REVIEWS

Stress-related mucosal disease in the

critically ill patient
Marc Bardou, Jean-Pierre Quenot and Alan Barkun
Abstract | Bleeding from stress-related mucosal disease in critically ill patients remains an important clinical
management issue. Although only a small proportion (1–6%) of patients admitted to an intensive care unit
(ICU) will bleed, a substantial proportion exhibit clinical risk factors (mechanical ventilation for >48 h and
a coagulopathy) that predict an increased risk of bleeding. Furthermore, upper gastrointestinal mucosal
lesions can be found in 75–100% of patients in ICUs. Although uncommon, stress-ulcer bleeding is a severe
complication with an estimated mortality of 40–50%, mostly from decompensating an underlying condition or
multiorgan failure. Although the vast majority of patients in ICUs receive stress-ulcer prophylaxis, largely with
PPIs, some controversy surrounds their efficacy and safety. Indeed, no single trial has shown that stress-ulcer
prophylaxis reduces mortality. Some reports suggest that the use of PPIs increases the risk of nosocomial
infections. However, several meta-analyses and cost-effectiveness studies suggest PPIs to be more clinically
effective and cost-effective than histamine‑2 receptor antagonists, without considerable increases in
nosocomial pneumonia. To help clinicians use the most appropriate strategy for treatment of patients in the
ICU, this Review presents the latest information on all aspects of stress-related mucosal disease.
Bardou, M. et al. Nat. Rev. Gastroenterol. Hepatol. 12, 98–107 (2015); published online 6 January 2015; doi:10.1038/nrgastro.2014.235

Introduction Epidemiology of stress-ulcer bleeding

Critically ill patients are at risk of bleeding from stress- Endoscopies performed within 72 h of the onset of illness
related mucosal disease (SRMD),1 which is an important have shown that 75–100% of critically ill patients exhibit
clinical problem.2,3 Patients who experience SRMD have gross gastric lesions.13 Lesions are most often diffuse sub-
a prolonged stay in intensive care and an increased risk of epithelial haemorrhages and erosions; however, if massive
mortality.4–6 Even though the true incidence is debated, bleeding occurs it suggests that an ulcer has developed.2
only a small proportion of critically ill patients will actu- Many definitions of stress-ulcer bleeding have been put
ally bleed.7,8 For example, a cohort study conducted in forward since it was first described in 1969.14 Although
Germany and published in 2014 found a bleeding rate occult bleeding, as defined by guaiac stool positivity in
of 0.9%.6 Nevertheless, the clinical factors that predict samples from nasogastric tube or emesis, has been pro-
a higher risk of bleeding are increasingly found among posed as a working definition of stress-ulcer bleeding,
patients admitted to an intensive care unit (ICU).3,7,8 the more relevant clinical presentation is that of overt
Stress-ulcer prophylaxis (SUP) has been found to be effi- bleeding manifesting as haematemesis, coffee ground
cacious and is thus considered as a standard of care in the emesis, melena, or bloody nasogastric aspirate. Pooled
ICU, included in the mnemonic FASTHUG (Feeding, results from 10 randomized trials of prophylactic therapy
Gastroenterology
Analgesia, Sedation, Thromboembolic prophylaxis, Head (versus placebo or no therapy) performed from 1980 to
and Hepatology of bed elevation, stress Ulcer prophylaxis, Glucose control) 1998 suggested an incidence of 17% for stress-ulcer bleed-
Department (M.B.), that was developed by ICU physicians to ensure that key ing.2 However, in reports from the past 10 years, the inci-
and Medical Intensive
Care Unit (J.‑P.Q.), aspects of care are addressed during each patient encoun- dence of SRMD-associated clinically significant bleeding
CHU de Dijon, France, ter.9 Histamine‑2 receptor antagonists (H2RAs) have been or stress-ulcer bleeding (variably defined to include overt
14 Rue Gaffarel
BP77908, 21079 Dijon
found to be efficacious in preventing SRMD bleeding in bleeding combined with transfusion, haemodynamic
Cedex, France. the ICU setting, as have PPIs.3,7,8,10 Their comparative effi- instability and/or the need for intervention) has markedly
Gastroenterology cacies, the possible development of nosocomial pneumo- decreased. Indeed, from a pooled estimate of 8% in older
Department, McGill
University Health nia and cost-effectiveness remain subjects of controversy, randomized trials,2 with the advent of improved care of
Centre, Montreal with disparate data in the literature.10–12 Therefore, uncer- critically ill patients, appropriate prophylactic therapy
General Hospital Site,
Room D7‑346,
tainty over whether routine SUP is indicated in critically and increasing use of enteral feedings,1,7,15,16 the reported
1650 Cedar Avenue, ill patients still exists among clinicians in 2014. In this pooled incidence in what are principally trials, published
Montréal, QC H3G 1A4, Review, the latest information on all aspects of SRMD are between 1993 and 2010, or retrospective analyses spanning
Canada (A.B.).
provided as an overview. over the same time frame, is now <1%.6,17,18 Nonetheless,
Correspondence to: SRMD and clinically significant bleeding remain impor-
M.B.
marc.bardou@ Competing interests tant clinical problems as the associated mortality is high,
u-bourgogne.fr The authors declare no competing interests. related primarily to the patients’ underlying illness.

