ABIC Score Calculator
ABIC Score Calculator
Age, Bilirubin, INR and Creatinine — a linear score that splits alcohol-related hepatitis into three risk classes at 6.71 and 9.00, with 90-day survival for each. Not a steroid-response score.
ABIC score
Four inputs → 3 classesAge 48, bilirubin 12.0 mg/dL, creatinine 1.1 mg/dL, INR 2.0
Formula
Below 6.71 class A · 6.71 to 8.99 class B · 9.00 or above class C
- linear, not logarithmic
- every other score in this family takes logarithms. ABIC takes the values raw, so a unit error scales rather than compresses: a bilirubin of 200 µmol/L entered as 200 mg/dL adds 15 points to a score whose three classes span about two and a half
- no floor, no cap
- no 1.0 floor, no creatinine cap, no score ceiling. Unlike MELD it keeps responding at the extremes — and will return a low number for a patient with normal laboratory values who does not have the disease it was derived in
- 6.71 and 9.00
- the two cut-offs. The lower is exclusive (below 6.71 is class A) and the upper is inclusive (9.00 or above is class C), which is how Dominguez’s classes are published and what this page implements
- a decade is a point
- the age coefficient is 0.1 per year, so ten years of age is worth one ABIC point — more than doubling a creatinine from 1.0 to 2.0 mg/dL, which is worth 0.3. ABIC is to a substantial degree an age score
- not a steroid-response score
- ABIC stratifies prognosis at presentation. Whether corticosteroids are working is what the Lille model answers, and only at day 7. Nor is it the instrument guidelines use to define severe disease, which is Maddrey’s function at 32
Worked example
Age 48, bilirubin 12.0 mg/dL, creatinine 1.1 mg/dL, INR 2.0
Age 48 × 0.1 = 4.80 · bilirubin 12.0 × 0.08 = 0.96
Creatinine 1.1 × 0.3 = 0.33 · INR 2.0 × 0.8 = 1.60
Total 7.69 — class B, 90-day survival 70% in Dominguez's cohorts
Note where that score came from: nearly two thirds of it is the age term. The same laboratory values in a 35-year-old give 6.39 — class A — and in a 70-year-old 9.89, class C. One patient's worth of liver disease, three classes
In SI: bilirubin 205.2 µmol/L and creatinine 97.2 µmol/L give the same 7.69. Enter 205.2 and 97.2 with the unit left on conventional and the score becomes 51.99, far off the top of the scale
At exactly 6.71 the class is B, because the lower cut-off is exclusive; at exactly 9.00 the class is C, because the upper one is inclusive
The three ABIC classes
| Class | ABIC | 90-day survival | Reported discrimination |
|---|---|---|---|
| A | Below 6.71 | 100% | High negative predictive value for early death |
| B | 6.71 to 8.99 | 70% | Least informative band; later work subdivides it |
| C | 9.00 or above | 25% | At a cut-off of 9: sensitivity 43.7%, specificity 93.5% for 28-day mortality, AUROC 0.78 |
The four scores used in alcohol-related hepatitis
| Score | What it answers | When | Threshold |
|---|---|---|---|
| Maddrey discriminant function | Is this severe disease? | At presentation | 32 or more |
| ABIC | Three prognostic classes | At presentation | 6.71 and 9.00 |
| GAHS | Prognosis, and steroid benefit | At presentation | 9 or more |
| Lille model | Are corticosteroids working? | Day 7 only — meaningless on day 0 | 0.45 |
A three-class prognostic score, and it is largely an age score
Alcohol-related hepatitis already had prognostic instruments when ABIC was published: Maddrey’s discriminant function, which defines severe disease at a value of 32 or more, the Glasgow alcoholic hepatitis score, and MELD. What ABIC added was three risk classes rather than two, from four numbers available at presentation, with 90-day survival attached to each — 100% in class A, 70% in class B and 25% in class C in Dominguez’s derivation and validation cohorts of patients with alcohol-related hepatitis.
