Thread by @ebtapper

A celebration of paracentesis: a QI #Tweetorial

For all my twitter friends who care for inpatients with #cirrhosis

Take home messages: paras for all, FFP for none
Keywords: #SBP and the #cirrhotic #coagulopathy

First - which hospitalized patient needs a diagnostic para?

Patients with ascites and:

Spontaneous bacterial peritonitis (SBP) is a deadly condition. Appropriate tx can prevent hepatorenal syndrome and death.

How many patients are asymptomatic?

Many.

In this study from Pinzello et al, 1 in 3 patients with SBP were "silent"
aasldpubs.onlinelibrary.wiley.com/doi/epdf/10.10…

A silent but deadly condition is a problem if delay in diagnosis is associated with worse outcomes.

Uh-oh ... patients with SBP definitely prefer early diagnosis

See this terrifying chart from Kim et al in #AmJGastro nature.com/articles/ajg20…

Do you agree that paracentesis is important to do and do it early?

Act 2:
But - you say - my patient has an INR so we can't 🙄

Well...maybe you can

For one, the American Association for the Study of Liver Diseases (#AASLD) recommends doing a para on everyone except those in DIC and not giving plts or FFP beforehand
aasld.org/sites/default/…

Why so confident?

First, bleeding is super rare and unrelated to platelet count and INR

Check out this summary of a few large-ish studies

Second, bleeding has little-to-nothing to do with PT/INR

Check out this graph of bleeding time vs PT - no relation.
This is a wild study where Dr. Ewe clocked bleeding time after a laparoscopic liver biopsy. PT does not predict!
Dig Dis Sci. 1981 May;26(5):388-93

If INR correlates with bleeding in patients on coumadin, what's the deal with #cirrhosis

The INR measures some BUT NOT ALL factors in the clotting cascade.

And its missing important parts of the story

Say what?

So this is a huge topic and I will summarize but first will refer to the masters including Tripodi and Caldwell:

nature.com/articles/ajg20…
nejm.org/doi/full/10.10…

I have a few points

One - all factors are altered in cirrhosis - including liver-derived pro AND anticoagulants.

Think of this picture - things are rebalanced in cirrhosis. But tenuous.

The patient can tip both ways (clot or bleeding) depending on the context

Clin Liver Dis. 2009 Feb;13(1):1-9

Two - the clotting factors from OUTSIDE the liver are higher.

Factor 8 says "dont forget about me!"

See this table from @AGA_Gastro sciencedirect.com/science/articl…

Three - and most importantly - PT assays generally lack thrombodulin; i.e. thrombin generation is not captured.

Thrombin generation is the business end of the clot cascade

And it turns out ... it is usually preserved in cirrhosis.

And sometimes it is HIGHER!

This is what happens when you put thrombomodulin in the mix

aasldpubs.onlinelibrary.wiley.com/doi/pdf/10.100…

And if you look at thrombin generation stratified by INR, you get an interesting graph, like these ones. Clotting tendency looks like Factor V Leiden by these assays!
J Thromb Haemost. 2010 Sep;8(9):1994-2000
Hepatology. 2010 Jul;52(1):249-55

So elevated INR is not by itself a contraindication to an important - potentially life saving procedure

But it is a window into the wonderful world of cirrhotic physiology!

Did you find this helpful?

You may have noticed that we never talked about thrombocytopenia.

Would you like to?