98 | FEBRUARY 2015 | VOLUME 12  www.nature.com/nrgastro

© 2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Key points production of prostaglandins, mucin glycoproteins,

water, bicarbonates, trefoil factor family peptides,
■■ Prevalence of clinically important stress-ulcer bleeding is low (~1%), but
phospho­lipids and heat-shock proteins (HSPs), par-
because of its severity it remains an important issue
■■ Prophylaxis for stress-ulcer bleeding should be restricted to patients with
ticularly HSP70.28–31 Prostaglandins stimulate mucosal
acknowledged risk factors because of the risk of nosocomial infection blood flow, and mucus and bicarbonate production,
■■ Despite controversies, PPIs seem to be the most clinically effective and thereby enhancing epithelial cell growth and repair.32
cost‑effective strategy The mucous bicarbonate barrier forms a physical barrier
■■ Randomized clinical trials and observational studies are needed to assess new against the influx of luminal acid and pepsin, and is also
preventive strategies and to adapt therapeutic management approaches of important in trapping bicarbonate anions that have dif-
upper gastrointestinal bleeding to the intensive care unit setting fused from the gastric epithelial cells.22 Bicarbonate is
■■ Stress-ulcer prophylaxis is being overused; two-thirds of patients without risk-
powerful enough to neutralize the pH within the mucus
factors receive treatment, two-thirds continue treatment upon transfer from the
intensive care unit and one-third are discharged home on treatment without a gel on the epithelial cell surface despite a luminal pH
new indication of 1.5–2.0.33,34 Gastric epithelial cell restitution (or re-
epithelization) is a defensive mechanism whereby rapid,
energy-dependent migration of epithelial cells occurs
Risk factors for SRMD towards areas of superficial damage, leading to the
Critically ill patients are at increased risk of developing restor­ation of epithelial integrity.22 Oxygen free-radical
SRMD and subsequent stress-ulcer bleeding as a result formation35 and physiological stress36 have been impli-
of both underlying disease and therapeutic interven- cated in reducing the rate of cellular proliferation and
tion. In a multicentre study of 2,252 patients, Cook gastric epithelial restitution.
et al.4 identified respiratory failure (defined as the need Among the aggressive physiological factors that might
for mechanical ventilation for at least 48 h) and coagulo­ act to break down the mucosal protective defences,
pathy (defined as a platelet count of <50,000 platelets splanchnic hypoperfusion (or gastric mucosal ischaemia)
per ml3, an international normalized ratio [INR] >1.5, is thought to be the major underlying cause of SRMD in
or a partial-thromboplastin time of more than two times critically ill patients.37 The shunt of blood away from the
the control value) as strong independent risk factors for viscera and skin is an adaptation to stress that aims to
SRMD, with an odds ratio (OR) of 15.6 and 4.3, respec- preserve perfusion to the vital organs during an insult.
tively. Another analysis from the same group published Temporary shunting of visceral blood flow is well toler-
in 1999 identified renal failure, as assessed by maximum ated; however, it becomes maladaptive, as it results in
