The New England Journal of Medicine NOVEMBER 28, 2018

Stress Ulcer Prophylaxis in the Intensive Care Unit (SUP-ICU) Trial

Pål Andre Krantz, et al. (The SUP-ICU Investigators)

Bottom Line

The SUP-ICU trial found that prophylactic intravenous pantoprazole in critically ill patients at risk for gastrointestinal bleeding did not reduce 90-day mortality compared to placebo.

Key Findings

1. The primary outcome of 90-day mortality was not significantly different between the pantoprazole group (31.1%) and the placebo group (30.4%) (hazard ratio, 1.02; 95% CI, 0.91-1.13).

2. Clinically important gastrointestinal bleeding occurred in 2.5% of patients in the pantoprazole group compared to 4.2% in the placebo group, although the trial was not powered to definitively assess this secondary outcome (hazard ratio, 0.58; 95% CI, 0.38-0.89).

3. There was no significant difference in the incidence of serious adverse events, including pneumonia, Clostridium difficile infection, or myocardial ischemia, between the two study arms.

4. Long-term follow-up at 1 year confirmed no survival benefit, with mortality rates of 37.3% in the pantoprazole group versus 37.0% in the placebo group (relative risk, 1.01; 95% CI, 0.92-1.10).

Study Design

Design

RCT

Double-Blind

Sample

3,298

Patients

Duration

90 days

Median

Setting

Multicenter, 6 countries

Population Acutely admitted adult ICU patients at risk for gastrointestinal bleeding (e.g., shock, mechanical ventilation, coagulopathy, or organ replacement therapy).

Intervention Intravenous pantoprazole 40 mg once daily.

Comparator Intravenous placebo (saline) once daily.

Outcome All-cause mortality at 90 days after randomization.

Study Limitations

• The study did not reach the originally targeted sample size of 3,550 patients, potentially limiting the power to detect small, clinically meaningful differences.

• The use of open-label acid suppression in a subset of patients who were excluded or withdrew consent may have introduced selection bias.

• The trial was conducted in a modern ICU setting where early enteral nutrition, which naturally provides some protective effects against stress ulcers, was already standard practice, potentially attenuating the baseline risk of bleeding.

• The composite secondary outcome measure was not fully reflective of long-term patient-centered outcomes, focusing instead on short-term markers.

Clinical Significance

The results suggest that routine pharmacological stress ulcer prophylaxis with PPIs may be overused in modern intensive care, as it does not translate into a mortality benefit and its role in preventing clinically significant bleeding must be weighed against its marginal absolute benefit and potential risks in stable patients.

Historical Context

Stress ulcer prophylaxis has been a standard of care in intensive care units for decades, based largely on older, smaller trials with high risks of bias. The SUP-ICU trial was designed to rigorously challenge this practice in the modern era of critical care, where advances in resuscitation and early enteral nutrition have already substantially reduced the incidence of stress-related mucosal bleeding.

Guided Discussion

High-yield insights from every perspective

Med Student

Medical Student

Explain the physiological mechanism of stress-related mucosal disease (SRMD) in the ICU and why researchers originally hypothesized that reducing gastric acid with pantoprazole would decrease mortality.

Key Response

Critically ill patients experience splanchnic hypoperfusion, which impairs the gastric mucosal barrier and increases susceptibility to acid-induced injury. The hypothesis was that preventing clinically significant gastrointestinal bleeding (CIGIB) would reduce complications such as hypotension, the need for transfusions, and subsequent multi-organ failure, thereby improving overall survival.

Resident

In the SUP-ICU trial, there was no difference in 90-day mortality, but there was a significant reduction in clinically important gastrointestinal bleeding. How should this influence your bedside decision-making when managing a patient with multiple risk factors such as mechanical ventilation and coagulopathy?

Key Response

While mortality is unchanged, the reduction in bleeding (2.5% vs 4.5%) suggests that prophylaxis provides a morbidity benefit. Clinicians must weigh the prevention of bleeding against potential risks (e.g., pneumonia, C. diff) and the patient's individual risk profile. For those at highest risk, the reduction in bleeding events may still justify the use of PPIs even without a mortality benefit.

Fellow

Analyze the potential confounding effect of enteral nutrition (EN) in the SUP-ICU trial, where approximately 80% of patients received EN. How does this impact the generalizability of the results to the subset of patients who are strictly NPO?

