Conventional point thresholds

All five components contribute at least 1, so the minimum total is 5 rather than zero. This calculator uses INR, not prothrombin-time prolongation. Do not count both.

Worked score of 7, Class B

The default total bilirubin is 1.8 mg/dL, below 2, for 1 point. Albumin is 3.2 g/dL, within 2.8–3.5, for 2. INR is 1.60, below 1.7, for 1. Those objective laboratory values total 4.

Mild or controlled ascites adds 2. No hepatic encephalopathy adds 1. The two clinical components total 3. Adding laboratory 4 and clinical 3 produces 7.

A total of 7 lies in the Class B range of 7–9. The adjacent boundary matters: improving one two-point component to one would produce 6, Class A; worsening three points could produce 10, Class C. The biology is continuous even though the labels change at integers.

Classes describe broad reserve bands

Published survival or operative-mortality tables vary by era, procedure, cohort, and care. This calculator does not attach a universal mortality percentage to a class.

How the score developed

Child and Turcotte originally grouped people with cirrhosis undergoing portosystemic shunt surgery by hepatic reserve. Pugh and colleagues modified the classification in 1973 while studying esophageal transection for variceal bleeding, replacing nutritional status with prothrombin time and formalizing a point approach.

The historical purpose matters. A score created around surgery for portal hypertension is now used in prognosis, study eligibility, procedure discussions, and drug-label hepatic-impairment categories. Performance is not identical across those uses.

AASLD teaching materials emphasize that ascites and encephalopathy are subjective and can change with treatment. The apparent precision of a 7 versus 8 must not obscure that judgment.

Ascites and encephalopathy require bedside assessment

Ascites severity depends on examination, imaging, symptoms, response to sodium restriction and diuretics, paracentesis history, and whether fluid is refractory. A small amount seen only on imaging may be categorized differently by a study or protocol.

Hepatic encephalopathy grading considers attention, sleep pattern, behavior, orientation, neurologic signs, and ability to perform tasks. Sedatives, infection, bleeding, kidney failure, electrolyte disorders, alcohol withdrawal, dementia, stroke, and other causes can mimic or worsen confusion.

Because treatment can suppress ascites or encephalopathy, record whether the selected category reflects treated status and follow the definition required by the clinical purpose. This public calculator cannot examine the patient.

INR can be confounded by anticoagulation

INR was standardized for vitamin K antagonist monitoring and can be elevated by warfarin independent of hepatic synthetic function. Direct oral anticoagulants can also affect some clotting assays variably. Laboratory reagents, vitamin K deficiency, transfusion, and acute illness add complexity.

If anticoagulant effect is selected, the calculator still shows the arithmetic but warns that interpretation is compromised. Do not substitute an assumed INR of 1 without a governing protocol and expert direction.

Stopping anticoagulation to obtain a “cleaner” score can cause thrombosis and is unsafe. Clinicians must interpret the indication, timing, medication, laboratory method, and alternatives.

Bilirubin cutoffs are not universal in cholestatic disease

The conventional table uses under 2, 2–3, and over 3 mg/dL. Some applications involving primary biliary cholangitis or other cholestatic disease use higher alternative bilirubin thresholds because bilirubin behavior differs.

This calculator does not implement an alternate scale. Selecting the cholestatic switch blocks a conventional result and directs the user to the exact protocol. Mixing a higher bilirubin cutoff with the remaining conventional fields without documentation creates a mislabeled score.

Use total bilirubin from the same clinical period as albumin and INR. Do not enter direct bilirubin in the total field.

Child–Pugh is not MELD or transplant allocation

MELD-based systems use laboratory variables and are central to U.S. liver transplant allocation. They answer different questions and are updated under national policy. Child–Pugh class alone does not determine wait-list priority.

MELD, Child–Pugh, portal-hypertension measures, frailty, kidney function, sodium, nutrition, complications, imaging, cancer status, and procedure-specific models may all contribute to assessment. A clinician selects the tool suited to prognosis, surgery, TIPS, transplant, or medication.

Do not use a favorable Child–Pugh class to delay referral when decompensation, hepatocellular carcinoma, recurrent variceal bleeding, refractory ascites, or other serious features are present.

Drug labels use class in drug-specific ways

FDA prescribing information may report pharmacokinetic studies in Child–Pugh A, B, or C and may recommend no adjustment, a different dose, avoidance, or state that evidence is unavailable. The rule for one medicine cannot be transferred to another.

A label’s studied population may exclude acute hepatitis, cholestasis, severe decompensation, transplant, or interacting drugs. Albumin binding, hepatic extraction, renal clearance, active metabolites, and therapeutic window matter beyond class.

Use the current label, indication, full medication list, renal function, and specialist or pharmacist review. This score does not calculate a dose or authorize treatment.

Timing and trends matter

Collect bilirubin, albumin, and INR from the same period and document transfusion, albumin infusion, vitamin K, anticoagulants, paracentesis, diuretic change, infection, bleeding, and encephalopathy treatment. Mixing a six-month-old albumin with today’s INR and ascites can create a fictitious state.

A trend can show recovery or decompensation, but small changes near cutoffs may reflect assay or assessment variation. Preserve the raw values and categories, not just the letter class.

A rapidly deteriorating patient needs urgent clinical action regardless of whether the integer total has crossed a class boundary.

Do not repurpose the score for acute liver failure

Acute liver failure is a time-critical syndrome with coagulopathy and encephalopathy in a person without established cirrhosis, and it needs urgent specialist and transplant-center evaluation. Child–Pugh was not designed as its triage calculator. A low albumin may not yet have developed, while INR and mental status can worsen rapidly.

Acute hepatitis, drug-induced liver injury, acetaminophen toxicity, ischemic injury, pregnancy-related liver disease, and other emergencies have cause-specific tests and prognostic systems. A mathematically obtainable Child–Pugh class must not delay poison consultation, transfer, antidotal therapy, intensive monitoring, or another indicated intervention.

A transparent scoring workflow

Child–Pugh checklist

1. Confirm the score is applicable to cirrhosis.
2. Use total bilirubin in mg/dL.
3. Use albumin in g/dL.
4. Use INR or PT, not both.
5. Check anticoagulant effect on INR.
6. Assess ascites clinically.
7. Assess encephalopathy clinically.
8. Document treatment-controlled findings.
9. Use one contemporaneous period.
10. Check for cholestatic alternate cutoffs.
11. Preserve raw values and component points.
12. Do not attach a universal mortality rate.
13. Do not infer transplant priority.
14. Verify medication-specific labeling.

It does not determine medical deductions, drug coverage, disability, liver reserve, prognosis, or treatment.

Frequently asked questions

References

• AASLD review of Child–Pugh and MELD
• Pugh et al., Transection of the oesophagus for bleeding oesophageal varices
• The Child–Turcotte Classification: From Gestalt to Sophisticated Statistics and Back
• 2026 AASLD-IDSA HCV guidance Child–Turcotte–Pugh table
• FDA prescribing information resources
• Design considerations for hepatic impairment pharmacokinetic studies