serum creatinine level, as an independent predictor lesions, when prolonged. MacLaren and colleagues 38
of upper gastrointestinal bleeding (UGIB) in patients showed that shock or hypotension is associated with a
requiring mechanical ventilation; enteral nutrition and small but significant increased risk of gastrointestinal
stress-ulcer prophylaxis with ranitidine were shown haemorrhage (OR, 1.17; 95% CI 1.04–1.33) whereas
to confer significant protective effects (RR = 0.30; 95% hypertension seems to be protective (OR, 0.77; 95%
CI 0.13‑0.67, P = 0.04 and 0.39; 0.17‑0.83, P = 0.24, for CI 0.67–0.87). Gastric mucosal perfusion decreases
enteral nutrition and ranitidine, respectively).19 However, early during critical illness39,40 owing to a number of
these risk factors were defined more than 20 years ago.4 factors: activation of the sympathetic nervous system;
Other factors have since been reported to be associated increased catecholamine release and vasoconstriction,
with increased risk of SRMD, such as age >50 years, notably within the gastrointestinal tract; hypo­volaemia;
maximum serum creatinine levels, acute and chronic decreased cardiac output; release of proinflam­matory
hepatic injury, sepsis, shock, or male gender. Indeed, cytokines; 41–44 and impaired production of nitric
numerous reviews of the literature have been published oxide.35,45 Decreased microcirculation in the gastro-
on this topic over the past 20 years.8,16,19–26 According to intestinal tract is even more likely to be observed in
the American Society for Health-System Pharmacists patients with sepsis,39 even if sepsis is not unanimously
guidelines, major risk factors—such as mechanical venti­ described as an independent risk factor for SRMD.4,38
lation and coagulopathy—are only found in a minority As a consequence of splanchnic hypoperfusion gastro-
of patients admitted to the ICU.27 The main risk factors intestinal motility is decreased,46 delaying the removal
for SRMD are summarized in Table 1. of acidic material and other irritants from the stomach,
prolonging exposure to gastric acid and resulting in a
Pathophysiology of SRMD corresponding increase in the risk of ulceration.37,43,44
SRMD is a multifactorial disease of which the underlying The factors favouring gastroduodenal mucosal damage
mechanisms are not fully understood. Briefly, an intricate in patients in the ICU are a decrease in the washout effect47
system of interactions ultimately causes a breakdown leading to local intramural acidosis, increased free-
of the mucosal protective defences, leading to injury of radical­ formation,48,49 decreased acid-buffering capac-
the gastrointestinal wall and ulceration from aggressive ity,50 a decrease in mucus and bicarbonate secretion,49,51–53
physio­logical factors (Figure 1).17,26 Generally, SRMD and possibly a defect in epithelial cell r­estitution.54 The
lesions are located in the acid-producing areas of the end result of these effects is an increase in H+ ion back-
stomach (that is, the upper body and fundus). diffusion­and disruption of the epithelial barrier,37,47,50,55,56
The structural integrity of the gastric mucosa is ultimately allowing injury to the gastric mucosa. Pepsin
maintained by defence mechanisms,28 including the might cause direct injury to the gastric mucosa, but

