Cirrhosis

Meet the patient

A 58-year-old man with a long history of heavy alcohol use is brought in by his daughter. His abdomen has been swelling for two months, he slept most of yesterday, and this morning he could not draw a clock. On examination he has spider naevi, palmar erythema, a flapping tremor at the wrist, and shifting dullness in the abdomen.^[1]

This is decompensated cirrhosis with hepatic encephalopathy, and three questions now run in parallel: what precipitated the decline?, is there an infection or a bleed hiding?, and how do I grade the severity and the transplant urgency? Hold those three and the rest of the topic slots in around them.^[1]

Compensated versus decompensated — the line that decides survival

Enlarge

One clinical line divides the whole disease and predicts the prognosis more than any scan. Cirrhosis is compensated until one of the four decompensation events appears, and the drop in survival across that line is steep.^[1]

• Compensated cirrhosis — the liver still performs its synthetic, detoxifying, and excretory functions despite architectural distortion. The patient is usually asymptomatic, detected by stigmata, abnormal liver tests, or incidental imaging. Median survival exceeds 10 to 12 years.
• Decompensated cirrhosis — defined by the development of one of four events: ascites, variceal haemorrhage, hepatic encephalopathy, or jaundice. Median survival without transplant falls to about 1.5 to 2 years.^[1]

^[1]

The number rule: 4 events, 1 line. Ascites, variceal bleed, encephalopathy, jaundice — any one moves the patient across the line. Ascites is the commonest first decompensation; name all four in a viva before you say anything else.^[1]

Etymology for viva gold: cirrhosis is from Greek kirrhos, "tawny-yellow" — Laennec coined it from the nodular, yellow-green appearance of the chronically diseased liver at autopsy. The word survived because the gross pathology it describes is unchanged two centuries on.^[1]

Baveno VII added a word that changes counselling: "recompensation". With sustained cause treatment — HCV cure, HBV suppression, alcohol abstinence — a decompensated patient who resolves ascites and encephalopathy off diuretics and off lactulose (or on stable low doses), with no further decompensation for at least 12 months, returns toward a compensated phenotype. The cirrhosis persists; the prognosis improves. This is not cure, but it is real.^[2]

The stigmata — the viva shopping list

A cirrhotic announces the diagnosis on the hands, the skin, and the face before you scan the abdomen. Examiners expect ten signs, clustered so they stay:^[1]

• Hands — palmar erythema (thenar and hypothenar, sparing the central palm), Dupuytren contracture, clubbing (especially in PBC), leuconychia (hypoalbuminaemia), asterixis if encephalopathic.
• Skin — spider naevi (a central arteriole with radiating vessels, confined to the drainage of the superior vena cava — above the nipple line), easy bruising.
• Face and head — parotid enlargement, fetor hepaticus (sweet faecal breath from mercaptans), xanthelasma in PBC, sallow complexion.
• Chest and endocrine — gynaecomastia (testosterone falls, oestrogen rises from reduced hepatic clearance), testicular atrophy, loss of secondary sexual hair.
• Abdomen — hepatosplenomegaly early, a small shrunken liver late, caput medusae (recanalised umbilical vein), ascites.
• General — sarcopaenia, especially temporal and proximal muscle wasting.^[1]

Aetiology clues at the bedside narrow the cause faster than a panel: tattoos or IV drug use (HBV, HCV); a Kayser-Fleischer ring (Wilson); bronze diabetes and heart failure (haemochromatosis); parotid enlargement and Dupuytren (alcohol); xanthomata and pruritus (PBC); inflammatory bowel disease with recurrent cholangitis (PSC).^[1]

Two master mechanisms — fibrogenesis and portal hypertension

Enlarge

The fibrogenesis cascade is the mechanism every candidate must recite as a chain. Each step earns its place because it explains a treatment.^[1]

