About the United Kingdom Model for End-Stage Liver Disease (UKELD)
A UKELD score of 49 or above is the minimum eligibility threshold for elective liver transplant listing in the United Kingdom, chosen because it corresponds to roughly a 9% predicted one-year mortality without transplantation — the point at which the procedure is judged to offer net survival benefit. The score combines four routine values on a natural-log scale: INR, creatinine, bilirubin and serum sodium, with creatinine and bilirubin expressed in µmol/L. It was derived specifically on UK elective waiting-list data and outperformed both MELD and MELD-Na at predicting list mortality in that population, which is why the UK uses it rather than MELD. A score of 60 corresponds to roughly 50% one-year mortality.
Formula
UKELD = 5.395 × ln(INR) + 1.485 × ln(creatinine µmol/L) + 3.13 × ln(bilirubin µmol/L) − 81.565 × ln(sodium mmol/L) + 435- INR
- International normalised ratio.
- creatinine
- Serum creatinine in µmol/L (multiply mg/dL by 88.4).
- bilirubin
- Total bilirubin in µmol/L (multiply mg/dL by 17.1).
- sodium
- Serum sodium in mmol/L, entering with a negative coefficient — the lower the sodium, the higher the score.
- The units are not optional. UKELD was published with creatinine and bilirubin in µmol/L, and feeding mg/dL values into the same coefficients produces a meaningfully wrong score.
- The constant of 435 exists to place the output on a convenient scale; it carries no clinical meaning on its own.
- Sodium's coefficient (−81.565) is by far the largest in magnitude, which is why hyponatraemia moves UKELD more than an equivalent proportional change in the other three terms.
Interpreting the result
A score of 49 or above meets the UK minimum for elective listing; below it, the predicted mortality without transplant is low enough that the risks of the operation are not considered justified, so listing is not offered on UKELD grounds alone. Crucially, reaching 49 establishes eligibility, not entitlement — it is a floor, and the decision to list still rests on a full multidisciplinary assessment of comorbidity, likely benefit and candidacy. Several variant syndromes are explicitly listed irrespective of UKELD, because the score does not capture their morbidity: these include hepatocellular carcinoma within agreed criteria, diuretic-resistant ascites, intractable pruritus, recurrent cholangitis and familial amyloidosis, among others. A score of 60 corresponds to roughly 50% one-year mortality. Rising scores also predict longer intensive-care and hospital stays after transplant, but notably not worse overall post-transplant survival.
| Score | Band | What it means | Action |
|---|---|---|---|
| < 49 | Below UK listing threshold | Predicted one-year mortality below the level at which transplant is judged to offer net benefit | Not eligible for elective listing on UKELD grounds; continue surveillance and reassess, and check whether a variant syndrome qualifies independently |
| ≥ 49 | Meets UK listing threshold | Approximately 9% or greater predicted one-year mortality without transplantation | Eligible for elective listing — refer to and discuss with the transplant centre; eligibility is a floor, not an automatic listing decision |
| ≈ 60 | High risk | Approximately 50% predicted one-year mortality | Urgent transplant-centre discussion; expect prolonged ICU and hospital stay post-transplant |
What the UKELD Score needs (4 inputs)
- INR
- International normalised ratio, entered directly.
- Creatinine
- Serum creatinine. The published formula requires µmol/L, so mg/dL values are converted before the log is taken.
- Bilirubin
- Total bilirubin, also required in µmol/L by the published formula.
- Sodium
- Serum sodium in mmol/L. It carries a large negative coefficient, so hyponatraemia raises the score substantially — the term MELD omits.
Units. UKELD is a µmol/L formula, unlike the mg/dL MELD family. To convert: creatinine mg/dL × 88.4 = µmol/L; bilirubin mg/dL × 17.1 = µmol/L. This calculator accepts mg/dL and converts internally, but if you compute UKELD by hand from a US-style lab report you must convert first.
What it returns
- UKELD score
- A continuous score. Typical listed patients fall in the high 40s to low 60s.
- Listing eligibility
- Whether the score reaches the UK minimum listing threshold of 49.
How it is calculated
Barber and colleagues built the model on 1,103 adults registered for a first elective liver transplant in the UK between 2003 and 2006, and tested it on a separate 452 patients registered in 2006–07. Candidate variables were assessed for their ability to predict death on the waiting list, and four survived into the final model: INR, creatinine, bilirubin and sodium. Each enters on a natural-log scale, so the score responds to proportional rather than absolute change — a creatinine rising from 60 to 120 µmol/L moves it as much as 120 to 240. The output was then scaled so that the policy threshold could be set as a round number, and 49 was chosen as the score corresponding to approximately 9% predicted one-year mortality without a transplant.
