About the Baveno VII Criteria for cACLD, CSPH and Screening Endoscopy
Two numbers — liver stiffness and platelet count — settle three separate questions, and Baveno VII is explicit that they do not all resolve at the same thresholds. Stiffness below 10 kPa rules out compensated advanced chronic liver disease, 10–15 kPa is suggestive and above 15 kPa is highly suggestive. For clinically significant portal hypertension, stiffness of 15 kPa or below with platelets of 150 ×10⁹/L or above rules it out, and 25 kPa or above rules it in. Screening endoscopy can be spared when stiffness is below 20 kPa and platelets above 150. Between the rule-out and rule-in lies a deliberate grey zone, part of which carries at least a 60% probability of portal hypertension.
Formula
cACLD: LSM < 10 ruled out · 10–15 suggestive · > 15 highly suggestive
CSPH: LSM ≤ 15 AND platelets ≥ 150 → ruled out; LSM ≥ 25 → ruled in
LSM 20–25 with platelets < 150, or LSM 15–20 with platelets < 110 → ≥ 60% probability
Endoscopy: LSM < 20 AND platelets > 150 → can be spared- LSM
- Liver stiffness measurement in kPa by transient elastography, fasted and reliable. The rule of five refers to the thresholds 10, 15, 20 and 25.
- Platelets
- Platelet count in ×10⁹/L. Note the two different cut-offs: 150 for the CSPH rule-out and the endoscopy rule, 110 for the lower arm of the high-probability subset.
- The rule of five is 10-15-20-25 kPa, and Baveno VII frames it as marking progressively higher relative risk of decompensation and liver-related death independently of aetiology — not merely as diagnostic cut-offs.
- The three determinations use different thresholds and routinely disagree. A patient at 18 kPa with platelets of 160 can spare endoscopy while their portal hypertension status remains uncertain, which is not a contradiction.
- The CSPH rule-out requires BOTH conditions. Stiffness of 12 kPa with platelets of 120 does not rule out portal hypertension.
- The endoscopy rule uses strict inequalities as published — stiffness below 20 and platelets above 150. A patient at exactly 150 platelets does not meet it.
- These criteria assume compensated disease. In a patient who has already decompensated, clinically significant portal hypertension is present by definition and the calculation adds nothing.
Interpreting the result
Read the three outputs separately, because they answer different questions. A cACLD result of 'ruled out' at under 10 kPa applies only in the absence of other clinical or imaging signs of advanced disease — a nodular liver on ultrasound overrides a reassuring stiffness. For portal hypertension, a rule-out is genuinely reassuring and a rule-in is strong enough to act on without measuring the pressure gradient. The high-probability subset matters more than its name suggests: at 60% or above, many hepatologists would start a non-selective beta-blocker, because the PREDESCI trial showed beta-blockade reduces decompensation in patients with clinically significant portal hypertension irrespective of whether varices are present. The grey zone is where additional information genuinely helps — spleen stiffness, repeat measurement after treating the underlying disease, or a pressure gradient where it would change management. Finally, none of this is static. Stiffness falls with sustained viral suppression or alcohol abstinence, and the assessment should be repeated rather than treated as a permanent label.
| Score | Band | What it means | Action |
|---|---|---|---|
| LSM ≤ 15 kPa AND platelets ≥ 150 | CSPH ruled out | Clinically significant portal hypertension excluded in compensated advanced chronic liver disease | No beta-blocker indicated on portal-pressure grounds. Reassess periodically, since stiffness and platelets both change |
| Neither rule met, outside the high-probability subset | Grey zone | Portal hypertension genuinely uncertain — the consensus declines to call it either way | Consider spleen stiffness, repeat assessment after treating the aetiology, or HVPG where it would change management |
| LSM 20–25 with plt < 150, or LSM 15–20 with plt < 110 | CSPH probable (≥ 60%) | At least 60% probability of clinically significant portal hypertension | Not a formal rule-in, but high enough that many would start a non-selective beta-blocker — carvedilol preferred |
| LSM ≥ 25 kPa | CSPH ruled in | Clinically significant portal hypertension established without measuring the hepatic venous pressure gradient | Start a non-selective beta-blocker to prevent decompensation, irrespective of whether varices are present |
What the Baveno VII Criteria needs (3 inputs)
- Liver stiffness by transient elastography (kPa)
- The median of a reliable fasted measurement by transient elastography. Stiffness rises with inflammation, cholestasis, hepatic congestion and food intake independently of fibrosis, so the same liver can read very differently depending on when it was measured — a value taken during a hepatitis flare overstates the fibrosis it implies.
