In summary, for patients who received PPI therapy:

Benefits in NNT

    No one was helped (no death, re-bleeding, or surgical intervention was prevented)
    No one was helped (no death, re-bleeding, or surgical intervention was prevented)

Harms in NNT

    Not reported
    Not reported

View As:

Source: Sreedharan A, Martin J, Leontiadis GI, Dorward S, Howden CW, Forman D, Moayyedi P. Proton pump inhibitor treatment initiated prior to endoscopic diagnosis in upper gastrointestinal bleeding. Cochrane Database Syst Rev. 2010;7:CD005415.

Study Population: 2223 adults with acute upper gastrointestinal bleeding, enrolled in six randomized controlled trials

Efficacy Endpoints: Mortality, need for surgical intervention, re-bleeding, blood transfusion requirements, length of hospital stay

Harm Endpoints: Not reported

Narrative: Upper gastrointestinal bleeding (UGIB) is common, with an annual incidence of approximately 67 to 150 per 100,000, with estimated mortality rates between 6% and 15%.1 Many patients require hospital admission and ultimately endoscopic evaluation for diagnosis and treatment of the hemorrhage. It is standard practice to start patients with undifferentiated UGIB on acid suppression therapy with an intravenous proton-pump inhibitor (PPI) in the emergency department prior to admission or endoscopy.2 This practice is based on data that the most common cause of UGIB is peptic ulcer disease. Intravenous PPI may create optimal conditions for clotting over arteries at ulcer bases, as neutralization of gastric acid leads to stabilization of blood clots.3, 4

The Cochrane systematic review discussed here5 included 6 randomized controlled trials (RCT) enrolling 2223 patients with undifferentiated UGIB evaluating the use of PPI therapy prior to endoscopy compared to placebo, H2-receptor antagonists (H2RA) or no treatment. Some trials included patients who had been admitted to the hospital for other reasons and subsequently developed UGIB. The trials were not confined to patients with peptic ulcer bleeding as three of the trials included patients with upper gastrointestinal bleeding due to esophageal varices. The included trials compared oral PPI to placebo, intravenous PPI to placebo, intravenous PPI to H2RA, or intravenous PPI to no treatment.5

This systematic review5 showed no statistically significant benefit in the primary outcomes of mortality, risk of re-bleeding, or need for surgical intervention. Secondary outcomes did show significantly reduced proportion of patients with stigmata of recent hemorrhage (visualizing lesions showing evidence of recent bleeding) at index endoscopy when comparing PPI to control (Odds ratio (OR): 0.67; 95% CI, 0.54 to 0.84, absolute risk difference [ARD]: 9.3%; Number-needed-to treat [NNT]: 11). The systematic review also showed significantly reduced need for endoscopic therapy at the index visit (OR: 0.68; 95% CI 0.50 to 0.93; ARD: 3.1%; NNT: 33). These secondary outcomes are at best surrogate outcomes that did not translate into patient-centered outcomes such as survival benefit. Therefore, we did not report them in the summary table.

There was not sufficient evidence to assess for amount of blood transfused or decrease in hospitalized days. The pre-planned subgroup analyses controlling for type of comparative (placebo or H-2 receptor antagonist, route of PPI administration and application of initial endoscopic hemostatic treatment) similarly revealed no significant differences in mortality, rebleeding or surgery.

A subsequent Cochrane meta-analysis of 22 randomized trials6 assessed the use of high-dose bolus PPI with continuous infusion compared to lower doses given by continuous infusion, intravenous bolus, or orally after an endoscopic evaluation of peptic ulcer bleeding. There was no significant difference in mortality, risk of re-bleeding, surgical interventions, length of hospital stay or blood transfusion requirement between higher versus lower dose regimens. Lastly, a 2014 systematic review by Sachar et al7 found that intermittent PPI therapy after successful endoscopic treatment of high-risk peptic ulcers was not inferior to high-dose PPI therapy plus continuous infusion in terms of re-bleeding within 7 or 30 days, mortality, and requirement for blood transfusion.

The systematic reviews did not report adverse events related to PPI use.

Caveats: The meta-analysis discussed here enrolled patients with undifferentiated upper gastrointestinal bleeding before endoscopic diagnosis. Patients with peptic ulcer and specifically high risk peptic ulcers might still benefit from PPI. However, the systematic review points out that the two trials that entirely focused on bleeding peptic ulcers, also did not report survival benefit or significant reduction in the risk of rebleeding.5

The Cochrane systematic review summarized here5 did not report significant statistical heterogeneity for the major outcomes of 30-day mortality, re-bleeding, or need for surgery. Length of hospital stay could not be pooled due to differences in reporting method for this outcome. Only one trial included in the Cochrane systematic review was rated high quality.5 However, the sensitivity analysis deemed the study results reliable.

The American College of Gastroenterology recommends the use of pre-endoscopic intravenous PPI (80mg bolus followed by 8mg/h infusion) for UGIB secondary to suspected or confirmed PUD. The guidelines also state that PPI can be discontinued after endoscopy if a non-ulcer cause of bleeding is found, and if endoscopy needs to be delayed then PPI should be continued until endoscopy to decrease the risk of further bleeding.8 These recommendations for the use of PPI in ulcer bleeding are most likely based on the data suggesting that this therapy decreases the stigmata of recent hemorrhage and need for endoscopic intervention.

One caveat worth mentioning is that in the trials that reported stigmata of recent hemorrhage, time to endoscopy varied significantly (within 24 hours in two trials, 24-48 hours in one trial, and mean time-to-endoscopy in the fourth). Additionally, definitions for re-bleeding were variable between the six trials.5

The data pertaining to the use of intermittent vs continuous infusion PPI therapy, mainly applies to patients after endoscopic treatment, and not prior to endoscopy. This strategy is not directly generalizable to emergency department populations with UGIB prior to endoscopy. Further research is needed to determine if intermittent PPI therapy is not inferior to bolus and infusion in undifferentiated UGIB in the emergency department.

In summary, intravenous PPI therapy in undifferentiated UGIB does not improve survival and does not reduce the risk of re-bleeding or need for surgery. Therefore, we have assigned a color recommendation of RED (No benefit) to this intervention. Large trials specifically targeting patients with documented peptic ulcers might provide different results. Since a large percentage of UGIB is not actually from peptic ulcer disease,7 the practice of intravenous PPI administration in all patients with undifferentiated UGIB may be unnecessary and costly. The risks related to adverse effects of PPI as well as cost analyses are important considerations, although beyond the scope of this review.

The original manuscript was published in Academic Emergency Medicine as part of the partnership between TheNNT.com and AEM.

See theNNT.com's previous reviews of this topic:
Proton Pump Inhibitors (PPIs) Given for Acute Upper Gastrointestinal Bleeding Given Prior to Endoscopic Diagnosis, January 7, 2010

Author: Mark Serpico, MD; Matthew Riscinti, MD
Supervising Editor: Shahriar Zehtabchi, MD

Published/Updated: December 16, 2019

  1. The Title Bar

    The title bar is color-coded with our overall recommendation.

    • Green: Benefits outweigh risks.
    • Yellow: Unclear risk/benefit profile.
    • Red: Benefits do not outweigh risks.
    • Black: Obvious harms, no clear benefits.
  2. Tip content...