Model for End-stage Liver Disease (MELD)

INTRODUCTION — Prognostic models are useful for estimating disease severity and survival and can serve as helpful medical decision-making tools for guiding patient care. These models are developed using statistical methodologies that involve determining the effects of variables of interest (eg, demographics, clinical data, and laboratory values) on specific outcomes, such as death. Machine learning is increasingly being used to derive these predictive models.

Several prognostic models are used in health care settings. Some focus on generalized health status such as the Acute Physiology and Chronic Health Evaluation System (APACHE III) [1], while others are disease-specific. Models that are used commonly in the care of patients with cirrhosis are the Child-Turcotte-Pugh score, the Model for End-stage Liver Disease (MELD) score, and the MELD-Sodium (MELD-Na) score [2-6].

This topic will review the use, impact, refinements, and limitations of the MELD score, particularly with regard to its use in allocating organs for liver transplantation. Other issues related to the selection of patients for liver transplantation are discussed separately. (See "Liver transplantation in adults: Patient selection and pretransplantation evaluation".)

MELD OVERVIEW — The original MELD score is a prospectively developed and validated chronic liver disease severity scoring system that uses a patient's laboratory values for serum bilirubin, serum creatinine, and the international normalized ratio (INR) for prothrombin time to predict three-month survival (original MELD score). In patients with cirrhosis, an increasing MELD score is associated with increasing severity of hepatic dysfunction and increased three-month mortality risk (figure 1) [7]. Given its accuracy in predicting short-term survival among patients with cirrhosis, MELD was initially adopted by the United Network for Organ Sharing (UNOS) in 2002 for prioritizing patients awaiting liver transplantation in the United States. (See 'Adoption of MELD for organ allocation' below and "Liver transplantation in adults: Patient selection and pretransplantation evaluation", section on 'Cirrhosis'.)

Development of the MELD score — MELD was originally developed to predict three-month mortality following transjugular intrahepatic portosystemic shunt (TIPS) placement and was derived using data from a population of 231 patients with cirrhosis who underwent elective TIPS placement. The model was subsequently validated in an independent cohort of patients from the Netherlands undergoing TIPS placement [4]. The original model included serum bilirubin, serum creatinine, INR, and etiology of the liver disease (cholestatic or alcohol-associated versus other etiologies).

The etiology of liver disease was subsequently removed from the model because it posed difficulties, such as how to categorize patients with multiple causes of liver disease. Modification of the MELD score by excluding etiology of liver disease did not significantly affect the model's accuracy in predicting three-month survival [8]. It has also been shown that the inclusion of portal hypertensive complications (eg, ascites, encephalopathy, variceal bleeding, and spontaneous bacterial peritonitis) in the MELD score does not substantially improve its predictive accuracy [5]. However, this does not imply that these portal hypertensive complications are not associated with decreased survival, but rather that these complications are more likely to be associated with advanced liver disease as determined by the MELD score.

Since survival following TIPS is primarily determined by the severity of the underlying liver disease, and since MELD was demonstrated as an accurate predictor of survival after TIPS, it was hypothesized that the MELD score might be useful as a prognostic indicator in a broader range of patients with advanced liver disease who were not undergoing TIPS placement. Subsequent studies demonstrated that the MELD score was useful in predicting mortality in several groups of patients, including patients on the waiting list for liver transplantation, hospitalized patients with hepatic decompensation, ambulatory patients with non-cholestatic liver disease, patients with primary biliary cholangitis, and a historic cohort of unselected patients with cirrhosis seen at Mayo Clinic Rochester at a time when liver transplantation was not available [5,9,10].

Calculating the MELD score — Several online tools are available for calculating the MELD score (original MELD score) [11]. The MELD equation that was initially used by UNOS beginning in 2002 for prioritizing allocation of deceased donor livers for transplantation is demonstrated below:

With this model, scores range from negative values to infinity. However, to avoid confusion, UNOS modified the MELD scoring system to eliminate negative values by setting to 1 any measured laboratory values that were less than 1. Thus, patients with the combination of an INR of ≤1, serum creatinine ≤1 mg/dL, and serum bilirubin ≤1 mg/dL will receive the minimum score of 6 MELD points. In addition, UNOS set an upper limit for the MELD score at 40 points.