- Platelet count (×10⁹/L)
- A falling platelet count reflects portal hypertension and splenic sequestration, which is why it adds information stiffness alone does not. Both thresholds used here, 150 and 110, are absolute counts.
- Aetiology of chronic liver disease
- The 25 kPa rule-in for portal hypertension was endorsed for viral, alcohol-related and non-obese MASLD disease. In obese MASLD the relationship between stiffness and portal pressure differs, so the rule-in is applied with more caution there.
What it returns
- cACLD status
- Ruled out, suggestive, or highly suggestive of compensated advanced chronic liver disease, from the lower two thresholds of the rule of five.
- CSPH status
- Ruled out, grey zone, probable at 60% or more, or ruled in. The grey zone is a real output, not a failure to compute one.
- Screening endoscopy
- Whether the Baveno VI criteria — retained by Baveno VII — allow screening endoscopy to be spared. This is a separate threshold from the portal hypertension call and frequently gives a different answer.
How it is calculated
The consensus works backwards from what portal pressure does rather than from what it is. Clinically significant portal hypertension is defined as a hepatic venous pressure gradient of 10 mmHg or more, the point above which varices, ascites and decompensation become likely — but since that measurement requires catheterisation, Baveno VII asked which non-invasive combinations predict it reliably enough to act on. Liver stiffness carries most of the signal because fibrosis and increased intrahepatic resistance are the same process; the platelet count adds the splenic and haemodynamic consequences of established portal hypertension, which stiffness alone does not capture. The thresholds were then set asymmetrically on purpose: a rule-out chosen for high negative predictive value, a rule-in chosen for high specificity, and an explicitly named grey zone between them where neither is achievable. That structure is the document's central methodological choice — it refuses to convert an uncertain measurement into a false binary.
Facts & figures
| LSM (kPa) | cACLD | Bearing on CSPH and endoscopy |
|---|---|---|
| < 10 | Ruled out | With platelets ≥ 150, rules CSPH out; endoscopy can be spared |
| 10–15 | Suggestive | With platelets ≥ 150, still rules CSPH out; endoscopy can be spared |
| 15–20 | Highly suggestive | CSPH grey zone; ≥ 60% probable if platelets < 110. Endoscopy sparable if platelets > 150 |
| 20–25 | Highly suggestive | ≥ 60% probable if platelets < 150. Screening endoscopy indicated |
| ≥ 25 | Highly suggestive | CSPH ruled in. Screening endoscopy indicated |
Baveno VII presents 10-15-20-25 as marking progressively higher relative risk of decompensation and liver-related death independently of aetiology, so the rule of five is a risk gradient as much as a set of diagnostic cut-offs.
| Question | Rule | Note |
|---|---|---|
| Does the patient have cACLD? | LSM 10 and 15 kPa | Absence of other clinical or imaging signs assumed |
| Can CSPH be ruled out? | LSM ≤ 15 AND platelets ≥ 150 | Both required — either alone is insufficient |
| Can CSPH be ruled in? | LSM ≥ 25 | Endorsed for viral, alcohol-related and non-obese MASLD disease |
| Can screening endoscopy be spared? | LSM < 20 AND platelets > 150 | The Baveno VI criteria, retained unchanged by Baveno VII |
Because the thresholds differ, the three answers frequently diverge on the same patient — most commonly a patient who can spare endoscopy while their portal hypertension status stays in the grey zone.
