Stress ulcer prophylaxis in critically ill patients: a randomized controlled trial.Hepatogastroenterology. 2004 May-Jun; 51(57):757-61.H
Critically ill patients especially who require mechanical ventilation or have coagulopathy are at increased risk for stress-related gastrointestinal hemorrhage. There are conflicting data on the efficacy and complication rates of various prophylactic regimens.
Our single-center randomized, placebo-controlled study included 287 patients with high risk for stress-related upper gastrointestinal hemorrhage (>48 h mechanical ventilation, coagulopathy). We compared 3 prophylactic regimens (proton pump inhibitor--omeprazole 40 mg i.v. once daily, n=72; H2 antagonists--famotidine 40 mg twice a day, n=71; and sucralfate 1 g every 6 hours, n=69) with placebo (n=75) in patients with trauma or after major surgery.
Of 287 assessable patients, clinically significant stress-related upper gastrointestinal bleeding was observed in 1%, 3%, 4%, and 1% of patients assigned to receive omeprazole, famotidine, sucralfate, and placebo, respectively (p>0.28). Bleeding developed significantly more often in patients with coagulopathy compared with the others (10% vs. 2%; p=0.006). The gastric pH (p>0.001) and gastric colonization (p<0.05) was significantly higher in the patients who received pH increasing substances when compared with the other 2 groups. Nosocomial pneumonia occurred in 11% of patients receiving omeprazole, in 10% of famotidine patients, in 9% of sucralfate patients and in 7% of controls (p>0.34). No statistically significant differences were found for days on ventilator, length of ICU stay, or mortality among all the 4 groups.
We could not show that omeprazole, famotidine, or sucralfate prophylaxis can affect already very low incidence of clinically important stress-related bleeding in high-risk surgical intensive care unit patients. Furthermore, our data suggested that especially gastric pH increasing medication could increase the risk for nosocomial pneumonia. Routine prophylaxis for stress-related bleeding even in high-risk patients seems not to be justified.