About the GastroAGI MASLD–MASH NIT Calculator
The GastroAGI MASLD–MASH NIT Calculator runs a whole metabolic liver disease work-up from one set of inputs. Age, sex, diabetes status, AST, ALT, platelets and a FibroScan reading give FIB-4, APRI, FAST and Agile 3+ outright; PRO-C3 adds ADAPT, and an ELF value obtained from the laboratory is banded against its published thresholds. Each score is reported with its own risk band and a one-line interpretation, and the calculator then states whether the first-line blood test and elastography actually agree — the judgement that running the scores on separate pages never produces.
Formula
FIB-4 = (Age × AST) ÷ (Platelets × √ALT)
APRI = (AST ÷ 40) ÷ Platelets × 100
NFS = −1.675 + 0.037·Age + 0.094·BMI + 1.13·Diabetes + 0.99·(AST/ALT) − 0.013·Platelets − 0.66·Albumin [not calculated here — BMI and albumin are not collected]
FAST = eˣ / (1 + eˣ), x = −1.65 + 1.07·ln(LSM) + 2.66×10⁻⁸·CAP³ − 63.3·AST⁻¹
Agile 3+ = eˣ / (1 + eˣ), x = −3.92368 + 2.29714·ln(LSM) − 0.00902·Platelets − 0.98633·(ALT/AST) + 1.08636·Diabetes − 0.38581·Sex + 0.03018·Age
ADAPT = exp[log₁₀((Age × PRO-C3) ÷ √Platelets)] + Diabetes- Age
- Years. In the FIB-4 numerator, the NAFLD Fibrosis Score, Agile 3+ and ADAPT.
- AST, ALT
- U/L. AST appears in four of the six expressions; the AST/ALT ratio is itself a term in the NAFLD Fibrosis Score, and its inverse in Agile 3+.
- Platelets
- ×10⁹/L, numerically identical to ×1,000/µL. In the denominator of FIB-4, APRI and ADAPT, and a negative term in the NAFLD Fibrosis Score and Agile 3+.
- BMI, Albumin
- kg/m² and g/dL. Terms of the NAFLD Fibrosis Score only. Neither is collected by this form, which is why NFS is shown as a dash rather than a number.
- LSM, CAP
- kPa and dB/m from vibration-controlled transient elastography. LSM enters FAST and Agile 3+ as a natural log, CAP as a cube.
- PRO-C3
- ng/mL. The N-terminal pro-peptide of type III collagen, and the only input ADAPT needs beyond age, diabetes and the platelet count.
- Sex, Diabetes
- Both binary, coded male = 1 and diabetes present = 1. Agile 3+ is the only score here that uses sex.
- Every expression is transcribed from the same publication as the standalone calculator on this site, so the composite and the individual page can never disagree.
- The NAFLD Fibrosis Score is printed above but not evaluated here. Its BMI and albumin terms have no input on this form, and running it with them treated as zero would return a figure roughly 3.6 lower than the truth at an average build — comfortably inside the band that reads as "advanced fibrosis ruled out". Use the standalone NAFLD Fibrosis Score page, which asks for both.
- APRI is reported against a fixed AST upper limit of 40 U/L, the conventional adult ceiling and the one its derivation used. A laboratory whose ceiling differs will read a different APRI from the same AST, so the value used is stated with the result.
- FIB-4's lower cut-off is 1.3 in MASLD, not the 1.45 from its original hepatitis C derivation, and AASLD raises it to 2.0 above age 65. Both are applied automatically.
- Agile 3+ takes the AST/ALT ratio *inverted* — what enters the model is ALT/AST. Getting that the wrong way round changes the answer without looking wrong, so it is pinned by a regression test against a published worked example.
- In ADAPT the diabetes term is added *outside* the exponential, not inside it. The source publications set the equation in a maths font that does not survive text extraction, and reading the term as inside would roughly double the score; the placement here is confirmed by the delimiter positions in two source PDFs and by a third paper that prints the formula as plain text, and it is the reading under which the published 6.3287 cut-off corresponds to a clean ratio of 70.
- The composite performs no meta-analysis across scores. Each is computed independently and shown with its own band; the overall verdict reports agreement, it does not average.
