MELD Score
On this page
Direct answer
A number between 6 and 40 decides who reaches the top of the transplant waiting list: the MELD score converts bilirubin, INR and creatinine into a three-month mortality estimate, later refined by serum sodium. Its logic is objective — hence allocation systems — and its anchors are quotable: MELD around 20 predicts roughly 20% three-month mortality, 30 around half, 40 above 70%. The ceiling rules matter as much as the formula: creatinine is capped at 4.0, and dialysis sets it there.
What you must remember
- The original formula: 3.78 × ln(bilirubin mg/dL) + 11.2 × ln(INR) + 9.57 × ln(creatinine mg/dL) + 6.43, rounded to a whole number, range 6–40 — know the components and their weights, not the logarithms.
- The renal cap: creatinine is capped at 4.0 mg/dL; renal replacement therapy at least twice within the prior week sets it to 4.0 — the two rules examiners check.
- MELD-Na: adds sodium because hyponatraemia independently predicts death — MELD(i) + 1.32 × (137 − Na) − [0.033 × MELD(i) × (137 − Na)], with Na bounded between 125 and 137.
- MELD 3.0: the current US allocation model adds albumin and a higher weight for female sex, refining accuracy in modern cohorts.
- Mortality anchors (commonly quoted): MELD ~20 → about 20% three-month mortality; 30 → about 50–53%; 40 → about 70–71%.
- Uses: deceased-donor liver allocation (highest MELD first), transplant benefit thresholds (15 is the commonly cited score where transplant beats waiting), perioperative and TIPS risk assessment, and HCC exception points (Milan-criteria tumours receive an assigned exception score).
- Limitations: INR varies by laboratory reagent; creatinine misleads in dehydration and muscle-wasted cirrhotics; the score says nothing about complications like intractable pruritus — hence exception systems.
- Indian context: deceased-donor allocation varies by state networks and donation rates remain low, so living-donor transplantation often outruns the MELD queue — but MELD remains the shared severity language in every Indian transplant conversation.
Calculating and interpreting a real score
Take the patient rather than the logarithm: a 47-year-old with alcoholic cirrhosis, bilirubin 6 mg/dL, INR 2.1, creatinine 1.8 mg/dL, sodium 128. Bilirubin and INR push him moderately; the creatinine pushes harder — renal function carries the heaviest single weight, the conceptual point behind the dialysis cap. His score computes in the mid-twenties; sodium correction lifts it further — at 128 mmol/L the MELD-Na arithmetic adds roughly two points, hyponatraemia being a death signal the original formula ignored. Now the two questions MELD exists to answer. How sick is he? A MELD in the twenties is roughly one-in-three three-month mortality — ward shorthand for "this patient needs a transplant conversation now". Where does he stand? Deceased-donor allocation follows the number, descending from 40; his HCC-free status means no exception points — a Milan-criteria patient beside him carries an assigned exception score precisely because tumour mortality does not fit the lab formula. The examinable reflexes: name the three labs, cap the creatinine, set it to 4.0 after two dialysis sessions, add sodium with the 125–137 bounds, and quote the 15–20–30–40 mortality ladder as the interpretation spine.
Why sodium had to join the formula
The exam wants the model's evolution as an argument. Original MELD assumed three objective labs capture liver failure; the field then noticed two cirrhotics with identical bilirubin, INR and creatinine dying at visibly different rates — the difference was sodium. Hyponatraemia marks advanced vasodilatory, water-retaining physiology, and adding it measurably improved death prediction — hence MELD-Na with its bounded sodium term. The dialysis cap next: without the 4.0 ceiling, one episode of renal failure would grant an unbounded score and monopolise every organ offer — allocation ethics expressed as arithmetic. Then the model's blind spots, which generate the exception system: refractory pruritus, recurrent cholangitis or hepatopulmonary syndrome may carry a modest MELD and an unarguable need — no laboratory triple captures everything. Finally, the Indian framing: with living-donor transplantation dominant, MELD works less as a queue and more as a shared risk language between donor discussions, intensivists and guidelines — the mortality anchors stay the same everywhere.
Frequently asked questions
Which variables enter the original MELD score?
Serum bilirubin, INR and creatinine — combined logarithmically into a score of 6–40, with creatinine capped at 4.0 mg/dL.
How does dialysis affect the score?
Renal replacement therapy at least twice within the preceding week sets the creatinine component to 4.0 mg/dL — the maximum contribution.
Why was sodium added?
Hyponatraemia independently predicts mortality in cirrhosis, and the MELD-Na modification (sodium bounded 125–137 mmol/L) improves death prediction.
What mortality does a MELD of 30 predict?
Roughly half — around 50% three-month mortality, compared with about 20% at a score of 20 and above 70% at 40.
How does HCC fit into a lab-based score?
It does not — Milan-criteria hepatocellular carcinoma receives assigned exception points, because tumour mortality is not reflected by the three labs.