“…Reichert and colleagues did demonstrate that if the INR was greater than 1.5, then the risk of bleeding was increased, but this was only for bleeding classified as of minor severity (Reichert et al, 2014) and stratification by type of anticoagulation has demonstrated that unfractionated heparin results in a higher bleeding risk than low molecular weight heparin in cirrhotic patients (Intagliata et al, 2014). Consistent with this, evaluation of a prophylactic dose of enoxaparin (4000 IU daily) to prevent PVT in patients with Child B/ C cirrhosis resulted in no increased risk of bleeding events in these patients when heparin was given for a period of 48 weeks when compared to a matched control group patients treated with placebo alone (Villa et al, 2012).It is probable that cirrhotics are more sensitive to unfractioned heparin (Potze et al, 2013), and therapeutic doses of this type of heparin also resulted in a significant drop in haemoglobin and platelet counts in cirrhotic patients (Fuentes et al, 2015); the authors of this study conclude that the drop in haemoglobin likely reflected 8 bleeding events during therapy, whilst the drop in platelet count could have represented heparin-induced thrombocytopenia. In contrast, administration of therapeutic doses of LMWH given over prolonged periods (mainly in the context of treating extrahepatic portal vein thrombosis) appears to be safe, with no significantly increased risk of bleeding (even in the presence of advanced fibrosis) when given alone (Amitrano et al, 2010; Francoz et al, 2005, Delagado et al 2012, Maruyama et al, 2012, Werner et al, 2013 or in conjunction with transjugular intrahepatic portosytemic shunt insertion (Senzolo et al, 2012).…”