Interpreting the result
Read the headline FIB-4 first, then the concordance statement, then the pathway. A FIB-4 below its cut-off with no other abnormal test is a genuine low-risk result: the work is metabolic, and the liver needs only interval retesting. A FIB-4 between the cut-offs is the commonest result and the least informative — it is precisely what the secondary test exists to resolve, and treating it as negative is the characteristic error. Concordantly abnormal tests need no third test to act on. The result that most often goes wrong is discordance: a raised FIB-4 with a low stiffness, or the reverse. Before treating discordance as a diagnostic puzzle, check the mechanics — an unreliable elastography examination, a non-fasted patient, a hepatitis flare, cardiac congestion or cholestasis all raise stiffness independently of fibrosis, and FIB-4 is inflated by age and by any cause of thrombocytopenia unrelated to liver disease. Genuine discordance that survives those checks is what MR elastography, MRI-PDFF, ELF and, occasionally, biopsy are for.
| Score | Band | What it means | Action |
|---|---|---|---|
| FIB-4 < 1.3 (< 2.0 if over 65) and LSM < 8 kPa | Low risk — concordant | Advanced fibrosis unlikely; both first-line and elastography rule it out | Manage cardiometabolic risk. Repeat FIB-4 every 1–2 years with diabetes or ≥ 2 risk factors, otherwise every 2–3 years |
| FIB-4 1.3–2.67, or LSM 8–12 kPa | Indeterminate | Neither ruled in nor ruled out; roughly a third of patients land here | Secondary test — VCTE or ELF after an indeterminate FIB-4; MR elastography, MRI-PDFF or ELF if both are already indeterminate |
| One test rules out while another rules in | Discordant | The disagreement is the finding; the underlying stage is genuinely uncertain | Check measurement validity first, then add a third modality and refer. Consider biopsy where the answer would change management |
| FIB-4 ≥ 2.67 or LSM > 12 kPa, nothing contradicting | High risk of advanced fibrosis | Advanced fibrosis likely | Refer to hepatology; complete the secondary test if not yet done |
| FIB-4 ≥ 2.67 and LSM > 12 kPa | High risk of advanced fibrosis — concordant | Concordant non-invasive evidence of advanced fibrosis; add cirrhosis risk if LSM ≥ 20 kPa | Hepatology referral, six-monthly HCC surveillance where cACLD is established, and screening endoscopy if LSM ≥ 20 kPa or platelets ≤ 150 ×10⁹/L |
Scroll the table sideways for every column.
What the MASLD–MASH NIT needs (6 inputs)
- Age and sex
- Age drives FIB-4, Agile 3+ and ADAPT, and moves the FIB-4 lower cut-off from 1.3 to 2.0 above 65. Sex is used by Agile 3+, which was derived with it as a binary term.
- Type 2 diabetes
- A term in both Agile 3+ and ADAPT, where it adds a flat point to the score. Diagnosed type 2 diabetes, or treatment for it.
- AST, ALT and platelet count
- The three required laboratory values. Together with age they are FIB-4; they also drive APRI, the inverted AST/ALT ratio in Agile 3+ and the platelet term in ADAPT, and AST is the third term of FAST. A normal ALT does not exclude advanced fibrosis — transaminases commonly fall as fibrosis progresses.
- Liver stiffness (LSM) and CAP
- Vibration-controlled transient elastography from a single fasted FibroScan examination. Stiffness completes Agile 3+, and stiffness with CAP and AST gives FAST. Stiffness is also read against the 8 and 12 kPa bands for the concordance verdict.
- MRI-PDFF and MR elastography (optional)
- Entered from an MRI already performed. MRI-PDFF is read against the 5% and 10% steatosis thresholds; MR elastography is combined with FIB-4 to give the MEFIB index. Both are reported as supporting measurements rather than as scores.
- ELF and PRO-C3 (optional)
- PRO-C3 in ng/mL computes ADAPT, using age, diabetes and the platelet count already entered. ELF is entered as a value obtained from the laboratory and interpreted against its published thresholds; its algorithm is proprietary and is not computed here.