Evidence
Derivation — Baveno VII consensus workshop
2022International consensus workshop on portal hypertension, revising Baveno VI. Statements were developed against the accumulated evidence on non-invasive assessment, with hepatic venous pressure gradient retained as the reference standard against which the non-invasive thresholds were calibrated.
Consensus-derived rather than fitted to a single cohort. The rule of five is presented as marking progressively higher relative risk of decompensation and liver-related death independently of the aetiology of chronic liver disease.
Precursor — Baveno VI endoscopy criteria
2015The 2015 Baveno VI workshop, which first defined the combination of liver stiffness below 20 kPa and platelets above 150 ×10⁹/L as identifying patients who can avoid screening endoscopy.
Established the endoscopy-sparing rule retained unchanged by Baveno VII, on the basis that patients meeting it carry a very low probability of varices needing treatment.
Risk modelling — ANTICIPATE study
2016Study developing non-invasive risk models for clinically significant portal hypertension and varices in compensated cirrhosis, using liver stiffness and platelet count against measured hepatic venous pressure gradient.
Provided the modelled probabilities of clinically significant portal hypertension across combinations of stiffness and platelet count that underpin the high-probability subset used in Baveno VII.
Treating CSPH changes outcomes — PREDESCI
2019Randomised, double-blind trial of non-selective beta-blockers versus placebo in patients with compensated cirrhosis and clinically significant portal hypertension, defined by hepatic venous pressure gradient.
Beta-blockade reduced decompensation in patients with clinically significant portal hypertension, which is what makes identifying CSPH non-invasively worth doing — it converts a diagnosis into a treatment decision.
Validation and expansion — Hepatology 2024
2024Validation study correlating the Baveno VII liver-stiffness and platelet categories with subsequent hepatic decompensation and death, and proposing further subdivision at the top of the range.
Risk of hepatic decompensation and death increased across the categories during follow-up, supporting the rule of five as a risk gradient. The authors proposed separating stiffness of 25–49.9 kPa from 50 kPa and above as 'critical' portal hypertension.
How it compares
Baveno VII Criteria vs Hepatic venous pressure gradient (HVPG)
HVPG remains the reference standard and Baveno VII does not claim otherwise — but it requires catheterisation, so the realistic comparison is against no measurement at all.
Clinically significant portal hypertension is defined as an HVPG of 10 mmHg or more, and every non-invasive threshold in Baveno VII was calibrated against it. The practical problem is availability: hepatic vein catheterisation is performed in a small number of centres, so outside those centres portal hypertension was historically diagnosed only once varices or ascites had appeared — that is, after the event the diagnosis exists to prevent. The non-invasive criteria trade some accuracy for a diagnosis that actually gets made. HVPG retains a role in the grey zone, in trials, and where the stakes of being wrong are high enough to justify the procedure.
Baveno VII Criteria vs EVendo score
Both aim to avoid unnecessary endoscopy, but Baveno VII needs elastography and EVendo needs only routine blood tests — availability is the deciding factor.
The Baveno VI criteria retained in Baveno VII require a transient elastography measurement, which not every service has. EVendo was derived to answer the same question — can this patient avoid screening endoscopy — using platelets, INR, AST, haemoglobin, BUN and the presence of ascites, all of which are already on the chart. Where elastography is available the Baveno criteria are the better-established rule and are what guidelines are written around; where it is not, EVendo covers the same decision without new equipment. The two are complementary rather than competing, and some services apply EVendo first and reserve elastography for those it does not clear.
Baveno VII Criteria vs FIB-4 index
Different stages of the same pathway — FIB-4 triages who needs elastography at all, and Baveno VII takes over once a stiffness measurement exists.