To avoid an unfair advantage for patients with intrinsic renal disease, the maximum serum creatinine level was originally set to 4 mg/dL, which is also the value that is automatically assigned to patients who have received hemodialysis at least twice, or continuous venovenous hemodialysis for 24 hours, in the preceding week. There is no modification in the score for patients receiving anticoagulation.

MELD-Na Score — In January 2016, Organ Procurement and Transplantation Network Policy 9.1 (MELD Score) was updated to include serum sodium as a factor in the calculation of the MELD score [12]. The MELD-Na score is used by UNOS for deceased donor liver allocation. The MELD-Na score can be calculated online.

Hyponatremia is a common problem in patients with cirrhosis, and the severity of the hyponatremia is a marker of the severity of the cirrhosis. (See "Hyponatremia in patients with cirrhosis", section on 'Prognosis'.)

Liver transplant candidates on the waiting list with hyponatremia may benefit from the MELD-Na liver allocation system. For example, a patient with a MELD score of 12 but a serum sodium level of 125 mmol/L, will have a MELD-Na score of 23. The specific policy states that for candidates with an initial MELD score greater than 11, the score is then recalculated as the MELD-Na score, as follows:

MELD-Na = MELD + 1.32 * (137-Na) - [0.033*MELD * (137-Na)]

In the MELD-Na equation, sodium values less than 125 mmol/L are set to 125, and values greater than 137 mmol/L are set to 137.

Serum sodium reflects the vasodilatory state in cirrhosis and predicts waitlist mortality independent of the MELD score [13]. There is a linear increase in mortality by 5 percent for each mmol decrease in serum sodium between 125 and 140 mmol/L [14]. Multiple studies have shown that the addition of serum sodium concentration improves the predictive accuracy of the MELD score in hyponatremic patients with low MELD scores who are awaiting liver transplantation [14-21]. Addition of serum sodium to the MELD model elevates the transplant priority for about 12 percent of listed patients [13].

A limitation of the MELD-Na score is that serum sodium levels may be vulnerable to alterations by diuretic use and intravenous fluid administration. (See 'Limitations of the MELD score' below.)

APPLICATIONS OF THE MELD SCORE — The primary use of the MELD and MELD-Na scores is in prioritizing patients on the waitlist for deceased donor liver transplantation based on liver disease severity and short-term mortality risk. However, as described above, the MELD score also predicts mortality following transjugular intrahepatic portosystemic shunt (TIPS) placement and has been demonstrated to have predictive value for outcomes in patients with cirrhosis undergoing nontransplantation surgical procedures [22]. Several other applications of the MELD score have been demonstrated and include, but are not limited to, predicting mortality in alcohol-associated hepatitis [22] and in acute variceal hemorrhage [23,24]. (See 'Development of the MELD score' above and 'Applications beyond organ allocation' below.)

Organ allocation — In 2002, the MELD score was adopted by United Network for Organ Sharing (UNOS) for use in deceased donor liver allocation for adults with cirrhosis. In 2016, MELD-Na was adopted in addition to the MELD score, for deceased donor liver allocation. MELD has also been adopted by multiple countries and regions worldwide [25]. In the United States, adoption of MELD, which serves as a marker of liver disease severity and mortality risk among patients awaiting liver transplantation, has been associated with decreased mortality among patients on the liver transplant waiting list [26]. (See 'Adoption of MELD for organ allocation' below.)

It is important to note that under the deceased donor liver allocation system, adult patients with acute liver failure (UNOS status 1A) are exempt from the MELD-based prioritization process. However, outside of the context of the UNOS allocation policy for status 1A patients, the MELD score may have some prognostic value in selected patients with acute liver failure [27-29]. (See 'Applications beyond organ allocation' below.)

Adoption of MELD for organ allocation — During the liver transplant community's search for a more equitable allocation system, the MELD score emerged as a more objective model for prioritizing patients based on liver disease severity. The MELD score was initially adopted by UNOS in 2002 for use in deceased donor liver allocation for adults with cirrhosis. MELD-Na was added into the deceased donor liver allocation system in 2016.

The utility of the MELD score for predicting three-month mortality among patients awaiting liver transplantation was initially demonstrated in a study that included 3437 adult liver transplantation candidates who were listed between 1999 and 2001 [9]. Of these, 412 died during the three-month follow-up period. Waiting list mortality was directly proportional to the MELD score at the time of listing, with mortality being 1.9 percent for patients with MELD scores less than 9, and 71 percent for patients with MELD scores ≥40 (figure 1).

