Live UNOS/OPTN liver transplant allocation score — logarithmic mortality model driving national waitlist priority (INR fixed 1.5, Na fixed 135 mEq/L)
The Model for End-Stage Liver Disease, sodium-adjusted (MELD-Na), is the objective, laboratory-derived score that the United Network for Organ Sharing (UNOS) and its regulatory body, the Organ Procurement and Transplantation Network (OPTN), use to rank adult candidates on the national liver transplant waitlist. Unlike earlier severity scales built on subjective clinical judgment, MELD-Na is calculated purely from four blood values, making it reproducible across every transplant center in the country and resistant to inter-observer variability.
MELD-Na is built from routine labs already drawn during hepatology follow-up, which is precisely why it scales to a national allocation system — no specialized testing or subjective exam is required:
• Serum bilirubin (mg/dL): reflects the liver's conjugation and excretory capacity. Rising bilirubin signals accumulating failure of hepatocyte function and biliary clearance.
• Serum creatinine (mg/dL): a surrogate for renal function, which deteriorates in advanced cirrhosis via hepatorenal physiology (splanchnic vasodilation, effective arterial underfilling, and renal vasoconstriction). Creatinine is capped between 1.0 and 4.0 mg/dL, and is set to the ceiling of 4.0 if the patient has received dialysis (or continuous veno-venous hemofiltration) at least twice in the prior week — since a dialyzed patient's measured creatinine no longer reflects true native renal failure.
• INR (international normalized ratio): reflects hepatic synthetic function, specifically the liver's production of clotting factors II, VII, IX, and X. A rising INR indicates the liver can no longer keep pace with coagulation factor turnover.
• Serum sodium (mEq/L): added in the MELD-Na revision because hyponatremia is an independent marker of advanced portal hypertension, ascites, and mortality risk not captured by the original three variables. Values are bounded between 125 and 137 mEq/L — extremely low sodium is truncated at 125 so that severe dilutional hyponatremia does not distort the score beyond its validated range.
The original MELD score, derived by Malinchoc and colleagues to predict survival after transjugular intrahepatic portosystemic shunt (TIPS) placement, uses natural-logarithm transforms of each lab value:
MELD = 3.78 × ln[bilirubin] + 11.2 × ln[INR] + 9.57 × ln[creatinine] + 6.43
The logarithmic transform is deliberate: mortality risk does not rise linearly with each lab value — it accelerates. Going from a bilirubin of 1 to 2 mg/dL matters less than going from 10 to 11 mg/dL in absolute terms, but on a multiplicative, log scale each doubling contributes a comparable increment to the score. By formula convention, any input value below 1.0 is rounded up to 1.0 before taking the logarithm, which keeps ln(x) from going negative and prevents the score from being reduced by supranormal lab values. Because ln(x) is null at x=0, this flooring rule is also what keeps the calculation numerically defined at the edges of the clinical range.
MELD's major methodological advance was replacing subjective clinical staging (as in the older Child-Pugh score, which incorporates ascites and encephalopathy graded by exam) with a fully laboratory-derived, continuous score. This makes it far harder to game and far easier to standardize across thousands of transplant candidates evaluated at hundreds of centers nationwide.
Hyponatremia is common in decompensated cirrhosis and tracks closely with the severity of portal hypertension and ascites, driven by non-osmotic antidiuretic hormone release and impaired free-water clearance. Studies showed sodium independently predicted waitlist mortality even after accounting for bilirubin, creatinine, and INR — so OPTN incorporated it into the score used for allocation, effective January 2016:
MELD-Na = MELD + 1.32 × (137 − Na) − [0.033 × MELD × (137 − Na)]
The adjustment is only applied when the base MELD exceeds 11; below that threshold, sodium is not incorporated, since the mortality benefit of the correction was validated primarily in more advanced disease. Sodium itself is bounded to 125–137 mEq/L before the subtraction, and the final MELD-Na value is bounded to the standard allocation range of 6 to 40 — a candidate whose calculated score falls below 6 is assigned a floor of 6, and any calculated value above 40 is capped at 40, since mortality risk at the extreme ceiling is already so high that further separation adds little discriminatory value.
MELD-Na is not a diagnosis and does not identify a specific disease process — it is a continuous, empirically calibrated predictor of short-term mortality in patients with chronic liver disease. Its power for organ allocation comes from validated studies correlating the score directly with observed 90-day death rates, which let UNOS rank candidates from every diagnosis and every transplant center on a single, comparable scale of urgency.