FIB-4 is a blood-based first-line triage: cheap, computable from tests already done, and designed to identify who warrants further fibrosis assessment. It does not address portal hypertension and its thresholds say nothing about varices or endoscopy. Baveno VII begins where FIB-4 ends, using stiffness plus platelets to determine advanced disease, portal hypertension and the need for endoscopy. In a well-organised pathway they run in sequence, and using FIB-4 to answer a portal hypertension question is applying it well beyond what it was derived for.
Baveno VII Criteria vs Child-Pugh score
Child-Pugh describes established liver dysfunction; Baveno VII identifies risk before dysfunction appears, in patients Child-Pugh would call class A.
Almost every patient the Baveno VII criteria are designed for is Child-Pugh class A — compensated, with normal or near-normal bilirubin, albumin and INR, and no ascites or encephalopathy. Child-Pugh therefore has essentially no discriminating power in this population, which is precisely the gap the non-invasive criteria fill. The two become relevant at different points in the disease: Baveno VII for stratifying compensated patients and deciding on prophylaxis, Child-Pugh once decompensation has occurred and the question shifts to prognosis and transplant assessment.
Pearls & pitfalls
- The CSPH rule-out needs both conditions. Stiffness of 12 kPa with platelets of 120 does not rule out portal hypertension, and reading the stiffness alone is the commonest error here.
- The endoscopy threshold and the CSPH thresholds are different numbers. A patient at 18 kPa with platelets of 160 can spare endoscopy while their portal hypertension status remains uncertain — that divergence is expected, not a contradiction.
- Measure fasted. A recent meal raises stiffness substantially, and a post-prandial reading can move a patient across a threshold.
- Stiffness is not a fibrosis meter. Hepatitis flares, cholestasis, hepatic congestion from right heart failure and infiltrative disease all raise it independently of fibrosis, and a value taken during any of those overstates the disease.
- The under-10 kPa rule-out for cACLD applies only in the absence of other clinical or imaging signs. A nodular liver, splenomegaly or a low platelet count overrides a reassuring stiffness.
- The 25 kPa rule-in was endorsed for viral, alcohol-related and non-obese MASLD disease. In obese MASLD the stiffness-to-pressure relationship differs and the rule-in is less secure.
- None of this applies to a decompensated patient. Once there is ascites, variceal bleeding or encephalopathy, clinically significant portal hypertension is established and the calculation adds nothing.
- Reassess rather than labelling. Stiffness falls with sustained viral suppression or alcohol abstinence, and a patient can move out of a risk category with treatment of the underlying disease.
- 'Probable CSPH' at 60% is a probability, not a diagnosis — but the PREDESCI evidence means it is frequently enough to justify starting a beta-blocker, so treating it as merely indeterminate under-serves the patient.
Critical actions
- Confirm the elastography measurement is reliable and fasted before acting on it, and repeat it if taken during a flare, in cholestasis or in cardiac congestion.
- Record all three outputs — cACLD status, CSPH status and the endoscopy decision — since they answer different questions and are needed for different purposes.
- Where CSPH is ruled in or probable, consider a non-selective beta-blocker, with carvedilol preferred, to prevent decompensation rather than waiting for varices to appear.
- Where endoscopy can be spared, repeat stiffness and platelets annually and perform endoscopy if stiffness reaches 20 kPa or platelets fall to 150 ×10⁹/L or below.
- Treat the underlying aetiology in every case — viral suppression, alcohol abstinence, metabolic management — and repeat the assessment afterwards, because the risk category is modifiable.
- In the grey zone, obtain additional information rather than defaulting to no action: spleen stiffness, repeat measurement, or hepatic venous pressure gradient where it would change management.
- Check for other signs of advanced disease on imaging before accepting a sub-10 kPa reading as ruling out cACLD.
