For healthcare professionals
Maddrey discriminant function calculator
The Maddrey discriminant function (DF) grades the severity of alcohol-associated hepatitis from the prothrombin time and the serum bilirubin. A score of 32 or more marks severe disease. Maddrey and colleagues introduced it in 1978, and the version used today was modified in 1989. This calculator is written for clinicians and is not a diagnosis.
Laboratory values
Use the prothrombin time in seconds, not the INR, and your laboratory's control value. Nothing you enter leaves this page.
What you will see
The discriminant function, how it was worked out, and where it falls against the cut point of 32.
Educational only, not medical advice. Not a diagnosis. Treatment decisions belong to the treating team.
Source
Maddrey WC and colleagues, 1978, modified in 1989. The formula and the cut point of 32 are taken from open-access (CC BY 4.0) reviews and studies: Rahimi and Pan, Biomarker Research 2015; Healey and Sterling, Biology 2022; Singeap and colleagues, Medicina 2024. The tests check the formula, the cut point and the effect of the control value.
Validated in
First derived in a placebo-controlled study of prednisolone in alcohol-associated hepatitis to identify patients at significant risk of early death (Rahimi 2015). It is used in adults with alcohol-associated hepatitis, not in other liver diseases.
Limits
Reported specificity for 30-day mortality is low (under 40% to 62%) and sensitivity 67% to 100% (Rahimi 2015). Scores under 32 have still been linked with a 28-day mortality of 16.7%. The result depends on which control prothrombin time is used (Healey 2022). In comparison studies, MELD predicted mortality as well as or better than the DF.
Formula
Modified discriminant function = 4.6 x (patient prothrombin time in seconds minus control prothrombin time in seconds) + total bilirubin in mg/dL (Rahimi 2015).
Worked example: prothrombin time 20 seconds, control 12 seconds, bilirubin 8 mg/dL gives 4.6 x 8 + 8 = 44.8.
The cut point of 32
A score of 32 or more indicates poor outcomes and is the threshold at which treatment with corticosteroids is considered (Healey 2022). Studies of severe alcohol-associated hepatitis define it as a score of 32 or more (Singeap 2024). Guidelines cited in one study recommend prednisolone for severe disease with a DF over 32 or a MELD over 20, when there is no contraindication (Healey 2022). The cut point has been debated: scores under 32 have still been linked with a 28-day mortality of 16.7% (Rahimi 2015).
At very high scores, results differ. An earlier study reported mortality near 100% above 54 even with steroids, while a more recent one found similar mortality of 21% and 23% above 54 with and without corticosteroids (Singeap 2024).
The control prothrombin time matters
Laboratories use different control values. In one study, the same patients were scored with control values of 12, 13.5 and 14.8 seconds: the share with a DF of 32 or more fell from 70% to 61% to 52%, and steroid use fell with it, while overall survival did not change (Healey 2022). In that study, MELD predicted 30-day survival better than the DF at every control value (AUROC 0.767 against 0.735) (Healey 2022).
Related calculators
For the MELD score, see the MELD score calculator. To grade cirrhosis, use the Child-Pugh score calculator. For alcohol withdrawal, see the CIWA-Ar calculator. Other bedside scores are listed under For healthcare professionals.
Sources
- Rahimi E, Pan JJ. Prognostic models for alcoholic hepatitis. Biomark Res. 2015;3:20. doi:10.1186/s40364-015-0046-z. PMID 26203357. Open access (CC BY 4.0). History and formula of the Maddrey discriminant function (1978, modified 1989), cut point 32, sensitivity and specificity, comparison with MELD. pmc.ncbi.nlm.nih.gov/articles/PMC4509754/
- Healey M, Sterling RK. Outnumbered: control prothrombin time in Maddrey's discriminant function impacts steroid use but not mortality in alcoholic hepatitis. Biology (Basel). 2022;11(12):1833. doi:10.3390/biology11121833. PMID 36552342. Open access (CC BY 4.0). pmc.ncbi.nlm.nih.gov/articles/PMC9775026/
- Singeap AM, Minea H, Petrea O, et al. Real-world utilization of corticosteroids in severe alcoholic hepatitis: eligibility, response, and outcomes. Medicina (Kaunas). 2024;60(2):311. doi:10.3390/medicina60020311. PMID 38399598. Open access (CC BY 4.0). pmc.ncbi.nlm.nih.gov/articles/PMC10890054/
Written and maintained by the uncovermyrisk.com editorial team. We do not have a medical reviewer; every claim links to its primary source above. Last reviewed: 2026-11-01. Spotted an error? See About.