Article Text
PDF
Abstract
Acute decompensated cirrhosis (DC) and acute-on-chronic liver failure are common reasons for hospital admission that have a high in-hospital mortality rate (10%–20%). Patients require a detailed assessment for precipitating factors and management of complications such as infections, ascites, acute kidney injury and hepatic encephalopathy. Multiple reports have demonstrated unwarranted variability in the care of patients with DC. In 2014, the British Society of Gastroenterology (BSG)/British Association for the Study of the Liver (BASL) DC care bundle (DCCB) was introduced to provide a structured approach for the management of patients with DC in the first 24 hours. Usage of the DCCB has been shown to improve care of patients with DC. However, despite evidence indicating the beneficial impact of the DCCB, overall usage across the UK was only 11.4% in a national audit. Our aim was to update the DCCB to incorporate recent advances in care and improve its usability and develop a strategy to improve its usage nationally. The updated bundle was developed by a multidisciplinary group of specialists from BSG, BASL and the Society for Acute Medicine with the quality of evidence supporting the bundle recommendations assessed using the Grading of Recommendation Assessment Development and Evaluation tool. Proposed minimum standards for audit were also developed. Finally, a strategy to promote usage of the bundle including education/training at a national and local level, improving accessibility for the bundle, and promotion of frameworks for use at an institutional level to improve and monitor utilisation of DCCB.
- CHRONIC LIVER DISEASE
- ASCITES
- HEPATIC ENCEPHALOPATHY
- OESOPHAGEAL VARICES
- CIRRHOSIS
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Introduction
Presentations of decompensated cirrhosis (DC), which include non-acute decompensation, acute decompensation and acute-on-chronic liver failure (ACLF), are common reasons for hospital admission and have a high in-hospital mortality rate (10%–20%).1 2 Unfortunately, the number of hospital admissions with DC has continued to rise over the last decade, being driven largely by admissions with alcohol-associated liver disease (ALD), which have increased by 57.4%.3 In addition, premature mortality from liver disease has also increased by 20% since the onset of the COVID-19 pandemic, having been stable for a decade.3
DC is a complex multisystem disorder that requires a detailed assessment for precipitating factors combined with management of complications such as infections, ascites, acute kidney injury (AKI), GI bleeding and hepatic encephalopathy (HE).2 4 5 In 2013, a National Confidential Enquiry in Patient Outcome and Death (NCEPOD) report found that medical care for many patients admitted with ARLD was inadequate, with only 47% receiving ‘good’ care.6 In response to this, the British Society of Gastroenterology (BSG)/British Association for the Study of the Liver (BASL) admission DC care bundle (DCCB) was developed to provide a structured approach for the medical management of this condition, focusing on the first 24 hours.7 Following the implementation of the care bundle at multiple sites, reports showed that the provision of evidence-based care for DC was better in nearly every domain when the DCCB was used.8 9 However, despite evidence indicating the beneficial impact of the DCCB, overall usage across the UK is disappointingly low, with a recent national audit showing that only 11.4% of people admitted with DC across the UK had the bundle used.9 A follow-up review on the organisation of alcohol-related liver disease services in 2022, ‘Remeasuring the Units’, showed that while some progress had been made, significant unwarranted variability in care was observed and there was a clear need for further improvement in service delivery and care.3
Therefore, our aim was to update the DCCB to incorporate recent advances in care and to develop a strategy to improve its usage across the UK, with the ultimate aim of improving outcomes for patients admitted with DC and reducing variation in care.
Methods
A group of specialists (hepatologists, gastroenterologists, acute medicine physicians, critical care physician, liver specialist nurse, pharmacists and dietitian) from BSG, BASL and the Society for Acute Medicine (SAM) was convened to update the bundle. The group was divided into subgroups to review each of the domains of the original DCCB. A literature search was conducted using PubMed/MEDLINE up to May 2024 to identify relevant original research papers and existing guidelines to support the recommendations made within the DCCB. The quality of evidence supporting the recommendations made in the DCCB was assessed using the Grading of Recommendation Assessment Development and Evaluation tool and is shown in table 1. A revised DCCB was developed to update the content and improve the usability of the document. Proposed minimum standards for audit are displayed in table 1. Draft versions of the DCCB were shown to a focus group of junior doctors to help gain feedback, and revisions were made to optimise the usability of the document. A strategy to promote the dissemination of the DCCB was also developed. This included a literature search and discussion within the team of ideas to promote the bundle and to increase its utilisation. The final manuscript was approved by the BSG liver section and Clinical Services and Standards committee, BASL and SAM.
