During the early years of critical care, stress related GIB was an important cause of morbidity and mortality. During the past two decades the rate of stress-related GIB has declined probably due to improved resuscitation and early enteral feeding of critically ill patients. To be useful SUP should affect clinical outcome. No clinical trial of stress ulcer prophylaxis has demonstrated a reduction in mortality or length of stay. Surprisingly the effect of SUP on the risk of bleeding is unclear. Cook and colleagues performed a meta-analysis of ten studies which randomized patients to receive a H2RA or placebo. The authors reported that H2RA reduced the risk of clinically significant bleeding (OR 0.44; 95 % CI 0.22–0.88), with a trend towards an increased risk of nosocomial pneumonia (OR 0.25; 95 % CI 0.78–2.0) with no effect on mortality [4]. It should be pointed out that these studies were performed in the 1980s when early enteral nutrition was not encouraged and many patients received parenteral nutrition (see below). Messori and colleagues performed a more recent meta-analysis of studies that compared ranitidine with placebo in ICU patients [12]. These authors concluded that “ranitidine is ineffective in the prevention of gastrointestinal bleeding in patients in the ICU and might increase the risk of pneumonia”.

Zandstra and Soutenbeek reported that 1 of 183 patients (0.6 %) receiving prolonged mechanical ventilation without any SUP developed stress ulcer related bleeding [13]. Erstad and colleagues conducted a prospective study on 543 patients and reported clinically significant GIB rates were similar for those patients with ineffective SUP and those with appropriate SUP [14]. Faisy and colleagues compared the rate of clinically significant GIB during two sequential time periods [15]. During the first phase all patients (n = 736) received SUP while SUP was withheld during the second period (n = 737). Although the patients’ during the second phase of the study were sicker (higher SAPS II score) the rate of overt (1.9 vs. 1.6 %) and clinically significant bleeding (1.4 vs. 1.1 %) as well as the use of blood products was similar between the two time periods. More recently, Kantorova and colleagues performed a randomized, placebo-controlled study in critically ill patients at high risk for stress related GIB (mechanical ventilation >48 h and coagulopathy) in which they compared three SUP regimens (omeprazole, famotidine and sucralfate) with placebo [16]. The overall bleeding rate was 1 % with no significant difference between treatment groups (placebo 1 %). Gastric pH and bacterial colonization was significantly greater in the patients who received acid suppressive therapy with a trend towards a higher incidence of VAP in these patients.

This data suggests that the rate of clinically significant bleeding from stress ulceration in critically ill ICU patients is currently very low and that SUP does not alter this risk or the natural history of this disease.

It has been suggested that patients receiving enteral alimentation have a lower incidence of stress ulceration than unfed patients [17]. In animal models, enteral alimentation has been demonstrated to protect the gastric mucosa from stress related gastric mucosal damage [18, 19]. It has been suggested that enteral nutrients buffer acid and may act as a direct source of mucosal energy, induce the secretion of cytoprotective prostaglandins and improve mucosal blood flow [18, 19]. Furthermore, mucosal immunity maybe supported via stimulation of the gut-associated lymphoid tissue. Because of duodenal-gastric reflux of liquid and increase in mesenteric blood flow due to small intestinal delivery, postpyloric feeding may offer protection against the development of stress ulceration.

Bonten and colleagues demonstrated that continuous enteral nutrition was more likely to raise gastric pH to >3.5 than patients receiving H2RA or PPI’s [20]. Two rat studies have evaluated the role of enteral nutrition in preventing stress ulceration [21, 22]. The results of both trials showed that continuous intragastric administration of elemental formulas significantly reduced the occurrence of macroscopic mucosal lesions compared with intragastric administration of an antacid or intravenous administration of cimetidine. In a retrospective analysis, Raff and colleagues demonstrated that early (within 12 h post-trauma) enteral nutrition was at least as effective as H2RA and/or antacids as stress-ulcer prophylaxis in a cohort of 526 severely burned patients [23]. Pingleton et al. reported a similar finding in 43 ventilated patients [17]. A review of the “historical” randomized controlled trials that studied the effectiveness of acid suppressive therapy in reducing the risk of bleeding, demonstrates that SUP was beneficial only in those patients who were NPO (received no gastric feeding). However, in those studies in which patients were fed enterally the risk of bleeding was equivalent in the treatment and placebo groups [16, 24–26].