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | FEBRUARY 2015 | 99

© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 1 | Risk factors for stress-ulcer bleeding from multiple regression analysis not. Of note, after the inclusion of old studies (16 of 20
were published before 1994), the meta-analysis mostly
Reference Risk factor for stress related OR (95% CI or
mucosal damage* P value)
assessed H2RAs and was unable to accurately compare
PPIs to H2RAs. Furthermore, in two meta-analyses
Cook et al. (1994)4 Respiratory failure requiring mechanical 15.6 (P <0.001)
ventilation for >48 h
published in 2012 and 2013—which included 13 RCTs
totalling 1,587 patients10 and 14 RCTs totalling 1,720
Laine et al. (2008)3, Coagulopathy (platelet count <50,000 4.3 (P <0.001)
Cook et al. (1994)4 platelets per ml3, INR >1.5, or
patients,12 respectively—PPIs were not found to be more
partial-thromboplastin time more than effective than H2RAs in reducing overall mortality (ICU
two times the control value) mortality OR = 1.19; 95% CI 0.84–1.68 and RR = 1.01;
Cook et al. (1999)20 Maximum serum creatinine level 1.16 (1.02–1.32) 95% CI 0.83–1.24, respectively). A study by Kotlyanskaya
MacLaren et al. (2014)27 Acute kidney injury 1.21 (1.02–1.43)
et al.60 (published in abstract form) was not included
in the meta-analysis by Barkun et al.10 and is the only
MacLaren et al. (2014)27 Acute respiratory failure 1.31 (1.10–1.56)‡
­difference between the two meta-analyses.
MacLaren et al. (2014) 27
Age >50 years (vs <40 years)
50–59 years 1.46 (1.18–1.83)‡
60–69 years 1.66 (1.26–2.19)‡ Data from observational cohorts
70–79 years 1.72 (1.27–2.34)‡ Few reports are available on the effect of SUP on overall
>80 years 2.04 (1.48–2.83)‡ or ICU mortality rates. A retrospective cohort study,
MacLaren et al. (2014)27 Acute hepatic injury 1.56 (1.29–1.88)‡ published in 2014, used information from the Colorado
MacLaren et al. (2014) 27
Chronic hepatic injury 1.85 (1.47–2.33)‡ Multiple Institutional Review Board and analysed data
from 35,312 patients requiring mechanical ventilation
MacLaren et al. (2014) 27
Sex: male versus female 1.17 (1.03–1.33)‡
for >24 h.38 Propensity-score-adjusted and propensity-
*Other risk factors that were identified by simple regression analysis, but lost statistical significance after
multiple regression include: hypotension; sepsis; hepatic failure; glucocorticoid administration; and score-matched multivariate regression models sug-
anticoagulation therapy. ‡Propensity score odds ratio. Abbreviation: INR, international normalized ratio. gested that PPIs were associated with an increased risk
of death in the ICU compared with H2RAs (15.3% versus
12.3%, respectively; P <0.001).38 Unfortunately, this study
becomes inactive if the pH is >4.5.37 In addition, pepsin cannot conclude on causation because of its design and
facilitates the lysis of clots, which could enhance bleeding the inability to completely remove confounding factors;
from SRMD.57,58 moreover, the authors did not assess patients in the ICU
who did not receive SUP and thus the net effect of PPIs