Quality of Evidence (using GRADE) supporting the recommendations and within the DC Care Bundle and proposed minimum standards to audit adherence with the recommendations
The care bundle
Acutely DC is defined as an acute deterioration in liver function in an individual with cirrhosis, and typically presents with symptoms including jaundice, ascites (new onset or worsening), HE, gastrointestinal bleeding, AKI or signs of systemic inflammatory response syndrome/sepsis.2 4 The DCCB (figure 1; an editable version is available in online supplemental file 1) should be completed for all patients presenting with DC and commenced within 6 hours of admission. The bundle provides a structured approach for management for the first 24 hours, when specialist gastroenterology/hepatology input may not be available. Individuals presenting with DC frequently have identifiable precipitants including gastrointestinal bleeding (variceal and non-variceal), infection/sepsis, alcohol-associated hepatitis, acute portal vein thrombosis, drugs, ischaemic liver injury (hypovolaemia/hypotension) or development of hepatocellular carcinoma. The admitting clinician should document the presenting symptoms and suspected precipitant(s) in the relevant section of the DCCB. Changes following the update of the bundle are summarised in table 2.
Supplemental material
The revised decompensated cirrhosis admission bundle. APTT, activated partial prothrombin time; CIWA, Clinical Institute Withdrawal Assessment; CRP, C reactive protein; FBC, full blood count; FFP, fresh frozen plasma; GMAWS, Glasgow Modified Assessment and Management of Alcohol tool; MCS, microscopy, culture and sensitivity; PCC, prothrombin complex concentrate; VTE, venous thromboembolism.
Summary of the changes to recommendations made in the updated DCCB
Baseline investigations
Patients presenting with DC should have a full history, clinical examination and investigations to look for the cause of decompensation and the presenting features.10 The National Early Warning Score should be recorded and used to monitor the patient’s physiological status. Blood tests should be taken including full blood count, urea and electrolytes, calcium, phosphate, magnesium, liver blood tests, coagulation profile, glucose and C reactive protein (CRP). Hepatitis B, hepatitis C and HIV serology should be taken unless recently performed and there are no new risk factors for these infections. A venous blood gas including lactate should also be obtained.
Patients should also be screened for infection including urinalysis and urine culture, chest X-ray and blood cultures. For patients with clinically detectable ascites, an ascitic tap should be performed at the earliest opportunity to exclude spontaneous bacterial peritonitis (SBP).5 11 Coagulopathy is not a contraindication to this procedure and does not require correction prior to the procedure.6 11 Ascitic fluid should be sent to the lab for analysis of the fluid polymorphonuclear (PMN) and white cell count, microscopy and culture, protein and albumin content.5 11 In addition, ascitic fluid should also be sent in blood culture bottles to aid organism identification and antibiotic sensitivities.11
An abdominal ultrasound including Doppler of the hepatic and portal veins should be requested and performed at the earliest opportunity (unless an ultrasound has been conducted in the last month).10 Ultrasound can help investigate for portal vein thrombosis or hepatocellular carcinoma as precipitants of decompensation and can also identify small volume ascites.
A blood liver aetiology screen should be requested for individuals presenting for the first time or those who have not previously had these investigations including: immunoglobulins, liver autoantibodies, ferritin, transferrin saturation, alpha-1 antitrypsin level, caeruloplasmin, lipids and alpha-fetoprotein.
Alcohol
Overall, 75% of patients admitted with DC in the UK have alcohol as the major aetiological factor.12 All patients should have an alcohol history documented, including current and previous alcohol consumption (in units/day; box 1), time of last drink and any symptoms of alcohol withdrawal.6 For patients with ALD who continue to consume alcohol, parenteral thiamine and other B vitamins (eg, Pabrinex: two pairs of vials three times daily for 3 days or parenteral thiamine (500 mg three times per day for those with features of Wernicke’s/alcohol related brain injury or 250 mg three times per day for prevention)) should be given to treat thiamine deficiency and reduce the risk of alcohol-related brain injury (Wernicke’s encephalopathy or Korsakoff’s syndrome).13 Patients who have symptoms of alcohol withdrawal or who are at high risk of alcohol withdrawal should be assessed with a validated tool such as the revised Clinical Institute Withdrawal Assessment–Alcohol scale or Glasgow Modified Assessment and Management of Alcohol tool and be treated with a symptom-triggered regimen of benzodiazepines.13–15 Consider using reduced dose chlordiazepoxide (25 mg) or lorazepam in patients with DC to treat alcohol withdrawal, as chlordiazepoxide is metabolised by hepatic enzymes and its half-life can be prolonged due to liver impairment (24–48 hours). Lorazepam is shorter acting (half-life 10–20 hours) and is metabolised by conjugation, so may be less likely to accumulate in patients with advanced liver disease.
Alcohol unit definitions
One unit of alcohol=10 mL of alcohol or 8 g of alcohol.
Example: a 500 mL can of 5% beer=25 mL of alcohol=2.5 units of alcohol.
The maximum recommended weekly alcohol consumption is 14 units for both men and women without liver disease.
Patients with cirrhosis of whatever aetiology should be advised to be completely abstinent.
Awareness of the possibility of alcohol-induced pancreatitis as a cause of decompensation and monitoring of blood sugar should also be considered.