Units. Every field takes one unit, the one its formula was derived in, and it is printed in the label. AST and ALT are in U/L worldwide, platelets are ×10⁹/L (numerically identical to ×1,000/µL), PRO-C3 is in ng/mL, liver stiffness and MR elastography are in kPa, CAP is in dB/m, and MRI-PDFF is a percentage. None has an alternative reporting convention that would change the number entered.
What it returns
- FIB-4 (headline)
- The first-line result, shown as the headline because that is the order the guidance puts it in. Banded against the MASLD cut-offs of 1.3 and 2.67, with 2.0 replacing 1.3 above age 65.
- Eight scores
- FIB-4, APRI, NFS, FAST, Agile 3+, Agile 4, ELF and ADAPT, each with its own value, risk band and one-line interpretation. FIB-4, APRI, FAST and Agile 3+ are always reported; ELF and ADAPT appear when their optional input is supplied and are absent otherwise, so nothing on screen implies a test was done that was not. NFS and Agile 4 report as a dash with the reason — see the limitations.
- Overall verdict
- Whether FIB-4 and liver stiffness are concordantly low, concordantly high, discordant, or indeterminate. This is the judgement that running the scores separately does not produce.
- Supporting measurements
- MRI-PDFF with its steatosis read, and the MEFIB index from MR elastography and FIB-4, shown only when those values were entered.
How it is calculated
The calculator is a sequence rather than a single model. Four scores come straight from the required sections: FIB-4 and APRI from age and routine bloods, FAST from liver stiffness with CAP and AST, and Agile 3+ from stiffness together with age, sex, diabetes and the blood values already entered. FIB-4 is reported first and as the headline, because that is where every current guideline starts. The optional sections then add the rest: PRO-C3 activates ADAPT, an entered ELF value is banded against its own published thresholds, and MRI-PDFF and MR elastography are reported as supporting measurements, the latter combined with FIB-4 as the MEFIB index. Two of the eight, the NAFLD Fibrosis Score and Agile 4, are shown as a dash carrying the reason they could not be produced rather than as a number. Finally the calculator compares the bands: concordantly high when FIB-4 and elastography are both in their rule-in bands, concordantly low when both rule out, and discordant when one rules out while the other rules in.
Facts & figures
| Test | Rule-out | Rule-in | Source |
|---|---|---|---|
| FIB-4 | < 1.3 (< 2.0 if age > 65) | ≥ 2.67 | AASLD 2023 guidance |
| VCTE liver stiffness | < 8 kPa | > 12 kPa (≥ 20 kPa cirrhosis) | AASLD 2023 guidance |
| ELF | < 7.7 | ≥ 9.8 (≥ 11.3 cirrhosis) | AASLD 2023 guidance |
| FAST | ≤ 0.35 | ≥ 0.67 | Newsome 2020 |
| Agile 3+ | < 0.451 | ≥ 0.679 | Sanyal 2023 |
| Agile 4 (not calculated here) | < 0.251 | ≥ 0.565 | Sanyal 2023 |
| NFS (not calculated here) | < −1.455 | > 0.676 | Angulo 2007 |
| APRI | < 0.5 | > 1.5 | Wai 2003 |
| ADAPT | ≤ 6.3287 | > 6.3287 | Daniels 2019 |
| MEFIB (MRE + FIB-4) | MRE < 3.3 and FIB-4 < 1.6 | MRE ≥ 3.3 and FIB-4 ≥ 1.6 | Jung 2021 |
| MRI-PDFF | < 5% | ≥ 10% | Conventional thresholds |
Scroll the table sideways for every column.
ADAPT has a single Youden-derived cut-off rather than a rule-out and a rule-in pair: at 6.3287 the negative predictive value is 96.6% but the positive predictive value only 48.4%, so a high ADAPT is a reason to confirm rather than to act.
| Score | Inputs beyond the required sections 1 to 4 |
|---|---|
| FIB-4 | None — always reported |
| APRI | None — always reported, against a 40 U/L AST ceiling |
| FAST | None — always reported |
| Agile 3+ | None — always reported |
| ADAPT | PRO-C3 |
| ELF | An ELF value from the laboratory |
| NFS | Not available — needs BMI and albumin, which this form does not collect |
| Agile 4 | Not available — its published equation cannot be transcribed |
| MEFIB | MR elastography stiffness |
| MRI-PDFF | An MRI-PDFF percentage |
Scroll the table sideways for every column.
