Hepatic Encephalopathy: Common Precipitants, Sneaky Precipitants, and Clinical Pearls
- Mar 11th, 2019
- Setareh Mohammadie
Authors: Setareh Mohammadie, MD (Emergency Medicine Resident) and Amy Zeidan, MD (@amyjwal), University of Kentucky, Department of Emergency Medicine // Edited by: Alex Koyfman, MD (@EMHighAK) and Brit Long, MD (@long_brit)
A 62-year-old male presents to the ED via EMS following a witnessed low-impact MVC. Patient was the restrained driver with no other passengers in the vehicle. He had no loss of consciousness and is GCS 14 on arrival. Patient has no obvious signs of trauma. Physical exam reveals a distended abdomen with large veins, gynecomastia, and palmar erythema. There are no focal neurologic deficits, but he is confused. No family is present to provide past medical history. Basic labs are remarkable for hemoglobin 10.9, platelets 89, Na 127, Cr 1.89, AST 90, ALT 47, total bilirubin 2.4, INR 1.4, negative serum alcohols, ammonia level 47, UDS negative. CT head is negative. What is your next step in workup of patient’s altered mental status?
Hepatic encephalopathy (HE) is a broad spectrum of neuropsychiatric symptoms in a patient with underlying liver disease or porto-systemic shunting1. Multiple clinical criteria such as the West Haven Criteria (WHC) and Critical Flicker Frequency (CFF) can grade the level of encephalopathy present, and will be discussed in more detail below1. Along with exploration of the mechanism of hepatic encephalopathy, this post evaluates commonly missed precipitants of HE in the cirrhotic patient and treatment pearls for the emergency physician.
Increased nitrogen load, decreased toxin clearance, and neurotransmitter alteration are significant contributors to HE. Ultimately, the pathophysiology of hepatic encephalopathy is thought to originate from toxic effects of ammonia, created by bacterial metabolism of protein breakdown in the gut. Ammonia is typically degraded by the liver and excreted by the kidneys. In liver dysfunction, ammonia is allowed to accumulate and readily crosses the blood-brain-barrier, where it is converted to glutamine. This causes astrocyte swelling and cerebral dysfunction2.
Morbidity and Mortality
Hepatic encephalopathy is a costly and prevalent disease in the United States and portends a poor prognosis. The Centers for Disease Control estimates 4.9 million Americans in 2016 were affected by chronic liver disease3. A study extrapolating information from the National Health And Nutrition Examination Survey data conducted between 1999 and 2010 concluded that over 630,000 adults in the United States during that time had cirrhosis4. With sweeping epidemics of hepatitis C related to IVDU over the last decade, these numbers likely represent an underestimation of data for 2018 and beyond. Clinically evident HE has been shown to affect 30-45% of patients with cirrhosis and is the most common reason for readmission of patients hospitalized for their cirrhosis5. Cirrhotic patients with HE have an observed increase in mortality independent of their MELD score6. Patients who have grade 3 or 4 HE at the time of their liver transplant have statistically significantly lower 1-year and 5-year survival rates than their counterparts without high grade HE7. Correct identification and management of HE can significantly impact the health outcomes of over half a million people.
Because HE is frequent, debilitating, recurrent, and associated with poor survival, it is important to highlight clinical pearls and remember that it is often a diagnosis of exclusion. Maintaining a broad differential is important, and keeping these pearls in the back of your mind may help you correctly identify HE.
There are a number of scores used for both HE and cirrhosis. Understanding terminology may help when communicating with consultants/inpatient teams and initiating appropriate disposition decisions. The West Haven criteria is the current gold standard for evaluating severity of HE, although there can be some subjectivity when grading8. A prior emDocs EM@3AM provides some great background information on grading of HE. For patients who are comatose or have stage 4 West Haven HE, the Full Outline of Unresponsiveness (FOUR) score has been used2. There are cognitive and motor testing strategies used by providers, typically not in the ED, that can help assess stages as well. The severity of liver disease is often assessed using the Model for End-Stage Liver Disease (MELD) score or Child-Turcotte-Pugh (CTP) criteria.