Why this score exists
The Baveno process has run since 1990 and its consistent method is to convert accumulating evidence into statements clinicians can act on without a research infrastructure. Baveno VII's particular contribution was to accept that non-invasive assessment had become good enough to replace hepatic venous pressure gradient measurement for most decisions — not because it is as accurate, but because the pressure gradient was in practice almost never measured, so the realistic comparison was against no assessment at all. Two design choices follow. The thresholds are deliberately asymmetric, chosen separately for ruling out and ruling in rather than as a single cut-off maximising overall accuracy, because the costs of the two errors differ. And the grey zone is named rather than eliminated: the faculty could have forced a binary by picking a midpoint, and chose instead to report uncertainty as uncertainty. The renaming of the disease itself — compensated advanced chronic liver disease rather than compensated cirrhosis — came from the same instinct, acknowledging that a stiffness measurement identifies a continuum the word cirrhosis implies is a state.
About the creator
Chair, Baveno workshops; first author of Baveno VI and VII
Led the Baveno consensus process across successive workshops, including the VII revision.
Baveno VII faculty; portal hypertension haemodynamics
Contributed much of the underlying work relating hepatic venous pressure gradient to clinical outcomes.
Baveno VII faculty; co-author of the AASLD guidance
Co-authored Baveno VII and the AASLD practice guidance that translated it for US practice.
Limitations
- Depends entirely on a reliable transient elastography measurement, which is unavailable in many services and unreliable in obesity, ascites and narrow intercostal spaces.
- Liver stiffness is raised by inflammation, cholestasis and hepatic congestion independently of fibrosis, so the same patient can cross a threshold for reasons unrelated to portal pressure.
- The 25 kPa rule-in was endorsed for viral, alcohol-related and non-obese MASLD disease; its performance in obese MASLD, cholestatic disease and vascular liver disease is less established.
- The grey zone is substantial in practice, and the consensus offers no single next step for patients who land in it.
- Consensus-derived rather than fitted, so the thresholds have no single published sensitivity and specificity — they were selected for asymmetric performance at each end.
- Platelet count is affected by hypersplenism, bone marrow disease, alcohol and drugs, so it is not a clean marker of portal pressure.
- Not applicable in decompensated disease, in acute liver failure, or after transplantation.
- The criteria address varices needing treatment and portal hypertension, not hepatocellular carcinoma surveillance, which continues on its own schedule regardless of the result.
If you are the patient
Baveno VII is a set of rules doctors use to work out how far liver scarring has progressed, and whether the pressure in the veins around the liver has risen — without needing an invasive test. It uses two things: a scan called FibroScan that measures how stiff the liver is, and your platelet count from a routine blood test. Those two numbers answer three questions. First, whether there is advanced scarring: under 10 means unlikely, over 15 means very likely. Second, whether the pressure in the liver's veins is high enough to need treatment: a stiffness of 15 or less with a normal platelet count makes it unlikely, while 25 or more makes it likely. Third, and this is the one patients notice, whether you need a camera test to look for enlarged veins in the gullet — if your stiffness is under 20 and your platelets above 150, that test can usually be skipped, and repeated blood tests and scans are enough. Two things are genuinely useful to know. The scan should be done fasted, because eating beforehand raises the reading and can push you into a higher category unnecessarily. And these numbers can improve — treating the underlying cause, whether that is a virus, alcohol or a metabolic problem, often lowers liver stiffness over time, so this is a picture of where things stand now rather than a permanent label.
Frequently asked questions
What is the Baveno VII rule of five?#
Liver stiffness thresholds of 10, 15, 20 and 25 kPa by transient elastography, which Baveno VII presents as marking progressively higher relative risk of decompensation and liver-related death independently of the cause of liver disease. The thresholds also anchor the specific diagnostic rules: 10 and 15 for compensated advanced chronic liver disease, 15 and 25 for portal hypertension, and 20 for the endoscopy decision.
What liver stiffness rules out clinically significant portal hypertension?#
A stiffness of 15 kPa or below together with a platelet count of 150 ×10⁹/L or above. Both conditions are required — a low stiffness with a low platelet count does not rule out portal hypertension, and this is the most frequent misapplication of the criteria.
