About the Child-Turcotte-Pugh Score (Child-Pugh)
The Child-Pugh score grades the severity of cirrhosis from five variables — total bilirubin, serum albumin, INR, ascites and hepatic encephalopathy. Each scores 1 to 3 points, so the total runs from 5 to 15 and sorts patients into class A (5–6), class B (7–9) or class C (10–15). Class A is well-compensated disease, class B significant functional compromise, and class C decompensated cirrhosis. Although MELD has replaced it for transplant allocation, Child-Pugh class remains the language of drug labels, trial eligibility criteria and procedural guidance.
Formula
Child-Pugh = bilirubin points + albumin points + INR points + ascites points + encephalopathy points
Class A = 5–6 Class B = 7–9 Class C = 10–15- Bilirubin points
- 1 if < 2 mg/dL, 2 if 2–3 mg/dL, 3 if > 3 mg/dL.
- Albumin points
- 1 if > 3.5 g/dL, 2 if 2.8–3.5 g/dL, 3 if < 2.8 g/dL.
- INR points
- 1 if < 1.7, 2 if 1.7–2.3, 3 if > 2.3.
- Ascites points
- 1 if none, 2 if mild/diuretic-responsive, 3 if moderate-severe/refractory.
- Encephalopathy points
- 1 if none, 2 if West Haven grade 1–2, 3 if grade 3–4.
- It is a simple additive point score, which is why it can be done at the bedside without a calculator — the reason it has outlasted more accurate models.
- Minimum 5, maximum 15. There is no 0.
- For SI units, divide bilirubin in µmol/L by 17.1 and albumin in g/L by 10, or use the unit toggles on this page.
Interpreting the result
Read the class first and the number second, because the class is what published guidance is written against. Then look at which categories are driving the total: a score of 8 made up of laboratory abnormalities is a different patient from an 8 driven by refractory ascites and encephalopathy, and only the second is likely to need urgent transplant discussion. Because two of the five items are subjective, treat a borderline score as borderline — a 6 and a 7 recorded by different clinicians on the same patient is common, and it is the difference between class A and class B.
| Score | Band | What it means | Action |
|---|---|---|---|
| 5–6 | Class A | Well-compensated disease. Around 95–100% one-year and 85% two-year survival in observational series | Surveillance (six-monthly ultrasound, variceal screening); treat the underlying aetiology |
| 7–9 | Class B | Significant functional compromise. Around 80% one-year and 60% two-year survival; markedly increased surgical risk | Transplant assessment; review every drug for hepatic dosing; treat ascites and encephalopathy actively |
| 10–15 | Class C | Decompensated disease. Around 45% one-year and 35% two-year survival | Urgent transplant evaluation if a candidate; expect poor tolerance of surgery and anaesthesia |
Scroll the table sideways for every column.
What the Child-Pugh Score needs (5 inputs)
- Total bilirubin (mg/dL)
- 1 point below 2 mg/dL, 2 points at 2–3, 3 points above 3. In primary biliary cholangitis and primary sclerosing cholangitis some centres use higher thresholds (4 and 10 mg/dL), because cholestatic disease raises bilirubin without the same prognostic weight.
- Serum albumin (g/dL)
- 1 point above 3.5 g/dL, 2 points at 2.8–3.5, 3 points below 2.8. Albumin is confounded by sepsis, nephrotic-range proteinuria and recent albumin infusion, none of which reflect synthetic liver function.
- INR
- 1 point below 1.7, 2 points at 1.7–2.3, 3 points above 2.3. Some versions of the score use prothrombin time prolongation in seconds instead; the classes are the same but the boundaries are not interchangeable.
- Ascites
- 1 point for none, 2 for mild or diuretic-responsive, 3 for moderate-to-severe or refractory. This is a clinical judgement, and it is the single largest source of disagreement between two clinicians scoring the same patient.
- Hepatic encephalopathy
- 1 point for none, 2 points for West Haven grade 1–2, 3 points for grade 3–4. Note this is not one point per grade — the five West Haven grades collapse into three scoring bands.
Units. Enter bilirubin in mg/dL and albumin in g/dL, or switch the unit toggles to µmol/L and g/L and the calculator converts exactly (bilirubin ÷ 17.1, albumin ÷ 10). If your service reports prothrombin time in seconds rather than INR, convert to INR before entering — the point thresholds here are the INR ones, and the seconds-based version of the score uses different boundaries.
What it returns
- Child-Pugh score (5–15)
- The sum of the five categories. Because every category scores at least 1, the minimum is 5 — a total below that indicates the inputs were mapped wrongly.
- Child-Pugh class (A, B or C)
- A for 5–6, B for 7–9, C for 10–15. The class is what most guidance is written against, so it usually matters more than the number.
How it is calculated
Child and Turcotte published the original classification in 1964 for patients undergoing portosystemic shunt surgery, using bilirubin, albumin, ascites, encephalopathy and nutritional status. Pugh's 1973 modification, in patients undergoing oesophageal transection for bleeding varices, replaced nutritional status — the most subjective item — with prothrombin time, and it is Pugh's version that is used today. The arithmetic is deliberately crude: each variable is compressed into three bands and the bands are simply summed, with no weighting. That means a patient can move from class A to class B on a single laboratory value crossing a threshold by 0.1, and two patients with the same total can be clinically very different.
