Free clinical calculators and scores for physicians and medical students · Browse all calculators
← All calculators and scores
Hepatology

Maddrey Discriminant Function (Maddrey Score) Calculator

Severity of alcoholic hepatitis; DF ≥ 32 typically indicates corticosteroids

Primary source: Maddrey WC et al. Gastroenterology 1978; 75: 193-198

Last clinical review: April 28, 2026

Online calculation

Patient prothrombin time
s
Control plasma PT (laboratory reference)
s
Total bilirubin
µmol/L

About this score

Maddrey's discriminant function (DF, Maddrey WC, Gastroenterology 1978) is a classic severity score for acute alcoholic hepatitis. Formula: DF = 4.6 × (patient PT − control PT) + total bilirubin (mg/dL).

DF ≥ 32 defines "severe" alcoholic hepatitis, a condition with 30-50% one-month mortality without treatment and an indication for glucocorticoids (prednisolone 40 mg/day × 28 days, with response assessed by the Lille score on day 7). DF < 32 is mild to moderate disease, usually managed with supportive care and alcohol cessation alone.

Alternative scores such as GAHS (Glasgow Alcoholic Hepatitis Score) and MELD can be used in parallel. Maddrey remains the standard tool for corticosteroid decision-making in most clinical protocols for alcoholic liver disease.

When to use

Parameters in detail

Patient prothrombin time

Prothrombin time of the patient's plasma, measured in seconds.

Control plasma PT (laboratory reference)

Prothrombin time of normal control plasma per the local laboratory reference, in seconds (typically about 12 s).

Total bilirubin

Total serum bilirubin. Enter the value in µmol/L; the calculator converts it to mg/dL for the formula.

Clinical example

Case

A 47-year-old man with a long history of alcohol use is admitted with jaundice and ascites after a two-week drinking binge. Bilirubin 9 mg/dL (~154 µmol/L), patient PT 20 seconds, control PT 12 seconds.

Calculation

DF = 4.6 × (20 − 12) + 9 = 36.8 + 9 = 45.8.

Interpretation and management

DF ≥ 32: severe alcoholic hepatitis. Management: assess contraindications to corticosteroids (active infection, GI bleeding, severe CKD); if none, start prednisolone 40 mg/day. On day 7, calculate the Lille score: < 0.45 continue therapy to 28 days, ≥ 0.45 discontinue (no response). In parallel: mandatory alcohol cessation, nutritional support (≥ 35 kcal/kg and 1.5 g/kg protein), B vitamins, and spontaneous bacterial peritonitis prophylaxis.

Limitations and cautions

Related calculators

Frequently asked questions

When should corticosteroids be started if DF ≥ 32?
After excluding active infection, GI bleeding, and severe CKD. Screening includes clinical assessment, procalcitonin, and cultures; upper endoscopy if bleeding is suspected. Therapy is usually started within the first 7 days of admission.
What counts as a "response" to corticosteroids?
A Lille score < 0.45 on day 7. If there is no response, corticosteroids are discontinued and N-acetylcysteine or transplantation (at specialized centers) may be considered.
Can pentoxifylline be used?
Per the STOPAH trial (NEJM 2015), pentoxifylline does not reduce mortality. Some protocols use it only when corticosteroids are contraindicated.
Maddrey vs MELD: which to use?
For the corticosteroid decision, use Maddrey (the historical standard). For prognosis and transplant listing, use MELD. Many clinicians use both scores together.

All clinical calculators in one place

Validated scores and formulas with full interpretation, clinical examples and source references. Open access, no registration required.

65+ calculators · scores and formulas · interpretation with management notes

АА
Aleksandr A. Aulov Physician, internal medicine, Sechenov School — independent medical education platform. Calculators are compiled from the original publications and current clinical guidelines. Interpretation thresholds follow the source study unless stated otherwise.
This calculator is intended for healthcare professionals. Do not use it for self-diagnosis or self-treatment. Management decisions are made by the treating clinician based on the full clinical picture of the individual patient.