As mentioned previously, detecting HE early can be critically important. Some of the first early clinical signs include mild personality changes like irritability, disinhibition, or apathy. Sleep disturbances can often be detected with excessive daytime sleepiness and yawning. Minor impairments can lead to falls, impaired driving, and overall decreased quality of life. Motor disturbances typically begin after this in stages 2-3. Asterixis is common but asterixis of other body parts can be detected as well including of the feet, legs, arms, tongue and eyelids. Extrapyramidal symptoms and Parkinsonian-like features may also occur. Although rare, focal neurologic deficits, most commonly hemiplegia, can occur with normal imaging8.
Workup of hepatic encephalopathy should be tailored to the patient’s presenting complaint and independent risk factors. A thorough history can often be the key and should include any new medications, history of medication compliance, changes in diet, vomiting/poor PO intake, symptoms of infection, and signs of GI bleeding. Physical exam should focus on findings consistent with HE detailed above, but also abdominal tenderness/ascites which could suggest SBP as a precipitant. We recommend diagnostic paracentesis, infectious workup, and FOBT if there is suspicion for occult GI bleed. Ammonia can be obtained but should be interpreted in the broader clinical context. Liver patients frequently exhibit coagulopathy and thrombocytopenia, so a low threshold to order CT head even in the absence of significant trauma is reasonable5,8,9.
The Use of Ammonia Levels
Though the pathophysiology of HE is thought to center heavily on accumulation of ammonia within the CNS, it is a common misconception that an elevated ammonia level can be used solely to diagnose HE. Rather, HE is a diagnosis of exclusion that should be made clinically after alternative explanations for altered mental status (AMS) are ruled out. The relation of ammonia elevation to degree to cerebral dysfunction has proven weak at best. Ammonia can lead to both false positive and false negative results. One study by Gundling and colleagues of 59 patients found results of ammonia testing were inaccurate in 40.2% of patients compared to WHC and 49.2% of patients by CFF10. In another study, elevated ammonia levels were found in 69% of a group of patients assessed to have no clinical evidence of HE. The same study found that a significant number of patients with grades 3 and 4 HE had normal or only mildly elevated ammonia levels11. The authors concluded that a single ammonia level had little utility in diagnosing HE. Yet another study found that of patients with clinical HE only 37.5% had ammonia values above the normal range, and 33.3% of patients without clinical HE had levels exceeding the upper limit of normal12. The method of collecting a blood sample for ammonia can also variably affect the sample13. Ammonia may be falsely elevated in the case of difficult venipuncture, not placing the sample immediately on ice, or delayed time to analysis due to spontaneous release of ammonia by red blood cells in standing samples14,15. Some liver experts even suggest that ammonia does not add any diagnostic or prognostic value to patients with HE, and therefore does not need to be routinely checked8. Rather than rely on ammonia level, we suggest thorough history and physical examination in conjunction with search for inciting factors for HE and alternative explanations for AMS.
Precipitants of HE
As ammonia is a significant player in the pathogenesis of HE, it’s no surprise that many of the common causes of HE can be divided into increased generation of ammonia or decreased clearance. Well known precipitants of hepatic encephalopathy include medication nonadherence, infection (SBP, UTI, PNA), GI bleed, electrolyte disturbances (hypokalemia, hyponatremia), hypoglycemia, renal failure, dehydration/intravascular hypovolemia, hepatic/portal vein thrombosis, constipation, meds (including sedatives), and large volume paracentesis16,17.
It is important to identify common precipitants andunderstand their pathogenesis in order to correctly identify, diagnose, and address the underlying precipitant. The most common precipitants of HE previously reported are GI bleed and infection18,19. A recent study by Pantham and colleagues sought to examine the frequency of most common precipitating factors resulting inhospitalizations. The most common precipitants included dehydration (46-76%), acute kidney injury (23-76%), lactulose nonadherence (50%), constipation (40%), and infections (20-42%)20. The most common causes of mortality in this group of patients was infection, GI bleed, and multi-organ dysfunction.
Patients with underlying chronic liver disease, particularly advanced liver disease, have a dysfunction of host defense mechanisms including immunological deficits in factors produced by the liver (specific complement factors and immunoglobulins)19. The body’s response to inflammation and infection is complex at baseline, and even more complex in those with liver disease9. It has been reported that neutrophil dysfunction and macrophage dysfunction occur in liver disease, contributing to a reduction in bacterial clearance by the liver19. A dysregulation of inflammatory cytokines has also been shown to exist in underlying liver disease, and are exacerbated by underlying infection9. In this setting, proinflammatory cytokines likely enhance ammonia production, contributing to the development of HE9.