Patients admitted with DC of any aetiology must be advised to abstain from alcohol, and the impact of continued alcohol consumption on prognosis must be explained to them.16 Patients with decompensated ALD still drinking should be reviewed by a member of the Alcohol Care Team and linked to community alcohol services for ongoing support postdischarge.17
Infections
Individuals with cirrhosis have impaired defence against bacteria due to immune dysfunction.18 19 Moreover, alcohol itself damages both innate and adaptive immune responses, increasing susceptibility to infection.20 Infections are common in acute DC, and as a result, bacterial infections are one of the most frequent reasons for hospital admission.21 22 Infections become more common in more severe liver disease, significantly increasing the risk of decompensation and worsening prognosis.23 24 Even in early liver disease, infections are an independent predictor of risk of decompensation and death.25
A detailed assessment for infection is critical in all patients admitted with DC, as prompt appropriate treatment with antibiotics improves prognosis.26–28 SBP, pneumonia and urinary tract infections are the most frequent infection sources. Patients with cirrhosis may not display typical signs of infection, such as fever or rise in CRP, so clinicians must have a high index of suspicion for infection. Procalcitonin, an inflammatory marker used in the diagnosis of infections, is frequently elevated in both patients with acute liver failure and advanced liver disease, even in the absence of infection.29 There is not, therefore, currently a clear role for this test in the diagnosis of infection or distinguishing between other causes of decompensation such as alcohol-associated hepatitis.
Gram-negative bacilli, such as Escherichia coli, are the most frequently encountered pathogens.30 Multidrug-resistant organisms are a growing challenge globally; these are associated with lower resolution rates, increased prevalence of septic shock and higher mortality.31 Broad-spectrum antibiotics should be prescribed in line with local epidemiology of bacterial infections, with appropriate microbiology sampling to allow more targeted therapy when the responsible pathogen has been identified to reduce future antibiotic resistance.32 Early de-escalation to appropriate narrow spectrum antibiotics should be considered, and antibiotics stopped if infection is subsequently excluded.
Patients with advanced liver disease are also at increased risk of invasive fungal infections, though these are less frequent than bacterial infections.33 One study of 2743 hospitalised patients with cirrhosis found that less than 5% were diagnosed with fungal infections, all of which were nosocomial.34 Fungal infections in patients with cirrhosis are associated with >50% mortality.34 35 There is little evidence for routine use of empirical antifungal treatment in all patients, though these may be considered in patients at higher risk, such as following intensive care admission or prolonged steroids for alcohol-associated hepatitis. When started, empirical antifungal therapy should be rapidly de-escalated if fungal infection is not confirmed; two negative 1,3-b-D-glucan assays can safely be used to discontinue antifungals.36 Overall, clinicians should have a high index of suspicion for fungal infections when patients are not responding to treatment, but there is no evidence for routine prophylaxis.
SBP is an infection of the ascitic fluid in the absence of a secondary cause, such as intestinal perforation, and occurs in approximately 10% of hospitalised patients with cirrhosis.21 37 SBP is frequently asymptomatic, and delayed diagnosis is associated with increased mortality.38 39 Therefore, all patients presenting with ascites should have a diagnostic ascitic tap to exclude SBP on admission to hospital, or if their clinical status deteriorates.5 11 Diagnosis of SBP is made when the absolute number of PMN cells is >250/mm3 (0.25×109/L) of ascitic fluid.40 Fluid should also be sent in blood culture bottles to aid organism identification and antibiotic sensitivities.11 SBP should be treated promptly with broad spectrum antibiotics according to hospital policy, with modifications according to culture results.5 11
Proton pump inhibitors (PPIs) have been associated with increased risk of development of SBP in some studies,41 42 but not all;43 where prescribed, the indication should be reviewed and the drug discontinued where possible.
Patients with SBP are at high risk of developing hepatorenal syndrome (HRS) and should have intravenous albumin administered to prevent worsening of renal function.5 11 One study showed that intravenous albumin 1.5 g/kg at diagnosis and 1 g/kg at 72 hours reduced the incidence of HRS from 30% to 10% and mortality from 29% to 10%.44 Routine use of albumin is not currently recommended in other sources of infection.5 45
Acute kidney injury
AKI occurs in approximately 20%–30% of hospitalised patients with cirrhosis and is associated with poor outcome.46–49 AKI in patients with cirrhosis is frequently multifactorial, with prerenal AKI the most common cause (45%), followed by acute tubular necrosis and glomerulonephritis (32%), HRS (23%) and post renal causes (<1%).49 A summary of the management of AKI in cirrhosis is shown in figure 2.
A summary of the management of acute kidney injury (AKI) in cirrhosis. HAS, Human Albumin Solution; ITU, intensive therapy unit; RRT, renal replacement therapy; NSAID, non-steroidal anti-inflammatory drug; SBP, spontaneous bacterial peritonitis.
AKI should be diagnosed according to the Kidney Disease: Improving Global Outcomes AKIN criteria50 51 where AKI is defined if any of the following are present: (1) an absolute rise in serum creatinine of ≥26 µmol/L within 48 hours; (2) a ≥50% increase in serum creatinine known or presumed to have occurred in the last 7 days and (3) a fall in urine output to <0.5 mL/kg/hour for 6 hours. It should be noted that patients with cirrhosis frequently have low baseline creatinine levels, as many have reduced muscle bulk and malnutrition. Therefore, some patients with creatinine levels within the normal range can have a significantly reduced glomerular filtration rate.
