Analysis of Severe Gastrointestinal Bleed in Patients on Anticoagulation for Atrial Fibrillation and Abnormal Endoscopic Evaluation

Background:

Widespread use of new oral anticoagulants (NOAG) for Atrial Fibrillation (A-Fib) has increased the annual incidence of gastrointestinal bleed (GIB) in recent years. These NOAG agents in our study include Rivaroxaban, Apixaban and Dabigatran. Convenience of use, quick onset, broad therapeutic window and eliminating need for frequent monitoring makes NOAGs attractive for A-Fib associated stroke prevention. Lack of reversal agents, however, does make NOAG associated severe GIB a rather difficult situation for physicians to encounter. Only supportive measures and blood transfusions exist as treatment options. This study was designed to quantify and compare all the encounters at our hospital of severe GIB in A-Fib patients on anticoagulation either with a NOAG or Coumadin.

Methods:

We first extrapolated data of all the patients discharged with diagnosis of Atrial Fibrillation and Gastrointestinal bleed at our facility from 2011 to 2015. Our inclusion criteria for study was then narrowed to only include patients on anticoagulation at home and record of colonoscopy or endoscopy performed during the admission. This left us with 208 encounters to analyze. After making a note of underlying GI pathology and anticoagulation, we looked at several variables to analyze severity of GIB. These included HAS-BLED score, hemodynamic status, hemoglobin level drop from baseline, length of stay, transfusion requirements, and reoccurrence.

Results:

While the study is still in progress, we report initial analysis of all 2014-2015 patient encounters that amount to a sample size of 76. Majority of the patients with severe GIB had underlying gastrointestinal disease (85%). Most common endoscopic findings included diverticulosis (67%), gastritis (33%), peptic ulcer disease (25%), hemorrhoids (21%) and colonic polyps (17%). Level of hemoglobin drop was 3.86 g/dL due to NOAGs vs 5.15 g/dL due to Coumadin. Length of hospital stay was 4.6 days for NOAGs versus 6.68 days for Coumadin. Among NOAGs, Dabigatran’s LOS was the highest at 6.1 vs Rivaroxaban at 4.2 and Apixaban at 3.5 days. About 36% of patients were hemodynamically unstable on admission with systolic BP < 100 mmHg and heart rate > 100 beats per minute. In this unstable group, only 20% of the cases were due to NOAGs while 80% were associated with Coumadin. Age of majority of patients with severe GIB was above 70 with only a couple in the 60s, who were mostly found to be on Coumadin. HAS-BLED score in NOAGs was on average 1.3 points lower than the Coumadin group.

Conclusions:

Overall, NOAGs appear to show a trend towards less risk of severe GIB when compared to Coumadin based on decreased length of stay, Hg drop from baseline, and less incidence in younger age group.  In Although it is difficult to reverse NOAGs associated GIB, supportive measures seem to be adequate as these agents are completely eliminated from the body after 5 half-lives or 1-2 days. Apixaban seems to be the safest of all with significantly decreased length of stay and lowest Hg drop from baseline. After diverticulosis, gastritis and peptic ulcer disease were the two most common GI pathologies we found to be associated with severe GIB due to anticoagulation. Our preliminary results confirm hospital wide safety and efficacy of NOAGs with recent data even better than expected when compared to Coumadin. Further analysis of 2011-2013 data will shed further light on these trends and allow for direct comparison of newer anticoagulants.