About the Revised Original Scoring System for the Diagnosis of Autoimmune Hepatitis (IAIHG 1999)
Thirteen parameters plus optional items, summed to a total that is read against different thresholds depending on when it is applied: before treatment, above 15 is definite and 10–15 probable; after treatment, above 17 is definite and 12–17 probable. The same total of 17 is therefore definite before treatment and only probable after it. Two features distinguish it from the simplified criteria: the histology items are additive, so interface hepatitis, a lymphoplasmacytic infiltrate and rosetting each score separately while their combined absence costs 5; and there are real penalties — a positive AMA costs 4 and a positive drug history costs 4 — which is what makes this the more sensitive score in atypical cases and also what most often sinks an otherwise convincing one.
Formula
Score = sex + ALP:AST ratio + globulins/IgG + autoantibody titre + AMA
+ viral markers + drug history + alcohol + histology (additive)
+ other autoimmune disease + optional items + response to therapy
Pre-treatment: definite > 15 · probable 10–15
Post-treatment: definite > 17 · probable 12–17- Histology (additive)
- Interface hepatitis +3, lymphoplasmacytic infiltrate +1, rosetting +1, none of these −5, biliary changes −3, other atypical features −3. Score each independently.
- Penalties
- AMA positive −4, drug history positive −4, viral markers positive −3, alcohol above 60 g/day −2, cholestatic ALP:AST ratio −2.
- Response to therapy
- Complete +2, relapse +3. Only applicable after treatment, which is why the post-treatment thresholds are two points higher.
- The thresholds differ by timing, and the difference matters: a total of 17 is DEFINITE before treatment but only PROBABLE after it, because the response items add points that the higher threshold offsets.
- The histology items are additive, not mutually exclusive. A biopsy showing interface hepatitis, a lymphoplasmacytic infiltrate and rosetting scores +5 in total.
- 'None of the three features above' carries −5 and is scored only when interface hepatitis, the infiltrate and rosetting are all absent — an eight-point swing against interface hepatitis alone.
- The score can go negative. A patient with a positive AMA, positive viral markers and a positive drug history starts 11 points down before anything positive is counted.
- Alcohol between 25 and 60 g/day scores 0, which is easy to miss — only the two extremes carry points.
- Relapse scores MORE than a complete response (+3 versus +2), because relapse on withdrawal is stronger evidence of a treatment-dependent autoimmune process.
Interpreting the result
Apply the right thresholds for the timing, since using the pre-treatment cut-offs on a post-treatment score will over-call the diagnosis. Before treatment, above 15 is definite and 10–15 probable; after treatment, above 17 is definite and 12–17 probable. When a score falls short, the productive question is which penalties are doing the work rather than which positives are missing. A positive AMA costs 4 points and a positive drug history another 4, and either can pull a clinically convincing case below the threshold — sometimes correctly, because that is the score working as intended, and sometimes not, because AMA can be positive at low titre in autoimmune hepatitis and because a drug history may be incidental. Both deserve a second look rather than automatic acceptance. A probable score is not a weak result: it supports a treatment trial in the right context, and the response to that trial is itself scored, which is the mechanism by which a probable pre-treatment case can become a definite post-treatment one.
| Score | Band | What it means | Action |
|---|---|---|---|
| > 15 (pre-treatment) | Definite autoimmune hepatitis | Meets the revised criteria for definite disease before treatment | Treat as autoimmune hepatitis; obtain histology if not already available to stage fibrosis |
| 10–15 (pre-treatment) | Probable autoimmune hepatitis | Supports the diagnosis and a treatment trial in the right clinical context | Consider corticosteroids with or without azathioprine; re-score after treatment using the higher thresholds |
| > 17 (post-treatment) | Definite autoimmune hepatitis | Meets the post-treatment criteria, with the response to therapy contributing | Continue treatment; plan maintenance and withdrawal strategy |
| 12–17 (post-treatment) | Probable autoimmune hepatitis | Meets the post-treatment probable criteria | Reassess alternative diagnoses alongside continuing treatment; consider overlap syndromes |
| Below the probable threshold | Criteria not met | Does not meet the revised criteria. Check whether the penalties rather than absent positives are responsible | Re-examine a positive AMA or drug history; test anti-SLA/LP, actin, LC1 and pANCA if not already done |
What the Revised Original AIH Score needs (12 inputs)
- Sex
- Female scores +2, reflecting the marked female predominance of the disease.
- Ratio of ALP to AST (or ALT)
- Both expressed as multiples of their upper limit of normal. Above 3 scores −2, 1.5–3.0 scores 0, and below 1.5 scores +2. A cholestatic ratio counts against the diagnosis.
- Serum globulins or IgG above normal
- Graded by multiple of the upper limit: below 1.0 scores 0, 1.0–1.5 scores +1, 1.5–2.0 scores +2, and above 2.0 scores +3.
