What the REVISE Trial Changed About Stress Ulcer Prophylaxis

The REVISE trial, reported in the New England Journal of Medicine in 2024, randomly assigned more than 4,800 invasively ventilated intensive care unit adults to daily intravenous pantoprazole or placebo, and it produced a clean, two-part answer. Pantoprazole lowered clinically important upper gastrointestinal bleeding from 3.5 percent to 1.0 percent, a hazard ratio of 0.30, while 90-day mortality was statistically unchanged at 29.1 percent versus 30.9 percent (hazard ratio 0.94). A drug that reduced a feared complication without a measurable survival penalty is what a decade of uncertain trials had been waiting for, and the result shifted the stress ulcer prophylaxis debate away from "does this help or harm" toward "in whom, and how much."

The question REVISE was built to answer

Critically ill patients on mechanical ventilation can develop stress-related mucosal bleeding, and acid-suppressing drugs have been given prophylactically for decades to prevent it. The practice drifted into routine long before the evidence was solid. By the late 2010s, two large trials had complicated the picture. The SUP-ICU trial, comparing pantoprazole with placebo, confirmed that proton pump inhibitors reduce bleeding but hinted at higher mortality among the most severely ill. The PEPTIC trial, comparing proton pump inhibitors with histamine-2 blockers, raised a similar shadow of a possible mortality signal.

The worry was mechanistic and plausible: suppressing gastric acid may permit bacterial overgrowth, raising the risk of pneumonia and Clostridioides difficile infection, and those harms might offset any bleeding benefit in the sickest patients. REVISE, led by Deborah Cook and an international collaboration spanning 68 ICUs in eight countries, was built to test that trade-off directly, with the statistical power the earlier studies lacked.

What REVISE found

On the efficacy side, the result was unambiguous. Clinically important upper gastrointestinal bleeding occurred in 25 of 2,385 pantoprazole patients (1.0 percent) and 84 of 2,377 placebo patients (3.5 percent), a hazard ratio of 0.30 with a 95 percent confidence interval of 0.19 to 0.47. The direction and magnitude were consistent with SUP-ICU, so REVISE both confirmed and sharpened the earlier finding, fixing the effect in place with precision. A broader measure the investigators called patient-important bleeding moved in the same direction.

On safety, the number everyone watched was mortality at 90 days: 696 of 2,390 patients (29.1 percent) with pantoprazole and 734 of 2,379 (30.9 percent) with placebo, a hazard ratio of 0.94 (95 percent confidence interval 0.85 to 1.04). The interval crosses one, so this is a neutral finding rather than proof of benefit, but the point that mattered is that it showed no signal of net harm. The feared downstream infections did not materialize either. There was no significant difference in ventilator-associated pneumonia, C. difficile infection, or new renal replacement therapy between the groups.

Why a mortality-neutral result carried weight

It is easy to read "no mortality difference" as a dull outcome. In this debate it was the opposite. The entire case for pulling back on prophylaxis rested on the suspicion that acid suppression might be quietly killing marginal patients through infection, and REVISE was large enough to detect a clinically meaningful mortality difference if one existed. Finding none, while confirming a real reduction in bleeding, tilts the benefit-risk balance back toward prophylaxis for ventilated patients who are not already receiving acid suppression. A companion updated meta-analysis published in NEJM Evidence, pooling twelve trials and more than 9,500 patients, reached the same overall conclusion: proton pump inhibitors reliably reduce clinically important bleeding with no overall effect on death.

The unresolved edge

REVISE did not erase the old worry entirely; it relocated it. Within the trial, a prespecified look at illness severity showed no statistically significant interaction, but the estimates hinted at the familiar pattern, with a possible mortality benefit in less severely ill patients and no benefit, perhaps the reverse, in the most severe. The companion meta-analysis described this as a subgroup observation of limited credibility. The possibility that proton pump inhibitors lower 90-day mortality in less sick patients and raise it in the sickest can be neither confirmed nor excluded from the current data. That is an honest scientific statement rather than a bedside instruction. It means the average ventilated patient benefits, and that reasonable clinicians will still individualize at the extremes of acuity.

How the picture changed

Before REVISE, a defensible reading of the evidence was that routine prophylaxis might be doing about as much harm as good and could often be withheld. After REVISE, the more defensible reading is that for invasively ventilated adults not already on acid suppression, a proton pump inhibitor prevents bleeding with no demonstrable survival cost, and the argument that remains concerns refinement at the margins of illness severity rather than the core practice itself. That is how strong trials tend to work. They rarely hand down a verdict on everything; more often they convert a noisy dispute into a narrower, more answerable one.

This article is educational and not medical advice. Prophylaxis decisions in critical illness depend on the individual patient and belong to the treating team.