenney Ng, Stanley Y. Shaw.
References
1. Kim WR, Brown RS Jr, Terrault NA, El-Serag H. Burden of liver disease in the United States: summary
of a workshop. Hepatology 2002; 36:227–42. https://doi.org/10.1053/jhep.2002.34734 PMID:
12085369

The MELD-Plus
2. Blachier M, Leleu H, Peck-Radosavljevic M, Valla DC, Roudot-Thoraval F. The burden of liver disease
in Europe: a review of available epidemiological data. J Hepatol 2013; 58(3):593–608. https://doi.org/
10.1016/j.jhep.2012.12.005 PMID: 23419824
3. Rakoski MO, McCammon RJ, Piette JD, Iwashyna TJ, Marrero JA, Lok AS, et al. Burden of cirrhosis on
older Americans and their families: analysis of the health and retirement study. Hepatology 2012; 55
(1):184–91. https://doi.org/10.1002/hep.24616 PMID: 21858847
4. Johnson KB, Campbell EJ, Chi H, Zheng H, King LY, Wu Y, et al. Advanced disease, diuretic use, and
marital status predict hospital admissions in an ambulatory cirrhosis cohort. Dig Dis Sci 2014; 59
(1):174–82. https://doi.org/10.1007/s10620-013-2832-5 PMID: 23990035
5. Younossi ZM, Henry L, Stepanova M. A new comorbidity model for predicting mortality in patients with
cirrhosis: does it work? Gastroenterology 2014; 146(1):19–24. https://doi.org/10.1053/j.gastro.2013.11.
026 PMID: 24287302
6. Kamath PS, Kim WR. The model for end-stage liver disease (MELD). Hepatology 2007; 45(3):797–805.
https://doi.org/10.1002/hep.21563 PMID: 17326206
7. Gotthardt D, Weiss KH, Baumgärtner M, Zahn A, Stremmel W, Schmidt J, et al. Limitations of the
MELD score in predicting mortality or need for removal from waiting list in patients awaiting liver trans-
plantation. BMC Gastroenterol 2009; 9(72):1–7.
8. Huo TI, Lin HC, Wu JC, Hou MC, Lee FY, Lee PC, et al. Limitation of the model for end-stage liver dis-
ease for outcome prediction in patients with cirrhosis-related complications. Clin Transplant 2006; 20
(2):188–94. https://doi.org/10.1111/j.1399-0012.2005.00463.x PMID: 16640525
9. Montano-Loza AJ, Duarte-Rojo A, Meza-Junco J, Baracos VE, Sawyer MB, Pang JX, et al. Inclusion of
Sarcopenia within MELD (MELD-Sarcopenia) and the prediction of mortality in patients with cirrhosis.
Clin Transl Gastroenterol 2015; 6(e102):1–8.
10. Kohane IS, Drazen JM, Campion EW. A glimpse of the next 100 years in medicine. N Engl J Med 2012;
367:2538–39. https://doi.org/10.1056/NEJMe1213371 PMID: 23268669
11. Liao KP, Kurreeman F, Li G, Duclos G, Murphy S, Guzman R, et al. Associations of autoantibodies,
autoimmune risk alleles, and clinical diagnoses from the electronic medical records in rheumatoid arthri-
tis cases and non-rheumatoid arthritis controls. Arthritis Rheum 2013; 65(3):571–81. https://doi.org/10.
1002/art.37801 PMID: 23233247
12. Doshi-Velez F, Ge Y, Kohane I. Comorbidity clusters in autism spectrum disorders: an electronic health
record time-series analysis. Pediatrics 2014; 133(1):e54–e63. https://doi.org/10.1542/peds.2013-0819
PMID: 24323995
13. Liao KP, Cai T, Savova GK, Murphy SN, Karlson EW, Ananthakrishnan AN, et al. Development of phe-
notype algorithms using electronic medical records and incorporating natural language processing.
BMJ 2015; 350:h1885. https://doi.org/10.1136/bmj.h1885 PMID: 25911572
14. Beam AL, Kartoun U, Pai JK, Chatterjee AK, Fitzgerald TP, Shaw SY, Kohane IS. Predictive modeling
of physician-patient dynamics that influence sleep medication prescriptions and clinical decision-mak-
ing. Sci Rep 2017; 7(42282); https://doi.org/10.1038/srep42282 PMID: 28181568
15. Malinchoc M, Kamath PS, Gordon FD, Peine CJ, Rank J, ter Borg PC. A model to predict poor survival
in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31(4):864–71.
https://doi.org/10.1053/he.2000.5852 PMID: 10733541
16. Kamath PS, Wiesner RH, Malinchoc M, Kremers W, Therneau TM, Kosberg CL, et al. A model to pre-
dict survival in patients with end-stage liver disease. Hepatology 2001; 33(2):464–70. https://doi.org/10.
1053/jhep.2001.22172 PMID: 11172350
17. Mulligan DC, Hirose R. OPTN / UNOS, Liver and Intestinal Organ Transplantation Committee. Report
to the Board of Directors. June 23–24, 2014. Richmond, Virginia.
18. Myers RP, Quan H, Hubbard JN, Shaheen AA, Kaplan GG. Predicting in-hospital mortality in patients
with cirrhosis: results differ across risk adjustment methods. Hepatology 2009; 49(2):568–77. https://
doi.org/10.1002/hep.22676 PMID: 19085957
19. Zou H. The adaptive lasso and its oracle properties. J Am Stat Assoc 2006; 101:1418–29.
20. Liao KP, Cai T, Gainer V, Goryachev S, Zeng-treitler Q, Raychaudhuri S, et al. Electronic medical rec-
ords for discovery research in rheumatoid arthritis. Arthritis Care Res (Hoboken) 2010; 62(8):1120–7.
21. Ananthakrishnan AN, Cai T, Savova G, Cheng SC, Chen P, Perez RG, et al. Improving case definition
of Crohn’s disease and ulcerative colitis in electronic medical records using natural language process-
ing: a novel informatics approach. Inflamm Bowel Dis 2013; 19(7):1411–20. https://doi.org/10.1097/
MIB.0b013e31828133fd PMID: 23567779
22. Corey KE, Kartoun U, Zheng H, Shaw SY. Development and validation of an algorithm to identify nonal-
coholic fatty liver disease in the electronic medical record. Dig Dis Sci 2016; 61(3):913–9. https://doi.
org/10.1007/s10620-015-3952-x PMID: 26537487