Pharmacological prophylaxis of SRMD or H2RA cannot be estimated.
Initiation of SUP using acid-suppressants in the ICU
setting is now established practice. Two landmark studies Prevention of bleeding
published in 1994 and 1999 by the Canadian Critical Results from randomized clinical trials
Care Trials Group serve as the benchmark for SUP in Results are more consistent in supporting a role for SUP
this context,4,20 and 1999–2010 estimates indicate that in reducing the risk of gastrointestinal bleeding. In the
~90% of patients in ICU receive some type of SUP.6 The Krag et al.59 meta-analysis, the risk of bleeding in the ICU
different approaches to prevent SRMD-related bleeding was reduced by 59% (fixed effect RR 0.41; 95% CI 0.31–
are based on the understanding of the pathophysiology 0.53) in patients treated with SUP compared with those
and recognized risk factors described earlier. treated with placebo or no prophylaxis. Interestingly,
the meta-analysis by Barkun et al.10 suggested the same
Effect on mortality magnitude of reduced risk of bleeding in patients treated
Results from randomized clinical trials with PPIs compared with those treated with H2RAs
No single study or meta-analysis has reported a (OR = 0.30; 95% CI 0.17–0.54), as did the meta-analysis
decreased overall mortality related to SUP approaches. by Alhazzani et al.,12 showing that both clinically impor-
To our knowledge, Krag and colleagues59 have published tant RR = 0.36; 95% CI 0.19–0.68) and overt bleeding (RR
the most recent meta-analysis of randomized clinical 0.35; 95% CI 0.21–0.59) were reduced in patients treated
trials comparing SUP versus placebo or no prophylaxis with PPIs compared with those treated with H2RAs. On
in 2014. In this analysis, 20 trials totalling 1,971 patients the contrary, another meta-analysis published in 2010
were included; all studies were judged as being at high did not find strong arguments in favour of PPIs over
risk of bias as advised by the Cochrane Collaboration. H2RAs for stress-related UGIB or for mortality.13
Among these 20 randomized controlled trials (RCTs),
16 were single-centre trials, 18 assessed H2RAs and two Data from observational cohorts
assessed PPIs. Using overall mortality as a primary end SUP has become the standard of care in the ICU, some-
point, no statistically significant difference was reported times irrespective of the presence of risk factors; the
between patients receiving SUP and those receiving no benefit of SUP using real-world data is therefore not easy
prophylaxis or placebo (fixed effect relative risk [RR] to estimate because of the absence of a control group.
1.00; 95% CI 0.84–1.20; P = 0.87; I2 = 0%).59 Similarly, A retrospective single-centre German study published
no difference was observed between treatment versus in 2014, in which 91.3% of the patients received SUP
placebo across all subgroup analyses; for example when (mostly PPIs), did not suggest any difference in terms
assessing H2RAs or PPIs, or patients fed enterally or of bleeding between no treatment and any of the SUP

