I have all the data I need for a comprehensive answer - including the original MELD formula, MELD-Na, MELD 3.0, the mortality table, and comparison with Child-Pugh. Let me compile the full answer now.
MELD Score (Model for End-Stage Liver Disease)
Overview
The MELD score was originally developed to predict 90-day (3-month) mortality in patients undergoing TIPS (transjugular intrahepatic portosystemic shunt) placement for complications of portal hypertension. It was later shown to accurately predict survival across a broad range of advanced liver disease states - cirrhosis with infections, variceal bleeding, cholestatic diseases, acute liver failure, and alcoholic hepatitis - and was adopted by UNOS (United Network for Organ Sharing) in 2002 as the primary tool for liver transplant prioritization, replacing the Child-Turcotte-Pugh (CTP) score. - Current Surgical Therapy 14e; Sabiston Textbook of Surgery
Versions of MELD
1. Original MELD (Pre-2016)
Uses 3 objective laboratory variables:
$$\text{MELD} = [0.957 \times \ln(\text{Cr}) + 0.378 \times \ln(\text{Bili}) + 1.120 \times \ln(\text{INR}) + 0.643] \times 10$$
| Variable | Notes |
|---|
| Serum creatinine (mg/dL) | Capped at 4.0 mg/dL; set to 4.0 if on hemodialysis (≥2 sessions or 24h CVVHD in prior week) |
| Serum bilirubin (mg/dL) | Total bilirubin |
| INR | Prothrombin time as INR |
| Constant | +0.643 |
Score range: 6-40 (UNOS caps at 40 for allocation purposes). Result is multiplied by 10 and rounded to the nearest integer. - Tietz Textbook of Laboratory Medicine, 7e
2. MELD-Na (UNOS Standard 2016-2021)
Adds serum sodium to address hyponatremia as an independent predictor of waitlist mortality. Multiple studies showed MELD-Na predicts waitlist mortality ~7% better than original MELD.
$$\text{MELD-Na} = \text{MELD} + 1.59 \times (135 - \text{Na})$$
- Sodium capped between 125-137 mEq/L
- Became the OPTN/UNOS standard in January 2016
3. MELD 3.0 (Current US Standard since 2022)
Developed to address sex-based disparities in organ allocation (women had historically lower waitlist-to-transplant rates). MELD 3.0 adds female sex and serum albumin as variables, plus interaction terms:
$$\text{MELD 3.0} = 1.33(\text{Female}) + 4.56\ln(\text{Bili}) + 0.82(137-\text{Na}) - 0.24(137-\text{Na})\ln(\text{Bili}) + 9.09\ln(\text{INR}) + 11.14\ln(\text{Cr}) + 1.85(3.5-\text{Alb}) - 1.83(3.5-\text{Alb})\ln(\text{Cr}) + 6$$
(Rounded to nearest integer)
New variables in MELD 3.0 vs. earlier versions:
- Female sex: +1.33 points
- Serum albumin (g/dL): incorporated with interaction term with creatinine
- Creatinine capped at 3.0 mg/dL (lowered from 4.0 in prior versions)
- Interaction terms between bilirubin-sodium and albumin-creatinine
MELD 3.0 is the current OPTN/UNOS standard in the United States. - Sabiston Textbook of Surgery; Kim WR et al., Gastroenterology 2021
3-Month Mortality by MELD Score
| MELD Score | 3-Month Mortality |
|---|
| 7 | ~1% |
| 20 | ~8% |
| 24 | ~10% |
| 26 | ~15% |
| 29 | ~20% |
| 31 | ~30% |
| 33 | ~40% |
| 35 | ~50% |
| ≥40 | >70% |
Broad risk categories:
-
MELD <10: Low risk
-
MELD 10-20: Intermediate risk
-
MELD >20: High risk - candidates strongly benefit from transplantation
-
Tietz Textbook of Laboratory Medicine, 7e; Current Surgical Therapy 14e
Clinical Uses
1. Liver Transplant Prioritization
Higher MELD = higher waitlist priority. Key threshold:
- MELD <15: Transplant may not be beneficial - post-transplant mortality can exceed waitlist mortality at this score. Transplantation generally deferred.
- MELD 15-20: Benefit of transplant starts to favor proceeding
- MELD ≥21: Clear benefit - transplantation strongly indicated
2. Perioperative Risk Stratification (Non-Transplant Surgery)
- MELD <16: Lower postoperative mortality
- MELD ≥16: Significantly higher postoperative mortality
- MELD increase positively correlates with postoperative mortality for any major surgery
- Studies show increased hospital cost and length of stay in cirrhotic patients even without portal hypertension - Sabiston Textbook of Surgery
3. TIPS Outcome Prediction
The original application. MELD <18 before TIPS associated with better outcomes.
4. Disease Monitoring
Serial MELD scores track disease progression. A rising MELD ("delta MELD") is associated with worsening prognosis.
MELD vs. Child-Turcotte-Pugh (CTP) Score
| Feature | MELD | CTP |
|---|
| Variables | Bilirubin, INR, Creatinine, Na (±sex, albumin in 3.0) | Bilirubin, INR, Albumin, Ascites, Encephalopathy |
| Type | Continuous mathematical model | Categorical (A/B/C) |
| Objectivity | Fully objective (lab-based) | Semi-subjective (ascites, encephalopathy grading) |
| Mortality prediction | Superior for short-term (90-day) | Less accurate |
| Allocation use | Current UNOS standard | Replaced (used historically) |
| Renal function | Captured (creatinine) | Not captured |
| Ceiling effect | Capped at 40 | Less precise at extremes |
MELD is superior to CTP for predicting short-term survival in cirrhosis. - Tietz Laboratory Medicine 7e
Exception Points (MELD Upgrades)
Some conditions confer mortality risk not fully captured by MELD lab values. UNOS allows "exception points" (MELD upgrade) for: - Sabiston Textbook of Surgery
- Hepatocellular carcinoma (HCC) - within Milan criteria
- Hepatopulmonary syndrome
- Portopulmonary hypertension (mean PAP <35 mmHg on treatment)
- Familial amyloid polyneuropathy
- Primary hyperoxaluria
- Cystic fibrosis
- Hilar cholangiocarcinoma (meeting specific criteria)
- Hepatic artery thrombosis within 14 days of transplant
Conditions that may qualify for exception points include refractory ascites, refractory hepatic encephalopathy, refractory variceal hemorrhage, and portal hypertensive gastropathy with chronic blood loss.
Special Populations
- Children <12 years: Use PELD (Pediatric End-Stage Liver Disease) score instead
- Hemodialysis patients: Creatinine automatically set to 4.0 mg/dL (original MELD) or 3.0 mg/dL (MELD 3.0) if on RRT
- HCC patients: MELD/PELD score is increased according to UNOS algorithm to account for cancer-related mortality risk not reflected in biochemical parameters
Regarding recent evidence: A 2024 review (
PMID 37751367) confirms MELD 3.0 improves upon MELD-Na with better sex equity and predictive accuracy in advanced chronic liver disease. The
2025 Endocrine Society Primary Aldosteronism guideline and
2025 AHA/ACC hypertension guideline do not affect MELD interpretation. No evidence currently contradicts the MELD 3.0 framework.
Sources: - Sabiston Textbook of Surgery | - Current Surgical Therapy 14e | - Tietz Textbook of Laboratory Medicine 7e | - Harrison's Principles of Internal Medicine 22e (2025) | - Barash Clinical Anesthesia 9e | Kim WR et al.,
MELD 3.0 original derivation paper,
Gastroenterology 2021