Facts & figures
| Variable | 1 point | 2 points | 3 points |
|---|---|---|---|
| Total bilirubin (mg/dL) | < 2 | 2–3 | > 3 |
| Serum albumin (g/dL) | > 3.5 | 2.8–3.5 | < 2.8 |
| INR | < 1.7 | 1.7–2.3 | > 2.3 |
| Ascites | None | Mild / diuretic-responsive | Moderate–severe / refractory |
| Encephalopathy (West Haven) | None | Grade 1–2 | Grade 3–4 |
Scroll the table sideways for every column.
Cholestatic disease (PBC, PSC) is sometimes scored with higher bilirubin thresholds of 4 and 10 mg/dL. Say which convention you used when recording the score.
Evidence
Original classification — Child & Turcotte
1964Described for patients with cirrhosis undergoing portosystemic shunt surgery, using bilirubin, albumin, ascites, encephalopathy and nutritional status. It was a surgical risk classification, not a general prognostic model.
Pugh's modification — oesophageal transection series
1973Pugh and colleagues applied the classification to patients undergoing transection of the oesophagus for bleeding varices, replacing nutritional status with prothrombin time. This is the version in universal use. The original paper has no abstract and the survival figures usually quoted for class A, B and C come from later observational series rather than from Pugh's own data.
Displacement by MELD for allocation
2003When MELD was compared with Child-Turcotte-Pugh for waiting-list prioritisation, MELD discriminated three-month mortality better and, being built only from laboratory values, could not be influenced by subjective assessment. US allocation moved to MELD in 2002.
In the VOCAL-Penn surgical-risk study, Child-Turcotte-Pugh had the weakest 30-day discrimination of every tool compared, with a C-statistic of 0.682 against 0.859 for a purpose-built surgical model.
How it compares
Child-Pugh Score vs MELD-Na
Use MELD-Na or MELD 3.0 for prognosis and transplant priority, and Child-Pugh whenever you need a class — the two answer different questions and cannot be converted into one another.
MELD-Na is continuous, fully objective, and validated for three-month mortality, which is why allocation moved to it. Child-Pugh contains two subjectively graded variables and compresses everything into three bands, so it discriminates less well. But drug labels, trial eligibility and procedural guidance are written in Child-Pugh classes, and no MELD score answers 'is this patient Child-Pugh B'. Most patients under assessment need both recorded.
Child-Pugh Score vs ALBI grade
ALBI is the better choice when you want liver function graded objectively in hepatocellular carcinoma — it uses only albumin and bilirubin, so it removes the subjectivity that limits Child-Pugh.
ALBI was developed specifically because Child-Pugh class A is heterogeneous in HCC populations and the subjective items make it unreliable for stratifying trial cohorts. ALBI splits patients on albumin and bilirubin alone and discriminates within Child-Pugh A. It says nothing about ascites or encephalopathy, so it is a complement rather than a replacement for general cirrhosis staging.
Child-Pugh Score vs CLIF-SOFA
Switch to CLIF-SOFA when the question is acute-on-chronic liver failure rather than chronic severity — Child-Pugh was not built for an acutely deteriorating patient and does not count organ failures.
Child-Pugh describes the chronic state of the liver. CLIF-SOFA counts failing organ systems, including renal, circulatory, cerebral and respiratory, and grades ACLF by how many have failed — which is what actually predicts 28-day mortality in a decompensating inpatient. A patient can be Child-Pugh B and have grade 3 ACLF.
Pearls & pitfalls
- The minimum score is 5, not 0 or 3. Every category scores at least 1 point, so a healthy person would score 5 if you scored them at all.
- Encephalopathy grades do not map one-to-one onto points. West Haven grades 1 and 2 both score 2 points, and grades 3 and 4 both score 3 — grade 4 does not score 4.
- Ascites and encephalopathy are subjective, and inter-observer disagreement on them is the main reason the same patient can be scored A by one clinician and B by another. Record what you observed, not just the total.
- Albumin is dragged down by sepsis, proteinuria and dilution, and pushed up by recent infusion. A patient given albumin for large-volume paracentesis yesterday will score better today without being better.
- Class A covers a wide range. A score of 5 and a score of 6 are both class A, but a 6 driven by a bilirubin of 2.5 is not the same patient as a 5 with normal labs.
- Bilirubin thresholds differ in cholestatic disease. If you are scoring PBC or PSC, note which convention you used, because the two give different classes for the same patient.
- Child-Pugh and MELD are not interconvertible in either direction. If guidance asks for a class, you need the class.
Critical actions
- Record which categories drove the score, not only the total — the composition changes management more than the number does.
- Check every prescription against hepatic dosing guidance once a patient reaches class B; several common drugs are contraindicated from class B onward.
- Calculate MELD 3.0 as well if transplant priority is the question, since Child-Pugh does not answer it.