The score's lineage traces through two landmark studies. Malinchoc et al. (2000) developed the original model at the Mayo Clinic specifically to predict 3-month survival after TIPS placement — a procedure used to relieve portal hypertension complications. The model performed unexpectedly well as a general predictor of mortality in cirrhosis, prompting Kamath et al. (2001) to validate it across a broader population of patients with chronic liver disease, independent of TIPS status. That validation study established MELD as a reliable, general-purpose severity index and paved the way for OPTN to adopt it for liver allocation in February 2002, replacing the older, more subjective Child-Turcotte-Pugh classification that had governed waitlist priority until then.
Analyses of UNOS waitlist data (Wiesner et al. and subsequent OPTN registry studies) established the mortality bands that anchor clinical interpretation of the score:
• MELD-Na <10: low risk — patients are typically followed in outpatient hepatology clinics rather than actively pursuing urgent transplant.
• MELD-Na 10–19: moderate risk, roughly 6% 90-day mortality — candidates are listed and monitored closely but organ offers at this range are less frequent outside high-donor-supply regions.
• MELD-Na 20–29: substantially elevated risk, approximately 19–20% 90-day mortality — most transplants in the modern allocation era occur to candidates in or above this range.
• MELD-Na 30–39: severe risk, roughly 52–53% 90-day mortality — organ offers accelerate sharply and status is reassessed as frequently as every 7 days.
• MELD-Na ≥40: critical risk, approximately 71% 90-day mortality — the ceiling of the score, representing the sickest candidates on the list who receive the highest allocation priority nationally.
MELD-Na operationalizes a core ethical principle of US organ allocation policy: medical urgency, not waiting time or diagnosis, should be the primary driver of who receives an organ first among similarly situated candidates. Because the score is diagnosis-agnostic — it does not matter whether a candidate's cirrhosis stems from alcohol-associated liver disease, viral hepatitis, autoimmune disease, or metabolic dysfunction-associated steatotic liver disease — it allows direct, apples-to-apples urgency comparison across the entire national candidate pool.
Because MELD-Na is a continuous mortality proxy rather than a categorical diagnosis, two candidates with identical scores but completely different underlying liver diseases are treated as equally urgent by the allocation algorithm — the score, not the label, drives priority.
Laboratory MELD-Na captures hepatic synthetic failure well, but it does not capture every cause of transplant urgency. The clearest example is hepatocellular carcinoma (HCC): a patient can have near-normal bilirubin, creatinine, and INR while carrying a tumor burden that will progress beyond curative transplant candidacy within months. Standardized MELD exception (SME) points exist to translate that non-laboratory risk into the same numeric priority scale.
A patient with compensated cirrhosis and a growing liver tumor may have preserved synthetic function — normal-range bilirubin and INR, stable creatinine — and therefore a low lab MELD-Na score, even though their tumor is approaching the threshold beyond which transplant is no longer curative. Left to a pure lab score, such a candidate would wait far too long, drop out of Milan criteria as the tumor grows, and lose transplant eligibility entirely. Exception points exist to correct this mismatch between measured liver synthetic function and actual clinical urgency.
The Milan criteria (Mazzaferro et al., 1996) define the tumor burden threshold within which liver transplant achieves outcomes comparable to transplant for non-malignant indications:
• A single tumor no larger than 5 cm in diameter, or • Up to three tumors, each no larger than 3 cm, and • No evidence of extrahepatic spread, and • No macrovascular (portal or hepatic vein) invasion
Candidates meeting Milan criteria historically received an automatic standardized MELD exception of 22 points at the time of listing — a score chosen to approximate the transplant priority of a moderately sick lab-MELD candidate — with incremental increases roughly every 3 months if the candidate remained untransplanted, mirroring how a real tumor's risk of progression accrues over time.
Other exception categories follow their own point schedules and criteria, including hepatopulmonary syndrome (abnormal oxygenation from intrapulmonary vascular shunting) and familial amyloid polyneuropathy (a hereditary transthyretin-mediated disease where transplant halts amyloid production), each reflecting a condition whose mortality or disability risk is poorly represented by standard liver labs.
Because exception points were originally granted automatically and could escalate every three months, registry data showed exception candidates were being transplanted faster, on average, than lab-MELD candidates with comparable calculated urgency — raising equity concerns. In response, OPTN policy has evolved: exception requests are now adjudicated by the National Liver Review Board rather than granted automatically, and the rate of point escalation for standing exceptions has been capped, aligning exception-candidate transplant rates more closely with the observed transplant rate of lab-MELD candidates at the same score.