The use of MELD to predict whether undergoing liver transplantation would provide a survival benefit compared with continued medical management has been investigated in several studies. It has been demonstrated that survival benefit increases with increasing MELD score and that at lower MELD scores, recipient mortality risk during the first post-transplantation year may be higher than for candidates who remain on the waiting list [30-32].

Prioritization for liver transplantation based on MELD score — Patients awaiting liver transplantation are ranked according to their MELD score and stratified by blood type. Patients have their MELD scores updated regularly with UNOS by the listing transplant center according to UNOS directives [33]. As a rule, patients with greater liver disease severity (ie, higher MELD scores) have their MELD scores updated more frequently compared with patients with less severe liver disease. Patients with a MELD score ≥25, for example, have their scores updated every seven days. Patients may have their MELD score updated more often if they experience a decline in health status (manifested by a rise in their calculated MELD score). Due to the dynamic nature of cirrhosis, the MELD score may either increase or decrease while patients await liver transplantation. (See "Liver transplantation in adults: Patient selection and pretransplantation evaluation", section on 'Cirrhosis'.)

Time spent on the liver transplant waiting list at a given MELD score is used to break ties among patients with the same blood type. If a patient's MELD score increases (indicating worsening liver disease severity), the waiting time clock is set to zero and restarted at the higher score. However, according to UNOS policy, if a patient's MELD score goes down, the time accumulated at the higher MELD score is maintained and added to the time accumulated at the lower score.

A policy governing organ allocation was approved in December 2018 but was implemented in February 2020 following legal delays. The updated system replaces the use of decades-old geographic boundaries of 58 donation service areas (DSAs) and 11 transplant regions. Medical urgency of liver transplant candidates and the geographic distance between the donor hospital and transplant hospitals are emphasized under this updated policy.

Standard MELD exceptions in liver transplantation — There are some conditions associated with chronic liver disease that may result in impaired survival, but that are not directly accounted for in the MELD scoring system [34]. Some of these conditions have been designated by the liver transplant community as "standard MELD exceptions." Patients who meet specific disease-related criteria for standard MELD exceptions may be eligible for an upgrade in MELD points provided that the patients continue to meet the specific disease-related criteria. These standard MELD exceptions were developed to more accurately represent the patient's mortality risk while awaiting liver transplantation. (See 'Hepatocellular carcinoma' below and 'Other standard MELD exceptions' below.)

Under the updated UNOS policy, standard MELD exception points are based upon the Median MELD at Transplant (MMaT) within a 250 nautical mile radius around the transplant hospital. Specifically, MELD exception points for all standard MELD exceptions (see list below) are assigned a score of MMaT-3 (ie, three points lower than the MMaT). There are two exceptions to the MMaT-3 score assignment for standard exceptions [35]:

●Primary hyperoxaluria is assigned a score equal to the MMaT, and

●Hepatic artery thrombosis occurring within the initial 14 days posttransplant (but not meeting criteria for status 1A) is assigned an exception score of 40 points

The MMaT for each liver transplant center in the United States is updated every 180 days.

Standard MELD exceptions include [36]:

●Hepatocellular carcinoma (HCC)

●Hepatopulmonary syndrome

●Portopulmonary hypertension

●Familial amyloid polyneuropathy

●Primary hyperoxaluria

●Cystic fibrosis

●Hilar/Perihilar cholangiocarcinoma (the liver transplant center must have a UNOS approved protocol for the work-up and management of patients with cholangiocarcinoma undergoing transplantation)

●Hepatic artery thrombosis (occurring within the initial 14 days after liver transplant surgery, but not meeting criteria for status 1A)

Hepatocellular carcinoma — HCC is the most common standard MELD exception. Whether patients receive exception points (or are candidates for transplant listing at all) depends on the extent of their HCC. Giving additional MELD points to patients with HCC initially led to a substantial increase in the proportion of listed patients with HCC [37]. Subsequently, the HCC exception point policy was modified to reflect more accurately the mortality risk due to HCC, and to not unduly favor patients with HCC relative to patients listed with their calculated, biologic MELD score. The decision to give exception points to patients with HCC was based on two observations (see "Liver transplantation for hepatocellular carcinoma", section on 'Indications for transplantation'):

●Patients with HCC who meet criteria for liver transplantation have a post-liver transplantation survival that is not worse than other patients undergoing liver transplantation.