- ANA, SMA or anti-LKM1 titre
- Below 1:40 scores 0, 1:40 scores +1, 1:80 scores +2, and above 1:80 scores +3. Unlike the simplified criteria this is a graded scale rather than a capped pair of options.
- Antimitochondrial antibody
- Positive scores −4, the heaviest single penalty in the score. A positive AMA points towards primary biliary cholangitis.
- Hepatitis viral markers
- Positive scores −3, negative scores +3 — a six-point swing.
- Drug history
- A positive history scores −4 and a negative one +1, a five-point swing that reflects how closely drug-induced liver injury mimics autoimmune hepatitis.
- Average alcohol intake
- Below 25 g/day scores +2, above 60 g/day scores −2, and the range between them scores 0.
- Liver histology (six additive items)
- Interface hepatitis +3, predominantly lymphoplasmacytic infiltrate +1, rosetting of liver cells +1, none of those three present −5, biliary changes −3, other atypical features −3. These accumulate rather than being alternatives.
- Other autoimmune disease
- In the patient or a first-degree relative, +2.
- Optional parameters
- Seropositivity for other defined autoantibodies (anti-SLA/LP, actin, LC1, pANCA) +2, and HLA DR3 or DR4 +1.
- Response to therapy
- Complete response +2, relapse +3. Scored only after treatment, and its availability is why the post-treatment thresholds are higher.
What it returns
- Total score
- A signed total that can be negative. Most diagnostic cases fall between 10 and 20.
- Diagnostic category
- Definite, probable, or criteria not met — read against pre-treatment or post-treatment thresholds as appropriate.
- Domain breakdown
- Including the net histology contribution and the optional parameters, so it is visible which section is driving the total.
How it is calculated
The revised score is a descriptive consensus instrument rather than a fitted statistical model: the International Autoimmune Hepatitis Group assembled the features that experienced hepatologists actually weigh, assigned point values by agreement, and then checked that the totals separated cases from non-cases. That origin explains its shape. It is long because the diagnosis genuinely rests on many strands; it contains large negative weights because autoimmune hepatitis is substantially a diagnosis of exclusion, and the two conditions it is most often confused with — primary biliary cholangitis and drug-induced liver injury — each get a dedicated 4-point penalty. The 1999 revision's main addition over the 1993 original was the response to therapy, which formalised something clinicians were already doing informally: treating a plausible case and taking the response as diagnostic evidence. Because those items can only add points, the post-treatment thresholds were raised by two to keep the categories comparable.
Facts & figures
| Finding | Points | Alternative diagnosis it points to |
|---|---|---|
| Antimitochondrial antibody positive | −4 | Primary biliary cholangitis |
| Drug history positive | −4 | Drug-induced liver injury |
| Hepatitis viral markers positive | −3 | Viral hepatitis |
| Biliary changes on histology | −3 | Cholangiopathy or overlap |
| Other atypical histological features | −3 | An alternative aetiology |
| ALP:AST ratio above 3 | −2 | Cholestatic rather than hepatitic disease |
| Alcohol above 60 g/day | −2 | Alcohol-related liver disease |
| No interface hepatitis, infiltrate or rosetting | −5 | Not autoimmune hepatitis histologically |
Total available penalties exceed 20 points. This is what makes the score more sensitive than the simplified criteria in atypical cases — it can actively rule alternatives out rather than only failing to rule them in.
| Total | Before treatment | After treatment |
|---|---|---|
| > 17 | Definite | Definite |
| 16–17 | Definite | Probable |
| 12–15 | Probable | Probable |
| 10–11 | Probable | Criteria not met |
| < 10 | Criteria not met | Criteria not met |
The post-treatment thresholds are two points higher because the response-to-therapy items can only add points. Applying the pre-treatment cut-offs to a post-treatment score systematically over-calls the diagnosis.
Evidence
Revised criteria — Alvarez et al., International Autoimmune Hepatitis Group
1999Consensus revision by the International Autoimmune Hepatitis Group of the scoring system first published in 1993, assembling the clinical, biochemical, serological and histological features bearing on the diagnosis and assigning point values by expert agreement.
Established the thirteen-parameter score with separate pre-treatment (definite > 15, probable 10–15) and post-treatment (definite > 17, probable 12–17) thresholds, and added response to therapy as a scored criterion over the 1993 version.
Simplified criteria — the score derived from this one
2008Development of the four-parameter simplified criteria within the same group, testing which of the revised original parameters retained independent diagnostic value.
Achieved about 88% sensitivity and 97% specificity at a cut-off of ≥6, trading the revised original's sensitivity in atypical cases for practicality in routine clinical use.