Patients with underlying liver disease are more susceptible to developing an infection, with a decreased ability to fight infection, and infections aggravate liver damage in an already damaged liver. Although outside the scope of an ED provider to understand the pathogenesis of infection in HE in detail, it is important to appreciate the complexity and thus value in early diagnosis and treatment considering the high risk of mortality due to infection in cirrhotic patients. Additionally, length of stay in patients with HE triggered by infection is longer, highlighting the impact on morbidity as well as mortality20.
Finally, a few studies have examined the most common infections identified in HE, and some debate exists. For the purpose of ED diagnosis, it is important to recognize the most commonly reported including respiratory tract infection (Pneumococci, Pseudomonas), urinary tract infection (E. coli), spontaneous bacterial peritonitis (E. coli), and intestinal tract (E. coli, C. diff)19.
The previously mentioned study by Pantham demonstrated that dehydration was the most commonly recognized precipitant of HE. The hydration status of patients with liver disease is challenging to balance but vital to maintain. Patients may experience significant ascites and/or edema yet have a relative intravascular hypovolemia. Triggers of dehydration include aggressive diuresis, lactulose related diarrhea, large volume paracentesis, poor PO intake, and overdiuresis post transjugular intrahepatic portosystemic shunt (TIPS) procedure20,21,22. Additionally, many patients with liver disease have diabetes as a comorbidity, placing them at greater risk of dehydration23. Dehydration can lead to acute kidney injury, which is critical in patients with liver disease who are at risk of developing hepatorenal syndrome. Further, even in mild cases of HE, fluid intake declines and maintaining appropriate oral intake can be challenging, contributing to dehydration. In patients with liver disease who experience dehydration, there is typically a reduction in renal perfusion and glomerular filtration rate, which lead to a decrease in excretion of ammonia (and urea) and thus increased circulating levels of ammonia21.
The kidneys can be overlooked in ammonia hemostasis but are an important contributor as they maintain acid-base hemostasis. For example, they play an important role in the pathogenesis of HE in the setting of hypokalemia. Decreasing levels of potassium cause an increase in the release of hydrogen ions, thus indirectly generating ammonia. Hypokalemia can be traced back to excessive stooling from lactulose in patients with liver disease and subsequent dehydration21. A study by Bajaj and colleagues demonstrated that lactulose induced diarrhea was associated with recurrent HE and was associated with prerenal azotemia and dehydration24.
The pathogenesis of HE in GI bleed is thought to be due to hyperammonemia, specifically due to an increase in production of proteins in the gut and thus increased production of nitrogenous products including ammonia. The exact mechanisms are still under review25,26.
Constipation contributes to HE as delayed excretion can lead to increase time for ammonia to be absorbed back into circulation. Dehydration, also common in HE and liver disease, can also exacerbate constipation. Common etiologies of constipation include medication noncompliance with lactulose, opioid use, and prolonged orocecal transit time20,27.
Patients with advanced liver disease often experience hyponatremia. This is a result of activation of ADH (vasopressin) secondary to decreased arterial volumes (relative hypovolemia) from splanchnic arterial vasodilatation. This can affect intracellular water regulation and a few studies have demonstrated a relationship between hyponatremia, cirrhosis and HE28. In a study by Angeli et al, they demonstrated that as serum sodium concentration decreases, the frequency of HE increases29. A similar study by Guevara concluded that the strongest predictor in the development of HE was hyponatremia29. The underlying mechanism at play is complex and still under investigation. A simplified explanation is related to intracellular swelling or osmotic imbalance of astrocytes contributing to HE. Swelling is indirectly related to excess ammonia crossing the blood brain barrier. It is thought that hyponatremia may also contribute to this osmotic imbalance contributing to HE28.
Recognizing common precipitants early in the setting of HE is critical to both treat the underlying cause and diagnose early in the disease process. Recognizing electrolyte imbalances and dehydration early is essential and can often be addressed with simple treatments. Early identification and treatment of infection and GI bleed can reduce progression to sepsis/hemorrhage and death.
Although less common, there are a few ‘sneaky’ precipitants worth highlighting.
Studies have demonstrated that patients with HE can have more than one trigger, citing multiple precipitants as the etiology20. This is important to highlight as identification of one trigger may end the diagnostic pathway prematurely.