AKI is staged according to percentage increase in creatinine from baseline:
Stage 1: increase in sCr≥0.3 mg/dL (≥26.5 µmol/L) or an increase in sCr≥1.5 fold to 2-fold from baseline.
*Stage 1 can be split between creatinine level <1.5 mg/dL and >1.5 mg/dL, with variation in management (stage 1A and 1B, respectively).
Stage 2: increase in sCr>2 fold to threefold from baseline.
Stage 3: increase of sCr>3 fold from baseline or sCr≥4.0 mg/dL (353.6 µmol/L) with an acute increase ≥0.3 mg/dL (≥26.5 µmol/L) or initiation of renal replacement therapy (RRT).
Circulatory changes mean that those with cirrhosis are at particularly high risk of prerenal AKI, for example, from hypovolaemia secondary to diuretics, bleeding or infection.52 Patients may also develop renal or postrenal causes of AKI, or a combination of causes.49 53 It is important to identify the cause of AKI, as the underlying cause affects prognosis.
AKI caused by HRS (HRS-AKI) carries a particularly poor prognosis with mortality rates approaching 100% without treatment. HRS is caused by reduced renal blood flow due to the haemodynamic changes seen in portal hypertension, alongside stimulation of vasoconstrictive factors54 and results in AKI that persists despite appropriate fluid replacement. While HRS only affects people with advanced cirrhosis and ascites, it can coexist with other causes of renal injury such as acute tubular necrosis or chronic kidney disease, the latter—alongside a degree of proteinuria—being a relatively common finding in people with liver disease, particularly metabolic dysfunction associated steatotic liver disease.55 56 Timely and robust investigation of potential causes of AKI is, therefore, important so that appropriate treatment for all contributing factors can be rapidly initiated. HRS-AKI is currently diagnosed according to the following criteria54:
Cirrhosis with ascites.
Diagnosis of AKI according to current criteria (increase in serum creatinine ≥0.3 mg/dL (26.5 µmol/L) within 48 hours or >50% from baseline value within past 7 days).
No response after two consecutive days of diuretic withdrawal and plasma volume expansion with albumin infusion (1 g/kg body weight per day).
Absence of shock.
No current or recent use of nephrotoxic drugs (non-steroidal anti-inflammatory drugs (NSAIDs), aminoglycosides or iodinated contrast media).
No signs of structural kidney injury, as indicated by proteinuria (>500 mg per day), microhaematuria (>50 red blood cells per high‐power field) and/or abnormal renal ultrasonography.
Initial management of AKI should include suspension of all diuretics and nephrotoxic drugs and assessment of intravascular volume.57 Patients should be fluid-resuscitated with crystalloid, giving boluses of 250 mL with regular volume status reassessment aiming to achieve euvolaemia with a urine output of >0.5 mL/kg/hour based on dry weight. There is no evidence that routine use of 5% human albumin solution (HAS) is beneficial in this scenario and can in fact increase adverse events such as pulmonary oedema.45 58 However, in AKI with creatinine >1.5 g/dL (133 µmol/L), or secondary to infection, European guidelines do recommend addition of fluid expansion with 20% HAS at the dose of 1 g of albumin/kg of body weight (with a maximum of 100 g of albumin) for two consecutive days.5
While there has been recent advice from the International Club of Ascites to reduce the time to diagnose HRS-AKI from 48 to 24 hours, and to remove the need to attempt plasma expansion with albumin for 48 hours, as some believe this will delay necessary treatment with vasoconstrictors and potentially increase risk of fluid overload,56 current guidelines continue to use the criteria above.
Accurate monitoring of fluid balance and weight is essential to titrate treatment effectively. Usually 1–2 L of intravenous crystalloid will correct most losses and restore euvolaemia. If after 6 hours the target urine output is not achieved, or the patient’s clinical parameters are worsening (such as with hypotension, hyperlactataemia or other organ failures), then escalation to high-dependency care for optimal fluid management with invasive monitoring, vasopressor support or RRT may be required. Patients with cirrhosis and AKI should be considered for a trial of RRT on a case-by-case basis, accepting that only a quarter will go on to have renal recovery.59
Where HRS is diagnosed, patients should be treated with intravenous albumin and terlipressin (0.5 mg four times daily initially, with dose titrated up to 2 mg four times daily according to response), which has been shown to reduce short-term mortality in patients with HRS, but mortality not after 90 days.5 60 However, there is emerging evidence that terlipressin use may be associated with increased risk of respiratory complications and infection.61 As a result, the UK Medicines and Healthcare Regulatory Authority released advice on the use of terlipressin in 2023, noting that terlipressin should be avoided in those with creatinine above 442 µmol, ACLF grade 3 or MELD>39, unless there is clear benefit.62 Options such as reduced albumin dose or giving terlipressin via continuous infusion63 can also be considered where appropriate. Beta-blockers should be suspended in individuals with HRS.
