Pre-Endoscopy Omeprazole for GI Bleeding Recommended

Published on: 

HONG KONG -- The pre-endoscopy use of high-dose infusions of proton-pump inhibitors improves outcomes for peptic-ulcer hemorrhage and other upper GI bleeding, investigators here confirmed.

HONG KONG, April 19 -- The pre-endoscopy use of high-dose infusions of proton-pump inhibitors improves outcomes for peptic-ulcer hemorrhage and other upper GI bleeding, investigators here confirmed.

Omeprazole (Prilosec) infusion, given after stabilization to urgent GI-bleeding patients, reduced the need for treatment during endoscopy the next day, such as epinephrine injection and pulses with a heater probe (19.1% vs. 28.4% placebo, P=0.007), said James Y. Lau, M.D., of Prince of Wales Hospital here, and colleagues.

In the placebo-controlled study of 638 consecutive patients (631 had endoscopy), pre-endoscopy omeprazole also shortened hospital stays, Dr. Lin and colleagues reported in the April 19 issue of the New England Journal of Medicine.

The findings reaffirmed that optimal acid suppression facilitates clot formation over arteries in bleeding peptic ulcers, and for those awaiting endoscopy stabilizes clots, and initiates healing, they wrote. "On endoscopy, fewer cases of actively bleeding peptic ulcers were seen among patients who had received omeprazole than among those who had received placebo," they added.

They did not propose using high-dose proton-pump inhibitors as a replacement for early endoscopy, they wrote. "In patients awaiting endoscopy, however, we recommend the preemptive use of high-dose intravenous omeprazole."

Some consensus guidelines already recommend post-endoscopy proton-pump inhibitor infusion to prevent bleeding recurrence. In clinical practice, though, infusion is often started before endoscopy despite little evidence in the literature for this approach, the investigators said.

So they conducted a double-blind, placebo-controlled trial of the patients who arrived at the emergency department of Dr. Lau's hospital from February 2004 to July 2005 with signs of upper-GI bleeding. These signs included black, tarry stools or bloody vomit with or without hypotension.

These with refractory hypertensive shock or long-term aspirin use for cardioprotection were excluded.

Patients were randomized to IV infusion of omeprazole or placebo. Each was given as an 80-mg bolus injection starting at admission followed by continuous infusion at 8 mg per hour until endoscopy the next morning. Afterward, for those who required ulcer hemostasis, omeprazole infusion was reinitiated at the same rate for 72 hours followed by 40 mg orally each day for eight weeks.

The source of bleeding for most patients was peptic ulcer (59.6% of the omeprazole group and 59.9% of the placebo group). Infusion duration before endoscopy was 14.7 hours and 15.2 hours in the respective groups.

For the primary endpoint, significantly fewer patients needed endoscopic therapy at the first endoscopic examination in the omeprazole than the placebo group (19.1% versus 28.4%, relative risk 0.67, 95% confidence interval 0.51 to 0.90, P=0.007).

The same was true among those with peptic-ulcer bleeding (22.5% versus 36.8%, relative risk with omeprazole 0.61, 95% CI 0.44 to 0.84, P=0.002).

These treatment differences included:

  • Lower mean volume of epinephrine injected to stop bleeding during endoscopy with omeprazole (9.26.0 ml versus 10.57.0 ml, P=0.31).
  • Lower median number of heater probe pulses used during endoscopic treatment with omeprazole (five versus. six, P=0.01).
  • Less frequent active bleeding on first endoscopy among patients with peptic ulcers on omeprazole (6.4% versus 14.7%, P=0.01).
  • More ulcers with clean bases in the omeprazole group (64.2% versus 47.4%, P=0.001).

Among the secondary findings, the researchers reported:

  • No difference in urgent endoscopy between groups (seven in the omeprazole group versus six in the placebo group, P=0.79).
  • No difference in blood transfusions (1.542.41 versus 1.883.44 units, P=0.12).
  • No difference in emergency surgery for failure to achieve hemostasis endoscopically (1.6% versus 2.1%, P=1.00).
  • Similar recurrent bleeding rates within 30 days of endoscopic therapy (3.5% versus 2.5%, P=0.49).
  • Shorter hospital stay with omeprazole (median three days for both, P=0.003, 60.5% versus 49.2% less than three days, P<0.005).
  • Identical 30-day all-cause mortality (2.5% for both, P=0.78).

No serious adverse events were judged to be related to either infusion.

However, the results may have limited generalizability because the study excluded potentially higher risk patients on long-term aspirin therapy (another study at the hospital was using this patient population). Also, they may be less applicable to other areas where variceal bleeding is a more prevalent cause of upper gastrointestinal bleeding than peptic ulcer.

Nonetheless, endoscopy within 24 hours of admission for patients with upper gastrointestinal bleeding "should be the standard of care in most hospitals," Dr. Lau and colleagues said.

They added that they expect pre-endoscopy omeprazole administration to be cost effective as post-endoscopy use has been shown to be in other studies.