Facts & figures
| UKELD | MELD | MELD-Na | |
|---|---|---|---|
| Variables | INR, creatinine, bilirubin, sodium | INR, creatinine, bilirubin | INR, creatinine, bilirubin, sodium |
| Derived on | UK elective waiting list (2003–06) | US cohort | US cohort |
| Units required | µmol/L | mg/dL | mg/dL |
| Policy role | UK minimum listing threshold (≥ 49) | US priority (superseded by MELD 3.0) | US priority component |
| List-mortality prediction in UK data | Best of the three | Outperformed by UKELD | Outperformed by UKELD |
The comparison is from the UKELD derivation study's own UK validation cohort, which is the population UKELD was built for — it is not a claim that UKELD outperforms MELD everywhere.
Evidence
Derivation and validation — Barber
2011 · n = 1,103A modelling dataset of 1,103 adults registered for a first elective liver transplant in the United Kingdom between 1 April 2003 and 31 March 2006, with a separate validation dataset of 452 patients registered between 1 April 2006 and 31 March 2007. The final model retained INR, serum creatinine, bilirubin and sodium.
UKELD accurately predicted waiting-list mortality in the validation cohort and was a better predictor than either MELD or MELD-Na. It was not associated with overall post-transplant survival, but higher scores were associated with longer intensive-care and total initial hospital stay.
UK selection and allocation framework — Neuberger
2008The UK guidelines for selecting patients onto the national liver transplant waiting list and for allocating donated livers, developed to base rationing on the risk of death without transplant and the procedure's ability to improve quality of life, for both acute liver failure and chronic liver disease.
Established the framework in which a minimum UKELD threshold operates as a transparent, evidence-based eligibility criterion, alongside the variant-syndrome categories that qualify for listing irrespective of score.
National Liver Offering Scheme — benefit-based allocation
2024UK Transplant Registry data on waiting-list registrations and transplants, used to build models of survival on the list and survival after transplantation, then applied to simulate allocation under the National Liver Offering Scheme introduced in 2018.
Moved UK allocation of most deceased-donor livers from geographic zones and estimated need to calculated transplant benefit — which is why UKELD now governs listing eligibility rather than the offering of an available organ.
How it compares
UKELD Score vs MELD-Na
In UK elective waiting-list data UKELD predicted list mortality better than MELD-Na, which is why the UK uses it — but MELD-Na remains the relevant score in US-based allocation, and the two are not interchangeable.
Both use INR, creatinine, bilirubin and sodium, so the difference is not which variables matter but how they are weighted and on what population. UKELD was fitted to UK elective registrations and validated against UK list mortality, where it outperformed both MELD and MELD-Na. MELD-Na was derived in a US cohort and feeds US priority. There is also a practical trap: UKELD expects µmol/L while MELD formulations expect mg/dL, so the same patient's numbers cannot be dropped from one into the other.
UKELD Score vs Child-Pugh score
Child-Pugh remains a useful bedside grade of cirrhosis severity, but UKELD is the instrument that determines UK transplant listing eligibility because it is continuous, objective and calibrated to waiting-list mortality.
Child-Pugh includes two subjective items — ascites and encephalopathy — and collapses patients into three broad classes, which limits its ability to rank candidates against one another. UKELD uses four objective laboratory values on a continuous scale, so it discriminates within the band where listing decisions are actually made. Child-Pugh still earns its place in wider prognostic work, including surgical risk and variceal management, where it is embedded in decades of literature.
Pearls & pitfalls
- The formula demands µmol/L for creatinine and bilirubin. Entering mg/dL values into the published coefficients gives a substantially wrong score — this is the single commonest UKELD error.
- Reaching 49 confers eligibility, not a listing decision. Comorbidity, candidacy and expected benefit are still assessed by the multidisciplinary team.
- Variant syndromes — hepatocellular carcinoma within criteria, diuretic-resistant ascites, intractable pruritus, recurrent cholangitis, familial amyloidosis and others — can qualify for listing irrespective of UKELD, so a score below 49 does not close the door.
- Sodium carries the heaviest coefficient, so correcting hyponatraemia with fluids or tolvaptan will lower a UKELD score without changing the underlying liver disease.
- UKELD predicts pre-transplant list mortality, not post-transplant survival — the derivation study found no association with overall survival after grafting.
- It is UK-specific. Do not use it to estimate priority in a MELD-based system, or to compare a UK patient with a US-listed one.
Critical actions
- Convert creatinine and bilirubin to µmol/L before calculating, or use a calculator that does it internally.
- Refer to the transplant centre once the threshold is met — and refer earlier if a variant syndrome applies, without waiting for UKELD to reach 49.
- Recalculate serially in listed patients; UKELD is a moving figure and a rising score signals deteriorating list survival.
