A Comparison of the Rate of Gastrointestinal Bleeding in Patients Taking Non-Vitamin K Antagonist Oral Anticoagulants or Warfarin

David J. Cangemi, Timothy Krill, Rick Weideman, Daisha J. Cipher, Stuart J. Spechler, Linda A. Feagins

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Objectives:Early reports suggested that the risk of gastrointestinal bleeding (GIB) was higher for patients on non-vitamin K antagonist oral anticoagulants (NOACs) than for those on warfarin. We compared the incidence of GIB in our patients on NOACs with those on warfarin.Methods:We used our VA pharmacy database to identify patients taking NOACs (dabigatran, rivaroxaban, and apixaban) or warfarin between January 2011 and June 2015, and used the VistA system to identify those who were hospitalized for GIB. We included only patients with clinically significant GIB, defined as documented GI blood loss with a hemoglobin drop ≥2 g/dl, hemodynamic instability, and/or need for endoscopic evaluation, angiography, or surgery.Results:We identified 803 patients on NOACs and 6,263 on warfarin. One hundred and fifty-eight patients on warfarin had GIB (2.5%), compared with only five patients (0.6%) on NOACs (odds ratio=4.13; 95% confidence interval: 1.69–10.09). Blood transfusion for GIB was significantly more common in patients on warfarin than on NOACs (64.6% vs. 20%, P=0.04). Within 90 days of GIB hospitalization, 12 patients (7.6%) in the warfarin group died, whereas there were no deaths in the NOAC group.Conclusions:In our patients, the incidence of GIB for those on warfarin was more than four times that for those on NOACs. Blood transfusions for GIB were more common in warfarin patients, and no NOAC patients died of GIB. In contrast to early reports, our findings suggest that the risk of GIB and subsequent complications is considerably lower for patients on NOACs than for patients on warfarin.Am J Gastroenterol advance online publication, 28 February 2017; doi:10.1038/ajg.2017.39.

Original languageEnglish (US)
JournalAmerican Journal of Gastroenterology
DOIs
StateAccepted/In press - Feb 28 2017

Fingerprint

Warfarin
Anticoagulants
Hemorrhage
Blood Transfusion
Incidence
Publications
Angiography
Hemoglobins
Hospitalization
Hemodynamics
Odds Ratio

ASJC Scopus subject areas

  • Gastroenterology

Cite this

A Comparison of the Rate of Gastrointestinal Bleeding in Patients Taking Non-Vitamin K Antagonist Oral Anticoagulants or Warfarin. / Cangemi, David J.; Krill, Timothy; Weideman, Rick; Cipher, Daisha J.; Spechler, Stuart J.; Feagins, Linda A.

In: American Journal of Gastroenterology, 28.02.2017.

Research output: Contribution to journalArticle

@article{19f02b7a5883426db2a91273c1232082,
title = "A Comparison of the Rate of Gastrointestinal Bleeding in Patients Taking Non-Vitamin K Antagonist Oral Anticoagulants or Warfarin",
abstract = "Objectives:Early reports suggested that the risk of gastrointestinal bleeding (GIB) was higher for patients on non-vitamin K antagonist oral anticoagulants (NOACs) than for those on warfarin. We compared the incidence of GIB in our patients on NOACs with those on warfarin.Methods:We used our VA pharmacy database to identify patients taking NOACs (dabigatran, rivaroxaban, and apixaban) or warfarin between January 2011 and June 2015, and used the VistA system to identify those who were hospitalized for GIB. We included only patients with clinically significant GIB, defined as documented GI blood loss with a hemoglobin drop ≥2 g/dl, hemodynamic instability, and/or need for endoscopic evaluation, angiography, or surgery.Results:We identified 803 patients on NOACs and 6,263 on warfarin. One hundred and fifty-eight patients on warfarin had GIB (2.5{\%}), compared with only five patients (0.6{\%}) on NOACs (odds ratio=4.13; 95{\%} confidence interval: 1.69–10.09). Blood transfusion for GIB was significantly more common in patients on warfarin than on NOACs (64.6{\%} vs. 20{\%}, P=0.04). Within 90 days of GIB hospitalization, 12 patients (7.6{\%}) in the warfarin group died, whereas there were no deaths in the NOAC group.Conclusions:In our patients, the incidence of GIB for those on warfarin was more than four times that for those on NOACs. Blood transfusions for GIB were more common in warfarin patients, and no NOAC patients died of GIB. In contrast to early reports, our findings suggest that the risk of GIB and subsequent complications is considerably lower for patients on NOACs than for patients on warfarin.Am J Gastroenterol advance online publication, 28 February 2017; doi:10.1038/ajg.2017.39.",
author = "Cangemi, {David J.} and Timothy Krill and Rick Weideman and Cipher, {Daisha J.} and Spechler, {Stuart J.} and Feagins, {Linda A.}",
year = "2017",
month = "2",
day = "28",
doi = "10.1038/ajg.2017.39",
language = "English (US)",
journal = "American Journal of Gastroenterology",
issn = "0002-9270",
publisher = "Nature Publishing Group",