●Many patients with HCC do not demonstrate the degree of hepatic synthetic dysfunction necessary to give them a competitive calculated MELD score, and thus would be given too low a priority based on their calculated MELD score alone. A low calculated MELD score in a patient with HCC would translate into increased waiting time, with a concomitant increased risk of tumor growth during the waiting period, resulting in increased morbidity and mortality.

The UNOS policy for HCC standard MELD exception points is discussed in more detail separately [33,38]. (See "Liver transplantation for hepatocellular carcinoma", section on 'Requirements for listing and management while on the wait list'.)

The approach to downstaging patients with advanced HCC by using neoadjuvant locoregional therapy is discussed separately. (See "Liver transplantation for hepatocellular carcinoma", section on 'Downstaging through neoadjuvant locoregional therapy' and "Localized hepatocellular carcinoma: Liver-directed therapies for nonsurgical candidates not eligible for local thermal ablation".)

Other standard MELD exceptions — The standard MELD exceptions other than HCC include the following [39-41]:

●Portopulmonary hypertension, characterized by an elevated mean pulmonary artery pressure in the presence of portal hypertension. However, since a mean pulmonary artery pressure >35 mmHg is associated with poorer outcomes after liver transplantation, to receive MELD exception points for portopulmonary hypertension, the mean pulmonary artery pressure must be maintained <35 mmHg with treatment. (See "Portopulmonary hypertension".)

●Hepatopulmonary syndrome, characterized by a PaO₂ <60 mmHg on room air and associated with portal hypertension and demonstration of intrapulmonary shunting in the absence of parenchymal lung disease.

●Familial amyloid polyneuropathy, characterized by the identification of the transthyretin (TTR) gene mutation (Val30Met versus non-Val30Met) by DNA analysis or mass spectrometry in a biopsy sample and confirmation of amyloid deposition in an involved organ.

●Primary hyperoxaluria with evidence of alanine glyoxylate aminotransferase deficiency. These patients must be listed for combined liver-kidney transplantation.

●Cystic fibrosis, characterized by a forced expiratory volume in one second (FEV1) <40 percent.

●Hilar/Perihilar cholangiocarcinoma, provided the transplant center listing the patient has a written and UNOS approved protocol detailing the work-up and management of patients with cholangiocarcinoma undergoing transplantation.

●Hepatic artery thrombosis (occurring within 14 days of liver transplantation, but not meeting criteria for status 1A).

Petitioning for additional MELD points — Some patients may have unique complicating medical conditions that are related to their liver disease, but that do not qualify for standard MELD exception points. Transplant centers may petition the National Liver Review Board (NLRB) for additional points for listed patients with complicating medical issues related to their liver disease if the patient's medical providers believe that the patient's biologic MELD score does not adequately reflect the patient's true liver-related morbidity and mortality. The NLRB was established by the OPTN in May 2019 to improve the consistency among transplant centers across the United States in determining MELD exception scores for patients listed for liver transplant.

The petition may be considered by the NLRB and either approved or denied, depending on the basis for the request and the objective data supporting the need for additional MELD points.

Impact of the MELD allocation system — The introduction of the MELD score for organ allocation in 2002 has, overall, improved liver allocation, though there have been other effects that were not anticipated. The impact of the adoption of the MELD score for organ allocation was examined in a study using data from UNOS [42]. Outcomes of deceased donor liver transplantation were compared between the pre-MELD era (era 1: February 27, 2001 to February 26, 2002) and the post-MELD era (era 2: February 27, 2002 to February 26, 2003).

Compared with the pre-MELD era, the post-MELD era was associated with:

●A 12 percent reduction in new patients added to the liver transplant waiting list (particularly patients with low MELD scores because accrual of waiting time was no longer advantageous in the MELD system).

●A higher mean MELD score at the time of transplantation (24 in the post-MELD era versus 18 in the pre-MELD era).

●A 10 percent increase in the number of deceased donor liver transplantations performed.

●A 3.5 percent decrease in the number of deaths on the liver transplantation waiting list.

In 2005, Argentina adopted MELD for organ allocation. Compared with the five years prior to the adoption of MELD, during the five years after the adoption of MELD, there were decreases in both waiting list mortality (29 versus 22 percent) and dropout rates (39 versus 29 percent) [43]. The number of deaths on the waiting list decreased from 273 per 1000 patient-years at risk in 2005 to 173 per 1000 patient-years at risk in 2010.