Hyponatraemia
Hyponatraemia is common in people with advanced liver disease and is associated with poorer prognosis, as well as increased risk of HE and HRS.54 64 Hypervolaemic hyponatraemia is the most frequently encountered, where there is an expansion of extracellular volume with ascites and peripheral oedema, leading to reduced effective circulating volume. This can be caused by the adverse effects of portal hypertension on salt and water balance mechanisms, or from excessive administration of hypotonic fluids such as 5% dextrose. Conversely, hypovolaemic hyponatraemia is usually seen without ascites and oedema and is often a result of overuse of diuretics or dehydration.5 54
A careful history, assessment of fluid status, medication review and review of other biochemistry are essential to determine the cause of hyponatraemia. Diuretics should be withheld. Consider withholding other medications that may cause hyponatraemia. People who are hypovolaemic should be fluid-resuscitated with 0.9% saline, whereas those with hypervolaemic hyponatraemia should be treated with fluid restriction (unless there is AKI, where this is not appropriate) aiming to reach negative water balance—approximately 1000 mL/day.54 It is important to note that some patients with significant ascites, peripheral oedema and low serum albumin levels who appear hypervolaemic are actually intravascularly deplete, and in this scenario, treatment with intravenous albumin may improve the hyponatraemia, but not mortality.65–67 Correction of hyponatraemia should be slow (<8 mmol/L increase per 24 hours), as rapid changes in serum sodium can precipitate osmotic demyelination syndrome. Hypertonic saline can aggravate fluid overload and increase the risk of demyelination, and use should, therefore, be limited to those with life-threatening sequelae, and in a critical care environment.
Gastrointestinal bleeding
Bleeding from oesophageal or gastric varices is another life-threatening complication of cirrhosis where prompt treatment reduces mortality. Upper GI bleeding in patients with cirrhosis should be regarded as variceal until proven otherwise. Acute variceal bleeding has a high mortality rate (15% at 30 days),68 frequently from associated complications such as aspiration, infection or AKI. Therefore, patients should be managed in a high dependency environment until they have been stabilised. Airway protection is vital, particularly in those with massive bleeding or HE. Patients should be fluid resuscitated according to pulse and blood pressure.
Terlipressin/octreotide
Terlipressin (2 mg stat then 1–2 mg four times a day; 1 mg if patients are <50 kg) is recommended as a first-line agent in the management of acute variceal bleeding unless absolute or relative contraindications are found.5 69–71 Terlipressin is a vasopressin analogue that induces splanchnic vasoconstriction and reduces portal pressure.72 Meta-analyses of studies conducted several years ago have shown reduced mortality, rebleeding rate and blood transfusion requirements in patients who receive terlipressin therapy for variceal bleeding.73 74 However, adverse events were more frequent in patients receiving terlipressin, particularly in those with ischaemic heart disease and vascular disease.74 A recent study comparing 1 day vs 3 days of terlipressin (2 mg, 4 hourly) found higher rates of adverse events in those treated for the longer duration (1 day=38% vs 3 days=56%).75 Diarrhoea was the most common adverse event and overall, 42% required a dose reduction.75 Therefore, the need for ongoing terlipressin should be reviewed at the time of endoscopy, and consider stopping terlipressin after endoscopic haemostasis is achieved, particularly in those with Child-Pugh A cirrhosis. Absolute contraindications to terlipressin are hypersensitivity, pregnancy, acute respiratory distress/hypoxia, septic shock and advanced renal failure (creatinine >442 µmol/L). Relative contraindications include age >70, peripheral arterial disease, prolonged QT, cardiac arrhythmia, uncontrolled hypertension, acute coronary syndrome or previous myocardial infarction. Beta-blockers should be suspended while individuals are treated with terlipressin.
Where contraindications are present, discuss patients with the on-call GI Bleed team and consider dose reduction or treatment with octreotide 50 µg bolus followed by 50 µg/hour infusion. Octreotide also induces splanchnic vasoconstriction and has similar efficacy to terlipressin but is short acting, so requires to be administered by continuous infusion.74
Antibiotics
Bacterial infections occur in 25%–65% of patients with cirrhosis following an episode of GI bleeding and are associated with increased mortality and prolonged hospital stay.76 The use of prophylactic antibiotics is associated with a reduction in mortality, rebleeding, septicaemia and SBP, particularly in patients with Child-Pugh C cirrhosis.77–80 One study showed that outcomes were better when patients were given prophylaxis with intravenous ceftriaxone compared with oral norfloxacin (not available in the UK).79 We, therefore, recommend that patients with cirrhosis presenting with GI bleeding are treated with prophylactic antibiotics according to local protocol. Consider stopping antibiotics after endoscopic haemostasis in Child-Pugh A patients with no signs of infection, as recent studies suggest they may not benefit from antibiotic prophylaxis.81 82
Proton pump inhibitors
PPIs are frequently used in patients presenting with upper GI bleeding with evidence of improved outcomes for patients with peptic ulcers with stigmata of recent bleeding. There is no evidence of benefit of the use of PPIs in cirrhotic patients with variceal bleeding. Moreover, the use of PPI has been associated with increased risk of SBP and HE.42 83 84 Therefore, PPIs should not be given to patients prior to endoscopy, but if stigmata of non-variceal bleeding are found, PPIs may be given where appropriate.