How it compares
MASLD–MASH NIT vs FIB-4 alone
FIB-4 alone is the correct first-line test and this calculator computes it identically — the difference is that the composite also tells you what to do with the answer, and applies the age-65 cut-off shift automatically.
Use the standalone FIB-4 page when FIB-4 is all you have or all you need. Use the composite when you have, or will have, more than four values: it adds the secondary test, the steatosis grade, the at-risk MASH probability and the portal hypertension assessment, and reports whether they agree. On the same inputs the two produce the same FIB-4 to two decimal places, because they share one transcription of the formula.
MASLD–MASH NIT vs FAST score
FAST answers one question — is this at-risk MASH — and answers it well; the composite includes FAST unchanged and places it alongside the fibrosis and portal hypertension questions that FAST does not address.
FAST is the right tool when a FibroScan has been done and the question is specifically whether the active fibrotic phenotype is present, for treatment or trial eligibility. It says nothing about cirrhosis or portal hypertension, and its grey zone swallows roughly a third of patients. The composite reports FAST with the same thresholds, and when FAST is indeterminate it can still return a usable fibrosis answer from stiffness, FIB-4 and the blood-based scores.
MASLD–MASH NIT vs NAFLD Fibrosis Score
The NAFLD Fibrosis Score is corroboration, not a first-line test — it needs albumin and BMI that FIB-4 does not, and its indeterminate zone is wider.
Both estimate advanced fibrosis from blood values. FIB-4 is preferred first because it needs only four routine values and is what current guidance specifies. The composite does not compute the NAFLD Fibrosis Score at all: its form collects neither BMI nor albumin, so use the standalone page when you want it. Running the two side by side is worth doing when they might disagree — a discordant pair of blood-based scores is an argument for elastography rather than for picking the more convenient one.
MASLD–MASH NIT vs Liver biopsy
Non-invasive testing has replaced biopsy for most staging decisions; biopsy remains useful when the tests are discordant or indeterminate and the answer would change management.
A concordant pair of non-invasive tests at the extremes is sufficient to act on, and that covers most patients. Biopsy retains a role where the composite reports discordance that survives a check of measurement validity, where a competing or additional diagnosis is suspected, and in trial settings where histology is the endpoint. This calculator is deliberately worded as probability of a stage rather than presence of a stage, because non-invasive tests and histology are not interchangeable in every situation.
Pearls & pitfalls
- FIB-4's MASLD cut-off is 1.3, not the 1.45 from its original hepatitis C derivation. Above age 65 it is 2.0 — applying 1.3 there produces mostly false positives, and this calculator switches automatically.
- Below age 35 FIB-4 has poor accuracy in either direction. A low result in a young patient with high metabolic risk or raised liver chemistries should not stop a secondary test.
- A normal ALT does not exclude advanced fibrosis. Transaminases often fall as fibrosis progresses, so the most advanced disease can have the least remarkable liver chemistry.
- Liver stiffness is raised by inflammation, cholestasis, cardiac congestion and a recent meal independently of fibrosis. An unfasted or acutely unwell patient can read several kPa high.
- An IQR/median ratio above 30% with a median of 7.1 kPa or more marks an unreliable FibroScan examination. This calculator does not see that ratio, so check it on the device report before entering a stiffness value.
- CAP is not stage-specific and shifts with the population — roughly 10 dB/m higher in MASLD and in diabetes, and about 4.4 dB/m per BMI unit. Here it is used only as the second term of FAST.
- APRI is reported against a fixed AST ceiling of 40 U/L. Where your laboratory's upper limit differs, the APRI shown here differs from one calculated against your own reference range, and FIB-4 — which needs no reference range at all — is the more portable of the two.
- Thrombocytopenia from any non-hepatic cause — hypersplenism, marrow disease, drugs — inflates FIB-4, APRI and ADAPT alike, since all three divide by the platelet count.