There had been initial concern that adoption of the MELD scoring system might result in poorer transplant outcomes if livers were being allocated to patients who were "too sick" (those with high MELD scores). However, there was no significant change in early (three-month) patient or graft survivals in Era 2 (post-MELD) in the United States compared with Era 1 (pre-MELD) [42]. Similarly, there was no change in one-year posttransplant survival after the adoption of MELD in the Argentinian study [43]. These data demonstrate that the MELD allocation system has been successful in de-emphasizing waiting time as a major factor in prioritizing patients for liver transplantation. In addition, adoption of the MELD scoring system has been associated with increased transplantation rates without concomitant increased mortality rates.

Adoption of MELD has also been associated with other effects, some of which were not completely anticipated:

●In the post-MELD era, race was no longer associated with the likelihood of receiving a liver transplant, risk of death on the liver transplant waiting list, or risk of being removed from the waiting list due to being too ill [44].

●There was an increase in the number of combined liver-kidney transplants being performed because of the emphasis of MELD on renal function [45].

●More high-risk deceased donor livers (ie, livers from older donors or donation after circulatory death [DCD] livers) were being steered toward patients with lower MELD scores [46,47]. In the pre-MELD era, high-risk donor livers were typically used for patients in most urgent medical need of liver transplantation. However, it should also be noted that since the adoption of the MELD system for liver allocation, there has also been a national initiative to increase the use of high-risk donor livers in order to increase the number of available organs. As such, it is likely that the adoption of the MELD score for prioritizing deceased donor livers for allocation is not the sole explanation for use of more DCD livers in patients with lower MELD scores [48].

●Female patients in the post-MELD era are more likely than males to die or become too ill for transplant, whereas in the pre-MELD era the likelihood for males and females were similar [44]. The precise reasons for this are not completely clear, but studies suggest that female patients may be somewhat disadvantaged in the MELD system due to their generally smaller body mass and, therefore, lower creatinine levels [49,50]. It is also likely that donor liver allocation to females is impacted by factors outside the purview of the MELD score, including matching of organ size to recipient body size and the lower prevalence of HCC in females [51,52].

Applications beyond organ allocation — The MELD scoring system has prognostic value in other clinical settings for patients with liver disease beyond its application in the deceased donor liver allocation process [10,23,24,27,28,53-60]:

●Selecting patients for TIPS placement – The MELD score was developed initially to predict three-month survival among patients undergoing TIPS placement [4]. Using the MELD score, as modified by UNOS, the best outcomes with TIPS occur among patients with a MELD score less than 14, although, TIPS may reasonably be performed in patients with slightly higher MELD scores as well. The probability of mortality following TIPS can be calculated online.

As a general rule, TIPS should be avoided if possible in patients with a MELD score greater than 24, unless the procedure is being used as salvage therapy to control active variceal bleeding [61,62]. When deciding whether to carry out TIPS placement in patients with MELD scores between 15 and 24, clinical judgment and extensive discussion with the patient, family members, and interventional radiology regarding the risks of hepatic decompensation are required.

Patients with ascites and MELD scores greater than 24 who are reasonable liver transplant candidates are probably best served by foregoing TIPS placement and waiting for liver transplantation [63]. However, the projected waiting time for a deceased donor liver to become available can be quite prolonged. As a result, clinical circumstances may arise that necessitate consideration of TIPS to control complications from portal hypertension in patients with high MELD scores. Such patients should ideally be evaluated for liver transplantation before TIPS in case they develop irreversible hepatic decompensation after the procedure.

●Alcohol-associated hepatitis – The Discriminant Function (DF) has traditionally been used to predict survival in patients with alcohol-associated hepatitis (calculator 1) [53]. However, the DF may be somewhat limited in its applicability as it relies primarily on measurement of the prothrombin time, which is subject to variability among different laboratories. Several studies have evaluated the MELD score as a prognostic index in alcohol-associated hepatitis, and a simple online tool is available in which the MELD score has been calibrated to predict 90-day mortality among such patients.

One study suggested that among several scoring systems, the MELD score was the most accurate predictor of mortality in patients with alcohol-associated hepatitis [64].

Patients with a MELD score >20 are classified as having severe alcohol-associated hepatitis and are potential candidates for treatment with glucocorticoids [65]. The benefit of glucocorticoid therapy for severe alcohol-associated hepatitis is seen in patients with MELD scores ranging from 25 to 39 [66]. The management of acute alcohol-associated hepatitis is discussed separately. (See "Management and prognosis of alcoholic hepatitis".)