1. Chronic hepatocellular injury — virus, alcohol, fat, cholestasis, autoimmune, metabolic — produces apoptotic hepatocytes and reactive oxygen species.
2. Injured hepatocytes, sinusoidal endothelium, and Kupffer cells release profibrogenic cytokines — TGF-beta1, PDGF, CTGF.
3. These activate quiescent hepatic stellate cells (the vitamin-A storing Ito cells of Disse space), which transdifferentiate into proliferating, contractile myofibroblasts.
4. Activated stellate cells secrete type I and III collagen into the space of Disse, express TIMPs that block matrix breakdown, and contract around sinusoids to raise intrahepatic resistance.
5. Sinusoidal endothelium loses its fenestrations (capillarisation), impeding oxygen and metabolite exchange; fibrous septa bridge portal tracts to central veins; regenerative nodules form.^[1]

The regression half of the story is why cause-treatment works. Remove the injurious stimulus — HCV cure, alcohol abstinence, HBV suppression — and activated stellate cells undergo apoptosis or senescence, collagenases (MMP-1) break down matrix, and fibrosis can regress. This is the biological basis of Baveno VII's "recompensation".^[2]

Portal hypertension has two components, and the split is therapeutically loaded. The structural component — fibrosis, nodules, distortion — is fixed. The dynamic component — stellate cell contraction driven by endothelin, thromboxane, and angiotensin-II against deficient intrahepatic nitric oxide — is reversible, and it is the rationale for NSBB (lowering portal inflow) and terlipressin (splanchnic vasoconstriction). Meanwhile excess splanchnic nitric oxide and glucagon dilate the splanchnic arteriolar bed, increasing portal inflow and worsening the pressure.^[1]

The HVPG thresholds are the single most reproduced numbers in hepatology exams.^[1]

^[1]

The hyperdynamic circulation explains ascites, hyponatraemia, and hepatorenal syndrome in one chain. Splanchnic vasodilation drops systemic vascular resistance and the effective arterial blood volume, which fires the RAAS, the sympathetic nervous system, and non-osmotic vasopressin. The consequences are avid sodium retention (ascites), free-water retention (dilutional hyponatraemia), and renal vasoconstriction (hepatorenal syndrome). Ascites itself needs three things together — sinusoidal portal hypertension (HVPG over 12), splanchnic vasodilation, and hypoalbuminaemia lowering oncotic pressure.^[1]

Hepatic encephalopathy — ammonia, inflammation, and the precipitants

Hepatic encephalopathy is reversible, and the single most important step is finding and treating the precipitant. The dominant hypothesis is ammonia modified by systemic inflammation: gut-derived nitrogenous toxins (chiefly ammonia) bypass hepatic clearance through portosystemic collaterals or the failing liver, cross the blood-brain barrier, and astrocytes metabolise ammonia to glutamine — which is osmotically active, swelling the astrocyte. Increased GABAergic tone and cytokine-driven blood-brain barrier permeability synergise with ammonia to depress neuronal function.^[1]

The precipitants are the list you hunt when a cirrhotic deteriorates — and they explain the management:^[1]

^[1]

The West Haven grades are reproduced verbatim, and the bedside test for grade I to II is the clock.^[1]

^[1]

The bedside exam for encephalopathy is asterixis and construction. Ask the patient to hold the hands dorsiflexed with fingers spread for 30 seconds — a liver flap (rapid flexion-extension at the wrist and MCP) appears bilaterally; it is also seen in uraemia and hypercapnia. Then ask them to draw a clock — failure indicates grade I to II. Overt HE is grades II to IV; covert and minimal HE predict overt HE and impair driving.^[1]

A deteriorating cirrhotic is not automatically encephalopathic — run the differential. Hyponatraemia (sodium under 125), sepsis or septic encephalopathy, an intracranial lesion (subdural haematoma from a fall and coagulopathy — CT head if any focal sign or unexplained deterioration), Wernicke encephalopathy (give IV thiamine before glucose), and drug or sedative accumulation all mimic HE.^[1]

Ascites and the SAAG — the single most testable ascites concept

Tap the ascites, then read the serum-ascites albumin gradient. The SAAG (serum albumin minus ascitic albumin, on the same day) classifies ascites by mechanism and is a viva staple:^[1]

^[1]