- Concordance is worth more than any single score. Two agreeing tests justify acting; two disagreeing tests justify a third, not a coin toss.
Critical actions
- Act on the pathway, not the headline number: an indeterminate FIB-4 is an instruction to test again, not a reassuring result.
- Where liver stiffness is 20 kPa or above, or platelets are 150 ×10⁹/L or below, arrange screening endoscopy and assume clinically significant portal hypertension until excluded.
- Where compensated advanced chronic liver disease is established, start six-monthly HCC surveillance ultrasound.
- Treat the cardiometabolic drivers at every risk level. Cardiovascular disease, not liver disease, remains the leading cause of death in MASLD, and fibrosis can regress when the driver is removed.
- Re-check the mechanics before working up discordance — fasting status, examination reliability, intercurrent illness and non-hepatic causes of thrombocytopenia explain a large share of it.
- Repeat non-invasive testing on a schedule rather than treating a single low result as a discharge.
Why this score exists
This composite is GastroAGI's own assembly, not a published score, and it deliberately claims nothing beyond the scores it contains. Each component was derived and validated by its own authors, cited below; what is added here is the sequence in which the guidance asks for them and the comparison between their results. No new coefficient, weighting or combined index has been invented — a composite that produced its own number would be an unvalidated score wearing the authority of eight validated ones.
Limitations
- This is a composite of published scores, not itself a validated score. Each component carries its own derivation and validation; the assembly does not, and no combined index or averaged probability is produced.
- Agile 4 is not computed, though Agile 3+ is. The liver-stiffness term of the Agile 4 equation is mangled in every obtainable rendering of its publication — the LSM variable itself is dropped by the typesetter — and the only reconstruction that fits dimensionally returns 0.97 on a worked example where the vendor's own tool prints 0.69. Rather than ship a cirrhosis probability that looks plausible and is wrong, it is shown as a dash. Where an Agile 4 value is available it comes from the FibroScan device itself, and its published cut-offs are 0.251 for rule-out and 0.565 for rule-in.
- Agile 3+ uses sex as a binary variable, coded male or female, because that is how it was derived. It has no coding for anyone outside that, and the score cannot be calculated more precisely than its own derivation allows.
- ELF is entered rather than calculated. The algorithm is proprietary; only its published thresholds are used here.
- The NAFLD Fibrosis Score is listed among the outputs but is not calculated here. Its published equation contains a BMI term and an albumin term, this form collects neither, and there is no defensible substitute for either — BMI cannot be inferred from age and sex, and albumin cannot be inferred at all. The standalone NAFLD Fibrosis Score page asks for both and computes it in full.
- Every component was derived in a specific population, and none was derived in this combination. Diagnostic accuracy figures quoted for each score apply to that score, not to the panel.
- The scores are not independent of one another. FIB-4, APRI and the NAFLD Fibrosis Score all share AST and the platelet count, and Agile 3+ shares liver stiffness with FAST, so agreement within those groups is weaker evidence than agreement between a blood-based score and elastography.
- None of this applies to acute hepatitis, decompensated cirrhosis, or non-metabolic liver disease, where the cut-offs shift or the scores lose meaning entirely.
- This calculator stages liver disease; it does not diagnose MASLD. The diagnosis requires steatosis plus at least one cardiometabolic risk factor, and alcohol intake and other causes must be assessed separately.
- ADAPT was derived and validated in NAFLD cohorts using the Nordic Bioscience PRO-C3 ELISA. A PRO-C3 result from a different assay may not be interchangeable with the value the published cut-off was set against.
- Nothing here replaces clinical assessment, and a result at any risk level should be read alongside the history, the examination and the alcohol history in particular.
If you are the patient
This tool estimates how much scarring, fat and inflammation there might be in the liver of someone with fatty liver disease linked to weight, diabetes, blood pressure or cholesterol — without a biopsy. It uses routine blood tests, and a scan called a FibroScan where one has been done. Most people who are tested turn out to have little or no scarring, and for them the important work is on the metabolic side: weight, blood sugar, blood pressure and cholesterol. A result in the middle range is common and does not mean something has been found — it means one test could not settle the question and another is needed. A high result means scarring is likely and a liver specialist should be involved. The tests measure probability rather than certainty, which is why two of them are often used together, and why a specialist may want a further scan when they disagree.
