Stress Ulcer Prophylaxis—Friend or Foe?*

    loading  Checking for direct PDF access through Ovid


Critically ill patients are at risk of developing gastrointestinal bleeding, which is associated to adverse outcome (1). As a result, three out of four adult critically ill patients in the ICU receive acid suppressants (2). The most common indication for acid suppressants is stress ulcer prophylaxis (SUP), and pantoprazole is the most frequently prescribed off-label drug in the ICU (3). In the recently published Surviving Sepsis Campaign 2016 guideline, use of SUP is recommended in high-risk patients (4). However, the rationale for using SUP has been challenged in recent years. First, the prevalence of clinically important gastrointestinal bleeding among ICU patients is low (2). Importantly, this includes all sources of gastrointestinal bleedings, also those not prevented by acid suppressants, for example, variceal bleedings (2). Second, the evidence base for use of SUP is limited. As outlined in a systematic review of SUP versus placebo, the quantity and quality of evidence supporting use of SUP is low, with no firm evidence for benefit or harm (5). Third, adjusted analyses have not been able to confirm the association between clinically important gastrointestinal bleeding and increased mortality, which may be explained by confounding, including confounding by indication (2). Finally, acid suppressants increase pH in the stomach, which may increase the risk of bacterial growth and serious infectious complications (6, 7). Consequently, there is clinical equipoise for use of SUP in critically ill patients.
In this issue of Critical Care Medicine, Alhazzani et al (8) report the results of the Reevaluating the Inhibition of Stress Erosions (REVISE) trial, an international, multicenter, randomized, blinded pilot trial of SUP with a proton pump inhibitor (PPI) versus placebo. The overall aim was to assess if withholding SUP in adult critically ill patients was feasible and safe. Ninety-one patients from eight ICUs in three countries were randomized to 40 mg of IV pantoprazole or placebo once daily. Inclusion criteria were adult patients admitted to the ICU and anticipated to receive invasive mechanical ventilation for at least 48 hours. Exclusion criteria included invasive mechanical ventilation for greater than 72 hours prior to randomization, use of PPIs due to active bleeding or increased risk of bleeding, use of dual antiplatelet therapy prior to randomization, palliative care or decision to withdraw advanced life support, previous enrollment in the REVISE trial or a related trial, pregnancy, trial participation declined by ICU treating physician, patient, or substitute decision maker, and receipt of two or more “daily dose equivalents” of prophylaxis with histamine-2-receptor antagonist or PPI during current ICU admission. The primary outcomes were feasibility outcomes, including consent rate, recruitment rate, and adherence to the protocol. Secondary outcomes were clinical outcomes: clinically important gastrointestinal bleeding, ventilator-associated pneumonia (VAP), Clostridium difficile infection, ICU and hospital mortality, and ICU and hospital length of stay. All a priori feasibility outcomes were achieved. Overall, 150 patients or substitute decision makers were asked for consent. In 59 cases, consent was unobtainable and 16 patients were not randomized for other reasons. The mean consent rate was 77.8%, the mean recruitment rate was 2.6 patients per center per month, and adherence to the protocol was 97.7%. No differences in the secondary clinical outcomes between the groups were reported: 6.1% versus 4.8% were diagnosed with clinically important gastrointestinal bleedings, 20.4% versus 14.3% with VAP, and 4.1% versus 2.4% with C. difficile infection in the PPI and placebo group, respectively.
The REVISE trial is a well-designed and conducted randomized multicenter feasibility trial. The methodology used is sound with resulting low risk of bias, and relevant feasibility outcomes were assessed. The trial and protocol proved feasible with no indication of harm by withholding SUP.
    loading  Loading Related Articles