Current issues of ACP Journal Club are published in Annals of Internal Medicine


Therapeutics

Omeprazole did not reduce mortality, rebleeding, surgery, or transfusion required in acute upper gastrointestinal bleeding

ACP J Club.1992 July-Aug;117:16. doi:10.7326/ACPJC-1992-117-1-016


Source Citation

Daneshmend TK, Hawkey CJ, Langman MJ, et al. Omeprazole versus placebo for acute upper gastrointestinal bleeding: randomised double blind controlled trial. BMJ. 1992 Jan;304:143-7.


Abstract

Objective

To determine if omeprazole reduces death, rebleeding, surgery, and transfusion requirements in patients with acute upper gastrointestinal bleeding.

Design

Randomized, double-blind, placebo-controlled trial of 40 months duration. The trial was designed to be able to detect a decrease in mortality from 10% to 5% in the treatment group.

Setting

General medical wards and intensive care units of 2 general hospitals, Nottingham, England.

Patients

1147 consecutive patients (729 men; mean age, 59 y) with acute upper gastrointestinal bleeding at admission were included. Patients were excluded if they were less than 18 years old, pregnant, or had severe disease not allowing treatment, potential drug interactions, bleeding requiring immediate surgery, trivial bleeding not requiring treatment, or bleeding that developed after admission. The placebo and omeprazole groups did not differ in possible prognostic factors.

Intervention

Endoscopy was done within 24 hours of admission. Patients were assigned to placebo (mannitol; n = 569) or to omeprazole treatment (n = 578). The initial dose was 80 mg intravenously as soon as possible after admission, then three 40-mg intravenous doses approximately 8 hours apart. Subsequently, 40-mg oral doses were given every 12 hours for 3 days. Patients were then followed for another 101 hours or until surgery, death, or discharge. Concurrent drug therapy was not allowed during the study period. Endoscopic treatment was permitted.

Main outcome measures

Transfusion requirements, rebleeding and surgery rates, and all-cause mortality. Analysis was by intention-to-treat.

Main results

53% of the placebo group and 52% of the treatment group received transfusions (95% CI for 1% difference -4% to 7%, P = 0.6). Rebleeding rates were 18% (placebo) and 15% (omeprazole) (CI for the 3% difference -1% to 7%, P = 0.2). The surgery rate was 11% for both groups (P = 0.9). All-cause mortality was 5.3% in the placebo group and 6.9% for omeprazole (95% CI for the 1.6% difference -1.2% to 4.4%, P > 0.2). An analysis using the 1049 patients who received their assigned treatment and analyses based on patients with confirmed upper gastrointestinal bleeding (n = 945) and with peptic ulcers (n = 503) also showed no differences between the groups. Omeprazole, however, was associated with a reduction in the incidence of endoscopic signs of bleeding (45% vs 33%, P < 0.001). {This absolute risk reduction of 12% means that 8 patients would need to be treated with omeprazole (compared with placebo) to prevent 1 additional patient from having endoscopic signs of bleeding, CI 6 to 16; the relative risk reduction was 27%, CI 15% to 37%.}*

Conclusion

Omeprazole treatment did not reduce mortality, rebleeding, surgery, or transfusion requirements for hospitalized patients with acute gastrointestinal bleeding.

Source of funding: Astra Clinical Research.

Address for article reprint: Dr. C.J. Hawkey, Department of Therapeutics, University Hospital, Nottingham NG7 2UH, United Kingdom.

*Numbers calculated from data in article.


Commentary

This important study was well designed and intelligently reported. The study includes an impressive sample size of patients with acute upper gastrointestinal tract bleeding. Because the study included all patients with acute upper gastrointestinal bleeding admitted to 2 general hospitals, albeit in 1 English city, its results are probably applicable to other centers.

The main conclusion of the study is of major interest to generalists as well as specialists. A high dose of the antisecretory drug, omeprazole, which would have been sufficient to cause a substantial increase in intragastric pH, did not influence the clinical outcome of patients with acute upper gastrointestinal tract bleeding. This finding was true for all causes of bleeding and, possibly more important, for peptic ulcer alone. Omeprazole is the most potent acid-suppressing drug currently available, so the results of this trial are important. As the authors rightly point out, no substantial evidence supports the widespread practice of giving intravenous antisecretory drugs, including the histamine-2-receptor antagonists, to patients with acute upper gastrointestinal tract hemorrhage. This lack of evidence may surprise some practitioners and is a point worth reiterating. The main implication for practice is that any further improvements in our management of patients with upper gastrointestinal tract bleeding are more likely to come from endoscopic rather than pharmacologic treatments.

The observation that intravenous omeprazole reduced the incidence of endoscopic signs of bleeding has no bearing on routine clinical practice but is certainly worthy of further detailed investigation.

Colin W. Howden, MD
University of South CarolinaColumbia, South Carolina, USA