Frequently asked questions
What is a MASLD or MASH NIT calculator?#
A non-invasive test (NIT) calculator estimates how much liver disease is present without a biopsy. This one is composite: from twelve inputs it reports FIB-4, APRI and FAST, plus Agile 3+ from the FibroScan reading and ADAPT where PRO-C3 is available, each with its own risk band — then states whether the first-line blood test and the FibroScan reading actually agree, which is the judgement running the scores on separate pages never produces.
Which test should be done first in MASLD?#
FIB-4. It needs only age, AST, ALT and the platelet count, all of which a routine blood test already produces, and current AASLD guidance uses it to define who needs anything further. Below 1.3 — or below 2.0 above age 65 — the risk of advanced fibrosis is low and no secondary test is required on that result alone.
What should I do with an indeterminate FIB-4?#
Arrange a secondary test: vibration-controlled transient elastography (FibroScan) or the ELF blood panel. An indeterminate FIB-4, between 1.3 and 2.67, is the single commonest result and is not a reassuring one — it is exactly the situation the secondary test exists to resolve. Treating it as negative is the most frequent error in the pathway.
What if FIB-4 and FibroScan disagree?#
Check the mechanics first. Liver stiffness is raised by a recent meal, inflammation, cholestasis and cardiac congestion independently of fibrosis, and an examination whose IQR/median ratio exceeds 30% is unreliable; FIB-4 is inflated by age and by any non-hepatic cause of a low platelet count. If the discordance survives those checks it is a real finding, and the next step is a third modality — MR elastography, MRI-PDFF or ELF, all of which this calculator accepts — with specialist assessment, and biopsy where the answer would change management.
What is at-risk MASH, and which test identifies it?#
At-risk MASH is steatohepatitis with a NAFLD activity score of at least 4 and fibrosis of at least F2 — the active, fibrotic phenotype that treatment and clinical trials target. FAST identifies it, combining liver stiffness, CAP and AST from a single FibroScan visit. At or below 0.35 it is effectively ruled out; at or above 0.67 it is ruled in. Fibrosis-only scores such as FIB-4 cannot answer this question, because they do not measure activity.
Does this calculator compute Agile 3+ and Agile 4?#
Agile 3+ yes, Agile 4 no. Agile 3+ is calculated automatically from values the form already collects — age, sex, diabetes, AST, ALT, platelets and liver stiffness — so it costs no extra input, and it is banded against the published thresholds of 0.451 and 0.679. Agile 4 is not calculated: the liver-stiffness term of its equation is illegible in every available rendering of the source paper, and the only sensible reconstruction disagrees with a published worked example. Rather than print a plausible-looking wrong cirrhosis probability, this calculator shows it as a dash. An Agile 4 value, where one is needed, is produced by the FibroScan device itself and read against cut-offs of 0.251 and 0.565.
What is a good Agile 3+ score?#
Below 0.451 is the rule-out result — advanced fibrosis (F≥3) is unlikely, at a threshold chosen for at least 85% sensitivity. At or above 0.679 is rule-in, chosen for at least 90% specificity, and advanced fibrosis is likely. Between the two it is indeterminate, though that zone is deliberately narrower than the one FIB-4 or liver stiffness leaves on its own — narrowing it is the reason Agile 3+ exists.
What is the ADAPT score and what does it need?#
ADAPT is a fibrosis score built around PRO-C3, a serum marker of type III collagen formation, combined with age, the platelet count and diabetes status — the name is an acronym of its four inputs. It is calculated here as exp[log₁₀((age × PRO-C3) ÷ √platelets)] + diabetes, with diabetes counting as one point. Above the published cut-off of 6.3287 advanced fibrosis is likely, though the positive predictive value at that threshold is only 48%; at or below it the negative predictive value is 97%. The only value you need to supply is PRO-C3 in ng/mL — everything else it uses is already on the form.
