Model for End-stage Liver Disease (MELD)
Kiran Bambha, MD
Patrick S Kamath, MD
UpToDate performs a continuous review of over 375 journals and other resources. Updates are added as important new information is published. The literature review for version 15.1 is current through December 2006; this topic was last changed on October 23, 2006. The next version of UpToDate (15.2) will be released in June 2007.
INTRODUCTION — Prognostic models are useful in estimating disease severity and survival and are used to make decisions regarding specific medical interventions. These models are developed using analytical methods that involve determining the effects of variables of interest (eg, demographic data and laboratory values) on specific outcomes such as death.
Several prognostic models are currently used in healthcare 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. Examples of the latter include models for predicting survival in patients with primary biliary cirrhosis (PBC), primary sclerosing cholangitis (PSC), and alcoholic liver disease [2-5]. There are two models that are used commonly in the care of patients with chronic liver disease: the Child-Turcotte-Pugh (CTP) score and the more recently described Model for End-stage Liver Disease (MELD) [6-10].
MELD OVERVIEW — MELD 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 for prothrombin time (INR) to predict survival. As will be discussed below, the MELD score has been modified slightly since its original development so that cause of liver disease is no longer routinely included in the model. Furthermore, the revisions eliminated the possibility of negative scores. The revised model is currently used by the United Network for Organ Sharing (UNOS) in prioritizing allocation of organs for liver transplantation. It is calculated according to the following formula:
MELD = 3.8[Ln serum bilirubin (mg/dL)] + 11.2[Ln INR] + 9.6[Ln serum creatinine (mg/dL)] + 6.4
Where Ln is the natural logarithm.
For ease of use, several on-line calculators are now readily available for calculating the MELD score (
http://www.unos.org/resources/MeldPe...r.asp?index=98) [11]. In patients with chronic liver disease, an increasing MELD score is associated with increasing severity of hepatic dysfunction and risk of death [12].
DEVELOPMENT OF MELD — MELD was developed initially using data from a population of 231 cirrhotic patients seen at four medical centers in the United States who underwent elective transjugular intrahepatic portosystemic shunt (TIPS) procedure [8]. Of these patients, 110 died during follow-up, with 70 of these deaths occurring within three months after the procedure. The MELD scoring system was developed from these data using Cox proportional hazards regression modeling to identify patients whose predicted survival post-procedure would be three months or less.
Serum bilirubin, serum creatinine, INR for prothrombin time and etiology of liver disease (classified as alcoholic, cholestatic, viral hepatitis, and "other") were strong predictors of three month survival. Patients with cirrhosis secondary to either alcoholic or cholestatic liver disease demonstrated a slightly better survival compared to patients with viral or "other" disease etiologies.
The formula used for calculation of the risk score in this study was: 3.8[Ln serum bilirubin (mg/dL)] + 11.2[Ln INR] + 9.6[Ln serum creatinine (mg/dL)] + 6.4[etiology: 0 if cholestatic or alcoholic, 1 if other etiologies]. Patients with cirrhosis related to non-alcoholic and non-cholestatic liver disease (that is, patients with viral, metabolic, and autoimmune hepatitis as a cause of cirrhosis) have a score that is higher than patients with cholestatic and alcoholic liver disease, even when the bilirubin, creatinine, and INR for prothrombin time are identical. This model was also validated in an independent population of patients from the Netherlands who had undergone placement of TIPS [8].
Survival following TIPS is primarily determined by the severity of the underlying liver disease. Thus, it was hypothesized that the MELD score might be useful as a prognostic indicator in patients with a broader range of liver disease severity and etiology. In an initial study, predictive accuracy was evaluated in patients with cirrhosis who did not have TIPS placed [9]. Four independent patient populations (total n=2278) were studied including (1) hospitalized patients with hepatic decompensation; (2) ambulatory patients with non-cholestatic liver disease; (3) patients with PBC; and (4) a historical cohort of unselected patients with cirrhosis seen at Mayo Clinic Rochester at a time when liver transplantation was not available. Patients with hepatocellular carcinoma (HCC), recent alcohol use, intrinsic renal disease, sepsis, and those patients admitted for liver transplantation were excluded.