The tension underlying exception policy is fundamental to allocation ethics: exception points must be generous enough to prevent HCC candidates from progressing beyond curative transplant while they wait, but not so generous that they systematically out-compete lab-MELD candidates whose actual 90-day mortality risk is equally high.
Even a perfectly calibrated urgency score is only half of allocation policy — organs must also be matched to candidates across geography, since donor livers have a limited window for successful transportation and transplantation. In February 2020, OPTN replaced its decades-old Donor Service Area (DSA) and regional boundary system with "acuity circles," a model built around concentric distance rings from the donor hospital rather than fixed administrative borders.
For decades, US liver allocation was organized around fixed Donor Service Areas and larger multi-DSA regions — a map inherited from the historic footprint of local organ procurement organizations rather than from any principled urgency or distance calculation. This produced substantial geographic disparity: a candidate with a given MELD-Na score might wait far longer for an organ in one DSA than an equally sick candidate in a neighboring DSA simply because of where they were listed, since livers were offered locally first regardless of relative urgency across the boundary.
Acuity circles replace those fixed boundaries with distance rings — typically 150, 250, and 500 nautical miles — drawn around the donor hospital itself. A liver becomes available for offer to candidates within successively larger circles, and within each circle, candidates are ranked first by medical urgency (status and MELD-Na) rather than by which side of an administrative line they happen to sit on.
Above the standard MELD-Na scale sit acute urgency statuses reserved for patients with fulminant, rapidly fatal liver failure who cannot survive the ordinary waiting process:
• Status 1A: adult candidates with acute liver failure and a life expectancy measured in hours to days without transplant — fulminant hepatic failure, primary non-function of a recent graft, hepatic artery thrombosis, or acute decompensated Wilson disease. Status 1A candidates receive the highest allocation priority nationally, ahead of any MELD-Na-ranked candidate.
• Status 1B: analogous acute urgency status for pediatric candidates.
Within each acuity circle, offers proceed first to Status 1A/1B candidates, then in descending order of MELD-Na/PELD score, before the offer pattern expands outward to the next, larger circle. This urgency-first, distance-secondary structure is the central design goal of the acuity circles model: reduce the penalty of geography while still respecting the practical reality that organ viability declines with transport time.
Post-implementation OPTN analyses of acuity circles have shown a narrowing of the gap in median MELD-Na score at transplant between historically organ-scarce and organ-rich regions, consistent with the policy's intended effect of broader, urgency-based sharing. The change followed years of public debate and litigation over whether geography or medical urgency should dominate allocation logic, and it remains one of the most closely studied policy transitions in US transplant history, informing parallel geographic reforms considered for other solid organs.
A MELD-Na score calculated at the moment of listing is only a snapshot. Liver disease is dynamic — patients decompensate, develop infections, or occasionally improve — so OPTN policy requires periodic laboratory recertification on a schedule tied to the candidate's own score, ensuring waitlist priority tracks current clinical status rather than a stale value from months earlier.
OPTN policy sets recertification intervals inversely proportional to a candidate's risk: the sicker the candidate, the more frequently labs must be redrawn and resubmitted. A candidate with MELD-Na ≥25 must recertify roughly every 7 days, reflecting how quickly severe hepatic decompensation can change from week to week. Candidates scoring 19–24 recertify every 30 days, those at 11–18 every 90 days, and the lowest-risk candidates at 10 or below only annually. If labs are not updated within the required window, the score reverts and the candidate loses priority until fresh values are submitted — a mechanism that prevents both stale under-scoring (a now-sicker patient still ranked at an old low score) and stale over-scoring (a since-improved patient still ranked at an old high score).
A higher recalculated MELD-Na score translates directly into higher waitlist priority — the core allocation logic remains "sickest first" throughout the candidate's time on the list. But raw waitlist mortality risk is not the only consideration transplant programs and OPTN increasingly weigh: transplant benefit models attempt to balance a candidate's risk of dying on the waitlist against their expected probability of survival after transplant, so that scarce donor organs are directed toward transplants most likely to convert into durable, long-term graft and patient survival rather than toward the single highest lab score in isolation.
This "survival benefit" framing does not replace MELD-Na-based prioritization, but it informs research into next-generation allocation policy, donor-recipient matching algorithms, and center-level decisions about organ acceptance — recognizing that optimal graft utility requires looking beyond urgency alone to the expected total years of life gained from each transplant.
The field is actively moving toward continuous distribution frameworks that would combine medical urgency, candidate biology, distance/travel efficiency, and post-transplant survival benefit into a single composite allocation score — extending the logic of MELD-Na and acuity circles into an even more granular, multi-factor ranking system.