About the AST to Platelet Ratio Index (APRI)
APRI estimates liver fibrosis from just two values — AST expressed as a multiple of your laboratory's upper limit of normal, and platelet count. Below 0.5, significant fibrosis is unlikely; above 1.5, it is likely; between them the result is indeterminate. A separate, higher threshold of about 2.0 is used for cirrhosis. In the original chronic hepatitis C cohort the AUROC was 0.80 for significant fibrosis and 0.92 for cirrhosis in the training set, and 0.88 and 0.94 respectively in validation.
Formula
APRI = [ (AST ÷ AST upper limit of normal) ÷ platelets (×10⁹/L) ] × 100- AST
- U/L, as measured.
- AST upper limit of normal
- Your laboratory's reference ceiling in U/L. Laboratory-specific, and sex-specific in many labs.
- platelets
- ×10⁹/L, identical to ×1,000/µL — no conversion between those two.
- The × 100 is a scaling convention so the result lands in a convenient numeric range; it has no clinical meaning.
- Because AST is expressed relative to a local reference range, the same patient can get slightly different APRI values at two hospitals. This is inherent to the score, not an error.
- There are no floors or caps — every value enters as measured.
Interpreting the result
Read the low and high bands as rule-out and rule-in, and the middle as an instruction to do something else. Two different threshold sets are in circulation, and the distinction matters: roughly 0.5 and 1.5 are used for significant fibrosis, while about 2.0 is used when the question is cirrhosis specifically. Because APRI shares two of its inputs with FIB-4, it is best read alongside it — concordant results are considerably more useful than either in isolation, and discordant results usually mean a non-hepatic cause is distorting the AST or the platelet count.
| Score | Band | What it means | Action |
|---|---|---|---|
| < 0.5 | Low probability | Significant fibrosis unlikely | Routine monitoring; recheck alongside FIB-4 over time |
| 0.5–1.5 | Indeterminate | The index does not discriminate in this range | Combine with FIB-4 and proceed to elastography or biopsy where available |
| > 1.5 | High probability | Significant fibrosis likely; cirrhosis becomes likely above about 2.0 | Refer to hepatology; confirm with elastography or biopsy |
Scroll the table sideways for every column.
What the APRI needs (3 inputs)
- AST (U/L)
- Aspartate aminotransferase. The absolute value matters less than its ratio to your laboratory's ceiling, which is why the next field exists.
- AST upper limit of normal (U/L)
- Your own laboratory's reference ceiling for AST. This is not a constant — it differs between laboratories and between sexes, commonly falling between about 30 and 40 U/L. Using the wrong figure rescales every result, and it is the single most common way APRI is calculated incorrectly.
- Platelet count (×10⁹/L)
- In the denominator, standing in for portal hypertension. As with FIB-4, any non-hepatic cause of thrombocytopenia inflates the score.
Units. AST and its upper limit of normal must be in the same units, which in practice is U/L for both. Platelets in ×10⁹/L and ×1,000/µL are the same number, so no conversion is needed. The one value you must look up rather than assume is your own laboratory's AST reference ceiling, which is commonly between 30 and 40 U/L and often differs between men and women.
What it returns
- APRI score
- A continuous value. Conventionally reported to two or three decimal places, since the cut-offs are close together at the low end.
- Risk band
- Low, indeterminate or high probability of significant fibrosis. A separate higher threshold is used when the question is specifically cirrhosis rather than significant fibrosis.
How it is calculated
APRI is the simplest of the fibrosis indices and its logic is transparent: as fibrosis progresses, hepatocellular injury raises AST while developing portal hypertension lowers the platelet count, so a ratio of the two moves in one direction. Wai and colleagues fitted it in patients with chronic hepatitis C who had undergone biopsy, and expressed AST as a multiple of the upper limit of normal rather than in absolute units specifically so the index would transfer between laboratories with different assays. That design choice is also its main practical hazard, because it makes the result depend on a number the user has to look up and frequently guesses.
Facts & figures
| Outcome | Training set (n = 192) | Validation set (n = 78) |
|---|---|---|
| AUROC — significant fibrosis | 0.80 | 0.88 |
| AUROC — cirrhosis | 0.92 | 0.94 |
Scroll the table sideways for every column.
In the training cohort, using optimised cut-offs allowed significant fibrosis to be predicted accurately in 51% of patients and cirrhosis in 81%.
| Question | Rule-out below | Rule-in above |
|---|---|---|
| Significant fibrosis (METAVIR F2+) | 0.5 | 1.5 |
| Cirrhosis (F4) | 1.0 | 2.0 |
Scroll the table sideways for every column.