}

TY - JOUR

T1 - A Comparison of the Rate of Gastrointestinal Bleeding in Patients Taking Non-Vitamin K Antagonist Oral Anticoagulants or Warfarin

AU - Cangemi, David J.

AU - Krill, Timothy

AU - Weideman, Rick

AU - Cipher, Daisha J.

AU - Spechler, Stuart J.

AU - Feagins, Linda A.

PY - 2017/2/28

Y1 - 2017/2/28

N2 - Objectives:Early reports suggested that the risk of gastrointestinal bleeding (GIB) was higher for patients on non-vitamin K antagonist oral anticoagulants (NOACs) than for those on warfarin. We compared the incidence of GIB in our patients on NOACs with those on warfarin.Methods:We used our VA pharmacy database to identify patients taking NOACs (dabigatran, rivaroxaban, and apixaban) or warfarin between January 2011 and June 2015, and used the VistA system to identify those who were hospitalized for GIB. We included only patients with clinically significant GIB, defined as documented GI blood loss with a hemoglobin drop ≥2 g/dl, hemodynamic instability, and/or need for endoscopic evaluation, angiography, or surgery.Results:We identified 803 patients on NOACs and 6,263 on warfarin. One hundred and fifty-eight patients on warfarin had GIB (2.5%), compared with only five patients (0.6%) on NOACs (odds ratio=4.13; 95% confidence interval: 1.69–10.09). Blood transfusion for GIB was significantly more common in patients on warfarin than on NOACs (64.6% vs. 20%, P=0.04). Within 90 days of GIB hospitalization, 12 patients (7.6%) in the warfarin group died, whereas there were no deaths in the NOAC group.Conclusions:In our patients, the incidence of GIB for those on warfarin was more than four times that for those on NOACs. Blood transfusions for GIB were more common in warfarin patients, and no NOAC patients died of GIB. In contrast to early reports, our findings suggest that the risk of GIB and subsequent complications is considerably lower for patients on NOACs than for patients on warfarin.Am J Gastroenterol advance online publication, 28 February 2017; doi:10.1038/ajg.2017.39.

AB - Objectives:Early reports suggested that the risk of gastrointestinal bleeding (GIB) was higher for patients on non-vitamin K antagonist oral anticoagulants (NOACs) than for those on warfarin. We compared the incidence of GIB in our patients on NOACs with those on warfarin.Methods:We used our VA pharmacy database to identify patients taking NOACs (dabigatran, rivaroxaban, and apixaban) or warfarin between January 2011 and June 2015, and used the VistA system to identify those who were hospitalized for GIB. We included only patients with clinically significant GIB, defined as documented GI blood loss with a hemoglobin drop ≥2 g/dl, hemodynamic instability, and/or need for endoscopic evaluation, angiography, or surgery.Results:We identified 803 patients on NOACs and 6,263 on warfarin. One hundred and fifty-eight patients on warfarin had GIB (2.5%), compared with only five patients (0.6%) on NOACs (odds ratio=4.13; 95% confidence interval: 1.69–10.09). Blood transfusion for GIB was significantly more common in patients on warfarin than on NOACs (64.6% vs. 20%, P=0.04). Within 90 days of GIB hospitalization, 12 patients (7.6%) in the warfarin group died, whereas there were no deaths in the NOAC group.Conclusions:In our patients, the incidence of GIB for those on warfarin was more than four times that for those on NOACs. Blood transfusions for GIB were more common in warfarin patients, and no NOAC patients died of GIB. In contrast to early reports, our findings suggest that the risk of GIB and subsequent complications is considerably lower for patients on NOACs than for patients on warfarin.Am J Gastroenterol advance online publication, 28 February 2017; doi:10.1038/ajg.2017.39.

UR - http://www.scopus.com/inward/record.url?scp=85014057415&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85014057415&partnerID=8YFLogxK

U2 - 10.1038/ajg.2017.39

DO - 10.1038/ajg.2017.39

M3 - Article

C2 - 28244496

AN - SCOPUS:85014057415

JO - American Journal of Gastroenterology

JF - American Journal of Gastroenterology

SN - 0002-9270

ER -