How often should non-invasive testing be repeated?#
In a patient at low risk on first-line testing, commonly every one to two years where diabetes or two or more cardiometabolic risk factors are present, and every two to three years otherwise. A low result is a scheduling decision rather than a discharge: MASLD progresses slowly, and the value of the pathway comes from repeating it.
Can this be used without a FibroScan?#
Not as it stands. Liver stiffness and CAP are required fields, because four of the eight scores — FAST and Agile 3+ directly, and the concordance verdict through the stiffness bands — cannot be produced without them. Where no FibroScan has been done, use the standalone FIB-4, APRI and NAFLD Fibrosis Score pages, which need blood values only, and return here once elastography is available.
Does a high score mean the patient has cirrhosis?#
No. Every output here is a probability of a stage, not the stage itself. Concordantly abnormal tests make advanced fibrosis likely enough to act on — referral, surveillance, endoscopy where indicated — but the wording throughout is deliberately probabilistic, because non-invasive tests and histology are not interchangeable in every clinical situation.
References
Guidance defining the pathway and thresholds
- Rinella ME, Neuschwander-Tetri BA, Siddiqui MS, et al. AASLD Practice Guidance on the clinical assessment and management of nonalcoholic fatty liver disease. Hepatology. 2023;77(5):1797-1835.
- EASL-EASD-EASO Clinical Practice Guidelines on the management of metabolic dysfunction-associated steatotic liver disease (MASLD). J Hepatol. 2024;81(3):492-542.
- Berzigotti A, Tsochatzis E, Boursier J, et al. EASL Clinical Practice Guidelines on non-invasive tests for evaluation of liver disease severity and prognosis – 2021 update. J Hepatol. 2021;75(3):659-689.
- Rinella ME, Lazarus JV, Ratziu V, et al. A multisociety Delphi consensus statement on new fatty liver disease nomenclature. Hepatology. 2023;78(6):1966-1986.
Primary references for the component scores
- Sterling RK, Lissen E, Clumeck N, et al. Development of a simple noninvasive index to predict significant fibrosis in patients with HIV/HCV coinfection (FIB-4). Hepatology. 2006;43(6):1317-1325.
- Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C (APRI). Hepatology. 2003;38(2):518-526.
- Angulo P, Hui JM, Marchesini G, et al. The NAFLD fibrosis score: a noninvasive system that identifies liver fibrosis in patients with NAFLD. Hepatology. 2007;45(4):846-854.
- Daniels SJ, Leeming DJ, Eslam M, et al. ADAPT: an algorithm incorporating PRO-C3 accurately identifies patients with NAFLD and advanced fibrosis. Hepatology. 2019;69(3):1075-1086.
- Madsen BS, Thiele M, Detlefsen S, et al. PRO-C3 and ADAPT algorithm accurately identify patients with advanced fibrosis due to alcohol-related liver disease. Aliment Pharmacol Ther. 2021;54(5):699-708.
- Newsome PN, Sasso M, Deeks JJ, et al. FibroScan-AST (FAST) score for the non-invasive identification of patients with non-alcoholic steatohepatitis with significant activity and fibrosis: a prospective derivation and global validation study. Lancet Gastroenterol Hepatol. 2020;5(4):362-373.
- Sanyal AJ, Foucquier J, Younossi ZM, et al. Enhanced diagnosis of advanced fibrosis and cirrhosis in individuals with NAFLD using FibroScan-based Agile scores. J Hepatol. 2023;78(2):247-259.
- Jung J, Loomba RR, Imajo K, et al. MRE combined with FIB-4 (MEFIB) index in detection of candidates for pharmacological treatment of NASH-related fibrosis. Gut. 2021;70(10):1946-1953.
Threshold sources for the entered tests
Further reading
- McPherson S, Hardy T, Dufour JF, et al. Age as a confounding factor for the accurate non-invasive diagnosis of advanced NAFLD fibrosis. Am J Gastroenterol. 2017;112(5):740-751.
- Kleiner DE, Brunt EM, Van Natta M, et al. Design and validation of a histological scoring system for nonalcoholic fatty liver disease (NAS). Hepatology. 2005;41(6):1313-1321.