Rebleeding in Variceal and Nonvariceal Gastrointestinal Bleeds in Cirrhotic Patients Using Vitamin K: The LIVER-K Study.

BACKGROUND

Gastroesophageal varices are the most common cause of upper gastrointestinal bleeding (UGIB) in patients with cirrhosis. Vitamin K is commonly administered to patients presenting with UGIB and elevated international normalized ratio, despite limited evidence to support this practice.

OBJECTIVES

The primary objective was to describe the incidence of rebleeding within 30 days after vitamin K administration in patients with cirrhosis and UGIB. The secondary objective was to describe prescribing patterns for vitamin K.

METHODS

This retrospective, descriptive multicentre study involved patients with cirrhosis and UGIB who were admitted to any of the 4 adult acute care hospitals in Calgary, Alberta, from January 1, 2014, to December 31, 2016. Patients were divided into 2 groups: those who received vitamin K and those who did not.

RESULTS

A total of 370 patients met the inclusion criteria, of whom 243 received vitamin K and 127 did not. Baseline characteristics were similar between the groups. Greater proportions of patients in the vitamin K group received transfusions of packed red blood cells, fresh frozen plasma, platelets, cryoprecipitate, or prothrombin concentrate during their admissions. There was no significant difference in the duration of octreotide and pantoprazole infusions. Among patients in the vitamin K group, there were more admissions to the intensive care unit and longer lengths of stay. More patients in the no vitamin K group had esophageal varices evident on endoscopy that required endoscopic treatment. Forty of the patients (16.5%) in the vitamin K group and 7 (5.5%) in the no vitamin K group had rebleeding within 30 days of the initial bleed. The median total vitamin K dose administered was 25 mg.

CONCLUSIONS

The study results suggest that vitamin K does not reduce the incidence of rebleeding within 30 days of the initial bleed in patients with cirrhosis and UGIB.