●Hepatorenal syndrome – A study of 105 consecutive patients with hepatorenal syndrome suggested that the MELD score may be a useful predictor of survival among patients with what was then defined as type 2 hepatorenal syndrome [54]. A MELD score of ≥20 was associated with significantly shorter transplantation-free survival compared with those patients with lower MELD scores (median survival of 3 versus 11 months, respectively). (See "Hepatorenal syndrome".)

●Acute liver failure (UNOS status 1A) – MELD is the scoring system used by UNOS for prioritizing organ allocation in adult patients with cirrhosis awaiting transplantation; however, the allocation process for patients with acute liver failure, designated as UNOS status 1A, is not based on the MELD score.

Although MELD is not currently used in clinical practice for UNOS status 1A patients, the accuracy of MELD in this patient population has been evaluated. In a study that included 720 adult status 1A liver transplantation candidates, patients with acute liver failure (non-acetaminophen induced) had the poorest overall survival on the liver transplant waiting list. Increasing MELD score was highly correlated with decreased survival in this group [27]. However, the mortality rates of patients with primary graft nonfunction and hepatic artery thrombosis were less tightly associated with the MELD score.

The MELD score has also been specifically assessed for its prognostic value in acetaminophen-induced hepatotoxicity [28]. In one study, a higher MELD score was predictive of the development of acute liver failure and hepatic encephalopathy, but once acute liver failure developed, the MELD score was not a more accurate predictor of survival than either the King's College Criteria or the international normalized ratio (INR) alone [28].

Similarly, in a study of patients with acute liver failure due to hepatitis A virus infection, MELD was only moderately accurate in predicting mortality risk (c-statistic 0.7) [56]. In a study utilizing UNOS data, it was demonstrated that patients who were designated status 1A had a mortality rate that was similar to that in patients with cirrhosis who had biologic MELD scores ranging from 36 to 40. Furthermore, the study demonstrated that status 1A patients were at lower mortality risk compared with patients with cirrhosis with biologic MELD scores >40. These investigators called into question the policy of prioritizing status 1A patients above all other liver transplant candidates [29]. Therefore, the relationship between MELD and survival in patients with acute liver failure is complex, and although some studies suggest that MELD may have some prognostic value in acute liver failure, there are likely other parameters that are important mortality predictors in this unique patient population. (See "Acute liver failure in adults: Etiology, clinical manifestations, and diagnosis" and "Acetaminophen (paracetamol) poisoning in adults: Treatment".)

●Acute variceal hemorrhage – Several studies have investigated the prognostic value of MELD in predicting mortality among patients presenting with acute variceal hemorrhage [23,24,58,59]. The studies suggest that higher MELD scores are associated with increased mortality rates. As an example, in a study of 178 patients with cirrhosis and acute variceal bleeding, a MELD score <11 was associated with a <5 percent mortality rate within six weeks, whereas a MELD score ≥19 was associated with a 20 percent mortality rate [24].

●Assessment of surgical mortality risk in liver disease – Patients with cirrhosis are at increased risk of perioperative morbidity and mortality, and the CTP score has traditionally been used to risk stratify these patients prior to surgical intervention. The MELD score has also been evaluated with respect to its ability to predict perioperative mortality in patients with cirrhosis undergoing various surgical procedures.

An online tool is available for determining the risk of postoperative mortality for several types of major surgery, including gastrointestinal, orthopedic, and cardiac surgery [22]. The choice of intervention for severe calcific aortic stenosis is discussed separately. (See "Choice of intervention for severe calcific aortic stenosis".)

LIMITATIONS OF THE MELD SCORE — The MELD score is vulnerable to variations in laboratory measurements. For example, despite being normalized for the sensitivity of thromboplastin, the international normalized ratio (INR) can vary across laboratories if thromboplastin derived from rabbit brain is used rather than recombinant thromboplastin, potentially leading to important differences in prioritization of patients according to MELD [67-69]. (See "Tests of the liver's biosynthetic capacity (eg, albumin, coagulation factors, prothrombin time)".)

The MELD score may also be influenced by the method by which serum creatinine is measured. Variability in serum creatinine measurement using different assays is particularly problematic in the presence of an elevated serum bilirubin concentration, although this can be circumvented by using an enzymatic method to measure serum creatinine, particularly when the total bilirubin level is >25 mg/dL [70,71]. Additionally, serum sodium levels may fluctuate with the use of diuretics or administration of free water.

