A detailed look at cirrhosis and acute-on-chronic liver failure as we typically see it in the ICU, with Dr. Elliott Tapper, gastroenterologist and transplant hepatologist, and director of the cirrhosis program at the University of Michigan in Ann Arbor.

Takeaway lessons

* When treating liver patients, think infection, infection, infection—and understand that with good care and reversal of the underlying cause, they can get better.* Lactulose is always the first line agent for hepatic encephalopathy. Rifaximin is a potential adjunct.* Ammonia levels are prone to sampling errors and correlate poorly with clinical encephalopathy. Although it may cause fights, there’s usually little value in monitoring it.* MELD is a useful tool for prognostication, but an admittedly blunt one that fluctuates in response to many factors.* AST tends to be higher than ALT in alcoholic hepatitis, partly because alcohol metabolism depletes B6, which is needed for ALT production. Cirrhosis also increases the halflife of AST. However, alcohol use is usually most easily detected from the history, not by comparing labs.* AST or ALT in the thousands is usually due to causes other than alcoholic hepatitis. In the hospital, rule out obstruction with an ultrasound, consider drug reaction, and consider clots (e.g. portal vein thrombosis), although this last isn’t the most common.* The most common cause of extrmely elevated liver enzymes in-hospital is probably cardiogenic shock, particularly some combination of system hypotension and venous congestion. Since portal vein pressures are intrinsically low, it doesn’t take much venous congestion to counter that flow gradient, leaving only the hepatic artery for perfusion. Unfortunately, the arterial circulation usually only provides about 20-30% of hepatic perfusion, which leaves the liver vulnerable to systemic hypotension.* Exacerbations of cirrhosis and liver failure most often present due to infection. Imbalances of fluid (i.e. hyper- or hypo-volemic) are common as well.* Diagnostic paracentesis is mandatory for these patients to rule out SBP. Every hour it’s delayed may increase mortality. Empiric treatment alone is not adequate (agents like ceftriaxone will often be the wrong choice). Overall, those who undergo paracentesis are less likely to die in the hospital.* The AGA (American Gastroenterology Association), AASLD (American Association for the Study of Liver Diseases), and SIR (Society of Interventional Radiology) all agree: not only is paracentesis generally safe in cirrhotic “coagulopathy,” providers should not even consider the INR when assessing risk. This number only reflects a portion of the clotting cascade, but does not show the “rebalancing” caused by other factors, such as increased platelet activity and increased factor VIII production. The net result is generally an increased tendency for clotting, not bleeding, unless other coagulopathic processes are present.* Bleeding from paracentesis is extremely rare, and generally due to mechanical reasons (usually puncturing a vessel), not coagulopathy. Just tap them!* When diagnosing new ascites, a SAAG >1.1 is usually due to portal hypertension, but can also be caused by heart failure (congestive hepatopathy). Total protein >2.5 in ascites is more suggestive of ascites.* >250 polys and positive culture in ascites is diagnostic of SBP. Innoculating culture bottles at the bedside increases the yield compared to just filling tubes.* Look outside the abdomen for infection too, with blood cultures, chest xrays, etc.* Alcoholic hepatitis per se may (maybe) be treated with steroids, but infection, renal injury, and other factors are relative contraindications. In any case, steroid treatment is not emergent and can be instituted after a day or two for the...