Correspondence

Omeprazole before Endoscopy in Patients with Gastrointestinal Bleeding

N Engl J Med 2007; 357:303-304July 19, 2007

Article

To the Editor:

Lau et al. (April 19 issue)1 suggest the infusion of high-dose omeprazole before endoscopy to reduce the need for endoscopic therapy in patients with non–aspirin-related bleeding peptic ulcers. We find it difficult to support this proposal, since this treatment was not associated with more favorable outcomes. Despite fewer endoscopic signs of bleeding, patients given omeprazole fared no better than those given placebo with respect to mortality, transfusion requirements, and the need for surgery.

The role of early use of a proton-pump inhibitor in unselected patients with upper gastrointestinal bleeding is an important clinical issue that should be addressed by trials with a statistical power that is sufficient to detect meaningful differences in key clinical end points. Until new data become available, we suggest that prompt resuscitation followed by early endoscopic assessment, with the selective use of high-dose proton-pump inhibitors guided by endoscopic findings, should remain the standard of care.

David A. Elphick, M.A.
Stuart A. Riley, M.D.
Northern General Hospital, Sheffield S5 7AU, United Kingdom
david.elphick@sth.nhs.uk

1 References

1. 1

   Lau JY, Leung WK, Wu JCY, et al. Omeprazole before endoscopy in patients with gastrointestinal bleeding. N Engl J Med 2007;356:1631-1640
   Full Text | Web of Science | Medline

To the Editor:

Lau et al. conclude that in their study the early administration of high-dose omeprazole before endoscopy reduced both the number of actively bleeding peptic ulcers and the need for endoscopic intervention. However, we disagree with their conclusion that the use of intravenous high-dose omeprazole before endoscopy prevented recurrent bleeding. In fact, there was an increased trend for recurrent bleeding with high-dose omeprazole as compared with placebo. Although there was an improvement with respect to endoscopic findings of active bleeding in the omeprazole group, the clinically important end points of the mean hospital stay, mortality, and the requirement for blood transfusion were not statistically different from those in the placebo group.

We believe that these results with respect to these clinical end points highlight the unclear role of endoscopic findings as a surrogate end point for long-term complications of ulcers. Perhaps a study design with a longer follow-up should have been planned to determine these more useful end points.

Henry A. Tran, M.D.
Eunice Kang, M.D.
Daniel Becker, M.D.
New York University, New York, NY 10016
hat7e@hotmail.com

Author/Editor Response

Contrary to Elphick and Riley's suggestion, we do not advocate infusion of a high-dose proton-pump inhibitor as a replacement for prompt resuscitation and early endoscopy in the management of upper gastrointestinal bleeding. In our trial, we excluded patients whose condition could not be stabilized while they were receiving fluid resuscitation. Patients in an unstable condition underwent urgent endoscopy and some underwent surgical intervention. We also provided 24-hour availability of emergency endoscopy in case recurrent bleeding developed after initial stabilization. Our trial findings indicate that in patients awaiting endoscopy, infusion of a proton-pump inhibitor would accelerate healing and resolve signs of bleeding. The early discharge of patients with ulcers that show low-risk signs or that have a clean base has implications for resource use. A cost analysis is required to determine the true merit of preemptive infusion of a proton-pump inhibitor. We alluded to the fact that its cost-effectiveness depends on the patient mix and specifically the proportion of patients with bleeding peptic ulcers.

In designing the trial, we hypothesized that infusion of a proton-pump inhibitor would reduce the need for endoscopic therapy. We hoped that given the large sample size, the trial would have statistical power to detect small differences in other clinical outcomes. We used active omeprazole infusion in both groups after achieving endoscopic hemostasis in patients with bleeding peptic ulcers.1 As a consequence, the rate of recurrent bleeding was low in both groups. We then followed patients until day 30 after randomization. As we report in our article, the mortality was low, and among the patients who died, the cause of death in most cases was coexisting illness rather than recurrent bleeding. Although clinically relevant, the design of a trial to detect the difference between groups with either death or recurrent bleeding as the primary end point would have been optimistic.

James Lau, M.D.
Joseph Sung, M.D.
Chinese University of Hong Kong, Hong Kong, China
laujyw@surgery.cuhk.edu.hk

1 References

1. 1

   Lau JY, Sung JJ, Lee KK, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med 2000;343:310-316
   Full Text | Web of Science | Medline

Citing Articles (1)

Citing Articles

1. 1

   Fernando Borda Celaya, Ana Borda Martín. (2008) Aproximación A. El tratamiento farmacológico de la hemorragia digestiva alta con inhibidores de la bomba de protones debería administrarse a todos los pacientes desde el mismo ingreso y antes de la endoscopia. Gastroenterología y Hepatología 31, 5-9
   CrossRef