Priority for liver transplantation is currently based on the Model for Endstage Liver Disease (MELD) score, a mathematical function which includes the following objective variables: bilirubin, creatinine (Cr), and international normalized ratio (INR). We have noted that specific laboratory methodologies may yield consistently higher values of bilirubin, Cr, and INR. Therefore, we performed a study to determine if higher MELD scores could be obtained by utilizing laboratory methodologies selected to return higher laboratory values than standard methodologies used in our hospital's clinical laboratory. Phlebotomy was performed for routine clinical indications in 29 consecutive patients listed for liver transplantation. MELD scores were calculated using bilirubin, Cr, and INR from laboratory methods in our hospital's clinical laboratory (designated Lab #1) and from 2 other clinical laboratories (designated Lab #2 and Lab #3). The mean MELD score in our hospital's clinical laboratory (Lab #1) (13.6) was not significantly different than in Lab #2 (14.7), but in Lab #3, it was significantly higher by 20% (17.1), P < .03. Virtually all of the difference in MELD score between our hospital's clinical laboratory (Lab #1) and Lab #3 could be attributed to the INR, which was significantly higher by 26% in Lab #3 (1.9) vs. Lab #1 (1.4), P < .00002. Using MELD scores calculated from our hospital's clinical laboratory, the average change in priority for liver transplantation was from the 58th percentile to the 77th percentile (compared to Lab #3), P = .01. In conclusion, patients listed for liver transplantation at our center achieved significantly higher MELD scores and therefore a higher priority for liver transplantation by using laboratory methodologies that yield higher INR values than our hospital laboratory. The selection of laboratory methodologies may have a significant impact on MELD score. (Liver Transpl 2004;10:995–1000.)
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
Priority for liver transplantation in the United States is based on the Model for Endstage Liver Disease (MELD) score, which is a mathematical function predictive of 3-month mortality.1, 2 Before implementation of the MELD score on February 28, 2002, waiting time as well as assessments of severity-of-illness (anticipated 7-day mortality, severity of ascites, and severity of encephalopathy) were important factors in the determination of priority for transplantation. Under this old allocation system (utilized until February 28, 2002), inequalities developed in the allocation of deceased-donor livers. Specifically, there were large disparities in access to transplantation across geographic regions, as well as difficulties in fairly prioritizing patients for transplantation because of the inclusion of criteria. As a result, the MELD score was developed as the primary means of prioritizing patients for transplantation to reduce the importance of waiting time and to permit only objective variables (bilirubin, Cr, and INR) to determine status for liver transplantation. The MELD score theoretically permits no subjective variation in priority for liver transplantation from center-to-center. That is, a patient with a MELD score of 20 in New York would have the same MELD score in Los Angeles, and therefore the same status on the transplant list. However, we have discovered that the laboratory methods used to determine the MELD score may introduce significant and clinically relevant subjective variation in the MELD score.
We performed a study to determine if higher MELD scores could be obtained for patients listed for liver transplantation by using selected laboratory methods that: 1) differ in terms of instrumentation or methodology from those used in our clinical laboratory, and 2) were specifically selected to return the highest values for bilirubin, Cr, and INR.
MELD, model for endstage liver disease; Cr, creatinine; INR, international normalized ratio.
Patients and Methods
A total of 29 consecutive patients listed for liver transplantation with endstage liver disease undergoing phlebotomy for routine clinical indications in our outpatient clinical laboratory at the University of Colorado Health Sciences Center were prospectively enrolled into this study, which was approved by the Colorado Multi-institutional Investigational Review Board (COMIRB). Blood samples from each patient were divided into two aliquots. The first aliquot was analyzed for total bilirubin, Cr, and INR in our hospital's (University Hospital of the University of Colorado Health Sciences Center) clinical laboratory, which is designated Lab #1. The second aliquot was frozen (at not less than −20°C for not more than 3 weeks) and then analyzed for total bilirubin, Cr, and INR at two other clinical laboratories. Labs #2 and #3 were selected because we anticipated slightly higher bilirubin values due to a slightly higher upper limit of normal (1.2 mg/dL for Lab #2 and 1.3 mg/dL for Lab #3), compared to our hospital's clinical laboratory (1.0 mg/dL, Lab #1). In addition, we anticipated different INR values in Labs #2 and #3 compared to our hospital's clinical laboratory (Lab #1), because of the use of different thromboplastin reagents. The specific methods for each assay are as follows, and are shown in Table 1. MELD scores were calculated using bilirubin, Cr, and INR from our hospital clinical laboratory (Lab #1) and the 2 other clinical laboratories, Labs #2 and #3.2, 3 Changes in priority for transplantation were measured based on change in percentile on the transplant list at our center, where the 0th percentile is the lowest position with the lowest MELD score and the 100th percentile is the highest position with the highest MELD score. Statistical analysis was performed using Student's t-test with Microsoft Excel (Microsoft Corporation, Redmond, WA).