The MELD score was a good predictor of three month survival in all four patient populations, with concordance (c)–statistics of 0.87 for hospitalized patients; 0.80 for ambulatory non-cholestatic liver disease; 0.87 for PBC; and 0.78 for the historical cohort of cirrhotics. The c-statistic is equivalent to the area under the receiver operating characteristic curve. It tests the probability that, of randomly selected patients, the patient with a greater risk score is more likely to die within a three month period compared to a patient with a lower risk score. In this context, a c-statistic may range from 0 to 1, with 1 corresponding to a perfect test with 100 percent accuracy and 0.5 corresponding to a clinically useless test that is accurate only 50 percent of the time (an accuracy rate that could be obtained by chance alone). A c-statistic greater than 0.7 generally indicates a clinically useful test, with a c-statistic greater than 0.8 indicating excellent accuracy. A c-statistic of 0.8, for example, indicates that 8 out of 10 times a patient with a higher MELD score is more likely to die within three months compared to a patient with a lower MELD score.
Inclusion of complications of portal hypertension (ascites, encephalopathy, variceal bleeding, and spontaneous bacterial peritonitis) in the models did not improve predicted survival compared with the MELD score alone [9]. This result does not suggest that these complications are unassociated with decreased survival. What it suggests is that these complications are more likely to be associated with advanced liver disease as determined by the MELD score. On the other hand, one study suggested that the presence of a low serum sodium and persistent ascites helped identify a subset of patients with a high mortality rate despite a low MELD score [13].
Finally, due to difficulties that may arise with the inclusion of liver disease etiology in the model (eg, how to handle multiple etiologies of liver disease, such as alcoholic liver disease and HCV, in a single patient), the validity of excluding diagnosis from the model was determined. The exclusion of etiology of liver disease did not significantly affect the c-statistic for three month survival. Even when compared to the CTP score based upon examination by a single observer, and biochemistry values from a single laboratory, MELD was statistically a slightly better predictor of three month mortality in the hospitalized patients (c-statistic of 0.84 and 0.87, respectively) [14].
COMPARISON OF MELD AND CHILD-TURCOTTE-PUGH SCORING SYSTEMS — One of the best known and widely published scoring systems of liver disease severity is the CTP system. It was initially developed by Child and Turcotte in 1964 to risk stratify patients undergoing shunt surgery for portal decompression [6]. The original Child-Turcotte classification system was empirically derived and incorporated serum albumin, serum bilirubin, "nutritional status", ascites, and encephalopathy.
The Child-Turcotte system was criticized due to the ambiguity of some of its parameters. As a result, the Child-Turcotte system was subsequently modified by Pugh in 1972 [7]. Points were assigned to each element such that an overall score could be calculated. Further, "nutritional status" was replaced by prolongation in prothrombin time, expressed in seconds.
The original Child-Turcotte scoring system modified by Pugh is now commonly referred to as the Child-Pugh score, or more accurately, as the Child-Turcotte-Pugh score (CTP). The current CTP scoring system is based upon five parameters: serum bilirubin, serum albumin, prothrombin time, ascites and encephalopathy. The sum of the points for each of these five parameters gives the total score. Patients with chronic liver disease are placed in one of three classes (A, B, or C) (show table 1).
Although the CTP scoring system was derived empirically, it has proven to be a good predictor of outcome in patients with complications of portal hypertension [15]. Because of the simplicity in its calculation, the CTP score has been the traditional scale used by many clinicians for assessing the risk of mortality in patients with cirrhosis.
However, the CTP classification has several shortcomings. It is based upon subjective parameters such as the degree of ascites and encephalopathy. The quantification of the degree of abnormality in these two physical findings may vary from observer to observer, may be altered substantially by medical intervention (eg, the use of diuretics for ascites or lactulose for encephalopathy), and also vary depending upon the means of detection. As an example, the CTP score was developed during a time when ultrasound quantification of ascites was not routinely available. However, currently the availability of ultrasound allows even minor degrees of ascites to be detected that might not have been detected on physical examination. Similarly, the degree of encephalopathy is highly subjective. (See "Clinical manifestations and diagnosis of hepatic encephalopathy").
