according to a randomized trial and a systematic review led by researchers at McMaster University, Hamilton, Ontario, Canada.
Patients in the intensive care unit (ICU) who need mechanical ventilation typically are given a PPI, such as pantoprazole, to prevent upper GI bleeding caused by stress-induced stomach ulcers, but some evidence suggested that their use might increase the risk for pneumonia and death in the most severely ill patients.
As a result, recent guidelines have issued only weak recommendations for stress ulcer prophylaxis, especially with PPIs, in critically ill patients at a high risk for bleeding, Deborah Cook, MD, professor of medicine at McMaster University, and colleagues noted.
Dr. Deborah Cook
To address clinical questions, they investigated the efficacy and safety of PPIs to prevent upper GI bleeding in critically ill patients.
Both the randomized trial in The New England Journal of Medicine and the systematic review in NEJM Evidence were published online in June.
Significantly Lower Bleeding Risk
The REVISE trial, conducted in eight countries, compared pantoprazole 40 mg daily with placebo in critically ill adults on mechanical ventilation.
The primary efficacy outcome was clinically important upper GI bleeding in the ICU at 90 days, and the primary safety outcome was death from any cause at 90 days.
A total of 4821 patients in 68 ICUs were randomly assigned to the pantoprazole group or placebo group.
Clinically important upper GI bleeding occurred in 25 patients (1%) receiving pantoprazole and in 84 patients (3.5%) receiving placebo. At 90 days, 696 patients (29.1%) in the pantoprazole group died, as did 734 (30.9%) in the placebo group.
No significant differences were found on key secondary outcomes, including ventilator-associated pneumonia and Clostridioides difficile infection in the hospital.
The authors concluded that pantoprazole resulted in a significantly lower risk for clinically important upper GI bleeding than placebo, and it had no significant effect on mortality.
Disease Severity as a Possible Factor
The systematic review included 12 randomized controlled trials comparing PPIs with placebo or no prophylaxis for stress ulcers in a total of 9533 critically ill adults. The researchers performed meta-analyses and assessed the certainty of the evidence. They also conducted a subgroup analysis combining within-trial subgroup data from the two largest trials.
They found that PPIs were associated with a reduced incidence of clinically important upper GI bleeding (relative risk [RR], 0.51, with high certainty evidence) and may have little or no effect on mortality (RR, 0.99, with low-certainty evidence).
However, the within-trial subgroup analysis with intermediate credibility suggested that the effect of PPIs on mortality may differ based on disease severity. The results also raised the possibility that PPI use may decrease 90-day mortality in less severely ill patients (RR, 0.89) and increase mortality in more severely ill patients (RR, 1.08). The mechanisms behind this possible signal are likely multifactorial, the authors noted.
In addition, the review found that PPIs may have no effect on pneumonia, duration of ICU stay, or duration of hospital stay, and little or no effect on C difficile infection or duration of mechanical ventilation (low-certainty evidence).
“Physicians, nurses, and pharmacists working in the ICU setting will use this information in practice right away, and the trial results and the updated meta-analysis will be incorporated into international practice guidelines,” Dr. Cook said.
Both studies had limitations. The REVISE trial did not include patient-reported disability outcomes, and the results may not be generalizable to patients with unassisted breathing. The systematic review included studies with diverse definitions of bleeding and pneumonia, and with mortality reported at different milestones, without considering competing risk analyses. Patient-important GI bleeding was available in only one trial. Other potential side effects of PPIs, such as infection with multidrug-resistant organisms, were not reported.
In an editorial accompanying both studies, Samuel M. Brown, MD, a pulmonologist and vice president of research at Intermountain Health, Salt Lake City, Utah, said that the REVISE trial was “well designed and executed, with generalizable eligibility criteria and excellent experimental separation.” He said the researchers had shown that PPIs “slightly but significantly” decrease the risk of important GI bleeding and have a “decent chance” of slightly decreasing mortality in less severely ill patients during mechanical ventilation. At the same time, he noted, PPIs “do not decrease — and may slightly increase — mortality” in severely ill patients.
Dr. Samuel Brown
Dr. Brown wrote that, in his own practice, he intends to prescribe prophylactic PPIs to patients during mechanical ventilation “if they have an APACHE II score of less than 25” or a reasonable equivalent. The APACHE II scoring system is a point-based system that estimates a patient’s risk of death while in an ICU.
“For sicker patients, I would probably reserve the use of proton-pump inhibitors for those who are being treated with antiplatelet agents, especially in the presence of therapeutic anticoagulants,” he added.
REVISE was supported by numerous grants from organizations in several countries. No funding was specified for the systematic review. Author disclosures and other supplementary materials are available with the full text of the article.
A version of this article first appeared on Medscape.com.