Within high-gradient (portal-hypertensive) ascites, the bedside setting splits cirrhosis from cardiac failure and Budd-Chiari. All three sit in the same high-gradient column of the table above — the discriminator is the clinical picture (the cardiac signs, the sudden painful hepatomegaly of Budd-Chiari), not the gradient itself.^[1]

The diagnostic paracentesis is mandatory in every cirrhotic with new or worsening ascites, and in every cirrhotic admitted to hospital with ascites — the commonest pitfall is failing to tap. Insert the needle in the left lower quadrant, 2 cm medial and 2 cm superior to the anterior superior iliac spine (avoiding the inferior epigastric vessels). Send fluid for cell count (PMN), albumin (for SAAG), culture in blood-culture bottles at the bedside (yield rises to 80 percent), total protein, and — if indicated — cytology, ADA or TB-PCR, amylase, and triglycerides.^[1]

Severity — Child-Pugh and MELD-Na

Two scores divide the labour, and reproducing their components verbatim is a viva staple. The Child-Pugh mixes labs with clinical findings and stratifies prognosis; the MELD-Na is a pure lab score that predicts 3-month mortality and drives transplant priority.^[1]

Child-Pugh — five components, each 1 to 3 points:^[1]

^[1]

Grades: A is 5 to 6 (1-year survival around 100 percent); B is 7 to 9 (1-year around 80 percent); C is 10 to 15 (1-year around 45 percent). Class C carries the highest mortality without transplant.^[1]

MELD predicts 3-month mortality from bilirubin, INR, and creatinine; MELD-Na adds sodium because hyponatraemia independently predicts death in cirrhosis.^[1]

MELD = 3.78 x ln(bilirubin mg/dL) + 11.2 x ln(INR) + 9.57 x ln(creatinine mg/dL) + 6.43^[1]

MELD-Na = MELD + 1.32 x (137 minus Na) minus [0.033 x MELD x (137 minus Na)]^[1]

Approximate 3-month mortality by MELD: 10 is around 6 percent, 20 around 20 percent, 30 around 53 percent, 40 around 71 percent. A MELD of 15 or more is the threshold to refer for transplant assessment.^[1]

Treat the cause — the most powerful long-term modifier

Before any complication bundle, the single intervention that transforms prognosis is treating the underlying cause. Name the cause, then the cause-specific therapy.^[1]

^[1]

The four complications and their bundles

Enlarge

Ascites — the stepwise ladder

Four rungs, in order — sodium first, diuretic combination second, albumin-covered paracentesis when refractory.^[7]

1. Dietary sodium restriction (2 g per day) — the first-line cornerstone.^[7]
2. Fluid restriction only if hyponatraemia — most patients with ascites need sodium control, not water control.
3. Diuretics — a loop plus a distal agent (furosemide plus spironolactone) — the combination is the first-line regimen; single-agent or nephrotoxic approaches fail. Watch for gynaecomastia, hyperkalaemia, and dehydration.^[7]
4. Refractory ascites — large-volume paracentesis with intravenous albumin 6 to 8 g per litre removed once 5 L or more is drained (to minimise post-paracentesis haemodynamic and renal dysfunction), TIPS when liver function is relatively preserved, and liver transplantation — the only effective definitive therapy.^[7]

^[7]

Spontaneous bacterial peritonitis — cefotaxime plus albumin, do not wait for culture

The diagnostic threshold is a PMN count of 250 cells/mm^3 or more, and the antibiotic starts the moment it is met.^[1]

• Treatment — cefotaxime 2 g IV every 12 hours (equally as effective as 2 g every 6 hours in the randomised dose-comparison trial)^[11], plus intravenous albumin 1.5 g/kg at diagnosis and 1 g/kg on day 3 — in the landmark trial albumin cut renal impairment from 33 to 10 percent and in-hospital mortality from 29 to 10 percent.^[3] A 5-day course is as effective as 10 days.^[12]
• Secondary prophylaxis (after one episode — recurrence risk is high) — norfloxacin 400 mg orally daily until the ascites resolves definitively or liver transplantation.^[8]
• Primary prophylaxis — for low-protein ascites (ascitic fluid total protein under 1 g/dL), norfloxacin 400 mg daily during hospitalisation; after a gastrointestinal bleed, antibiotic prophylaxis is also indicated (see the variceal bundle below).^[8]