Patient demographic data is depicted in Table 2, and is typical for patients listed for liver transplantation. The bilirubin, Cr, INR, and MELD scores from Labs #1, #2, and #3 for each patient are shown in Figs. 1–4. The mean bilirubin, Cr, INR, and MELD score for the three clinical laboratories are shown in Table 3, along with the standard deviation and minimum and maximum values for the cohort. Comparisons of mean values between our hospital's laboratory values (Lab #1) and the other two laboratories (Labs #2 and #3) were performed. Comparisons of values from Labs #1 and #2 found no difference between bilirubin, Cr, INR, or MELD score. Comparisons of values from Labs #1 and #3 found no difference between bilirubin or Cr. However, the INR from Lab #3 (1.9) was significantly higher by 26% compared to Lab #1 (1.4), P = .00002 and the MELD score from Lab #3 (17.1) was significantly higher by 20% compared to Lab #1 (13.6), P < .03. Comparisons were made to determine how many patients had a change in MELD score based on values from Lab #1 vs. Labs #2 and #3. Compared to MELD scores at Lab #1, the MELD scores at Lab #2 were higher in 16 patients, the same in 7 patients, and lower in 6 patients. Compared to MELD scores at Lab #1, the MELD scores at Lab #3 were higher in 27 patients, and the same in 2 patients. Using MELD scores calculated from our hospital's clinical laboratory (Lab #1), the average change in priority for liver transplantation was from the 58th percentile to the 77th percentile (compared to Lab #3), P = .01.
Table 2. Patient Demographics
Mean age (mean ± SD)
52.4 ± 8.5
Etiology of liver disease
Primary sclerosing cholangitis
Primary biliary cirrhosis
Table 3. Mean Bilirubin, Creatinine, INR, and MELD vs. Clinical Laboratory
These results demonstrate that patients listed for liver transplantation at our center achieved higher MELD scores and therefore a higher priority for liver transplantation using laboratory methods that were specifically selected to yield higher values for bilirubin, Cr, and INR than standard methods used in our hospital's clinical laboratory. If these results are confirmed to be applicable to all patients listed for transplantation, then the selection of specific laboratory methodologies could have an important widespread impact on MELD scores and thus on priority for liver transplantation. The most obvious implication is that a patient could gain a relatively higher priority for transplantation based on the selection of laboratory methodologies. In addition, posttransplantation outcomes are frequently measured based on severity-of-illness as determined by the MELD score at the time of transplantation. Our results indicate that MELD scores could vary significantly from center-to-center based on the selection of laboratory methodologies used to determine the MELD score at each center. As a result, comparisons of outcomes between transplant centers, based on MELD score, could be complicated by the potential variation in MELD score from center-to-center. However, there is no current analysis to determine whether MELD scores vary from center-to-center and the extent to which this variation might occur.
Of the 3 laboratory variables measured (bilirubin, Cr, and INR), the INR exhibited the greatest variability based on laboratory methodology. In fact, virtually all of the difference in the MELD score between our clinical laboratory (Lab #1) and Lab #3 may be attributed to differences in the INR. Hepatologists and transplant surgeons may be surprised by this finding. However, the INR was devised primarily to standardize the anticoagulation effects of warfarin, and was not designed to provide a reproducibly precise assessment of severity-of-illness in patients with endstage liver disease. Standardization of therapeutic anticoagulation became necessary when clinicians noted that the various thromboplastin reagents used to determine prothrombin time differed markedly in their responsiveness to the anticoagulant effects of warfarin.4 To rectify this problem, the World Health Organization proposed the INR, which is a correction formula that adjusts for the variable sensitivities of different thromboplastin reagents.5 The INR is defined by the following equation: INR = (patient's prothrombin time/mean normal prothrombin time)x, where x = International Sensitivity Index. The INR allows for standardization of prothrombin time ratio (determined by any thromboplastin reagent) to a reference World Health Organization thromboplastin standard. With this standardization, the INR allows safe and effective dosing of warfarin, independent of the sensitivity of the thromboplastin reagent.