In addition, the CTP score is limited in its discriminatory capacity due to both a "ceiling " and a "floor" effect. A patient with a serum bilirubin level of 4 mg/dL, for example, would be assigned the same number of points as a patient with a bilirubin of 30 mg/dL, even though the degree of elevation in serum bilirubin level is an important prognostic indicator in patients with chronic liver disease. Furthermore, a serum albumin of 2.8 or 1.5 g/dL (a difference that is potentially clinically important) are assigned the same score. Finally, the CTP score divides patients into one of only three classes (A, B, or C) and cannot distinguish among patients within a single class.
Another problem with the CTP score is the use of the prothrombin time and serum albumin levels. Laboratories vary with regard to the sensitivity of the thromboplastin reagents used for measuring prothrombin time. The normal values for serum albumin may also differ across different medical centers. There may be enough variation in the reporting of the prothrombin time and serum albumin such that there may be a significant point change in the CTP score due to laboratory techniques alone [9]. On the other hand, the MELD score is not entirely immune to this problem. Despite being normalized for the sensitivity of thromboplastin, the 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 [16]. (See "Tests of the liver's biosynthetic capacity").
These shortcomings are of great significance in clinical practice, especially in stratifying patients from different centers according to mortality risk. A desirable disease severity scoring system that will have practical utility should have the following characteristics [17]: Rely on a few, readily available, objective variables Generally applicable to a heterogeneous group of patients with chronic liver disease Able to distinguish disease severity along a broad continuum Validated in diverse subsets of patients with chronic liver disease
The MELD score meets all of these criteria. It is based upon three laboratory parameters (serum bilirubin, serum creatinine, and INR for prothrombin time), which are easily and reliably measured in almost all medical laboratories. Most importantly, MELD, which was developed in patients with chronic liver disease undergoing elective TIPS, has subsequently been validated as an accurate survival predictor in other subsets of patients with chronic liver disease [18]. Studies comparing the accuracy of MELD to the CTP score for predicting survival of patients on the transplant waiting list have shown inconsistent results [19]. Nevertheless, the relatively objective criteria used in the MELD score have led to its widespread adoption as a predictor of mortality in patients awaiting liver transplantation.
USE FOR SELECTION OF PATIENTS FOR TIPS — As noted above, the MELD score was developed initially in patients undergoing TIPS [8].
It is preferable to use the original MELD score (which included etiology of liver disease) when making predictions in patients undergoing TIPS since clinical considerations are different than in the transplant setting. Consider a patient with alcoholic liver disease, for example. Those listed for liver transplantation are required to have had a period of alcohol abstinence (usually at least six months). Thus, ongoing alcohol use is not a factor in predicting their likelihood of dying. In contrast, alcohol abstinence is not a requirement in patients undergoing TIPS in whom mortality after TIPS may be influenced by ongoing alcohol consumption. Alcoholic patients undergoing TIPS have a better prognosis than patients with viral hepatitis who have identical values for prothrombin time, creatinine and bilirubin [8]. This probably reflects the potential for improved hepatic function in patients who actually do stop drinking following TIPS.
The value of the MELD score can be illustrated by the following examples. A patient with hepatitis C related cirrhosis with a creatinine of 1.0 mg/dL, bilirubin of 1.1 mg/dL, and INR for prothrombin time of 1.2, has a MELD score of 9. The risk of mortality at three months is approximately 33 percent if this patient undergoes elective TIPS. In contrast, the MELD score is 39 and risk of mortality within three months is 94 percent if a patient with alcoholic liver disease has a creatinine of 2.3 mg/dL, bilirubin of 17.4 mg/dL, and INR for prothrombin time of 3.4.
When using the original MELD score formula (which includes liver disease etiology), patients with a MELD score of less than eight generally do very well post TIPS, while those with a score greater than 18 have a poor outcome. When using the MELD score as modified by UNOS, best outcomes are seen in patients with a MELD score less than 14. TIPS should generally be avoided in patients with the MELD score greater than 24 unless the procedure is used as salvage therapy to control active variceal bleeding. Clinical judgment and extensive discussion with the patient and family members is required in deciding whether to carry out TIPS in patients with MELD scores between 15 and 24. Patients with MELD score greater than 24 are probably best served by liver transplantation. This approach to the selection of patients for placement of TIPS, whether the procedure is carried out for refractory variceal bleeding or for refractory ascites, has been validated independently [20].