^{[11] [3] [12]}

Acute variceal bleed — the time-critical bundle

The variceal bleed bundle starts at first suspicion, before endoscopy confirms varices — every component is examinable.^[6]

• Vasoactive drug — terlipressin, octreotide, or somatostatin, started at the time of presentation and continued for up to 5 days (ESGE strong recommendation).^[6] A terlipressin regimen from the bolus-versus-infusion study: 2 mg IV bolus, then 1 mg IV every 4 hours.^[18]
• Antibiotic — ceftriaxone 1 g IV per day for up to 7 days for all patients with advanced chronic liver disease and acute variceal haemorrhage; the randomised trial showed ceftriaxone 1 g/day for 7 days beat norfloxacin for preventing infection after a bleed (11 versus 33 percent).^[6][13]
• Endoscopy — endoscopic evaluation within 12 hours of presentation (once resuscitated); endoscopic band ligation for oesophageal variceal haemorrhage, cyanoacrylate injection for gastric cardiofundal varices.^[6]
• Restrictive transfusion — transfuse when the haemoglobin falls below 7 g/dL (ESGE threshold 70 g/L, post-transfusion target 70 to 90 g/L); in the restrictive-strategy trial the liberal threshold of 9 g/dL raised the portal-pressure gradient and worsened survival.^[6][10]
• Rescue and pre-emptive TIPS — urgent rescue TIPS for bleeding that persists despite vasoactive and endoscopic therapy; pre-emptive TIPS within 72 hours (preferably 24) for high-risk patients — Child-Pugh C up to 13 points, or Child-Pugh B over 7 with active bleeding at endoscopy. In the early-TIPS randomised trial this cut 1-year failure or rebleeding to 3 percent versus 50 percent and improved 1-year survival (86 versus 61 percent).^[6][9]

Hepatic encephalopathy — fix the precipitant, lactulose, then rifaximin

• Identify and treat the precipitant — infection, bleeding, constipation, electrolyte disturbance, sedatives — the step that decides the episode.
• Lactulose remains the backbone — over 90 percent of patients in the rifaximin prevention trial were on concomitant lactulose.^[4]
• Rifaximin 550 mg orally twice daily — added for secondary prevention of overt HE: over 6 months, breakthrough encephalopathy fell from 45.9 to 22.1 percent and HE-related hospitalisation from 22.6 to 13.6 percent versus placebo.^[4]

Hepatorenal syndrome — recognise early, terlipressin, then transplant

Diagnosis (ADQI/ICA consensus): HRS-AKI is a specific form of AKI in patients with advanced cirrhosis and ascites, and it carries a particularly high mortality. Early recognition matters because splanchnic vasoconstrictors can reverse the AKI and serve as a bridge to liver transplantation — the only curative option.^[14]

• Terlipressin plus albumin is the pharmacologic bridge — in the CONFIRM phase 3 trial, verified HRS reversal occurred in 32 versus 17 percent with placebo, but with more adverse events (including respiratory failure, which caused death within 90 days in 11 versus 2 percent) — monitor for ischaemia and fluid overload.^[5]
• Renal support and transplantation — vasoconstrictors buy time; liver transplantation is the only curative option for HRS-AKI.^[14]

Prevent decompensation — NSBB, HCC surveillance, and the Baveno rule-out

In compensated cirrhosis the goal is to prevent the first decompensation, and two interventions do the heavy lifting.^[2]

Primary prophylaxis of first variceal bleed for clinically significant portal hypertension (HVPG over 10 mmHg or liver stiffness over 25 kPa): a non-selective beta-blocker, preferably carvedilol (ESGE strong recommendation), or, when NSBBs are not tolerated and high-risk varices are found, endoscopic band ligation every 2 to 4 weeks until eradication.^[6] Direct carvedilol evidence: 12.5 mg once daily versus band ligation in the randomised trial halved the first bleed (10 versus 23 percent).^[15] Secondary prophylaxis after a bleed is NSBB (propranolol or carvedilol) plus endoscopic therapy combined — the ESGE standard.^[6]