100 | FEBRUARY 2015 | VOLUME 12  www.nature.com/nrgastro

© 2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Critical illness infections,10,65–67 and C. difficile infection and recur-

rence,68–73 are sources of ongoing controversy because of
Increased Shock or hypotension Hypovolaemia
methodological limitations in study design.
catecholamines
Results from randomized clinical trials
Cardiac output Proinflammatory
Vascoconstriction
cytokine release Two meta-analyses published in 2012 and 2013 of RCTs
comparing SUP agents did not provide evidence of an
Splanchnic and mucosal increased risk of nosocomial pneumonia attributable to
hypoperfusion PPIs. Indeed, the ORs were 1.05 (95% CI 0.69–1.62)10 and
1.06 (95% CI 0.73–1.52).12 Interpretation of these results
must be performed with caution as only seven studies
Bicarbonate Mucosal Gastrointestinal Protective factors Acid back- reported on nosocomial pneumonia. Interestingly, the
secretion blood flow motility e.g. HSP and TFF diffusion
meta-analysis that compared prophylaxis to placebo
or no prophylaxis did not suggest any increased risk of
no­socomial pneumonia in the prophylaxis group (RR
Acute stress-ulcer 1.16; 95% CI 0.84–1.58).59 One RCT conducted from
2010 to 2011 in Iran, in which 137 patients on mechani-
Figure 1 | Pathophysiology of stress-ulcer bleeding.| Gastroenterology
Abbreviations: HSP,
& heat shock
Nature Reviews Hepatology cal ventilation were randomly assigned treatment with
proteins; TFF, trefoil factor family peptides. Adapted with permission from Elsevier
© Stollman, N. & Metz, D, J. Crit. Care 20, 35–45 (2005).
sucralfate or pantoprazole, found a significant increase
in the incidence of pneumonia in the pantoprazole group
(36.4% versus 14.1%, respectively; P <0.001).74 Again,
agents that were used (including combinations of PPIs results from this study must be interpreted with caution
and H2RAs, PPIs and sucralfate or any other combi- because of methodol­ogical issues. No evidence exists
nation).6 The retrospective cohort study by MacLaren regarding the risk of C. difficile infection related to SUP
et al.,38 described earlier, suggested that PPIs were associ- as this a­ssociation was not a reported outcome in any of
ated with a greater risk of gastrointestinal haemorrhage the RCTs.
than H2RAs (34% versus 30.7%, respectively; P <0.001).
As this finding contradicts the hypothesis that stronger Data from observational cohorts
acid suppression reduces gastrointestinal bleeding, the A 1 day prevalence study conducted in 17 European
study authors suggest that whereas both drug classes countries in which 1,417 ICUs provided reports from
inhibit acid production, H2RAs might also limit reper­ 10,038 patients showed that 4,501 (44.8%) patients
fusion injury, possibly reducing oxidative stress after had infections, 2,064 (20.6%) of which were ICU-
mucosal injury.38 To conclude, in patients with recog- acquired infections, with Enterobacteriaceae (34.4%),
nized risk factors, good evidence supports the use of PPIs Staphylociccus aureus (30.1%) and Pseudomonas aerugi-
to prevent stress-ulcer bleeding. nosa (28.7%) being the three most frequent.75 Pneumonia
was the most frequent site of infection, accounting for
Evidence for adverse events caused by SUP 47% of the patients. SUP was an independent risk factor
The acidity of gastric contents represents the major for infection (OR 1.38; 95% CI 1.20–1.60).75 Buendgens
control factor for bacterial survival in the stomach. et al.6 found that PPIs were associated with the occurrence
Gastric and duodenal bacterial overgrowth frequently of pneumonia in univariate analysis (OR 1.79; 95% CI
occurs in conditions characterized by diminished 1.47–2.02) but not in multivariate analysis (OR 1.28; 95%
gastric acid secretion.61 Thorens et al.62 conducted an CI 0.95–1.73). Sucralfate was confirmed to have a protec-
RCT in which 47 outpatients were randomly allocated tive effect against infection (OR 0.68; 95% CI 0.50–0.73).6
to receive 4 weeks of 20 mg omeprazole or 800 mg cime- MacLaren et al.38 only compared PPIs to H2RAs, finding
tidine, daily. Bacterial overgrowth, assessed by culture that PPIs were associated with a significant increase in
of gastroduodenal contents, was present in 53% of the ICU-acquired pneumonia incidence (34% versus 30.7%,
patients receiving omeprazole and 17% of those receiv- respectively; P <0.001). A case–control study of 110
ing cimetidine. A subsequent study also showed that patients with ventilator-associated ­pneumonia and
both PPIs and H2RAs induced bacterial overgrowth, 110 ventilated controls without pneumonia did not find
with a substantially more-pronounced effect with PPIs SUP to be a significant risk factor for pneumonia (OR
than H2RAs.63 Furthermore, leucocyte function might 1.01; 95% CI 1.00–1.02; P = 0.08).76
be impaired in patients taking PPIs because of an inhi- Cohort or case–control studies published within the
bition of reactive oxygen intermediate production with past 10 years almost unanimously suggest an increased
consequential reduction of neutrophil bactericidal risk of acquiring C. difficile infection in patients receiv-
activity.64 The two types of infections most frequently ing SUP in the ICU, even if the absolute risk is small
discussed with the use of acid-suppressive agents are compared with that for nosocomial pneumonia. A case–
nosocomial (or ventilator-associated) pneumonia and control study including 67 patients and 134 controls
diarrhoea associated with Clostridium difficile infection. (20 controls and 40 patients with C. difficile infecction
The association between the use of acid-suppressive were ICU patients) reported duration of PPI use to be
agents and c­ommunity-acquired or hospital-acquired positively associated with the risk of developing C. difficile