This data suggests that in those patients receiving enteral nutrition SUP is not required and indeed may increase the risk of complications (see below). As early enteral nutrition (as opposed to delayed enteral nutrition or parenteral nutrition) has been demonstrated to reduce the morbidity and mortality of critically ill patients, [27, 28] enteral nutrition should be initiated within 24 h of admission to the ICU unless an absolute contraindication exists (bowel obstruction, short gut syndrome).

Acid-suppressive therapy (PPI’s and H2RB’s) should be avoided in patients with cirrhosis as acid-suppressive therapy has been demonstrated to increase the risk of SBP [29]. Acid suppressive therapy has been associated with increased colonization of the small bowel. Presumably this leads to increased bacterial translocation with an increased risk of SBP.

It would appear to be no accident of natural selection that the gastric mucosa of mammalian species secretes acid. Acid plays an important role in protein digestion, but more importantly sterilizes the upper gastro-intestinal tract. Acid suppressive therapy is associated with increased colonization of the upper gastrointestinal tract with potentially pathogenic organism. This may be of critical importance in ICU patients where protocols of oral and enteric decontamination (with non-absorbable antibiotics/anti-microbials) have been demonstrated to reduce the incidence of ventilator associated pneumonia (VAP) [30, 31]. As an extension of these observations, acid suppressive therapy has been demonstrated to increase the gastric colonization and the risk of VAP [12, 16]. In a large prospective pharmaco-epidemiologic cohort study involving non-ICU hospitalized patients, Herzig and colleagues demonstrated that acid-suppressive medication was associated with a 30 % increased odds ratio of hospital-acquired pneumonia [32]. In a subset analysis, these authors demonstrated that this risk was related to the use of PPI’s and not H2RB. Furthermore, the use of gastric suppressive therapy together with the use of broad spectrum antibiotics has been associated with an increased risk of Clostridia difficile infection [33–35]. Gastric acidity may be important is destroying ingested C. difficile spores while broad spectrum antibiotics reduce colonization resistance. PPI’s have been shown to significantly raise gastric pH compared with H2RAs, which may result in a greater risk of C. difficile and pneumonia. Dial and colleagues demonstrated that PPI’s doubled the risk of hospitalized patients developing C. difficile colitis, whereas H2RA did not increase this risk [34]. These authors subsequently demonstrated that the use of both PPI’s and H2RA increased the risk of community acquired C. difficile, however the risk was greater with PPIs [36]. The rapid increase in the incidence of C. difficile colitis in hospitalized patients may be causally related to the exploding use of PPI’s. Furthermore the use of a PPI during incident CDI treatment is associated with a significantly increased risk of recurrence [37].

MacLaren et al. performed a large pharmaco-epidemiological cohort study in adult patients requiring mechanical ventilation for 24 h or more and who were administered either an H2RA or PPI [38]. In this study PPI’s were associated with a greater risks of GI hemorrhage, pneumonia, and Clostridia difficile infection than H2RA’s. Furthermore, ICU LOS and ICU mortality was higher with PPI’s as were hospital costs. Alhazzani et al. performed a meta-analysis comparing the risk of GI bleeding with H2RA’s or PPI’s [39]. In this study PPI’s were more effective than H2RA’s in reducing clinically important upper gastrointestinal bleeding (RR 0.36; 95 % CI 0.19–0.68; p = 0.002). However a meta-analysis by Lin et al. did not find evidence that PPI’s were different from H2RA’s in terms of stress-related upper gastrointestinal bleeding [40].

Krag and colleagues performed a systematic review and trial sequential analysis comparing SUP versus placebo or no prophylaxis in critically ill patients [39]. Trial sequential analysis and sub-group analysis demonstrated no difference in the risk of bleeding between SUP and no-SUP groups. The authors concluded that “both the quality and the quantity of evidence supporting the use of SUP in adult ICU patients is low”.

These data clearly demonstrate that SUP does not reduce mortality or ICU LOS. Furthermore, the risk of significant bleeding from stress ulceration appears to be low and it is unclear if this risk is altered by SUP. SUP with both H2RB’s and PPI’s increase the risk of pneumonia and Clostridia difficile infection. There is conflicting data as to the benefits and risks of PPI’s compared to H2RB’s. Based on these data we do not recommend routine SUP. However, SUP should be considered in very high risk patient’s i.e. on mechanical ventilation for >48 h and DIC. Treatment with a PPI or H2RB is indicated in patients with overt GI bleeding and those with an unexplained drop of hemoglobin of >2 g/dL.