AASLD practice guidance
2020The 2019 AASLD practice guidance and guidelines on autoimmune hepatitis in adults and children.
Retains the revised original score for atypical presentations and research case definition, with the simplified criteria recommended for routine diagnosis.
EASL clinical practice guidelines
2015EASL clinical practice guidelines on autoimmune hepatitis.
Recommends the revised original score where the presentation does not fit the typical pattern captured by the simplified criteria, particularly in suspected overlap syndromes.
How it compares
Revised Original AIH Score vs Simplified AIH criteria
Use the simplified criteria first in routine practice, and this score when they fall short but suspicion persists — more sensitive, at the cost of being far longer.
The simplified criteria use four parameters and reach about 88% sensitivity and 97% specificity; the revised original uses thirteen plus optional items. The extra parameters are not padding — the AMA penalty, the drug history penalty, the alcohol term and the response to therapy have no counterpart in the simplified score, and they are exactly what makes the difference in seronegative disease, overlap syndromes and drug-induced mimics. The two scores disagree in a meaningful minority of cases, and applying both is reasonable when the answer changes management.
Revised Original AIH Score vs RUCAM
Complementary and often needed together — this score penalises a drug history by 4 points, and RUCAM quantifies how likely that drug really is to be responsible.
Drug-induced liver injury with autoimmune features is the hardest differential in this area, and the two instruments approach it from opposite directions. The revised AIH score docks points for any positive drug history regardless of plausibility; RUCAM assesses whether the temporal relationship, dechallenge and alternative causes actually support that drug as the cause. A high RUCAM alongside a borderline AIH score argues for drug-induced injury; a low RUCAM suggests the drug history is incidental and the 4-point penalty may be misleading.
Revised Original AIH Score vs R factor
Upstream — the R factor classifies the biochemical pattern, which informs whether the ALP:AST term here will help or hurt.
A cholestatic R factor corresponds to the ALP:AST ratio above 3 that costs 2 points in this score, and should also prompt consideration of primary biliary cholangitis or an overlap syndrome. A hepatocellular pattern is consistent with typical autoimmune hepatitis. Calculating the pattern first makes the ratio term easier to score correctly.
Pearls & pitfalls
- Use the thresholds that match the timing. A total of 17 is definite before treatment but only probable after it, and applying the pre-treatment cut-offs to a post-treatment score over-calls the diagnosis.
- The histology items are additive. Interface hepatitis, a lymphoplasmacytic infiltrate and rosetting each score separately, together contributing +5.
- 'None of the three histological features' carries −5, an eight-point swing against interface hepatitis alone. It is scored only when all three are genuinely absent.
- When a score falls short, look at the penalties first. A positive AMA (−4) or drug history (−4) can pull a convincing case below the threshold — sometimes correctly, sometimes not.
- A low-titre AMA can occur in autoimmune hepatitis. The −4 penalty is appropriate as a default but deserves a second look before the diagnosis is abandoned.
- Alcohol between 25 and 60 g/day scores zero. Only intake below 25 or above 60 carries points.
- Relapse scores more than a complete response (+3 versus +2), which surprises people — relapse on withdrawal is stronger evidence of a treatment-dependent process.
- The optional items are genuinely optional. Anti-SLA/LP, actin, LC1 and pANCA are worth +2 and HLA typing +1, but their absence is not penalised — so a score is not invalid for omitting them.
- This score does not stage disease. A definite result says nothing about fibrosis, which comes from the biopsy itself.
Critical actions
- Establish whether you are scoring before or after treatment, and apply the corresponding thresholds explicitly.
- Obtain liver histology — it carries the largest single block of points and is where the score is most often incomplete.
- Take a careful drug and supplement history, since a positive one costs 4 points and drug-induced liver injury is the principal mimic.
- Test AMA, and interpret a low-titre positive result in context rather than treating it as automatically exclusionary.
- Test the extended autoantibody panel — anti-SLA/LP, actin, LC1 and pANCA — which contributes 2 optional points and is frequently omitted.
- Where a treatment trial is undertaken, document the response formally and re-score afterwards using the post-treatment thresholds.
- Consider overlap with primary biliary cholangitis or primary sclerosing cholangitis where the biochemistry is cholestatic or the AMA is positive, and arrange cholangiography where indicated.
Why this score exists
The score's most revealing feature is how much of it is negative. More than twenty points of penalty are available, against roughly twenty-five points of positive findings, and the two heaviest single items — a positive AMA and a positive drug history, at −4 each — are both about something other than autoimmune hepatitis. That balance encodes the group's view that this is substantially a diagnosis of exclusion, and it is the structural reason the revised score outperforms the simplified criteria in atypical cases: it can actively argue against an alternative rather than merely failing to find support for the diagnosis. The other telling choice is that relapse scores higher than a complete response. A patient who responds and stays well might have had something self-limiting; a patient who relapses when treatment is withdrawn has demonstrated a treatment-dependent process, which is stronger evidence — and the group weighted it accordingly.
