Does "Auto-anticoagulation" Protect Against VTE in Patients with Liver Disease?

Abstract & Commentary

By David J. Pierson, MD, Professor, Pulmonary and Critical Care Medicine, Harborview Medical Center, University of Washington, Seattle. Dr. Pierson reports no financial relationships relevant to this field of study. This article originally appeared in the July 2010 issue of Critical Care Alert. It was peer reviewed by William Thompson, MD. Dr. Thompson is Associate Professor of Medicine, University of Washington; he reports no financial relationships relevant to this field of study.

Synopsis: In this retrospective study of patients hospitalized because of severe chronic liver disease, venous thromboembolism was relatively common and "auto-anticoagulation" in the form of an elevated INR had no apparent protective effect.

Source: Dabbagh O, et al. Coagulopathy does not protect against venous thromboembolism in hospitalized patients with chronic liver disease. Chest. 2010;137:1145-1149.

This study sought to determine whether the coagulopathy associated with chronic liver disease — specifically the elevated International Normalized Ratio (INR) frequently present in patients with advanced disease — is protective against venous thromboembolism (VTE) in hospitalized patients. The investigators performed a retrospective chart review of all patients with diagnosis codes for chronic liver disease and cirrhosis who were admitted to their tertiary care university hospital during an eight-year period. Excluding patients on anticoagulants at the time of admission, those with previously known VTE, and subsequent admissions in individuals hospitalized more than once, they recorded the highest INR measured during admission and also reviewed the findings of all investigations for VTE (spiral computed tomography, lower-extremity venous Doppler ultrasound, ventilation-perfusion scanning, or pulmonary angiography) that were carried out in the patients during admission. The diagnosis of chronic liver disease could be either biopsy-proven or clinical, and only hospitalizations that were primarily for this condition were included.

Of 193 patients meeting the entry criteria, three were excluded for incomplete data, leaving 190 patients in the cohort. Most of them had alcohol-related liver disease. They were divided into quartiles according to the highest recorded INR during the admission: < 1.4 (n = 47), 1.4-1.7 (n = 61), 1.7-2.2 (n = 38), and > 2.2 (n = 44). Although, as expected, patients in the higher quartiles had increasing serum bilirubin and lower albumin and platelet counts, as well as progressively worse Child-Pugh scores, the patients were similar demographically and there were no interquartile differences in either VTE risk assessment or history of VTE. Forty-three percent of the patients underwent one or more diagnostic test for VTE during hospitalization, of which venous ultrasound and spiral CT were most often used. Prophylaxis against deep venous thrombosis (DVT), either mechanical or pharmacologic, was administered in 25% of the patients, without differences in distribution among the quartiles.

In-hospital VTE was diagnosed in 12 patients (6.3%), without detectable differences according to whether DVT prophylaxis was used. One case (4.2% of 24 patients) occurred in a patient who was in Child-Pugh stage A, three cases (4.6% of 66 patients) in stage B, and eight cases (8% of 100 patients) in stage C; these differences in incidence, according to the severity of liver disease, were not statistically significant. Noting that the overall incidence of diagnosed VTE in this cohort was higher than that previously reported in cirrhotic patients, the authors conclude that, consistent with the findings of previous studies, the "auto-anticoagulation" manifested by an elevated INR in patients with chronic liver disease does not protect against VTE.


This is not the first study to demonstrate a lack of protective effect against VTE by the coagulopathy of chronic liver disease. However, it is a well-done study, and the authors make a number of good points in their discussion. This is the first study to assess the relationship between severity of coagulopathy and incidence of VTE. That an elevated INR in patients hospitalized for liver problems is not protective is emphasized by the findings that more than half of the new cases of VTE occurred in patients with INR > 1.6, and that the risk persisted even at values exceeding 2.2. The authors also documented a low rate of DVT prophylaxis, which seems inappropriate given the relatively high rate of incident VTE. And the latter was likely an underestimation, given that 57% of the patients had no evaluation for VTE.

The lesson from this study might seem obvious, but it still needs emphasis. A target-range INR is protective against DVT when it is achieved by warfarin therapy, but not when it occurs as a manifestation of liver disease- associated coagulopathy. This is a case of "bad news and bad news": An elevated INR in a patient with cirrhosis predisposes to bleeding and also does not protect against thrombosis.