The arithmetic is unusual for this family and worth reading closely. ABIC is linear in bilirubin, creatinine and INR, where every MELD variant and the Lille model take logarithms. It bounds nothing: no floor of 1.0, no creatinine cap, no score ceiling. And the age coefficient of 0.1 per year makes age the dominant term across most of the realistic range. In the worked example above, nearly two thirds of a score of 7.69 is the age term, and the same laboratory values put a 35-year-old in class A and a 70-year-old in class C. That is not a flaw — age genuinely predicts death here — but it means ABIC should not be read as a measure of how badly the liver is inflamed.
Because it is linear and unbounded, a unit error is catastrophic rather than merely wrong. A bilirubin of 200 µmol/L typed in as 200 mg/dL adds fifteen points to a score whose three classes span about two and a half, so the result is obviously absurd — which is the good case. A bilirubin of 60 µmol/L entered as 60 adds about four and a half points, which is not obviously absurd and will move most patients two classes. The unit selector above exists for that.
Two limits on what it is for. ABIC stratifies prognosis at presentation and does not assess treatment response; the Lille model is the day-7 instrument for that, and it is meaningless on day 0 because what it measures is the fall in bilirubin over the first week of corticosteroids. And ABIC was derived in patients with alcohol-related hepatitis, a population whose median age is in the forties and fifties; applied to an elderly patient, or to liver disease of another cause, it is being used outside the data it was built on. A score is not a diagnosis and a risk derived from a cohort is not a probability for one patient. This states what the number predicted in a named study; the clinician in front of the patient decides what follows.
Frequently asked questions
What is the ABIC score formula?
ABIC = (age in years × 0.1) + (bilirubin in mg/dL × 0.08) + (creatinine in mg/dL × 0.3) + (INR × 0.8). It is linear in all four terms, with no floors, no caps and no ceiling — which makes it unlike every MELD variant.
What are the ABIC cut-offs?
Below 6.71 is class A, 6.71 to 8.99 is class B, and 9.00 or above is class C. The reported 90-day survival in Dominguez’s cohorts was 100%, 70% and 25% respectively. The lower cut-off is exclusive and the upper one inclusive.
Is ABIC better than the Maddrey discriminant function?
They answer different questions. Maddrey’s function is what guidelines use to define severe alcohol-related hepatitis, at 32 or more; ABIC offers three prognostic classes instead of a binary. In one head-to-head cohort ABIC reached an area under the curve of 0.78 for both 28-day and 90-day mortality.
Can ABIC tell me whether corticosteroids are working?
No. ABIC is calculated at presentation and stratifies prognosis. The Lille model is the instrument for treatment response, and it is a day-7 instrument — it reads the fall in bilirubin over the first week of steroids, so it is meaningless on day 0. The GAHS and MELD are the other two used at presentation.
Why does age matter so much in ABIC?
The coefficient is 0.1 per year, so a decade of age is worth a whole ABIC point — more than doubling the creatinine from 1.0 to 2.0 mg/dL, which is worth 0.3. The same laboratory values can put a 35-year-old in class A and a 70-year-old in class C.
Related calculators
References
- Dominguez M, Rincón D, Abraldes JG, et al. A new scoring system for prognostic stratification of patients with alcoholic hepatitis. Am J Gastroenterol. 2008;103(11):2747–56.
- Kim HY, Chang Y, Park JY, et al. Characterization of acute kidney injury in critically ill patients with severe coagulopathy and prognostic scores in alcoholic hepatitis. World J Gastroenterol. 2016;22(41):9205 (Table 1, prognostic score equations).
- Altamirano J, Miquel R, Katoonizadeh A, et al. A histologic scoring system for prognosis of patients with alcoholic hepatitis. Gastroenterology. 2014;146(5):1231–9.
Not medical advice. For healthcare professionals and education. Reference intervals vary by laboratory and assay — always use your own laboratory's. Never base a dose or a treatment decision on this page alone. Full disclaimer at calcengines.com/disclaimer/