Quoted thresholds vary between publications and guidelines. State which pair you used.
Evidence
Derivation — chronic hepatitis C, biopsy-controlled
2003 · n = 270Patients with chronic hepatitis C who had undergone liver biopsy, split into a training cohort of 192 and a validation cohort of 78. AST expressed as a multiple of the local upper limit of normal, divided by platelet count.
AUROC 0.80 for significant fibrosis and 0.92 for cirrhosis in the training set; 0.88 and 0.94 respectively in validation. With optimised cut-offs, significant fibrosis was predicted accurately in 51% and cirrhosis in 81% of the training cohort.
WHO adoption for hepatitis B in resource-limited settings
2015APRI has been recommended in World Health Organization hepatitis B guidance as a means of assessing fibrosis where transient elastography and biopsy are unavailable, on the basis that two routine laboratory values are obtainable almost anywhere.
Accepted as adequate for identifying patients who need treatment in settings without access to better tests, rather than as equivalent to elastography.
Performance outside chronic hepatitis C
2023Applied to chronic hepatitis B, HIV/HBV co-infection and metabolic liver disease, where reported accuracy is more variable than in the derivation setting, and where age-specific cut-offs have been proposed for chronic hepatitis B.
Comparative studies against transient elastography generally find APRI and FIB-4 similar to each other and both inferior to elastography.
How it compares
APRI vs FIB-4
FIB-4 is the usual first-line choice in guideline pathways because it needs no local reference range — but APRI is a worthwhile cross-check, and it was the better performer for cirrhosis specifically in its derivation cohort.
Both indices combine AST and platelets. FIB-4 adds age and ALT and is specified in most MASLD and primary-care pathways, partly because it can be computed retrospectively from any historic blood panel with no extra lookup. APRI requires your laboratory's AST upper limit of normal, which makes automated calculation across a health system harder and hand calculation more error-prone. In the Wai cohort APRI reached an AUROC of 0.92–0.94 for cirrhosis, better than FIB-4's 0.765 for advanced fibrosis — though the two figures answer different questions in different populations and should not be compared directly.
APRI vs Transient elastography (FibroScan)
Elastography is more accurate and is preferred wherever it exists; APRI's role is in settings where it does not, which is why WHO guidance names APRI rather than elastography for hepatitis B assessment in resource-limited care.
Liver stiffness measurement outperforms both APRI and FIB-4 in head-to-head comparisons. It requires equipment, a trained operator and a fasted patient, and is confounded by obesity and ascites. APRI needs an AST, a platelet count and a reference range — available essentially everywhere — which is a different kind of advantage and the reason it remains in international guidance.
Pearls & pitfalls
- The AST upper limit of normal is not 40 U/L by default. It differs by laboratory and often by sex, commonly between 30 and 40, and using the wrong value rescales every result proportionally.
- APRI and FIB-4 share AST and platelets, so they are not independent tests. Agreement is reassuring but not as strong as two genuinely independent methods agreeing.
- Two threshold sets are in circulation — roughly 0.5/1.5 for significant fibrosis and 1.0/2.0 for cirrhosis. Quoting a band without saying which question you asked is ambiguous.
- Non-hepatic thrombocytopenia inflates the score exactly as it does with FIB-4. Check the platelet count has a hepatic explanation.
- Acute hepatitis, recent alcohol and muscle injury all raise AST without fibrosis, and the index cannot distinguish those causes.
- Platelets in ×10⁹/L and ×1,000/µL are the same number. Converting between them introduces hundred-fold errors.
- Accuracy for cirrhosis was better than for significant fibrosis in the derivation study, which is the opposite of the intuition that harder questions are answered worse.
Critical actions
- Look up your own laboratory's AST upper limit of normal rather than assuming a value — this is the difference between a correct and a misleading result.
- Calculate FIB-4 alongside it and act on the more concerning of the two, investigating any discordance.
- Treat an indeterminate result as an instruction to escalate, not as a normal finding.
- Exclude non-hepatic causes of a low platelet count or a raised AST before referring on a high score.
- Do not use it during acute hepatitis or soon after heavy alcohol intake.
- Where elastography is available, use APRI to triage rather than to conclude.
