MELD Score Calculator (Original Model)
MELD Score Calculator (Original Model)
The original MELD — creatinine, bilirubin and INR, with the lower bound of 1.0, the creatinine cap at 4.0 mg/dL and the dialysis rule that make it agree with a transplant centre’s number.
MELD score (original model)
Three labs → 6–40Bilirubin 3.2 mg/dL, INR 1.8, creatinine 1.6 mg/dL, no dialysis
Formula and the bounds that come with it
Creatinine, bilirubin and INR below 1.0 are set to 1.0 · creatinine capped at 4.0 mg/dL · dialysis sets creatinine to 4.0 · score floored at 6 and capped at 40
- mg/dL, always
- the equation was derived and is implemented in conventional units. Divide a µmol/L bilirubin by 17.104 and a µmol/L creatinine by 88.4, or set the unit above and let the calculator do it
- the floor of 1.0
- applies to all three laboratory values, so no logarithm is negative. Its side effect is that the score cannot distinguish a bilirubin of 0.4 from one of 0.9
- the cap of 4.0 mg/dL
- creatinine only, and the clearest marker of which MELD you are using. Original MELD and MELD-Na cap at 4.0; MELD 3.0 caps at 3.0
- the dialysis rule
- two or more dialysis treatments in the preceding 7 days, or 24 hours of continuous veno-venous haemodialysis, SETS creatinine to the cap. Setting is not capping: it raises a low creatinine as well as lowering a high one
- the ×10 form
- often written as 0.957, 0.378, 1.120 and 0.643 with the sum multiplied by ten. Algebraically identical; mixing them — tenths coefficients without the ×10 — gives a tenth of the true score
- what is not in it
- ascites, encephalopathy, variceal bleeding, hepatocellular carcinoma, sodium, albumin, sex and age. The original 2001 model also carried a term for the cause of the cirrhosis, which UNOS removed on adopting the score
Worked example
Bilirubin 3.2 mg/dL, INR 1.8, creatinine 1.6 mg/dL, no dialysis
No value is below the floor and the creatinine is below the cap, so all three enter unchanged
9.57 × ln(1.6) = 4.4989 · 3.78 × ln(3.2) = 4.3967 · 11.2 × ln(1.8) = 6.5832 · constant 6.43
Sum 21.9079 → MELD 22
The same patient scores 25 on MELD-Na at a sodium of 132 and 25 on MELD 3.0 as a man with an albumin of 3.0 — and 26 on MELD 3.0 as a woman. Three instruments, four numbers, one patient
Set the units to SI and enter bilirubin 54.7 µmol/L and creatinine 141.4 µmol/L and the answer is the same 22
Now switch dialysis on. The creatinine is SET to 4.0, giving 9.57 × ln(4.0) = 13.2694 and a total of 30.6768 → MELD 31. Nine points, from one checkbox
Raise the creatinine to 6.0 instead. It is capped back to 4.0, so the score is again 31 — the score cannot see the difference between a creatinine of 4 and one of 6
The three MELDs, side by side
| MELD (original) | MELD-Na (2016) | MELD 3.0 (2023) | |
|---|---|---|---|
| Inputs | Creatinine, bilirubin, INR | and sodium | and sodium, albumin, sex |
| Creatinine cap | 4.0 mg/dL | 4.0 mg/dL | 3.0 mg/dL |
| Lower bound on labs | 1.0 | 1.0 | 1.0 |
| Sodium bounds | — | 125–137 mmol/L | 125–137 mmol/L |
| Albumin bounds | — | — | 1.5–3.5 g/dL |
| Sex term | — | — | +1.33 if female |
| Score range | 6–40 | 6–40 | 6–40 |
| US allocation | 2002–2016 | 2016–2023 | July 2023 onward |
What the bounds do to a real patient
| Creatinine entered | Creatinine used | MELD (bili 3.2, INR 1.8) |
|---|---|---|
| 0.6 mg/dL | 1.0 (floored) | 17 |
| 1.0 mg/dL | 1.0 | 17 |
| 1.6 mg/dL | 1.6 | 22 |
| 4.0 mg/dL | 4.0 | 31 |
| 6.0 mg/dL | 4.0 (capped) | 31 |
| 1.1 mg/dL on dialysis | 4.0 (set) | 31 |
Which MELD is this, and why the bounds decide the answer
MELD began as a model to predict survival after elective placement of a transjugular intrahepatic portosystemic shunt, and was adapted into a general prognostic score for end-stage liver disease. In the initial validation its c-statistic for three-month survival was 0.87 against 0.84 for the Child-Turcotte-Pugh score, and it had the advantage that every input is a laboratory number rather than a clinician’s judgement of how much ascites there is. UNOS adopted it for deceased-donor liver allocation in 2002, dropping the aetiology term that the original model carried.
