2.1.3. Additional Measures

#### No Need for Correction of Coagulopathy

Hospitalized patients with decompensated cirrhosis have severe coagulopathy [21–23]. However, a prolonged prothrombin time does not reflect an increased bleeding tendency in these patients [21,23], and correction of INR by fresh frozen plasma should not be performed [5–8]. Not surprisingly, administration of recombinant FVII, which can correct prolongation of INR, was not associated with additional benefit compared with standard of care in an individual patient data meta-analysis of two RCTs [24]. Administration of plasma to correct coagulopathy in cirrhosis with bleeding is a very common practice [25]; however, this practice not only is ineffective, but is also likely harmful [26]. In a recent, multicenter cohort study administration of fresh frozen plasma in cirrhosis with VH was independently associated with increased risks of 42-day mortality (primary outcome, OR: 9.41, 95% CI: 3.71–23.90), failure to control bleeding at five days (OR: 3.87, 95% CI: 1.28–11.70) and length of stay (adjusted OR: 1.88, 95% CI: 1.03–3.42) (secondary outcomes) [27]. No specific data exist regarding the management of severe thrombocytopenia in the setting of VH, and therefore, no recommendation can be made. In patients without chronic liver disease, desmopressin increases levels of plasmatic Von Willebrand factor/procoagulant Factor VIII, and its use was associated with reduced bleeding time in an old study including compensated patients [28]. However, in a subsequent RCT no difference in control of VH was observed between patients randomized to terlipressin alone vs. patients treated with terlipressin plus desmopressin [29]. Therefore, desmopressin is not currently recommended.

Limited Usefulness of PPIs

As peptic ulcers are the source of bleeding in ~30% of patients with cirrhosis presenting with GI bleeding [30], intravenous PPIs should be initiated as soon as possible. However, when portal hypertensive bleeding is confirmed at endoscopy, discontinuation of PPIs may be considered as they have shown no efficacy in this clinical setting. Limited evidence suggested that a short-term use (10 days) of PPIs might reduce banding ulcer size [31], however, this was not associated with a significant reduction of bleeding risk. PPIs in decompensated cirrhosis are associated with significantly increased risks of hepatic encephalopathy, bacterial infection, and readmission at 30-days [32–34]. In a landmark analysis including 1198 patients from three RCTs evaluating the use of satavaptan in patients with cirrhosis and ascites, Dam et al. demonstrated that the use of PPIs was associated with a significantly increased risk of encephalopathy (OR: 1.88, 95% CI: 1.21–1.91) and spontaneous bacterial peritonitis (OR: 1.72, 95% CI: 1.10–2.69) during follow-up [33]. Recent data with extended period of follow-up confirmed that regular use of PPIs not only is associated with increased risk of spontaneous bacterial peritonitis, but also predicts liver-related mortality independent of MELD and stage of cirrhosis (OR: 2.01, 95% CI: 1.38–2.93) [34]. Therefore, their use should not be extended past discharge.