Diet plays a critical and sometimes overlooked role in individuals with chronic liver disease. Appropriate protein balance can be challenging, as diets too high or too low in protein can lead to elevated levels of ammonia21. Poor oral intake in general can lead to dehydration, which is a trigger for HE21. Additionally, poor PO intake and poor dietary absorption can lead to malnutrition and nutritional deficiencies, which play a role in muscle wasting21. There is some thought that muscle mass depletion is related to elevated ammonia which can worsen HE1,21. Related to diet is glucose control. Hypoglycemia has been demonstrated as a trigger for HE. This can be related to poor dietary intake, glycogen depletion from extensive cirrhosis, or underlying infection2,30. A study by Liu et al found that patients with cirrhosis and diabetes were more likely to experience decompensation events, including hepatic encephalopathy, highlighting the importance of glucose control in cirrhosis23.
In patients with cirrhosis, any exogenous ingestion of medicines metabolized by the liver need to be considered. For example, certain benzodiazepines and opioids are both metabolized by the liver and both can trigger HE20. Additionally, the role of GABA/benzodiazepine receptors in cirrhosis is currently being investigated. Some suggest there may be an increase in a benzodiazepine-like compound in cirrhosis which plays a role in HE21. As mentioned previously, constipation is a common trigger of HE, and opioids routinely cause constipation20. It may be helpful to review medications with a liver specialist before initiating, or reviewing hepatotoxic affects at https://livertox.nlm.nih.gov/.
Large volume paracentesis is sometimes performed in the ED. An important complication to be aware of is paracentesis induced circulatory dysfunction (PICD), which can lead to hyponatremia and renal impairment, both triggers of HE31. The pathology is likely multifactorial due to fluid shifts and resulting hypovolemia from arteriolar vasodilation, with subsequent downstream affects thereafter. It is generally recommended that albumin is given if >5L is removed to assist with volume expansion and therefore reduce negative side effects of PICD31.
Although not a precipitant, it is important to recognize that the both the incidence and prevalence of HE are related to severity of liver failure8. The Pantham study evaluating common precipitants of HE demonstrated that all cases of SBP were diagnosed in patients with MELD > 18, supporting a previously documented summary by Gayatri et al that patients with cirrhosis who have a MELD score of >18 are at increased risk of SBP32. Interestingly, the Pantham study found no correlation between infection and clinical severity of HE20. In patients with end stage liver disease, maintain a high index of suspicion for HE.
Finally, one may consider whether triggers of HE differ depending on the underlying cause of liver disease. A study by Wang identified GI bleed as being more common in patients with HBV cirrhosis compared to patients with alcoholic cirrhosis19. Studies by Borzio and Wang found no difference in incidence of infection regardless of cirrhosis etiology33. However few studies provided depth or clarity on this issue.
Hepatic encephalopathy may not resolve until the inciting event is treated1. Avoid benzodiazepines and opiates, as these can potentiate hepatic encephalopathy21. In the setting of current alcoholism, consider Wernicke-Korsakoff syndrome as a mimic of HE and treat empirically with thiamine. Lactulose and rifaximin are standard treatment in the ED and can be started in the clinically stable patient prior to admission1. If a patient is being discharged, caregivers need to understand to titrate lactulose to 2-4 bowel movements per day1. Avoiding hepatotoxic drugs is important, whether the patient is discharged or admitted.
There is a large spectrum of disease in HE, and the patient’s comorbidities and social situation should be taken into account when determining their disposition. If a patient is grade 1 HE, has appropriate caregivers and follow-up, they may be discharged home. For grades 3 and 4, admission to the hospital for further management is appropriate. For grade 2, the broader clinical context must be considered. Appropriate counseling to patients and caregivers on driving restriction while encephalopathy is present is crucial, though no formal guidelines exist currently for scenarios to prohibit driving8. HE was classically thought to be a reversible cognitive dysfunction, but some studies suggest that neurologic deficits can persist even after appropriate therapy21. Patients who struggle with severe HE may be persistently encephalopathic and fluctuate with symptoms from day to day, similar to dementia. It is important to counsel family members on expectations and goals of treatment, which will vary on a case to case basis.