Transfusion: red blood cells
Transfusion of red blood cells is an important treatment for some patients with variceal bleeding, but overtransfusion can increase portal pressure and increase the risk of rebleeding.85 86 There is good evidence that a restrictive strategy (transfuse with haemoglobin <7 g/L) is associated with better outcomes than more liberal approaches in haemodynamically stable patients with upper GI bleeding.87 A recent meta-analysis showed that restrictive transfusion was associated with lower rebleeding risk (RR=0.53 95% CI 0.30 to 0.94) and mortality (RR=0.52 95% CI (0.29 to 0.94) in patients with cirrhosis.88 We, therefore, recommend a restrictive transfusion target of 70–80 g/L in haemodynamically stable patients with upper GI bleeding. However, in the case of haemodynamic instability, the transfusion strategy will depend on the clinical scenario.
Correction of haemostasis
Conventional markers of clotting (prothrombin time (PT), activated partial PT (APTT) and platelets) do not reflect bleeding risk in patients with cirrhosis. Despite many patients with cirrhosis having blood tests suggesting they are anticoagulated (prolonged PT), there is actually increased risk of thromboembolic events.89 One study has evaluated the use of thromboelastography (TEG) to measure bleeding risk and guide transfusion. In that study, the use of TEG was associated with the use of significantly less unnecessary blood products, while overall outcomes were similar.90 Moreover, the infusion of large volumes of FFP and other blood products may increase portal pressure, potentially increasing the risk of rebleeding. Given the lack of evidence of benefit of transfusion of blood products to correct haemostasis in patients with cirrhosis and the attendant risk associated with these treatments, EASL89 recommends against the routine use of blood products (including platelets and FFP and prothrombin complex concentrate) targeted at improving haemostasis in haemodynamically stable patients. In a haemodynamically unstable patient, discussion with the endoscopist and haematologist regarding transfusion targets may be warranted—in patients with life-threatening bleeding, activation of the major haemorrhage protocol may be appropriate.
Tranexamic acid does not reduce mortality in upper GI bleeding and increases the risk of venous thromboembolism (VTE), so should not be used in patients with cirrhosis presenting with GI bleeding.91
Endoscopy
Endoscopy is an important procedure to make an accurate diagnosis of the source of upper GI bleeding and to deliver endotherapy. Endoscopy should be conducted once patients have been adequately resuscitated. The optimal timing of endoscopy has been a matter of debate. Overall data suggest that outcomes are poorer in patients waiting >24 hours for endoscopy.92 However, there is no benefit from an early endoscopy (<6 hours) compared with 6–24 hours.93 Prognostic scores for GI bleeding such as Glasgow-Blatchford score, AIM65, MAP, modified Glasgow-Blatchford Score and Lino score have poor positive predictive value at determining the need for emergency endoscopy.94 While specific randomised controlled data are lacking in the variceal population, repeatedly the only predictors of need for urgent endoscopy are liver disease severity (MELD score) and haemodynamic instability.95–98 We, therefore, suggest that all patients where there is a suspicion of portal hypertensive bleeding should have inpatient endoscopy within 24 hours. Discuss patients with the on-call GI bleed team who are clinically unstable, have advanced liver disease (Child Pugh B/C) or where there is other clinical concern.
Treatment with prokinetics, such as erythromycin, can be considered pre-endoscopy as this may improve visualisation at endoscopy, reduce the need for repeat endoscopy and shorten hospital stay.99
Hepatic encephalopathy
HE is another life-threatening complication of cirrhosis.100 Common precipitants include infection, electrolyte disturbance, occult bleeding, constipation or sedative drugs, and these should be sought through history, clinical examination and investigations.10 101 Serum ammonia levels are usually elevated in patients with acute HE, but this investigation is not required in patients with clinical signs of HE. In patients with delirium, a normal ammonia reliably excludes HE.100 Grade 4 HE (coma) is a medical emergency that carries a high risk of aspiration, so consider orotracheal intubation and management in a critical care unit. In conscious patients, oral lactulose (20 mL–30 mL four times per day) should be administered aiming for two soft stools per day.10 For patients with a reduced conscious level, phosphate enemas or lactulose administered by naso-gastric tube should be considered. Rifaximin treatment does not improve time to resolution of acute HE in critically ill patients.102 Where confusion is unexplained, consider a CT head examination to exclude intracranial pathology, such as subdural haematoma.