OPTIMIZING THE MELD SCORE

Goals — Refinements of the MELD score continue to be proposed and are driven by the liver transplantation community's desire to increase the efficiency of allocation of donated livers. The proposed refinements of the MELD scoring system are promising steps toward evidence-based improvements in the liver allocation process [13,72,73]. These models require further vetting and assessment of the impact on waiting list mortality and transplantation outcomes before being considered for formal adoption. Additionally, as the field of liver transplantation evolves and the transplant patient population changes from a majority with hepatitis C virus-related cirrhosis towards more alcohol-associated cirrhosis and metabolic-associated fatty liver cirrhosis, the predictive accuracy of the MELD score may be reduced. Specifically, from 2002 to 2016, the concordance between the MELD score and three-month mortality decreased for patients with alcohol-related cirrhosis and those with cirrhosis related to metabolic-associated fatty liver [74]. This evolution over time in the liver transplant population opens further opportunities for refinement of the liver allocation process.

MELD 3.0 — An optimized version of the MELD score, named MELD 3.0, has been developed and includes additional variables of patient sex and serum albumin along with updated variable model coefficients and a revised upper limit for creatinine (3.0 mg/dL) [72]. In a large registry study of patients on the liver transplantation waiting list in the United States, MELD 3.0 had relatively high predictive accuracy and slightly outperformed MELD-Na (c-statistic 0.869 versus 0.862) in predicting 90-day waiting list mortality. Among 514 decedents who were reclassified (ie, either up- or down-categorized) using MELD 3.0, a net gain of 45 decedents (9 percent) had higher scores with the possible impact of a higher chance for receiving a donor liver [72]. Importantly, MELD 3.0, with the inclusion of patient sex in the model, addresses the sex disparity in liver transplantation whereby females have historically been less likely than males to receive a donor liver. This study suggests that MELD 3.0 shows promise for lowering waiting list mortality and improving organ allocation [75].  

SUMMARY AND RECOMMENDATIONS

●Background – The Model for End-stage Liver Disease (MELD) score is a prospectively developed and validated cirrhosis severity scoring system that uses a patient's laboratory values for serum bilirubin, serum creatinine, and the international normalized ratio (INR) to predict three-month survival. In patients with cirrhosis, an increasing MELD score is associated with increasing severity of hepatic dysfunction and increasing three-month mortality risk (figure 1). (See 'MELD overview' above.)

●MELD-Na score – The MELD-Na score includes serum sodium as a factor in the calculation of the MELD score and has been used by United Network for Organ Sharing (UNOS) for prioritizing allocation of deceased donor livers for transplantation. (See 'MELD-Na Score' above.)

●Clinical applications – Patients awaiting liver transplantation are ranked according to their MELD-Na score and stratified by blood type. Patients have their MELD-Na scores updated regularly with UNOS by the listing transplant center. The MELD-Na score may either increase or decrease while patients await liver transplantation. Time spent on the waiting list at a given MELD-Na score is used to break ties among patients with the same blood type. (See 'Prioritization for liver transplantation based on MELD score' above.)

There are some conditions associated with chronic liver disease that may result in impaired survival, but that are not directly accounted for in the MELD scoring system. Some of these conditions have been designated as standard MELD exceptions. Patients who meet specific disease-related criteria may be eligible for standard MELD exceptions and, as such, may receive an upgrade in MELD points. (See 'Standard MELD exceptions in liver transplantation' above.)

Standard MELD exceptions include:

•Hepatocellular carcinoma

•Hepatopulmonary syndrome

•Portopulmonary hypertension

•Familial amyloid polyneuropathy

•Primary hyperoxaluria

•Cystic fibrosis

•Hilar/Perihilar cholangiocarcinoma (the liver transplant center must have a UNOS approved protocol for the work-up and management of patients with cholangiocarcinoma undergoing transplantation)

•Hepatic artery thrombosis (occurring within the 14 days after liver transplant surgery, but not meeting criteria for status 1A)

The MELD scoring system has prognostic value in other clinical settings for patients with liver disease beyond its application in the deceased donor liver allocation process. (See 'Applications beyond organ allocation' above.)

●Optimizing the MELD score – Refinements of the MELD score have been proposed and are driven by the liver transplantation community's desire to increase the efficiency of allocation of donated livers. An updated model, termed MELD 3.0, may further improve predictive accuracy. (See 'Optimizing the MELD score' above.)