The patient’s family arrives and confirms decompensated liver disease secondary to former alcohol abuse. His wife reports he is typically well controlled on lactulose, but several family members have been sick at home with vomiting for the past few days. The patient has had poor PO intake and has a history of esophageal varices. He has been behaving normally for the past few days other than seeming sleepy during the day. FOBT, diagnostic paracentesis, UA, and UDS are negative. Despite a minimally elevated ammonia level, he is suffering from hepatic encephalopathy induced by volume depletion and appears to have AKI likely secondary to dehydration. He remains hemodynamically stable and is admitted to medicine for correction of electrolytes, AKI resolution, and improvement in his level of encephalopathy.
- HE is common in patients with cirrhosis and is the most common reason for readmission. Patients with cirrhosis who experience HE have an increase in mortality demonstrating the seriousness of this common disease.
- Clinicians should rely on the history, exam, and clinical suspicion when diagnosing HE, rather than ammonia levels.
- Causes of HE are typically due to increased generation of ammonia or decreased clearance. The most common precipitants include infection, GI bleed, dehydration, AKI, lactulose nonadherence, and constipation.
- Patients with cirrhosis are more susceptible to developing infections due to a dysfunction of host defense mechanisms as well as dysregulation of inflammatory cytokines.
- Diet and hydration status are critical in patients with cirrhosis. Dehydration is a common precipitant of HE due to use of diuretics, lactulose, large volume paracentesis, poor PO intake, and difficult medication titration post TIPS.
- Less common precipitants to keep in mind when evaluating for HE include concomitant precipitants, such as infection and dehydration or electrolyte abnormality.
- Be wary of new medications in cirrhotic patients. Consider reviewing side effects of all medications used in patients with cirrhosis before prescribing to assess for hepatotoxicity: https://livertox.nlm.nih.gov/.
- Suraweera D, Sundaram V, Saab S. Evaluation and management of hepatic encephalopathy: Current status and future directions. Gut and Liver 2016;10(4):509–19.
- Wijdicks EF. Hepatic Encephalopathy. New England Journal of Medicine 2016;375:1660–70.
- Center for Health Statistics N. Table A-4. Selected diseases and conditions among adults aged 18 and over, by selected characteristics: United States, 2016. 2016;8:1–9. Available from: https://www.cdc.gov/nchs/data/series/sr_02/sr02_175.pdf
- Scaglione S, Kliethermes S, Cao G, Shoham D, Durazo R, Luke A, Volk ML. The epidemiology of cirrhosis in the United States a population-based study. Journal of Clinical Gastroenterology 2015;49(8):690–6.
- Elwir S, Rahimi RS. Hepatic Encephalopathy: An Update on the Pathophysiology and Therapeutic Options. Journal of Clinical and Translational Hepatology 2017;5:142–51.
- Stewart CA, Malinchoc M, Kim WR, Kamath PS. Hepatic Encephalopathy as a Predictor of Survival in Patients With End-Stage Liver Disease. Liver Transplantation 2007;13:1366–71.
- Wong RJ, Aguilar M, Gish RG, Cheung R, Ahmed A. The Impact of Pretransplant Hepatic Encephalopathy on Survival Following Liver Transplantation. Liver Transplantation 2015;21(873–880).
- Vilstrup H, Amodio P, Bajaj J, Cordoba J, Ferenci P, Mullen KD, Weissenborn K, Wong P. Hepatic Encephalopathy in Chronic Liver Disease: 2014 Practice Guideline by the European Association for the Study of the Liver and the American Association for the Study of Liver Diseases. Journal of Hepatology 2014;61(3):642–59.
- Atluri DK, Prakash R, Mullen KD. Pathogenesis, Diagnosis, and Treatment of Hepatic Encephalopathy. Journal of Clinical and Experimental Hepatology 2011;1(2):77–86.
- Gundling F, Zelihic E, Seidl H, Haller B, Umgelter A, Schepp W, Dodt C. How to diagnose hepatic encephalopathy in the emergency department. Annals of Hepatology 2013;12(1):108–14.
- Ong JP, Aggarwal A, Krieger D, Easley KA, Karafa M, Van Lente F, Arroliga AJ, Mullen KD. Correlation between ammonia levels and the severity of hepatic encephalopathy. American Journal of Medicine 2003;114(3):188–93.
- Kundra A, Jain A, Banga A, Bajaj G, Kar P. Evaluation of plasma ammonia levels in patients with acute liver failure and chronic liver disease and its correlation with the severity of hepatic encephalopathy and clinical features of raised intracranial tension. Clinical Biochemistry 2005;38(8):696–9.