Other considerations
VTE risk
Cirrhosis is associated with alterations in the haemostatic system, which are apparent on standard coagulation tests in many patients (prolonged PT, prolonged APTT and reduced platelet count).89 Historically, these changes were believed to increase bleeding risk. However, detailed studies of coagulation factors in patients with cirrhosis indicate that there are changes in both procoagulant and anticoagulant factors, which may result in some individuals becoming hypercoagulable.103 Indeed, individuals with cirrhosis have a 1.7-fold increase in risk of VTE compared with the general population.104 Unfortunately, studies assessing outcomes of the use of VTE prophylaxis in patients with cirrhosis have been of low quality. A recent meta-analysis of available data did not demonstrate a reduction in VTE events with low-molecular-weight heparin (LMWH), but bleeding risk was not increased.105 Given the potential increased risk of VTE and significant associated morbidity, recent guidance from EASL recommends that thromboprophylaxis with LMWH can be used in patients with cirrhosis who have risk factors for VTE as it has a reasonable safety profile.89 We, therefore, recommend thromboprophylaxis with LMWH for patients admitted with DC unless they have active bleeding.
Nutrition
Malnutrition and nutritional deficiency are common in hospitalised patients with cirrhosis106 107 and are associated with increased morbidity and mortality.108 Moreover, reduced caloric intake in hospitalised patients with cirrhosis is an independent risk factor for short-term mortality.106 Therefore, malnutrition and nutritional deficiency should be identified and managed. Although widely used, common methods of nutritional screening that rely on body mass index and weight changes are not accurate in cirrhosis due to fluid retention and can also miss sarcopenic obesity which is prevalent in patients with liver disease.109 Cirrhosis-specific and active malnutrition screening such as the Royal Free Hospital Nutrition Prioritising Tool,110 upper arm anthropometric measures and indirect calorimetry demonstrate high prevalence of malnutrition and are predictive of mortality.111 Ideally, a cirrhosis-specific malnutrition screening tool should be used and those at risk, referred to a dietitian for nutrition support.111 In practice, this may not be feasible. It is reasonable, therefore, to consider all patients with DC at risk of malnutrition. Oral intake should be monitored, and if energy and protein requirements are not being met, oral nutrition supplements should be used to bridge the gap. If this remains inadequate, then enteral or parenteral nutrition should be used.111 Where possible, all patients should be referred to a dietitian for detailed nutritional assessment and care planning.
Close monitoring of electrolytes (phosphate, magnesium, potassium and calcium) is mandatory as refeeding syndrome is common, particularly in those with alcohol and substance misuse disorders.112 Low electrolyte levels should be corrected promptly.113
Ongoing patients review and treatment escalation planning
Patients with DC often have multiple, complex medical needs with an associated high risk of mortality. Therefore, a consultant review should occur at the earliest opportunity, ideally within 6 hours, but no more than 12 hours.6 In addition, a review by a specialist gastroenterologist/hepatologist should occur within 24 hours, but no later than 72 hours after admission into hospital.6 Innovative practice in some areas allows for appropriately trained healthcare professionals such as clinical nurse specialists or advanced care practitioners to deliver these specialist reviews.114 To facilitate this, it is recommended that every acute hospital should have a liver lead or ‘champion’ to try and improve care for patients with liver disease and improve pathways of care.115
Treatment escalation plans should be reviewed at the earliest opportunity. Given the high risk of mortality in patients presenting with acute DC, services should consider managing these patients in ‘Enhanced Care Beds’, which provide higher level ward-based care.116 Escalation to a higher level of care needs to be considered in all patients not responding to treatment when reviewed after 6 hours, particularly in patients with WHO performance status 0–1 prior to the recent illness, those with first presentation and individuals listed for transplant or currently under consideration/referred for liver transplant assessment.6 A discussion between senior decision-makers is recommended. Prognostic scores such as the Sepsis-related Organ Failure Assessment score (SOFA), assessed at baseline and at 48 hours, can be effective in predicting mortality in patients with cirrhosis who are admitted to ICU and may have a role in selecting patients who may benefit from ICU care.117
Some patients with advanced liver disease will have advanced care plans in place, and these should be considered when planning care. Consider early palliative care input (after taking into account patient’s and carer wishes) for patients who are not responding to treatment and escalation to higher level care would be futile.118 Patients with advanced liver disease have significant physical and psychosocial symptom burden, and early palliative care improves quality of life.53
Strategies for promotion of the bundle to improve utilisation
In the 2022 NCEPOD survey ‘Remeasuring the Units’, 69% of hospitals reported that they used the BSG/BASL DCCB, with an additional 15% reporting use of a local guideline.3 Despite these reported high usage rates, a national audit of patients admitted with DC in November 2019 found that only 11.4% had a DCCB completed, indicating a clear disparity between hospital-reported performance and real-world practice.9 The authors of that work hypothesised that reasons for low utilisation could include a lack of education around the recognition of liver disease and a lack of engagement of key stakeholders providing acute care.9 It was also suggested that the original DCCB standards were set by gastroenterologists and hepatologists without the inclusion of other care providers such as emergency medicine and acute medicine, potentially including standards that were not feasible to complete in the first 24 hours.9 Difficulties accessing the bundles and the non-user friendly interface on some electronic patient records may also be a factor.119 Use of care bundles has been shown to improve care in other areas such as sepsis, but even despite these being widely publicised, utilisation and adherence to these bundles is well below expected standards.120 Simple measures like automated reminders can modestly increase adherence with bundles.120 However, to have a major impact, implementation requires ‘full institutional support’ with the whole team behind the intervention, including regular organised education for the whole team, local implementation leaders (‘Champions’) and regular feedback on performance.121 Moreover, bundles must be fully embedded in electronic systems to automate the process. High bundle usage (90%) for the original DCCB was achieved in Newcastle with full institutional support and a financial incentive for implementation and achieving performance targets. However, full institutional support (particularly ongoing education and feedback on performance) must be maintained; otherwise, performance may later fall. Box 2 shows a list of potential strategies to promote and improve bundle utilisation with the aim of the DCCB being routine standard practice for all patients admitted with DC
Potential strategies to promote the decompensated cirrhosis care bundle (DCCB) and improve utilisation
Education
National
Royal Colleges of Physicians
Include in national and regional acute medicine teaching events days organised by the colleges.
