Also known as Child-Pugh, Child-Turcotte-Pugh, CTP
Cirrhosis severity class from five clinical and laboratory items.
Applies to / doesn’t apply to6
Applies to: Adults with established cirrhosis. Pugh’s 1973 derivation was 38 patients undergoing emergency oesophageal transection for bleeding varices; its use as a prognostic class in broader cirrhosis cohorts, and in primary sclerosing cholangitis (Shetty 1997), rests on later validation.
- Acute liver failure — no validation located.
- Acute-on-chronic liver failure — dedicated scores exist for short-term mortality.
- Non-cirrhotic liver disease, and isolated abnormal liver tests without cirrhosis.
- Paediatric patients.
- During or shortly after an acute decompensating event — the score then reflects that event, not baseline reserve.
- To predict an individual drug’s clearance. The score categorises liver disease; it does not measure elimination capacity.
Under 2.0 mg/dL (34.2 µmol/L) +1 · 2.0–3.0 mg/dL (34.2–51.3 µmol/L) +2 · over 3.0 mg/dL (51.3 µmol/L) +3. A value of exactly 2.0 or exactly 3.0 scores +2. In primary biliary cholangitis, Pugh’s separate row applies instead (answer the next question).
Yes applies the bilirubin row Pugh’s original table gives for primary biliary cirrhosis (now primary biliary cholangitis): under 4 mg/dL +1 · 4–10 mg/dL +2 · over 10 mg/dL +3, with exactly 4 and exactly 10 scoring +2. Not sourced for primary sclerosing cholangitis or other cholestatic disease — answer No for those. Scores no points itself.
Over 3.5 g/dL (35 g/L) +1 · 2.8–3.5 g/dL (28–35 g/L) +2 · under 2.8 g/dL (28 g/L) +3. Higher is better, so a value of exactly 3.5 or exactly 2.8 scores +2. Serum, not urine. A recent albumin infusion raises serum albumin, and intravenous fluid can dilute it, without any change in hepatic synthetic function.
Enter INR here, or PT prolongation in the next field instead — one, not both. Under 1.7 +1 · 1.7–2.3 +2 · over 2.3 +3. A value of exactly 1.7 or exactly 2.3 scores +2. Pugh’s original table used prothrombin-time prolongation in seconds; INR is a later substitution, and the two can assign different points for the same sample. INR is also raised by anticoagulants and by vitamin K deficiency, independent of liver synthetic function.
Instead of INR, not with it: how many seconds the patient’s prothrombin time exceeds the laboratory’s control. Pugh’s original measure — 1–4 s +1 · 4–6 s +2 · over 6 s +3; exactly 4 or exactly 6 scores +2. It depends on the laboratory’s control time. Never converted to or from INR.
How much ascites is present — a judgement of volume, not of how well it responds to treatment. Grade what is there now, not whether it is controlled on diuretics.
Graded per Trey, Burns & Saunders (1966), the grading Pugh used. Grade 1: mild change in affect, cognition or sleep rhythm. Grade 2: lethargy, disorientation or inappropriate behaviour, usually with asterixis. Grade 3: stuporous but rousable, marked confusion, incoherent speech. Grade 4: coma.
What this score estimates
Classifies the severity of chronic liver disease — current hepatic functional reserve — from five items: total bilirubin, serum albumin, coagulopathy (INR, or prothrombin-time prolongation in seconds as Pugh used), ascites and hepatic encephalopathy. Each scores 1, 2 or 3 points, so the total runs from 5 to 15, grouped into Class A (5–6), Class B (7–9) and Class C (10–15). It is an ordinal class, not a probability: the outcome shown for each class is the one-year cumulative survival observed for that class in a pooled analysis of 118 studies and 23,797 patients with cirrhosis (D’Amico 2006).
Pearls & pitfalls
19
- Survival by class depends on the cohort and the endpoint. The same pooled analysis reports two-year cumulative survival of 90%, 70% and 38% for Classes A, B and C (D’Amico 2006). Among 30,840 US Veterans, one-year transplant-free survival was 94.6%, 81.8% and 51.3% — a different endpoint, and notably higher at Class C (Kaplan 2015); that cohort’s classes were assigned by an algorithm from records rather than at the bedside.
- The same pooled analysis found that the distinction between compensated and decompensated cirrhosis separates patients more sharply than Class A/B/C does — 95% against 61% one-year survival (D’Amico 2006). The class shown here is the weaker of the two stratifications in that paper.
- Two of the five items — ascites and encephalopathy — are clinician-judged, so the same patient can score differently between assessors. No published inter-rater agreement figure exists for either item. Encephalopathy grading is described as having limited interobserver reliability, especially at the lowest grade (AASLD/EASL 2014).