- Ge PS, Runyon BA. Serum ammonia level for the evaluation of hepatic encephalopathy. Journal of the American Medical Association 2014;312(6):643–4.
- Hawke L. Ammonia. The Association for Clinical Biochemistry and Laboratory Medicine. 2018;
- Barsotti R. Measurement of ammonia in blood. Journal of Pediatrics 2001;138:19–20.
- Smyrnios NA, Irwin RS. Current concepts in the pathophysiology and management of near-drowning. Journal of Intensive Care Medicine 1991;6(1):26–35.
- Cichoz-Lach H, Michalak A. Current pathogenetic aspects of hepatic encephalopathy and noncirrhotic hyperammonemic encephalopathy. World Journal of Gastroenterology 2013;19(1):26–34.
- Devrajani BR, Shah SZA, Devrajani T, Kumar D. Precipitating factors of hepatic encephalopathy at a tertiary care hospital Jamshoro, Hyderabad. Journal of the Pakistan Medical Association 2009;59(10):683–6.
- Wang QM, Ji Q, Duan ZJ, Zhang M, Chang QY. A study on the position and etiology of infection in cirrhotic patients: A potential precipitating factor contributing to hepatic encephalopathy. Experimental and Therapeutic Medicine 2013;6:584–90.
- Pantham G, Post A, Venkat D, Einstadter D, Mullen KD. A New Look at Precipitants of Overt Hepatic Encephalopathy in Cirrhosis. Digestive Diseases and Sciences 2017;62(8):2166–73.
- Frederick R. Current Concepts in the Pathophysiology and Management of Hepatic Encephalopathy. Gastroenterology & Hepatology 2011;7(4):222–33.
- Jalan R, Kapoor D. Reversal of diuretic-induced hepatic encephalopathy with infusion of albumin but not colloid. Clinical Science 2004;106(5):467–74.
- Liu TL, Trogdon J, Weinberger M, Fried B, Barritt AS. Diabetes Is Associated with Clinical Decompensation Events in Patients with Cirrhosis. Digestive Diseases and Sciences 2016;61(11):3335–45.
- Bajaj JS, Sanyal AJ, Bell D, Gilles H, Heuman DM. Predictors of the recurrence of hepatic encephalopathy in lactulose-treated patients. Alimentary Pharmacology and Therapeutics 2010;31(9):1012–7.
- Olde Damink SWM, Jalan R, Deutz NEP et al. The kidney plays a major role in the hyperammonemia seen after simulated or actual GI bleeding in patients with cirrhosis. Hepatology 2003;37(6):1277–85.
- Olde Damink SWM, Dejong CHC, Deutz NEP, Van Berlo CLH, Soeters PB. Upper gastrointestinal bleeding: An ammoniagenic and catabolic event due to the total absence of isoleucine in the haemoglobin molecule. Medical Hypotheses 1999;52(6):515–9.
- Fukui H, Wiest R. Changes of Intestinal Functions in Liver Cirrhosis. Inflammatory Intestinal Diseases 2016;1(1):24–40.
- Guevara M, Baccaro ME, Torre A, et al. Hyponatremia is a risk factor of hepatic encephalopathy in patients with cirrhosis: A prospective study with time-dependent analysis. American Journal of Gastroenterology 2009;104(6):1382–9.
- Angeli P, Wong F, Watson H, et al. Hyponatremia in cirrhosis: Results of a patient population survey. Hepatology 2006;44(6):1535–42.
- Viplove S. Recurrent hypothermia and hypoglycemia as the initial presentation of hepatitis C. American Journal of Gastroenterology 2010;105:S262.
- Lindsay AJ, Burton J, Ray CE. Paracentesis-Induced Circulatory Dysfunction : A Primer for the Interventional Radiologist. 2014;1(212):276–8.
- Gayatri AA, Suryadharma IG, Wibawa ID. The relationship between a model of end stage liver disease score (MELD score) and the occurrence of spontaneous bacterial peritonitis in liver cirrhotic patients. Acta medica Indonesiana 2007;39(2):75–8.
- Borzio M, Salerno F, Piantoni L, Cazzaniga M, Angeli P, Bissoli F, Boccia S, Colloredo-Mels G, Fornaciari G, Marenco G, Pistara R, Salvagnini M, Sangiovanni A. Bacterial infection in patients with advanced cirrhosis: A multicentre prospective study. Digestive and Liver Disease 2001;33(1):41–8.