USE OF MELD IN ALLOCATION OF ORGANS FOR LIVER TRANSPLANTATION — In the 1990s, as the number of patients listed for liver transplantation increased and the number of available deceased donor livers remained stable, the number of patients dying while awaiting transplantation rose linearly. The time spent on the waiting list became an important, though undesirable, deciding factor in the allocation process. It was demonstrated that time spent on the liver transplant waiting list did not correlate with the risk of death while awaiting transplant [21]. Because of increasing concerns regarding this disparity in liver allocation, the Department of Health and Human Services issued a mandate in 1998 that deceased donor livers for transplantation be prioritized in a more equitable manner with a de-emphasis on time spent on the waiting list [22,23].
The United Network for Organ Sharing (UNOS) initially adopted the Child-Turcotte-Pugh (CTP) scoring system for its liver transplantation prioritization system. However, it soon became apparent that the CTP score was not sufficient to resolve the dominance of waiting time as a deciding factor.
The failure of the CTP score was due to several factors. First, the allocation system adopted by UNOS defined only three categories of disease severity in patients with chronic liver disease (status 2A, 2B, and 3). The usual listing criteria for liver transplantation were based upon a CTP score seven. Patients with a CTP score of seven to nine were placed in status 3, while those with a CTP score greater than 10 were status 2B. Patients were placed in status 2A if it was determined that they were at risk of dying within seven days.
The prevailing allocation system was thus limited in its ability to discriminate among patients with regard to disease severity. Each category contained numerous patients and therefore waiting time continued to be a major, albeit unintended, factor in the allocation process. Furthermore, determining whether a patient would die within seven days was not based upon validated criteria.
During the liver transplantation community's search for a more equitable prioritization system, the model for end-stage liver disease (MELD) score emerged as an ideal candidate for a liver disease severity scoring system. An illustrative study included 3437 adult patients who were listed for liver transplantation between 1999 and 2001 [24]. 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. Mortality was 1.9 percent for patients with MELD less than 9 and a 71.3 percent for patients with a MELD 40. The c-statistic for MELD's accuracy in predicting three-month mortality on the waiting list was 0.83, compared with a c-statistic of 0.76 for the CTP score.
In February 2002, MELD was adopted by UNOS as the basis for deceased donor liver allocation for adult patients (18 years). Once a patient has been listed for liver transplantation by a transplant center, the MELD score is then used to determine that individual patient's immediate need for transplantation. The MELD score has also been considered for adoption by transplant programs in several other countries or regions around the world.
The original MELD score was a continuous numbering system with scores ranging from negative values to positive infinity. In practice, most patients have a score between 0 and 60. Because a negative score might create confusion in the minds of patients and their families (a negative score, in fact, indicates that the patient has a better survival than the rest of the cohort), UNOS modified the MELD scoring system to eliminate negative values.
To avoid a negative MELD score, any laboratory values less than 1.0 are set to 1.0. Thus, patients with a combination of an international normalized ratio for prothrombin in time (INR) of 1 or less, a serum creatinine of 1mg/dL or less, and serum bilirubin of 1 mg/dL or less will receive the minimum score of six points. To avoid an unfair advantage to patients with intrinsic renal disease, the maximum serum creatinine level is set to 4.0 mg/dL, which is also the value that is automatically assigned to patients on dialysis. The upper limit of MELD is capped at 40 points by UNOS and thus, MELD scores for prioritization of organ allocation for liver transplantation range between 6 and 40.
Patients have their MELD scores updated and forwarded regularly to UNOS by the listing transplant center according to UNOS directives [11]. As a general rule, patients who are most severely ill will have their MELD scores updated more frequently than patients whose illness is less acute. Patients with a MELD score 25, for example, have their scores updated weekly. Patients may have their MELD score updated more often if they experience a decline in health status.