What liver stiffness rules clinically significant portal hypertension in?#
25 kPa or above. Baveno VII endorsed this for viral, alcohol-related and non-obese MASLD compensated advanced chronic liver disease, where it is considered sufficient to diagnose portal hypertension without measuring the hepatic venous pressure gradient. In obese MASLD the relationship between stiffness and portal pressure differs and the rule-in is applied more cautiously.
When can screening endoscopy be avoided under Baveno VII?#
When liver stiffness is below 20 kPa and the platelet count is above 150 ×10⁹/L. These are the Baveno VI criteria, retained unchanged, and patients meeting them have a very low probability of varices needing treatment. Repeat stiffness and platelets annually and perform endoscopy if stiffness reaches 20 kPa or platelets fall to 150 or below.
What is compensated advanced chronic liver disease (cACLD)?#
The term Baveno adopted in place of 'compensated cirrhosis', to reflect that non-invasive testing identifies a continuum of severe fibrosis and cirrhosis rather than a single state, and that the two cannot be reliably separated without biopsy. Stiffness below 10 kPa rules it out in the absence of other clinical or imaging signs, 10–15 kPa is suggestive, and above 15 kPa is highly suggestive.
What does the Baveno VII grey zone mean?#
That the combination of stiffness and platelets meets neither the rule-out nor the rule-in criteria for portal hypertension. It is a deliberate output rather than a failure — the faculty chose to report uncertainty instead of forcing a binary at an arbitrary midpoint. In the grey zone, spleen stiffness, repeat measurement after treating the underlying disease, or a hepatic venous pressure gradient are the usual next steps.
What is the 60% probability subset in Baveno VII?#
Stiffness of 20–25 kPa with platelets below 150 ×10⁹/L, or stiffness of 15–20 kPa with platelets below 110 ×10⁹/L. These combinations correspond to at least a 60% probability of clinically significant portal hypertension. It is not a formal rule-in, but given that beta-blockade reduces decompensation in patients with portal hypertension, many clinicians would start treatment at this level.
Do the Baveno VII criteria apply to decompensated cirrhosis?#
No. They are built for compensated advanced chronic liver disease. Once a patient has developed ascites, variceal bleeding or hepatic encephalopathy, clinically significant portal hypertension is present by definition and there is nothing for the criteria to determine.
Can liver stiffness improve, and does the category change?#
Yes to both. Stiffness falls with sustained viral suppression in hepatitis, with alcohol abstinence, and with metabolic improvement in MASLD, and patients can move into a lower risk category as a result. That is why the assessment should be repeated after the underlying disease has been treated rather than being recorded once and treated as a permanent label.
References
Original / primary reference
- de Franchis R, Bosch J, Garcia-Tsao G, Reiberger T, Ripoll C; Baveno VII Faculty. Baveno VII — Renewing consensus in portal hypertension. J Hepatol. 2022;76(4):959-974.
- de Franchis R; Baveno VI Faculty. Expanding consensus in portal hypertension: report of the Baveno VI Consensus Workshop. J Hepatol. 2015;63(3):743-752 (source of the endoscopy-sparing criteria).
Validation and evidence
- Abraldes JG, Bureau C, Stefanescu H, et al. Noninvasive tools and risk of clinically significant portal hypertension and varices in compensated cirrhosis: the 'Anticipate' study. Hepatology. 2016;64(6):2173-2184.
- Villanueva C, Albillos A, Genescà J, et al. β blockers to prevent decompensation of cirrhosis in patients with clinically significant portal hypertension (PREDESCI): a randomised, double-blind, placebo-controlled, multicentre trial. Lancet. 2019;393(10181):1597-1608.
- Validation and expansion of Baveno VII criteria for cACLD and CSPH based on liver stiffness and platelet count: correlation with risk of hepatic decompensation and death. Hepatology. 2024;82(2):422-437.