The Model for End-Stage Liver Disease (MELD) turns three routine lab values into a single number that predicts short-term mortality in advanced liver disease. Because it is objective and reproducible, it became the basis for ranking patients on transplant waiting lists, later refined with serum sodium as MELD-Na.
How it works
The original MELD is a weighted sum of the natural logs of three labs:
MELD = 0.957 x ln(creatinine)
+ 0.378 x ln(bilirubin)
+ 1.120 x ln(INR)
+ 0.643
MELD = round(MELD x 10)Before the logs are taken, any value below 1.0 is set to 1.0, creatinine is capped at 4.0 mg/dL, and dialysis forces creatinine to 4.0. The result is bounded between 6 and 40.
The sodium correction is applied when MELD is 12 or higher:
MELD-Na = MELD + 1.32 x (137 - Na)
- 0.033 x MELD x (137 - Na)with sodium clamped to the range 125 to 137 mEq/L.
Why each variable was chosen
The three labs in the original MELD score each capture a distinct axis of liver dysfunction:
Creatinine reflects renal function. In the setting of cirrhosis, worsening renal function — often a component of hepatorenal syndrome — is one of the strongest predictors of short-term mortality. Creatinine is capped at 4.0 mg/dL because the marginal prognostic value of further elevation beyond that point is small, and dialysis forces the cap regardless of the measured value.
Bilirubin reflects hepatic excretory function. Rising bilirubin indicates that the liver is losing its ability to process waste products, a direct measure of hepatocellular dysfunction. Even modest elevation contributes meaningfully to the score.
INR (International Normalised Ratio) reflects the liver’s synthetic function. The liver produces the vast majority of clotting factors; an elevated INR in a cirrhotic patient means the synthetic capacity is severely compromised. INR received the highest coefficient in the formula (1.120) because synthetic failure was found to be the most prognostically powerful of the three inputs.
Sodium was added to create MELD-Na because hyponatraemia independently predicts worse outcomes in cirrhosis at lower MELD scores. Its correction nudges the score upward for patients whose absolute labs look moderate but whose sodium is low, correctly prioritising them on the waitlist.
Reading a MELD score in practice
| MELD Score | Approximate 90-day mortality |
|---|---|
| 6–9 | ~2% |
| 10–19 | ~6% |
| 20–29 | ~20% |
| 30–39 | ~50% |
| 40 | ~70% |
These mortality bands are population-level estimates derived from validation studies of the original formula. Individual prognosis varies with the underlying aetiology of liver disease, the presence of acute-on-chronic decompensation, infections, malnutrition, and access to care. The score performs best as a relative rank for waitlist prioritisation rather than as a precise mortality prediction for a specific patient.
Worked example and notes
A patient with creatinine 1.9 mg/dL, bilirubin 4.0 mg/dL, and INR 1.8 produces a MELD in the high teens. Adding a low serum sodium of 128 mEq/L pushes MELD-Na several points higher via the correction formula, reflecting the worse prognosis of hyponatraemia — and correctly elevating the patient’s waitlist priority relative to someone with the same underlying labs but normal sodium.
The score is a prognostic estimate, not a treatment decision, and should always be interpreted alongside the full clinical picture, local transplant centre policy, and the most current UNOS allocation guidelines.