Patient selection for liver transplantation also depends upon blood type, the number of other patients listed for transplant in the local area, the severity of illness of those other patients as determined by their MELD scores, and the number of deceased donor livers available in the local area. Among patients waiting in the same locale with identical blood type and MELD score, time spent on the waiting list at that score is used to break ties. In addition, if a patient experiences an increase in the MELD score (worsening of liver disease severity), the waiting time clock is set to zero and started anew at the higher score. However, based upon current UNOS policy, if a patient experiences a decline in the MELD score (decreased liver disease severity), the time accumulated at the higher MELD score is maintained and added to however much time is accumulated at the lower score.
Under the current deceased donor liver allocation system, patients with acute liver failure (UNOS status 1) are exempt from the standard prioritization process outlined above. These patients are given priority over all other patients listed for liver transplantation due to the high acuity and mortality of their illness.
MELD exceptions in liver transplantation selection — There are some special conditions, including hepatocellular carcinoma (HCC), hepatopulmonary syndrome, and select systemic metabolic diseases that are associated with chronic liver disease and result in impaired survival, but are not directly accounted for in the MELD scoring system. Thus, the calculated MELD score for such patients may not accurately reflect their increased mortality. As a result, additional MELD points may be assigned to patients with these conditions in order to adjust for their increased mortality.
Patients with hepatocellular carcinoma — Patients with HCC are assigned specific MELD scores based upon tumor burden and the associated predicted three month survival for HCC, rather than on the traditional MELD parameters (ie, serum bilirubin, serum creatinine, and INR). The decision to give exceptional status to patients with HCC is based upon two observations. First, patients with HCC who meet criteria for liver transplantation have a post-procedure survival rate that is similar to other patients undergoing liver transplantation. (See "Liver transplantation for hepatocellular carcinoma").
Second, many patients with HCC do not demonstrate the degree of hepatic synthetic dysfunction necessary to give them a competitive MELD score and thus would be given lower priority based on their calculated MELD score alone. A low calculated MELD score would translate into increased waiting time with concomitant increased risk of tumor growth during the waiting period with its associated morbidity and mortality. As a result, a decision was made by UNOS to assign a pre-determined MELD score to selected patients with HCC so that they would receive an early transplant.
Prioritization scores for patients with HCC are based upon tumor size and number. The staging schema used by the American Liver Tumor Study Group (ALTSG) underlying prioritization differs slightly (show table 2) from the standard TNM staging system (show table 3). Under the current UNOS policy for allocation of deceased donor livers for transplantation, patients with ALTSG stage II HCC (single HCC between 2 and 5 cm, or two to three lesions, none greater than 3 cm) who are potential liver transplant candidates are assigned a MELD score of 22. The assignment of this relatively high MELD score reflects both an estimated three month mortality rate of 15 percent in such patients, and the risk of the tumor progressing to a level that would render patients ineligible for transplantation within three months of listing.
Since it is not always feasible or desirable to biopsy hepatic lesions suspected of being HCC, the following criteria are acceptable to support the diagnosis of HCC: vascular blush corresponding to the area of suspicion on imaging studies, an alpha-fetoprotein (AFP) concentration >200 mg/L, or an arteriogram that confirms presence of a tumor. Patients with chronic liver disease who have an increasing AFP concentration of 500 mg/L may be listed with a MELD score equivalent to an 8 percent mortality risk, even if the imaging studies do not show evidence of a tumor. However, the predictive accuracy of this cutoff has been questioned [25].
After patients have been on the waiting list for three months with a MELD score of 22, the score is increased to reflect a corresponding increase in the estimated three-month mortality rate. For every subsequent three months spent on the waiting list, patients receive additional points corresponding to an estimated increased mortality of 10 percent. As an example, MELD scores of 30 and 32 points in an HCC patient correspond to three-month mortality rates of 35 and 45 percent, respectively.
Under the current UNOS liver allocation policy, patients with HCC characterized by a single lesion less than 2 cm in diameter (ALTSG stage I, show table 2) are not assigned a higher MELD priority score. Rather, their calculated MELD score is used for listing priority. Similarly, patients with more advanced HCC, as characterized by a single tumor greater than 5 cm in diameter or more than three tumors, are not assigned higher MELD scores either. Liver transplantation centers may consider these patients with larger HCC tumor burden for transplant listing on a case-by-case basis but the patients will be UNOS listed at their calculated MELD scores with no additional priority given for the diagnosis of HCC.