HCC surveillance is for life, in every cirrhotic, regardless of cause or viral control. 6-monthly ultrasound plus or minus serum alpha-fetoprotein; a nodule with arterial hypervascularity and washout on multiphase CT or MRI is hepatocellular carcinoma by imaging criteria (LI-RADS) without biopsy.^[1]

The Baveno VI rule-out avoids unnecessary endoscopy. A liver stiffness under 25 kPa and platelets over 110 x 10^9/L makes high-risk varices unlikely, and screening endoscopy can be safely skipped.^[2]

Liver transplantation and HCC

Liver transplantation is the definitive therapy for decompensated cirrhosis. Indications: decompensation with MELD 15 or more, a first episode of SBP, a first episode of overt HE, refractory ascites, or HCC within Milan criteria (a single tumour under 5 cm, or up to 3 nodules each under 3 cm, no vascular invasion, no extrahepatic spread). Absolute contraindications: active alcohol or substance misuse, uncontrolled extrahepatic sepsis, severe cardiopulmonary disease, uncontrolled extrahepatic malignancy. Survival is around 85 to 90 percent at 1 year and 70 to 75 percent at 5 years.^[1]

HCC is staged and treated by the BCLC framework. Very early or early disease (single tumour, preserved function): resection, thermal ablation, or transplant within Milan criteria. Intermediate (multinodular, preserved function): transarterial chemoembolisation. Advanced (portal invasion, metastases): systemic therapy — atezolizumab plus bevacizumab first-line, or sorafenib or lenvatinib. Terminal: best supportive care.^[1]

How cirrhosis patients come to harm — the preventable list

• Failing to perform a diagnostic tap in a cirrhotic with new or worsening ascites — SBP missed.^[1]
• Over-transfusing a variceal bleed to a haemoglobin of 9 or 10 — portal pressure rises and bleeding worsens.^[1]
• Giving NSAIDs, ACE-inhibitors, aminoglycosides, or iodinated contrast — precipitating hepatorenal syndrome.^[1]
• Protein-restricting an encephalopathic patient — worsening malnutrition and mortality.^[1]
• Missing HCC surveillance in a stable compensated cirrhotic.^[1]
• Rapidly correcting hyponatraemia — osmotic demyelination.^[1]
• Forgetting the antibiotic and vasoactive drug in the acute variceal bleed bundle.^[1]
• Forgetting thiamine before glucose in the alcoholic hypoglycaemic patient — Wernicke encephalopathy.^[1]

Acute-on-chronic liver failure — the precipitant-driven emergency

When organ failure accompanies acute decompensation, the syndrome is acute-on-chronic liver failure (ACLF), and 28-day mortality is 20 to 60 percent. It is graded ACLF-1, -2, or -3 by the number of organ failures on the CLIF-C organ-failure score — liver (bilirubin), kidney (creatinine), brain (West Haven), coagulation (INR), circulation (vasopressors), lung (PaO2 over FiO2). Management is treat the precipitant (infection, bleed, alcohol binge), organ support in ICU, terlipressin and albumin if HRS, and early transplant in selected patients.^[1]

The trials and scores that set the standard

^{[1] [2]}

Special populations — what changes

Alcohol-related cirrhosis. When acute alcoholic hepatitis supervenes, severity and steroid decisions run on numbers with trial evidence behind them: severe disease is a Maddrey discriminant function of 32 or more — the entry criterion of the steroid-response cohorts.^[17] Prednisolone is the steroid to know: in the STOPAH factorial trial 28-day mortality was 14 versus 17 percent with placebo (odds ratio 0.72, P 0.06 — not significant), and prednisolone carried more serious infections (13 versus 7 percent).^[16] Stop futile steroids early — a day-4 Lille score over 0.45 identifies non-responders (28-day survival 90 versus 66 percent for responders).^[17]