- Do not use Child-Pugh class alone to refuse surgery — use a model built for operative risk, which accounts for the specific procedure.
- Reassess after treating ascites or encephalopathy: the class can improve, and a class recorded during an acute decompensation may not represent the patient's baseline.
- Start hepatocellular carcinoma surveillance and variceal screening in every class, including class A — good compensation is not an absence of risk.
Why this score exists
Both versions of this score were built to answer a surgical question, not a prognostic one. Child and Turcotte were assessing who could survive a portosystemic shunt; Pugh was assessing who could survive oesophageal transection. The score became a general-purpose severity classification afterwards, by adoption rather than by design, which explains its weaknesses — nobody set out to build a model for drug dosing or trial eligibility, and the coarse three-band structure that makes it easy to do by hand is the same structure that makes it imprecise. Pugh's specific contribution is instructive: he removed nutritional status precisely because it was too subjective to score reliably, yet left ascites and encephalopathy in, and those remain the score's weakest links.
About the creator
Originator of the 1964 classification
With Jeremiah Turcotte, published the original A/B/C grading of hepatic reserve in patients undergoing portosystemic shunt surgery. He is not an author of the 1973 paper cited here.
First author, 1973 modification
Modified the classification for use in transection of the oesophagus for bleeding varices, replacing nutritional status with prothrombin time — the version in use today.
Limitations
- Two of the five variables are graded subjectively, so the same patient can be scored differently by different clinicians, and a borderline case can cross a class boundary on opinion alone.
- The three-band structure discards information: a bilirubin of 2.1 and one of 3.0 score identically, and a value 0.1 either side of a threshold changes the class.
- Class C spans 10 to 15, so it groups patients with very different prognoses together.
- Albumin and INR are both affected by things other than liver synthesis — infusion, sepsis, proteinuria, anticoagulation.
- It was derived for surgical decisions in the 1960s and 1970s, in populations and with treatments that no longer resemble current practice, and the survival figures quoted for each class come from later observational series rather than the original papers.
- It says nothing about renal function, which is a major determinant of outcome in cirrhosis and is the reason MELD includes creatinine.
If you are the patient
The Child-Pugh score is a way of describing how well your liver is working, using three blood tests and two things your doctor examines you for — fluid in the abdomen (ascites) and whether liver problems are affecting your concentration or alertness. The total puts you in class A, B or C. Class A means the liver is coping well despite scarring; class B means it is struggling in ways that need active management; class C means it is failing and transplant is usually part of the conversation. Two things are worth knowing. First, the class is used to decide which medicines are safe for you, which is why your team records it even when they also use other scores. Second, it can improve — treating fluid build-up or an infection can move someone from class B back toward class A, so a class measured when you were acutely unwell is not necessarily your baseline. Ask which parts of the score are highest for you, because that points at what can be treated.
Frequently asked questions
What is the Child-Pugh score?#
The Child-Pugh score grades the severity of cirrhosis from five variables — total bilirubin, serum albumin, INR, ascites and hepatic encephalopathy — each scoring 1 to 3 points. The total runs from 5 to 15 and gives class A (5–6), class B (7–9) or class C (10–15).
What are the Child-Pugh classes A, B and C?#
Class A is 5–6 points and represents well-compensated cirrhosis. Class B is 7–9 points and represents significant functional compromise. Class C is 10–15 points and represents decompensated cirrhosis. Reported survival falls across the classes, from roughly 95–100% one-year survival in class A to around 45% in class C in observational series.
Is Child-Pugh class B bad?#
Class B means the liver has lost meaningful reserve and needs active management — it is the point at which many drugs require dose adjustment or become contraindicated, surgical risk rises substantially, and transplant assessment is usually appropriate. It is not the end stage; class C is more advanced. Class B can also improve if a treatable driver such as ascites, infection or continued alcohol use is addressed.
What is the lowest Child-Pugh score?#
Five. Every one of the five categories scores at least 1 point, so 5 is the minimum and there is no score of 0, 3 or 4. If a calculator returns less than 5, its category mapping is wrong.
Is Child-Pugh or MELD better?#
For prognosis and transplant priority, MELD — it discriminates three-month mortality better and is fully objective, which is why US allocation moved to it in 2002. For anything expressed in Child-Pugh classes, such as hepatic drug dosing or trial eligibility, Child-Pugh is the only option, because the two scores cannot be converted into each other.
How is hepatic encephalopathy scored in Child-Pugh?#
In three bands, not five: no encephalopathy scores 1 point, West Haven grade 1–2 scores 2 points, and grade 3–4 scores 3 points. Grade 4 encephalopathy does not score 4 points, because no Child-Pugh category exceeds 3.
Can the Child-Pugh score improve?#
Yes. Ascites and encephalopathy both respond to treatment, and albumin and bilirubin can improve when a precipitant such as infection or alcohol is removed. A score recorded during an acute decompensation often overstates the patient's baseline, which is why it is worth repeating once they are stable.
Does Child-Pugh include kidney function?#
No. There is no creatinine or renal term, which is one of its main limitations — renal dysfunction is a major determinant of outcome in cirrhosis. MELD includes creatinine specifically for that reason.