Other MELD exceptions — Patients diagnosed with hepatopulmonary syndrome and a PaO2 <60 mmHg on room air may be approved for additional MELD points [26]. Similarly, patients with familial amyloidosis or primary oxaluria may also receive additional MELD points. There are a variety of other special situations in which individual liver transplantation centers may submit a petition to a regional review board for additional MELD points in order to give the patient a score that better reflects their risk of death without transplantation. The merits of these cases are decided individually.
IMPACT OF MELD ON LIVER ALLOCATION FOR TRANSPLANTATION — The introduction of the MELD score has improved liver allocation. A study using data from UNOS compared the outcomes of deceased donor liver transplantation in the pre-MELD era (2/27/01 to 2/26/02, Era 1) to the post-MELD era (2/27/02 to 2/26/03, Era 2) [27]. In the post-MELD era there was a 12 percent reduction in new patients registered onto the liver transplant waiting list, particularly for patients with the lowest MELD scores. There was a 10.2 percent increase in the number of deceased donor liver transplantations performed and a 3.5 percent decrease in the number of deaths on the waiting list in Era 2. The mean MELD score at the time of transplant was also higher in Era 2 compared to Era 1.
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" (ie, with high MELD score). However, there was no significant change in early (three month) patient and graft survival in Era 2. Not surprisingly, the number of patients with HCC who underwent liver transplantation was also increased in Era 2 due to the exceptional status given to these patients. The authors concluded that the new 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 appears to have been associated with increased transplantation rates without concomitant increased mortality rates.
FURTHER APPLICATIONS OF THE MELD SCORING SYSTEM — The MELD scoring system has prognostic value in populations of patients with liver disease outside of the liver transplantation setting [18].
Transjugular intrahepatic portosystemic shunts (TIPS) — The MELD scoring system was initially developed as a prognostic model for patients undergoing elective TIPS, a setting in which it performed better than the CTP score in predicting three-month mortality. (See "Treatment of diuretic-resistant ascites in patients with cirrhosis"). Thus, MELD can identify a subgroup of high-risk cirrhotics who, due to high post-procedure mortality, would not be suitable candidates for TIPS unless they are expected to undergo early liver transplantation. As a result, unless TIPS is indicated for emergency purposes, patients with high MELD who are near the top of the list for liver transplantation should generally not receive TIPS. TIPS would be best utilized in patients who have MELD scores associated with liver transplant list waiting times of greater than three months [28]. Consistent with these recommendations are the results of a study of 119 patients undergoing elective TIPS procedure demonstrating an increased 30-day mortality rate post-TIPS in patients with MELD score >24 compared to patients with MELD 24 (60 versus 5 to 25 percent, respectively) [8].
Alcoholic hepatitis — The Discriminant Function (DF) has traditionally been used to predict survival in patients with alcoholic hepatitis [5]. However, the DF may be somewhat limited in its applicability as it relies primarily upon measurement of the prothrombin time, which may be subject to variability among different medical laboratories. The prognostic value of MELD in patients with alcoholic hepatitis has suggested that it may be a clinically useful predictor of 30- and 90-day mortality. (See "Treatment of alcoholic liver disease").
Hepatorenal syndrome — A study of 105 consecutive patients with hepatorenal syndrome suggested that the MELD score may be a useful predictor of survival [29]. Among patients with type 2 hepatorenal syndrome, a MELD score of 20 was associated with significantly shorter transplant-free survival compared with those with lower values (median survival of 3 versus 11 months, respectively). (See "Diagnosis and treatment of hepatorenal syndrome").
Cirrhotics with sepsis unrelated to spontaneous bacterial peritonitis — The MELD score was the only prognostic marker useful for predicting mortality in patients with cirrhosis and sepsis unrelated to spontaneous bacterial peritonitis in one report [30]. Three month survival was greater than 90 percent in patients with MELD 20 compared to approximately 60 percent in those with MELD 20. Patients who developed renal failure associated with sepsis had worse overall mortality even if the renal failure was reversible. Interestingly, the MELD score was a more accurate predictor of survival than the development of renal failure alone; this may be explained, in part, by the fact that the MELD score reflects prognosis related to both renal and liver failure.