Viral hepatitis B or C. Treat HCV with DAAs when feasible (curative); suppress HBV lifelong with tenofovir or entecavir to reduce decompensation and HCC risk. HCC surveillance continues regardless of viral control — antiviral cure does not eliminate HCC risk. Vaccinate for HAV and HBV if non-immune.^[2]

MASLD or cryptogenic cirrhosis. Look for the metabolic syndrome; many present as cryptogenic (burnt-out MASLD). Cardiovascular disease is the leading cause of death — manage aggressively with statins (safe and beneficial in cirrhosis), antihypertensives (avoid ACE-inhibitors and NSAIDs), tight glycaemic control, and weight loss of 7 to 10 percent.^[1]

Pregnancy with cirrhosis. High risk of variceal bleed in the second trimester and at delivery (increased blood volume and intra-abdominal pressure); screen and band varices before conception if known, prefer regional anaesthesia, and plan delivery in a tertiary centre.^[1]

The mantra

The mantra: treat the cause, prevent the first decompensation with a beta-blocker, tap every new ascites, give the antibiotic to every cirrhotic who bleeds — and watch for the cancer for life.^[1][2]

Ward-round test — three stems, thirty seconds each

Stem 1 — the man from the top of the topic (answer)ShowHide

The 58-year-old with alcohol use, abdominal swelling, sleepiness, and an inability to draw a clock, with spider naevi, palmar erythema, asterixis, and shifting dullness. What is your first move? Model: This is decompensated cirrhosis with hepatic encephalopathy. The first move is to find and treat the precipitant — send cultures, check the sodium, do a diagnostic paracentesis (any cirrhotic with ascites who deteriorates gets a tap), a rectal exam for melaena, and review the drug chart. Start lactulose (the backbone — over 90 percent of patients in the prevention trial were on it)^[4] and, for secondary prevention after a second episode, rifaximin 550 mg twice daily (breakthrough HE 22.1 versus 45.9 percent over 6 months).^[4] If the tap shows SBP, treat with cefotaxime plus albumin 1.5 g/kg at diagnosis and 1 g/kg on day 3.^[3]

Stem 2 — the cirrhotic who bleeds (answer)ShowHide

A known cirrhotic presents with massive haematemesis. What do you start before the endoscopy? Model: Start the variceal bleed bundle at first suspicion — do not wait for endoscopy to confirm varices.^[6] A vasoactive drug at presentation, continued for up to 5 days (e.g. terlipressin 2 mg IV bolus then 1 mg every 4 hours)^[18], ceftriaxone 1 g IV daily for up to 7 days (superior to norfloxacin after a bleed)^[13], protect the airway, and resuscitate restrictively — transfuse below a haemoglobin of 7 g/dL (liberal transfusion raised the portal-pressure gradient)^[10]. Book endoscopic evaluation within 12 hours — band ligation for oesophageal varices, cyanoacrylate for gastric cardiofundal varices.^[6] If bleeding is uncontrolled, urgent rescue TIPS; consider pre-emptive TIPS within 72 hours (preferably 24) for Child-Pugh C up to 13 or Child-Pugh B over 7 with active bleeding.^[6][9]

Stem 3 — the rising creatinine in tense ascites (answer)ShowHide

A decompensated cirrhotic with tense ascites on spironolactone and furosemide has a rising creatinine. What is the pathway? Model: First exclude the reversible causes — SBP (diagnostic tap), hypovolaemia from over-diuresis, nephrotoxins. If the creatinine keeps rising, treat as HRS-AKI: terlipressin plus albumin — in CONFIRM, verified reversal in 32 versus 17 percent with placebo, at the cost of more adverse events including respiratory failure (fatal within 90 days in 11 versus 2 percent) — then liver transplantation, the only curative option.^[5][14] The recurring error is reaching for more diuretics without excluding SBP first.

