Challenges in the Diagnosis of Hepatic Encephalopathy

Arun B. Jesudian, MD: Perfect lead in into this spectrum of manifestations that patients might have. Elliot, maybe you can talk about that. Neuropsychiatric type symptoms, what do you see, what do patients complain of?

Elliot B. Tapper, MD: I think the key thing is to listen to the patients with cirrhosis and to actively look for cues that they could be developing hepatic encephalopathy [HE]. And one of the first things that patients are going to tell you is that there’s a reversal of their sleep/wake cycle. They’re having difficulty sleeping, but they’re sleeping during the day. And then another thing [that] is exceedingly common and in fact, highly predictive of the presence of hepatic encephalopathy on formal neuropsychiatric testing, is stumbling or falls. Patients with hepatic encephalopathy have a prevalence of falls in a given year of 40%. And when they fall, they’re going to break their hip, and when they break their hip, it’s a disaster for them. So you’re looking for difficulties with balance, you’re looking for difficulties with sleep, they’re going to be more irritable and impatient with themselves, eating less than usual. Those sorts of changes in patient reported outcomes or executive function are key to those early stages. As we progress through the later stages, something perhaps worth talking about are the kinds of things that you would pick up more likely in somebody with waxing and waning mental status, [who] might end up in the emergency [department].

Arun B. Jesudian, MD: In that type of patient when you’re concerned, this is a very common question I get, should you be sending an ammonia level in the blood work because you mentioned ammonia can be implicated in the pathogenesis of HE? Is that a helpful test, do you think?

Elliot B. Tapper, MD: In practice, ammonia is not a helpful test and there’s several reasons for that. One, recall that there’s other things going on in the body of a patient with cirrhosis in addition to that ammonia level. So they can be fully encephalopathic and have a relatively normal level if they’re in the setting of high inflammatory burden with an infection like a UTI [urinary tract infection] or cellulitis. Remember, too, how do you get that ammonia level? Well, you can’t use a tourniquet. If you use a tourniquet, you’re going to increase the ammonia level 10%, 15% and your values are useless. You’ve got to put it immediately on ice. Preferably, you’ve taken it out of the artery, and then the [laboratory] has to stop everything that they’re doing and have to run it immediately. Does this happen in practice? No. So, for that reason, ammonia is more of an academic thought. In practice, [it] just doesn’t help us.

Steven L. Flamm, MD: The guidelines are very clear about this, Arun. Serum ammonia levels are not to be checked routinely in practice. I’ll give you an example. I had a patient not long ago come to me who is an accountant, never had any issues with his mentation, no work-related issues, nothing. And he told me his doctor for 2 years every week was checking an ammonia level, and he was on lactulose. And patients exaggerate sometimes. Every week; it was probably 4 times a year. I didn’t have the records. I get the records about a couple days later, and he literally had over 100 checks of serum ammonia levels over the last few years. It was every week. And when the ammonia level was a little bit high, even though he had no symptoms, he was put on lactulose and his lactulose dosage was adjusted.

So the first thing I did, of course, was stop his lactulose, which made him very happy. But this is an example of how ammonia levels should not be used. The only time they say consider them, us getting ammonia levels, is when a patient is confused, and you’re confused as to [why] the patient is confused. So, if both of you are confused, that’s a good time to get an ammonia level or to consider it because then if the ammonia level is markedly elevated, it might push you more towards a diagnosis of HE. For instance, if a patient has seizures and you’re not sure if at home they’re having postictal periods or if they’re having HE, maybe draw an ammonia level then. But the bottom line is in a cirrhotic patient, it’s not necessary to check ammonia levels and really it shouldn’t be done.

Arun B. Jesudian, MD: It’s also important, as we focus on higher value health care, [to realize] it’s an unnecessary test with a cost attached to it, so that’s one way we can minimize unnecessary spending.