UNOS status 1 — Although MELD is now the scoring system used by UNOS for prioritizing organ allocation in patients with chronic liver disease awaiting transplantation, the allocation process for patients with acute liver failure (UNOS status 1) remained unchanged and currently does not use the MELD score. In UNOS status 1 patients, deceased donor livers are allocated according to the patient's blood type, waiting time on the liver transplant list, and geography. Patients categorized as status 1 are severely ill with a very high short-term mortality. They may be divided into two subsets: (1) fulminant hepatic failure, and (2) early graft failure after liver transplantation, including graft primary non-function and hepatic artery thrombosis occurring in the immediate post-transplant setting requiring urgent retransplantation.
Although MELD is not currently used in clinical practice for UNOS status 1 patients, the accuracy of MELD in this patient population has been evaluated in a study that included 720 adult status 1 transplant candidates [31]. Patients with fulminant hepatic failure (not due to acetaminophen toxicity) had the poorest overall survival on the liver transplant waiting list. Increasing MELD score was highly correlated with decreased survival. However, the mortality of patients with graft primary non-function did not appear to be associated with the MELD score. Thus, the results of this study suggest that MELD may have prognostic value in UNOS status 1 with fulminant hepatic failure, perhaps warranting higher priority compared to patients with early graft failure post-transplantation.
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 been traditionally used to risk stratify these patients prior to surgical intervention. More recently, the MELD score has been evaluated with respect to its ability to predict perioperative mortality in certain surgical settings. (See "Assessing surgical risk in patients with liver disease").
SUMMARY — MELD 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 INR to predict survival. The MELD score, as currently used by UNOS in prioritizing allocation of organs for liver transplantation, is calculated according to the following formula:
MELD = 3.8[Ln serum bilirubin (mg/dL)] + 11.2[Ln INR] + 9.6[Ln serum creatinine (mg/dL)] + 6.4
For ease of use, several on-line calculators are now readily available for calculating the MELD score (
http://www.unos.org/resources/MeldPe...r.asp?index=98) [11].
The MELD score also has prognostic value in several clinical settings outside of liver transplantation including predicting mortality associated with: alcoholic hepatitis, hepatorenal syndrome, acute liver failure, sepsis in cirrhosis, surgical procedures in chronic liver disease patients, and the TIPS procedure (although it may be preferable to use the original MELD score, which includes the etiology of liver disease, when making predictions about patients undergoing TIPS).
Use of UpToDate is subject to the Subscription and License Agreement. REFERENCES 1.
www.openclinical.org/aisp_apache.html (Accessed 6/12/06).
2. Dickson, ER, Grambsch, PM, Fleming, TR, et al. Prognosis in primary biliary cirrhosis: Model for decision making. Hepatology 1989;10:1.
3. Grambsch, PM, Dickson, ER, Kaplan, M, et al. Extramural cross-validation of the Mayo primary biliary cirrhosis survival model establishes its generalizability. Hepatology 1989;10:846.
4. Kim, WR, Therneau, TM, Wiesner, RH, et al. A revised natural history model for primary sclerosing cholangitis. Mayo Clin Proc 2000; 75:688.
5. Maddrey, WC, Boitnott, JK, Bedine, MS, et al. Corticosteroid therapy of alcoholic hepatitis. Gastroenterology 1978; 75:193.
6. Child III, CG, Turcotte, JG. Surgery and portal hypertension. In: Child III CG, ed. The Liver and Portal Hypertension, Saunders, Philadelphia 1964. p.50.
7. Pugh, RN, Murray-Lyon, IM, Dawson, JL, et al. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg 1973; 60:646.
8. Malinchoc, M, Kamath, PS, Gordon, FD, et al. A model to predict survival in patients undergoing transjugular intrahepatic portosystemic shunts. Hepatology 2000; 31:864.