However, to our knowledge, there has never been a formal demonstration of reproducibly precise INR values (with various thromboplastin reagents used in different clinical laboratories) in patients with endstage liver disease. In fact, there is evidence to the contrary. Three studies have shown a wide variation in the INR in liver patients based on the selection of the thromboplastin reagent used by the clinical laboratory. Robert and Chazouilleres6 documented the mean INR and range of INR for 29 patients on anticoagulants and in 27 patients with liver disease for 7 types of thromboplastin reagents. They found a large and statistically significant variation in the mean INR for the 27 liver patients between 7 specific thromboplastins. The range of the mean INR for the thromboplastin reagents was 2.3–4.1 (a 40% difference) for the patients with liver disease, P = .007. As expected, the difference in INR for the anticoagulation patients was much smaller, range 3.0–3.6 (a 16% difference), P = not significant. Another study using 3 different thromboplastins found similar variation in the mean INR in 29 liver patients, but no difference in 31 patients on warfarin.7 The range in mean INR for the liver patients was significantly different: 1.88 (recombinant human thromboplastin) to 2.63 (rabbit brain thromboplastin) (a 29% difference), P < .0001. However, the range in INR for the anticoagulation patients was not significantly different, at 2.63–2.75, which is a 4% difference, P = not significant. Denson et al.8 compared INR values in 20 patients with liver impairment using human and rabbit thromboplastin and reported a difference in INR between the two reagents of 25%. The variation in INR noted in these 3 studies is 25 to 40%, and is similar to our cohort in which the range was 1.4–1.9 or 26%. The reason(s) for the wide variation in INR for liver patients compared to patients receiving warfarin (Coumadin; DuPont, Wilmington, DE), is not known. One explanation may be the different mechanisms of prothrombin time elevation caused by warfarin and liver disease. Warfarin causes prolonged prothrombin time through inhibition of the vitamin K-dependent gamma-carboxylation of coagulation factors II, VII, IX, and X. In liver disease, the elevated prothrombin time is due in large part to decreased production of coagulation factors, including the vitamin K-dependent factors II, VII, IX, and X.9
The increase in MELD score that we have demonstrated in this study (approximately 4 MELD points) has different implications for specific patients, based on their relative position on the transplant list. For patients with low MELD scores (less than 14 points), an increase in MELD score by 4 points is unlikely to be clinically relevant at most transplant centers. Because a MELD score of at least 22–24 is necessary to receive a deceased-donor organ at most transplant centers, an increase in MELD score of 4 points (from 12 to 16 points) would have no significant impact on the patient's likelihood of transplantation. However, for patients with high MELD scores (>18 points), an increase in MELD score (from 20 to 24 points) could increase priority such that transplantation could become imminent. Currently, only 7% of the nearly 18,000 patients listed for liver transplantation have a MELD score greater than of 18.10 For these patients, the findings of our study could be especially relevant.
The potential effect of freezing the samples used in this study must be considered relative to differences in the results of subsequent assays for bilirubin, Cr, and INR. Of these 3 assays, the INR would seem to be the most vulnerable to the effects of freezing, because of the potential for deterioration in the function and/or concentration of clotting factors. However, there are numerous studies that have evaluated the effect of freezing on prothrombin time, INR, and clotting factor stability. One study found that the coefficient of variation for 5 deep-frozen pooled plasmas from coumarin-treated patients was only 2.6 to 3.3%.11 Another study evaluated the effect of freezing blood samples in 84 patients on anticoagulation therapy with INR between 2.0–4.5.12 The samples were frozen at −40°C for less than 1 month. They found no difference in INR between freshly assayed or frozen samples. Woodhams et al.13 evaluated the stability of clotting factors and prothrombin time in 6 normal volunteers whose blood samples were frozen at both −24 and −74°C. After 2 months there was less than 5% difference in the prothrombin times for samples stored at either temperature. Finally, Iazbik et al.14 measured prothrombin time in plasma samples frozen for 1 week from 9 dogs. They found no significant difference in prothrombin time in the frozen vs. freshly assayed samples. Therefore, we do not believe that sample freezing could be responsible for the differences in INR values between the clinical laboratories in this study.
In summary, we found that patients listed for liver transplantation at our center may achieve a higher MELD score and therefore a higher priority for liver transplantation by using laboratory methodologies selected to return the highest possible bilirubin, Cr, and INR values. Almost all of the difference in MELD score could be attributed to differences in INR. While the MELD score is determined only by “objective variables,” we have shown that there may be considerable and clinically relevant variation in the MELD score based on the selection of laboratory methods for the determination of bilirubin, Cr, and INR. Finally, physicians attending patients listed for liver transplantation may find the results of this study potentially useful to obtain the highest possible transplantation priority for their patients.