- Three items are confounded by common circumstances: albumin by albumin infusion or intravenous fluid, INR by anticoagulants and vitamin K deficiency, and bilirubin by Gilbert’s syndrome — each independent of hepatic synthetic function (FDA 2026).
- Pugh’s original table used prothrombin-time prolongation in seconds over the laboratory’s control (1–4, 4–6, over 6), which this tool accepts in place of INR. The INR thresholds (under 1.7, 1.7–2.3, over 2.3) are a later substitution, the two can assign different points for the same sample, and INR calibration does not normalise results in cirrhosis — INR from the same cirrhotic plasma differed across thromboplastins (Tripodi 2007). The derivation of the 1.7 and 2.3 cut-offs is not sourced. The tool never converts one form into the other; PT prolongation depends on the laboratory’s control time.
- It omits renal function and serum sodium, both material to short-term mortality in cirrhosis.
- Discrimination is moderate: a Harrell’s C of 0.701 for five-year transplant-free survival among 30,897 patients (Kaplan 2016).
- The class boundaries (5–6, 7–9, 10–15) do not appear in Pugh’s own point table; they are universally attributed to Pugh 1973 and restated by later sources (Garcia-Tsao 2016). A value exactly on an item’s boundary — bilirubin 2.0 or 3.0 mg/dL, albumin 2.8 or 3.5 g/dL, INR 1.7 or 2.3 — is scored in the middle band, as MDCalc scores it. Pugh’s printed PT ranges overlap at exactly 4 and 6 seconds; those are scored in the middle band too, by the same convention.
- The primary-biliary-cholangitis bilirubin row is printed in Pugh’s table with ranges that overlap at exactly 4 and 10 mg/dL. Both are scored in the middle band here, by the same rule as the standard rows. The row is not sourced for primary sclerosing cholangitis or other cholestatic disease.
- The derivation cohort was 38 surgical patients in 1973, with six-month post-operative survival as its outcome. Everything the score is used for today rests on later validation.
Risk factors the score does not count:
- Renal function
- Serum sodium
- Age and sex
- Portal pressure
- Variceal size or bleeding history
- Infection
- Aetiology of liver disease
- Sarcopenia
- Nutritional status — deliberately removed by Pugh from the Child-Turcotte original
What the guidelines say
4
Uses the Child-Pugh score to categorise hepatic function in pharmacokinetic studies and reproduces its point table, citing Pugh 1973.
View guideline source ↗Reproduces the Child-Pugh point table, including the separate bilirubin row for primary biliary disease.
View guideline source ↗Reports, across 118 studies, that a compensated/decompensated distinction stratifies survival more sharply than Child-Pugh class.
View guideline source ↗Describes the grading of hepatic encephalopathy, which underlies one item, as having limited interobserver reliability, particularly at the lowest grade.
View guideline source ↗Guideline positions are stated as published. Curie.MD Calc does not publish treatment direction.
Where this comes from
11
How it’s scored
What was observed
References
- Pugh RNH, Murray-Lyon IM, Dawson JL, Pietroni MC, Williams R. Transection of the oesophagus for bleeding oesophageal varices. Br J Surg. 1973;60(8):646-9. View source ↗
- D'Amico G, Garcia-Tsao G, Pagliaro L. Natural history and prognostic indicators of survival in cirrhosis: a systematic review of 118 studies. J Hepatol. 2006;44(1):217-31. View source ↗
- Garcia-Tsao G. Dig Dis Sci. 2016;61(11):3102-4. View source ↗
- Trey C, Burns DG, Saunders SJ. Treatment of hepatic coma by exchange blood transfusion. N Engl J Med. 1966;274(9):473-81. View source ↗
- Kaplan DE, Dai F, Aytaman A, et al. Development and performance of an algorithm to estimate the Child-Turcotte-Pugh score from a national electronic healthcare database. Clin Gastroenterol Hepatol. 2015;13(13):2333-41. View source ↗
- Kaplan DE, Dai F, Skanderson M, et al. Dig Dis Sci. 2016;61(11):3309-20. View source ↗
- Shetty K, Rybicki L, Carey WD. The Child-Pugh classification as a prognostic indicator for survival in primary sclerosing cholangitis. Hepatology. 1997;25(5):1049-53. View source ↗
- Tripodi A, et al. Hepatology. 2007. doi:10.1002/hep.21732. View source ↗
- Murray KF, Carithers RL Jr. Hepatology. 2005;41(6):1407-32. View source ↗
- Vilstrup H, Amodio P, Bajaj J, et al. Hepatology. 2014;60(2):715-35. View source ↗
- US Food and Drug Administration, CDER. Pharmacokinetics in Patients with Impaired Hepatic Function: Study Design, Data Analysis, and Impact on Dosage. Draft guidance for industry, September 2026. View source ↗