The MELD-Plus
23. Ihaka R. & Gentleman R. R: a language for data analysis and graphics. J Comput Graph Stat 1996;
5:299–314.
24. The IBM Explorys Platform / Solution Brief. IBM Watson Health. IBM Corporation 2016.
25. Singal AG, Rahimi RS, Clark C, Ma Y, Cuthbert JA, Rockey DC, et al. An automated model using elec-
tronic medical record data identifies patients with cirrhosis at high risk for readmission. Clin Gastroen-
terol Hepatol 2013; 11(10):1335–41. https://doi.org/10.1016/j.cgh.2013.03.022 PMID: 23591286
26. Volk ML, Tocco RS, Bazick J, Rakoski MO, Lok AS. Hospital readmissions among patients with decom-
pensated cirrhosis. Am J Gastroenterol 2012; 107(2):247–52. https://doi.org/10.1038/ajg.2011.314
PMID: 21931378
27. Ravnskov U, Diamond DM, Hama R, Hamazaki T, Hammarskjöld B, Hynes N, Kendrick M, Langsjoen
PH, Malhotra A, Mascitelli L, McCully KS, Ogushi Y, Okuyama H, Rosch PJ, Schersten T, Sultan S,
Sundberg R. Lack of an association or an inverse association between low-density-lipoprotein choles-
terol and mortality in the elderly: a systematic review. BMJ Open 2016; 6(6):e010401. https://doi.org/10.
1136/bmjopen-2015-010401 PMID: 27292972
28. Sacks JJ, Gonzales KR, Bouchery EE, Tomedi LE, Brewer RD. 2010 National and State Costs of
Excessive Alcohol Consumption. Am J Prev Med 2015; 49(5):e73–9. https://doi.org/10.1016/j.amepre.
2015.05.031 PMID: 26477807

Liver Committee Minutes Plus

Hochgeladen von

Dokumentinformationen

Originaltitel

Copyright

Verfügbare Formate

Dieses Dokument teilen

Dokument teilen oder einbetten

Freigabeoptionen

Stufen Sie dieses Dokument als nützlich ein?

Sind diese Inhalte unangemessen?

Copyright:

Verfügbare Formate

Liver Committee Minutes Plus

Hochgeladen von

Copyright:

Verfügbare Formate

OPTN/UNOS Liver and Intestinal Organ Transplantation Committee

Julie Heimbach, MD, Chair

Percentage 55% 45% 0%

Percentage 37% 63% 0%

In favor of In favor of Abstain

Percentage 55% 45% 0%

Percentage 74% 11% 16%

Percentage 83% 11% 6%

The MELD-Plus: A generalizable prediction risk

OPEN ACCESS Background and aims

Editor: Sheng-Nan Lu, Chang Gung Memorial Methods

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 1 / 15

Abbreviations: AUROC, Area under the receiver

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 2 / 15

MELD Score ¼ þ6:43 þ 9:57 lnðCreatinineÞ

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 3 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 4 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 5 / 15

Logistic regression model

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 6 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 7 / 15

MELD Plus ðexcluding length of stay and total cholesterolÞ ¼

Prediction of 90-day mortality after a cirrhosis-related admission

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 8 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 9 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 10 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 11 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 12 / 15

In conclusion, we describe an unbiased and well-validated score to estimate 90-day mortal-

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 13 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 14 / 15

PLOS ONE | https://doi.org/10.1371/journal.pone.0186301 October 25, 2017 15 / 15

Das könnte Ihnen auch gefallen