9. Kamath, PS, Wiesner, RH, Malinchoc, M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology 2001; 33:464.
10. D'Amico, G, Garcia-Tsao, G, Pagliaro, L. Natural history and prognostic indicators of survival in cirrhosis: A systematic review of 118 studies. J Hepatol 2006; 44:217.
11.
www.unos.org (Accessed 6/12/06).
12. Freeman, RB Jr, Wiesner, RH, Harper, A, et al. The new liver allocation system: moving toward evidence-based transplantation policy. Liver Transpl 2002; 8:851.
13. Heuman, DM, Abou-Assi, SG, Habib, A, et al. Persistent ascites and low serum sodium identify patients with cirrhosis and low MELD scores who are at high risk for early death. Hepatology 2004; 40:802.
14. Wiesner, RH, McDiarmid, SV, Kamath, PS, et al. MELD and PELD: application of survival models to liver allocation. Liver Transpl 2001; 7:567.
15. Christensen, E, Schlichting, P, Fauerholdt, L, et al. Prognostic value of Child-Turcotte criteria in medically treated cirrhosis. Hepatology 1984; 4:430.
16. Trotter, JF, Brimhall, B, Arjal, R, Phillips, C. Specific laboratory methodologies achieve higher model for endstage liver disease (MELD) scores for patients listed for liver transplantation. Liver Transpl 2004; 10:995.
17. Institute of Medicine. Analysis of waiting times. In: Committee on Organ Transplantation. Assessing current policies and the potential impact of the DHHS final rule. Washington DC: National Academy Press, 1999;57.
18. Said, A, Williams, J, Holden, J, et al. Model for end stage liver disease score predicts mortality across a broad spectrum of liver disease. J Hepatol 2004; 40:897.
19. Cholongitas, E, Marelli, L, Shusang, V, et al. A systematic review of the performance of the model for end-stage liver disease (MELD) in the setting of liver transplantation. Liver Transpl 2006; 12:1049.
20. Ferral, H, Gamboa, P, Postoak, DW, et al. Survival after elective transjugular intrahepatic portosystemic shunt creation: prediction with model for end-stage liver disease score. Radiology 2004; 231:231.
21. Freeman, RB Jr, Edwards, EB. Liver transplant waiting time does not correlate with waiting list mortality: implications for liver allocation policy. Liver Transpl 2000; 6:543.
22. Institute of Medicine. Analysis of waiting times. In: Committee on Organ Transplantation. Assessing current policies and the potential impact of the DHHS final rule. National Academy Press, Washington DC 1999. p.57.
23. Organ Procurement and Transplantation Network--HRSA. Final rule with comment period. Fed Regist 1998; 63:16296.
24. Wiesner, R, Edwards, E, Freeman, R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology 2003; 124:91.
25. Kemmer, N, Neff, G, Kaiser, T, et al. An analysis of the UNOS liver transplant registry: High serum alpha-fetoprotein does not justify an increase in MELD points for suspected hepatocellular carcinoma. Liver Transpl 2006; 12:1519.
26. Swanson KL, Wiesner RH, Krowka MJ. Natural History of Hepatopulmonary Syndrome: Impact of Liver Transplantation. Hepatology 2005; 41:1122.
27. Freeman, RB, Wiesner, RH, Edwards, E, et al. Results of the first year of the new liver allocation plan. Liver Transpl 2004; 10:7.
28. Rosado, B, Kamath, PS. Transjugular intrahepatic portosystemic shunts: an update. Liver Transpl 2003; 9:207.
29. Alessandria, C, Ozdogan, O, Guevara, M, et al. MELD score and clinical type predict prognosis in hepatorenal syndrome: Relevance to liver transplantation. Hepatology 2005; 41:1282.
30. Terra, C, Guevara, M, Torre, A, Gilabert, R. Renal Failure in Patients With Cirrhosis and Sepsis Unrelated to Spontaneous Bacterial Peritonitis: Value of MELD Score. Gastroenterology 2005; 129:1944.
31. Kremers, WK, van IJperen, M, Kim, WR, et al. MELD score as a predictor of pretransplant and posttransplant survival in OPTN/UNOS status 1 patients. Hepatology 2004; 39:764.
New Search Contents My UpToDate CME 5.0 Help
©2007 UpToDate® •
www.uptodate.com