References21ShowHide

1. [1]Kaplan DE, Ripoll C, Thiele M, et al. AASLD Practice Guidance on risk stratification and management of portal hypertension and varices in cirrhosis Hepatology, 2024.PMID 37870298
2. [2]de Franchis R, Bosch J, Garcia-Tsao G, et al. Baveno VII - Renewing consensus in portal hypertension J Hepatol, 2022.PMID 35120736
3. [3]Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis N Engl J Med, 1999.PMID 10432325
4. [4]Bass NM, Mullen KD, Sanyal A, et al. Rifaximin treatment in hepatic encephalopathy N Engl J Med, 2010.PMID 20335583
5. [5]Wong F, Pappas SC, Curry MP, et al. Terlipressin plus Albumin for the Treatment of Type 1 Hepatorenal Syndrome N Engl J Med, 2021.PMID 33657294
6. [6]Gralnek IM, Camus Duboc M, Garcia-Pagan JC, et al. Endoscopic diagnosis and management of esophagogastric variceal hemorrhage: European Society of Gastrointestinal Endoscopy (ESGE) Guideline Endoscopy, 2022.PMID 36174643
7. [7]Garcia N, Sanyal AJ Ascites Curr Treat Options Gastroenterol, 2001.PMID 11696279
8. [8]Hillebrand DJ Spontaneous Bacterial Peritonitis Curr Treat Options Gastroenterol, 2002.PMID 12408785
9. [9]García-Pagán JC, Caca K, Bureau C, et al. Early use of TIPS in patients with cirrhosis and variceal bleeding N Engl J Med, 2010.PMID 20573925
10. [10]Villanueva C, Colomo A, Bosch A, et al. Transfusion strategies for acute upper gastrointestinal bleeding N Engl J Med, 2013.PMID 23281973
11. [11]Rimola A, Salmerón JM, Clemente G, et al. Two different dosages of cefotaxime in the treatment of spontaneous bacterial peritonitis in cirrhosis: results of a prospective, randomized, multicenter study Hepatology, 1995.PMID 7875666
12. [12]Runyon BA, McHutchison JG, Antillon MR, et al. Short-course versus long-course antibiotic treatment of spontaneous bacterial peritonitis. A randomized controlled study of 100 patients Gastroenterology, 1991.PMID 2019378
13. [13]Fernández J, Ruiz del Arbol L, Gómez C, et al. Norfloxacin vs ceftriaxone in the prophylaxis of infections in patients with advanced cirrhosis and hemorrhage Gastroenterology, 2006.PMID 17030175
14. [14]Nadim MK, Kellum JA, Forni L, et al. Acute kidney injury in patients with cirrhosis: Acute Disease Quality Initiative (ADQI) and International Club of Ascites (ICA) joint multidisciplinary consensus meeting J Hepatol, 2024.PMID 38527522
15. [15]Tripathi D, Ferguson JW, Kochar N, et al. Randomized controlled trial of carvedilol versus variceal band ligation for the prevention of the first variceal bleed Hepatology, 2009.PMID 19610055
16. [16]Thursz MR, Richardson P, Allison M, et al. Prednisolone or pentoxifylline for alcoholic hepatitis N Engl J Med, 2015.PMID 25901427
17. [17]Garcia-Saenz-de-Sicilia M, Duvoor C, Altamirano J, et al. A Day-4 Lille Model Predicts Response to Corticosteroids and Mortality in Severe Alcoholic Hepatitis Am J Gastroenterol, 2017.PMID 27922027
18. [18]Şenkaya A, Çelik F, Kurtulmuş Akgün İ, et al. Terlipressin Treatment for Acute Esophageal Variceal Bleeding: Bolus or Infusion? Turk J Gastroenterol, 2025.PMID 41669923
19. [19]Caraceni P, Riggio O, Angeli P, et al. Long-term albumin administration in decompensated cirrhosis (ANSWER): an open-label randomised trial. Lancet, 2018.PMID 29861076
20. [20]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.PMID 30910320
21. [21]Garcia-Tsao G, Abraldes JG, Rich NE, et al. AGA Clinical Practice Update on the Use of Vasoactive Drugs and Intravenous Albumin in Cirrhosis: Expert Review. Gastroenterology, 2024.PMID 37978969