Why this score exists
The most deliberate decision in APRI's design is the one users most often undo. Wai and colleagues expressed AST as a multiple of the upper limit of normal, rather than in absolute U/L, so that the index would be comparable across laboratories running different assays with different reference ranges. That is why the calculator asks for a value most clinicians have to look up — and why entering a remembered 40 U/L when the local ceiling is 31 quietly rescales the result by about 30%. The score was also built in biopsy-controlled chronic hepatitis C, at a time when biopsy was routine and the aim was to reduce how often it was needed; it was never presented as a staging tool.
About the creator
First author, 2003 derivation study
Derived APRI in chronic hepatitis C, combining AST with platelet count to estimate fibrosis without biopsy.
Senior author
Led the Michigan hepatology programme in which APRI was developed and validated.
Limitations
- Derived in biopsy-controlled chronic hepatitis C; performance in metabolic liver disease and hepatitis B is more variable, and age-specific cut-offs have been proposed for chronic hepatitis B.
- Depends on a laboratory-specific AST upper limit of normal, so results are not strictly comparable between services and are easy to compute wrongly.
- Only two variables, so it carries less information than indices that include age or other markers.
- Cannot stage fibrosis — it estimates the probability of crossing a threshold.
- Distorted by any non-hepatic cause of raised AST or low platelets, with no way for the score to signal that.
- Competing threshold sets for significant fibrosis and for cirrhosis create real ambiguity in how results are reported.
If you are the patient
APRI is a simple calculation from two things in a routine blood test: a liver enzyme called AST, and your platelet count. As liver scarring develops, AST tends to rise and platelets tend to fall, so the ratio between them gives an estimate of how likely significant scarring is. A low result makes significant scarring unlikely; a high result makes it likely and means you will be referred to a liver specialist; a middle result means the test could not tell and another test is needed. It is often calculated together with a similar score called FIB-4, and your team pays most attention when the two agree. One caveat worth knowing: a low platelet count from a cause unrelated to your liver — some medicines, or a blood condition — will push this score up without your liver being the problem, so your team will check for that before drawing conclusions.
Frequently asked questions
What is the APRI score?#
APRI is the AST to Platelet Ratio Index, a non-invasive estimate of liver fibrosis calculated as [(AST ÷ AST upper limit of normal) ÷ platelet count] × 100. Below 0.5 significant fibrosis is unlikely, above 1.5 it is likely, and about 2.0 is used as the threshold for cirrhosis.
How is the APRI score calculated?#
Divide the AST by your laboratory's upper limit of normal for AST, divide that result by the platelet count in ×10⁹/L, then multiply by 100. Expressing AST relative to the local reference range is deliberate, so the index transfers between laboratories with different assays.
What AST upper limit of normal should I use?#
Your own laboratory's, which is typically between 30 and 40 U/L and often differs by sex. It is not a fixed constant, and substituting a remembered value for the local one rescales the whole result — this is the commonest error in calculating APRI.
What APRI score indicates cirrhosis?#
Around 2.0 is the conventional rule-in threshold for cirrhosis, with about 1.0 used as the rule-out. These are higher than the 0.5 and 1.5 thresholds used for significant fibrosis, so it matters which question you are asking. In the derivation study APRI reached an AUROC of 0.92–0.94 for cirrhosis.
Is APRI or FIB-4 better?#
They perform similarly and are often used together. FIB-4 is specified more often in guideline pathways because it needs no local reference range and can be computed automatically from existing blood results. APRI is simpler and performed strongly for cirrhosis in its derivation cohort. Where the two disagree, look for a non-hepatic cause affecting AST or platelets.
How accurate is the APRI score?#
In the original chronic hepatitis C cohort the AUROC was 0.80 for significant fibrosis and 0.92 for cirrhosis in the training set, and 0.88 and 0.94 in validation. Accuracy is more variable outside chronic hepatitis C, and it is consistently less accurate than transient elastography.
Can APRI be used in hepatitis B?#
Yes, and World Health Organization guidance has recommended it for assessing fibrosis in hepatitis B where transient elastography and biopsy are not available. Reported performance in hepatitis B is more variable than in hepatitis C, and age-specific cut-offs have been proposed.
Why does a low platelet count raise the APRI score?#
Platelets are in the denominator, standing in for portal hypertension — as cirrhosis develops, splenic sequestration and reduced thrombopoietin lower the count. That is the intended signal, but any other cause of thrombocytopenia produces the same arithmetic effect without any liver fibrosis, which is why the cause of a low count should be established before acting on a high APRI.