Suboptimal use of pharmacological venous thromboembolism prophylaxis in cirrhotic patients

Cirrhosis was previously perceived as a haemorrhagic disease state due to frequent associations with coagulopathy and bleeding. However, the coagulopathy of cirrhosis is complex with defects in both procoagulant and anticoagulant factors. Derangements in common laboratory indices of coagulation do not accurately reflect bleeding risk or protection from thrombotic events.

To assess the rate of pharmacological prophylaxis for venous thromboembolism (VTE) among hospital inpatients with cirrhosis and analyse factors associated with prophylaxis being inappropriately withheld.

A retrospective cohort study was performed in a tertiary teaching hospital. Patients included were admitted for greater than 48 h with discharge diagnosis codes corresponding to chronic liver disease and/or cirrhosis. The use of VTE chemoprophylaxis with enoxaparin was assessed in cirrhotic patients and non-cirrhotic controls. Patient data collected included contraindications to prophylaxis, known high-risk varices, international normalised ratio (INR), creatinine, bilirubin, haemoglobin and platelet count.

Of 108 patients with cirrhosis eligible for VTE prophylaxis, 61 (56.5%) received prophylaxis compared to 104 (96.3%) non-cirrhotic patients. Platelets and INR were significantly different between those who did and did not receive VTE prophylaxis. On multivariate analysis, platelet count and INR were independent predictors for VTE not being administered.

The administration of chemoprophylaxis in accordance with the hospital guidelines was suboptimal in patients with cirrhosis. Platelet count and INR were independent predictors of prophylaxis use. Our results suggest persistent misperceptions that prolonged INR and thrombocytopenia predict bleeding risk in cirrhosis.