Key Response

Enteral nutrition acts as a natural buffer for gastric acid and provides trophic support to the mucosa. The high prevalence of EN in this trial may have mitigated the baseline risk of stress ulcers, potentially diluting the treatment effect of pantoprazole. This suggests that the trial's 'negative' mortality result might not apply as strongly to patients with prolonged NPO status or those with primary gastrointestinal pathology.

Attending

The SUP-ICU trial is often cited alongside the later REVISE trial. How do these findings collectively change your approach to 'PPI inertia'—the tendency to continue stress ulcer prophylaxis after the acute phase of critical illness has passed?

Key Response

Both trials suggest that the absolute benefit of PPIs in the modern ICU (where EN and resuscitation are optimized) is smaller than previously thought. This evidence supports a more restrictive approach to starting PPIs and provides a strong teaching point for aggressive de-escalation once the primary risk factors (ventilation, coagulopathy) resolve, preventing the unnecessary continuation of these meds into the post-ICU setting.

Scholarly Review

Critical appraisal through the lens of expert reviewers and guideline development

PhD

The SUP-ICU trial was powered to detect a 5% absolute risk reduction in 90-day mortality. Critically evaluate whether mortality is an appropriate primary endpoint for a prophylaxis trial, or if a composite endpoint including transfusion requirements and infectious complications would have been more statistically sensitive.

Key Response

Mortality in the ICU is multifactorial and often driven by the primary illness rather than secondary complications like GI bleeding. A 5% mortality reduction for a single prophylactic drug is an ambitious effect size (high bar). Using a composite endpoint might increase power but complicates interpretation due to the varying clinical significance of its components (e.g., a transfusion vs. death).

Journal Editor

The trial reported a low incidence of C. difficile and pneumonia in both groups, which contradicts previous observational data. What features of the study design or ICU practice environment might have contributed to this finding, and does it settle the debate regarding the infectious risks of PPIs?

Key Response

Observational studies are prone to 'confounding by indication,' where sicker patients get PPIs and are also more likely to get infections. As a randomized controlled trial, SUP-ICU provides higher-quality evidence that PPIs may not be as significant a driver of VAP or C. diff in the acute setting as previously feared. However, the trial may still be underpowered for these secondary safety outcomes.

Guideline Committee

The Surviving Sepsis Campaign guidelines traditionally issued a 'weak recommendation' for stress ulcer prophylaxis in patients with risk factors. Does the SUP-ICU trial's failure to show a mortality benefit necessitate a shift toward recommending against routine prophylaxis in patients receiving enteral nutrition?

Key Response

Current guidelines (like SSC and the 2020 SCCM/ESICM guidelines) suggest prophylaxis for patients with high-risk factors (ventilation >48h, coagulopathy). While SUP-ICU shows no mortality benefit, the reduction in bleeding maintains the rationale for prophylaxis in high-risk subsets. The committee must decide if the recommendation should be narrowed specifically to patients where the risk of CIGIB outweighs the cost and potential (though statistically insignificant in this trial) risk of infections.

Clinical Landscape

Noteworthy Related Trials

1998

Cook et al. Study

n = 1,238 · NEJM

Tested

Ranitidine (H2RB)

Population

Critically ill ICU patients

Comparator

Sucralfate

Endpoint

Clinically important gastrointestinal bleeding

Key result: Ranitidine was more effective than sucralfate in preventing clinically important gastrointestinal bleeding in critically ill patients.

2020

PEPTIC Trial

n = 26,982 · JAMA

Tested

Proton pump inhibitors (PPIs)

Population

Adult ICU patients receiving invasive mechanical ventilation

Comparator

Histamine-2 receptor blockers (H2RBs)

Endpoint

90-day in-hospital mortality

Key result: There was no significant difference in 90-day in-hospital mortality between patients receiving PPIs and those receiving H2RBs.

2023

REVISE Trial

n = 3,446 · NEJM

Tested

Pantoprazole 40 mg daily

Population

Critically ill adults receiving invasive mechanical ventilation

Comparator

Placebo

Endpoint

Clinically important upper gastrointestinal bleeding at 90 days

Key result: Pantoprazole did not result in a lower rate of clinically important upper GI bleeding compared to placebo in critically ill patients.

Tailored to your role

Want this tailored to you?

Add your specialty or training stage to get role-specific takeaways and more questions.

Personalize this analysis