Promote on the Medical Care-Driving Change website.
The Society for Acute Medicine
Promote at the Annual Conferences.
Deliver webinars.
Promote on website.
Promote bundle as part of acute medicine training.
British Society of Gastroenterology and British Association for the Study of the Liver
Promote at the Annual meetings.
Deliver webinars.
Promote on website.
Gastroenterology Training Program Directors to ensure the DCCB is part of teaching for doctors in training.
Incorporate DCCB into the Quality Standards Framework.
BHPG
Dissemination through the network of liver pharmacists.
Promotion at BHPG educational meetings.
BLNA
Dissemination through the network of liver nurses.
Webinar targeting nursing staff.
Royal College of Emergency Medicine
Promote via the Quality in Emergency Care Committee https://rcem.ac.uk/quality-in-emergency-care-committee-qecc/
Engage with bodies, such as the Care Quality Commission, Commissioners (such as National Health Service England), Getting it Right First Time and Health Improvement Scotland.
Develop a series of easily accessible short videos describing the bundle accessible via a QR code and giving detail around the specific aspects of the bundle (including diagnostic ascitic tap).
Promotion via social media.
Patient advocacy
Charities and third sector support, for example, British Liver Trust to promote DCCB as something patients and carers should expect to see as part of good hospital care.
Local
Incorporate the DCCB into medical student education at university.
Regular education for ED and acute medical teams, particularly when new staff rotate into a department.
An ad hoc education of staff by cirrhosis specialist nurses and liver champions.
Institutional support
Local liver champions in key places (ED, acute medical unit, gastroenterology/liver wards).
Regular audit of performance with feedback.
Incentives for performance (eg, Best Practice Tariff).
Accessing the bundle
‘Off the shelf’ DCCB bundles developed for the main electronic medical record providers.
Posters in ED and acute medical units signposting to the DCCB and supporting guidance with the use of prominently displayed QR codes.
Development of automated prompts to promote completion of the DCCB on electronic medical records.
BHPG, British Hepatology Pharmacy Group; BLNA, British Liver Nurses Association; ED, Emergency Department.
Conclusions
Multiple reports have shown that there is variability in the care of patients with DC. Use of the BSG/BASL DCCB is associated with improved patient care, but levels of usage are low across the UK. We have updated the DCCB with recent advances and improved its usability. Moreover, a strategy has been developed to promote utilisation of the bundle. We hope that by increasing use of the DCCB, this will improve the care of patients with DC and reduce variation in care.
Ethics statements
Patient consent for publication
References
Footnotes
X @stumcp, @hoganbrianj, @drdina_mansour, @sheiybani
Contributors All authors contributed by writing the original draft, reviewing and editing the final draft of the manuscript and revising the care bundle. SM was the project lead.
Disclaimer These BSG guidelines represent a consensus of best practice based on the available evidence at the time of preparation. They may not apply in all situations and should be interpreted in the light of specific clinical situations and resource availability. Further controlled clinical studies may be needed to clarify aspects of these statements, and revision may be necessary as new data appear. Clinical consideration may justify a course of action at variance to these recommendations, but we suggest that reasons for this are documented in the medical record. BSG guidelines are intended to be an educational device to provide information that may assist in providing care to patients. They are not rules and should not be construed as establishing a legal standard of care or as encouraging, advocating, requiring or discouraging any particular treatment.
Competing interests SM has received consultancy/speakers' fees from Abbvie, Allergan, BMS, Gilead, Intercept, MSD, Novo Nordisk, Norgine, Novartis and Sequana; LC has received consultancy/speakers' fees from Ipsen, Novo Nordisk and Intercept. JH has shared in Pfizer and Seres Therapeutics; ODT has received consultancy fees from Chiesi. MEDA declares research funding support from GSK. The other authors have none to declare.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.