About the creator
First author, 1999 revised criteria
Led the International Autoimmune Hepatitis Group report revising the diagnostic criteria.
Co-author of both the revised original and the simplified criteria
A principal contributor to autoimmune hepatitis diagnostic criteria across both scoring systems.
Co-author of the 1999 International Autoimmune Hepatitis Group report.
Limitations
- Consensus-derived rather than statistically fitted, so the point values reflect expert agreement rather than regression weights and have not been re-derived against modern outcome data.
- Long enough that it is rarely completed fully in routine practice, and partial completion systematically lowers the total because most optional items only add points.
- Includes HLA typing, which is not routinely available in many settings and contributes only 1 point when it is.
- The response-to-therapy items create circularity — treatment is given on the basis of a suspected diagnosis, and the response is then used as evidence for that diagnosis.
- Performs less well in acute severe and fulminant presentations, where autoantibodies and IgG are often normal and histology may be dominated by necrosis.
- The drug history penalty is binary and does not distinguish a plausible temporal relationship from an incidental exposure, which RUCAM exists to assess.
- Not designed for children, in whom autoimmune sclerosing cholangitis overlaps closely and requires cholangiography rather than scoring to identify.
- Says nothing about severity, fibrosis stage or prognosis.
If you are the patient
Autoimmune hepatitis is diagnosed by weighing up many pieces of evidence rather than by one test. This scoring system is the longer and more thorough of the two in use — doctors usually try a shorter four-part version first, and turn to this one when the diagnosis is less clear-cut or the presentation is unusual. It adds points for things that support the diagnosis, such as certain antibodies, raised immune proteins, characteristic findings on a liver biopsy, and other autoimmune conditions in the family. Just as importantly, it subtracts points for things that suggest a different cause — an antibody associated with a different liver condition, a medication that could be responsible, evidence of a viral infection, or heavy alcohol use. That is why it can pick up cases the shorter version misses. The threshold for a confident diagnosis is higher after treatment has started than before, because the response to treatment itself adds points.
Frequently asked questions
What is the revised original AIH score?#
The 1999 International Autoimmune Hepatitis Group scoring system: thirteen parameters plus optional items covering sex, biochemistry, autoantibodies, viral markers, drug and alcohol history, histology, other autoimmune disease, HLA type and response to therapy.
What score confirms autoimmune hepatitis?#
It depends on timing. Before treatment, above 15 is definite and 10–15 is probable. After treatment, above 17 is definite and 12–17 is probable. A total of 17 is therefore definite before treatment and only probable after it.
Why are the post-treatment thresholds higher?#
Because the response-to-therapy items can only add points — a complete response scores +2 and a relapse +3. Raising the thresholds by two keeps the diagnostic categories comparable before and after treatment.
Why does relapse score more than a complete response?#
Because it is stronger evidence. A patient who responds and remains well might have had a self-limiting process; a patient who relapses when treatment is withdrawn has demonstrated a treatment-dependent autoimmune disease.
How are the histology items scored?#
Additively, not as alternatives. Interface hepatitis scores +3, a predominantly lymphoplasmacytic infiltrate +1 and rosetting of liver cells +1, so a biopsy showing all three contributes +5. If all three are absent, that scores −5 instead.
My patient has a positive AMA — does that exclude autoimmune hepatitis?#
Not automatically, though it costs 4 points and points towards primary biliary cholangitis. A low-titre AMA can occur in autoimmune hepatitis, and overlap syndromes exist. Treat the penalty as a strong prompt to reconsider rather than as an exclusion.
When should I use this instead of the simplified criteria?#
When the simplified criteria fall below 6 but suspicion persists, in atypical or seronegative presentations, in suspected overlap syndromes, and in research or case-definition settings. The simplified criteria are the right first choice in routine practice.
Can the score be negative?#
Yes. Over twenty points of penalties are available, so a patient with a positive AMA, positive viral markers and a positive drug history begins 11 points down before any positive finding is counted.
References
Original / primary reference
Related scoring and guidelines
- Hennes EM, Zeniya M, Czaja AJ, et al. Simplified criteria for the diagnosis of autoimmune hepatitis. Hepatology. 2008;48(1):169-176.
- Mack CL, Adams D, Assis DN, et al. Diagnosis and Management of Autoimmune Hepatitis in Adults and Children: 2019 Practice Guidance and Guidelines from the American Association for the Study of Liver Diseases. Hepatology. 2020;72(2):671-722.
- European Association for the Study of the Liver. EASL Clinical Practice Guidelines: Autoimmune hepatitis. J Hepatol. 2015;63(4):971-1004.