- Interpret a low sodium carefully: check whether it reflects genuine circulatory dysfunction or a reversible iatrogenic cause before treating the score as fixed.
- Do not use UKELD to counsel a patient about their chance of surviving transplantation — it was not built for that and does not predict it.
Why this score exists
The model exists because the UK declined to import a US allocation score without testing it. MELD had become the international default, but it was derived elsewhere, for a different decision, and the UK question was narrower: among patients being considered for an elective graft here, who dies waiting? Building the model on UK registration data answered that question directly and, as it turned out, better than MELD or MELD-Na did. The choice to retain sodium reflects the same reasoning — hyponatraemia is a marker of the circulatory dysfunction of advanced cirrhosis, and excluding it discards information a UK waiting list was demonstrably sensitive to. Setting the threshold at a predicted mortality rather than a percentile was a deliberate policy decision: it ties access to an explicit statement about when transplantation does more good than harm.
About the creator
First author, 2011 derivation study
Derived UKELD on UK elective waiting-list registrations, where it outperformed both MELD and MELD-Na at predicting list mortality.
Co-author, and lead author of the UK selection guidelines
Co-authored the derivation and led the UK framework in which a minimum UKELD threshold operates as a listing criterion.
Senior author
Co-authored both the derivation and the UK selection and allocation guidelines.
Limitations
- It predicts waiting-list mortality only, and shows no association with overall survival after transplantation, so it cannot inform how well a patient will do with a graft.
- It is calibrated to the UK elective list and to UK practice; its thresholds carry no meaning in other allocation systems.
- Like all MELD-family scores it under-represents complications that dominate quality of life without deranging these four values — refractory ascites, intractable pruritus, recurrent encephalopathy — which is exactly why variant-syndrome listing routes exist.
- Sodium is manipulable, and treating hyponatraemia lowers the score without altering the liver disease, creating a theoretical route to score reduction that has nothing to do with improved prognosis.
- It excludes patients with acute liver failure, who are assessed under entirely separate super-urgent criteria.
- Since 2018 it no longer drives which patient receives an available organ in the UK; benefit-based allocation does, so UKELD alone no longer describes a listed patient's realistic prospects of being offered a graft.
If you are the patient
UKELD is a score used in the UK to help decide whether someone with long-term liver disease should be put on the waiting list for a liver transplant. It is worked out from four routine blood results: your clotting time (INR), your kidney function (creatinine), your bilirubin (which reflects jaundice), and your blood sodium level. A score of 49 or more means the risk of becoming seriously unwell over the next year is high enough that a transplant is likely to do more good than harm, so you can be considered for the waiting list. A score below 49 does not always mean a transplant is off the table — some specific liver problems, such as certain tumours, severe fluid build-up or relentless itching, allow someone to be listed whatever their score. Being eligible is not the same as being listed: the transplant team still assesses your overall health and whether an operation is right for you.
Frequently asked questions
What UKELD score is needed for a liver transplant in the UK?#
49 or above is the minimum score for elective listing. It was set at that level because it corresponds to roughly a 9% predicted one-year mortality without transplantation, the point at which transplant is considered to offer net benefit.
What is the difference between UKELD and MELD?#
UKELD adds serum sodium to MELD's INR, creatinine and bilirubin, uses different coefficients, expects creatinine and bilirubin in µmol/L rather than mg/dL, and was derived on UK elective waiting-list data — where it predicted list mortality better than either MELD or MELD-Na.
Can you be listed with a UKELD below 49?#
Yes. A number of variant syndromes qualify for listing irrespective of UKELD, including hepatocellular carcinoma within agreed criteria, diuretic-resistant ascites, intractable pruritus, recurrent cholangitis and familial amyloidosis, because the score does not capture their impact.
What does a UKELD of 60 mean?#
Approximately 50% predicted one-year mortality without transplantation. It also predicts a longer stay in intensive care and in hospital after transplant, though not worse overall post-transplant survival.
Does UKELD decide who gets the next available liver?#
No, not since 2018. The National Liver Offering Scheme allocates most deceased-donor livers in the UK by calculated transplant benefit. UKELD now sets eligibility for listing rather than priority for an offer.
Which units does the UKELD formula need?#
Creatinine and bilirubin must be in µmol/L, and sodium in mmol/L. Using mg/dL values with the published coefficients produces a significantly incorrect score.
References
Original / primary reference
UK policy context
- Neuberger J, Gimson A, Davies M, et al. Selection of patients for liver transplantation and allocation of donated livers in the UK. Gut. 2008;57(2):252-257.
- Allen J, Attia Z, Rowe I, et al. Transplant benefit-based offering of deceased donor livers in the United Kingdom. J Hepatol. 2024;81(3):471-478.