This page implements that original three-variable model. It is not the instrument the United States allocates on: MELD-Na replaced it in January 2016 and MELD 3.0 replaced MELD-Na in July 2023. It is still worth computing, because a great deal of the literature a reader will meet — survival curves, transplant-benefit thresholds, trial inclusion criteria written before 2016 — is expressed in original MELD, and because the three versions give different numbers for the same patient.
The bounds are where reimplementations fail. All three laboratory values are raised to 1.0 if they are lower, so the logarithms cannot go negative and no score is below zero. Creatinine is capped — at 4.0 mg/dL in the original and in MELD-Na, at 3.0 mg/dL in MELD 3.0 — and a patient on dialysis twice in the preceding week, or on 24 hours of continuous veno-venous haemodialysis, has creatinine set to that cap whatever it measures. The finished score is then floored at 6 and capped at 40. Get the coefficients right and the caps wrong and the score agrees with the transplant centre for ordinary patients and disagrees by several points for the sickest, which is exactly the wrong way round.
The second trap is units. Every published MELD equation takes bilirubin and creatinine in mg/dL. A reader working in µmol/L who types their own laboratory’s numbers in gets a score that is badly wrong and perfectly plausible. The unit selector above handles it; the conversions are a division by 17.104 for bilirubin and 88.4 for creatinine. A score is not a diagnosis and a risk derived from a cohort is not a probability for one patient. This states what the number predicted in a named study; the clinician in front of the patient decides what follows.
Frequently asked questions
Which version of MELD is this?
The original three-variable model — creatinine, bilirubin and INR — as UNOS implemented it from 2002, without the aetiology term the 2001 paper carried. MELD-Na added sodium in 2016 and MELD 3.0 added albumin and sex and lowered the creatinine cap in 2023. If you need the number a United States transplant centre is working from today, that is MELD 3.0.
What units does MELD take?
Bilirubin and creatinine in mg/dL, always. Divide a µmol/L bilirubin by 17.104 and a µmol/L creatinine by 88.4, or set the unit selector above. Entering µmol/L values as mg/dL is the commonest error here: bilirubin 90 µmol/L is 5.3 mg/dL, and typing 90 instead adds about 11 points.
Why is creatinine capped, and at what value?
The cap limits how much a non-hepatic problem can drive a liver allocation score. The original MELD and MELD-Na cap at 4.0 mg/dL; MELD 3.0 caps at 3.0 mg/dL. That difference is the easiest way to tell which equation a calculator has actually implemented.
What happens to the score on dialysis?
Two or more dialysis treatments in the previous seven days, or 24 hours of continuous veno-venous haemodialysis, means creatinine is set to the cap — 4.0 mg/dL for the original model. Set, not capped: a dialysed patient with a measured creatinine of 1.1 is scored at 4.0, because dialysis has removed the very substance the score reads renal failure from.
Does a MELD of 40 mean the same thing for every patient?
No. Forty is a ceiling, not a measurement: the unrounded model value beside the result keeps climbing past it, so 41 and 52 are both reported as 40. The score also ignores ascites, encephalopathy, variceal bleeding and hepatocellular carcinoma, which is why allocation systems carry exception points alongside the number.
Related calculators
References
- Kamath PS, Kim WR. The model for end-stage liver disease (MELD). Hepatology. 2007;45(3):797–805.
- Kamath PS, Wiesner RH, Malinchoc M, et al. A model to predict survival in patients with end-stage liver disease. Hepatology. 2001;33(2):464–70.
- Wiesner R, Edwards E, Freeman R, et al. Model for end-stage liver disease (MELD) and allocation of donor livers. Gastroenterology. 2003;124(1):91–6.
- OPTN. Policy Guidance: Improving Liver Allocation — MELD, PELD, Status 1A and Status 1B. Organ Procurement and Transplantation Network, 2023.
- Merck Manual Professional Edition. MELD Scores and Mortality. Rahway, NJ: Merck & Co. (tabulating Kim 2021 and Lim J et al, Gut Liver 2025;19(3):427–37).
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/