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | FEBRUARY 2015 | 101

© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

infection (OR 1.14; 95% CI 1.02–1.27).77 Buendgens et al.6 For patients receiving enteral nutrition or medication
reported an OR of 3.11 (95% CI 1.11–8.44) for patients of any type there is no reason to prescribe intravenous
receiving PPIs versus no SUP, which is much stronger PPIs, enteral route is the most appropriate. SUP should
than the risk associated with the use of third-generation not be given to low-risk patients because of the potential
cephalosporins (OR 1.8; 95% CI 1.0–3.23) or fluoro­ increased risk of nosocomial infection.
quinolone (OR 1.87; 95% CI 1.1–3.14); no such risk was
described for H2RAs. MacLaren et al.38 also suggested Enteral nutrition for prophylaxis of SRMD
the risk of C. difficile infection to be higher in patients The use of acid blockers or antisecretory agents in
on PPIs than H2RAs (3.4% versus 2.6%, respectively; patients in the ICU has been questioned because of
P = 0.002). Interestingly, an online survey conducted uncertain efficacy and safety concerns. Furthermore, the
among Australian and New Zealand intensive care phy- rationale for PPI or H2RA use might be considered weak
sicians suggests that C. difficile is perceived as less of a as decreased mucosal blood flow with subsequent tissue
threat (14% of the physicians perceived C. difficile as ischaemia is thought to be the mechanism responsible for
a threat) than ventilator-associated pneumonia (21% of stress-induced gastropathy, with an amplification loop
the physicians).78 driven by subsequent activation of inflam­matory and
vasoconstrictive mediators, a mechanism that PPIs
Risk:benefit ratio of SUP and/or H2RA are unlikely to have an effect on. Enteral
Defining the risk:benefit ratio of SUP in patients who are nutrition has been suggested to improve mucosal blood
critically ill or on mechanical ventilation is not easy, as flow and reverse the generation of these inflammatory
both bleeding and ICU-acquired infections carry a risk mediators.82 However, clinical studies evaluating the
of increased death in ICU (OR 1.59; 95% CI 1.06–2.41 effectiveness of enteral nutrition versus acid-suppressiv­e
C. difficile-associated diarrhoea; OR 1.82; 95% CI 1.52– medications have shown variable results.83 No RCTs to
2.18 nosocomial infection; OR 2.5; 95% CI 1.19–5.25 date have assessed enteral nutrition as SRMD-related
UGIB).6 Overall, mortality might be the best reflec- bleeding prophylaxis in critically ill patients. One
tion of a risk:benefit analysis. Meta-analyses from RCTs RCT published in the early 1990s only described that
report no effect on mortality, whereas data from observa- additional prophylaxis in the form of sucralfate or
tional studies suggest—although not unanimously—that ranitidine added no protective effect to total parenteral
PPIs might increase mortality compared with H2RAs.38 nutrition alone.84
Nevertheless, as there are strong arguments in favour of a Marik et al.1 performed a systematic review of the lit-
protective effect of PPIs in the prophylaxis of stress-ulcer erature to investigate the benefits and risks of SUP in the
bleeding, with no clearly proven increased risk of adverse ICU setting and the possible influence of enteral nutri-
events, the risk:benefit ratio of SUP in high-risk patients tion. A total of 1,836 patients across 17 studies were
can be considered positive. included in the analysis and they found that the reduced
risk of gastrointestinal bleeding with H2RAs (OR 0.47;
Actual trends in SRMD prophylaxis 95% CI 0.29–0.76) was only observed in patients not
Despite the vast number of publications on the topic that receiving enteral nutrition. Furthermore, whereas
have defined the most relevant risk factors for SRMD H2RAs apparently did not carry an increased risk of
and discussed the use of sucralfate, PPIs or H2RAs, nosocomial pneumonia in fasting patients, a signifi-
epidemiological studies published in 2014 highlight cant increase in risk was observed among enterally fed
that the proportion of patients in ICU receiving SUP is patients (OR 2.81; 95% CI 1.2–6.56; P = 0.02).1 Similarly,
>90% and although most of them receive PPIs, a sub- hospital mortality was also found to be increased in
stantial proportion (up to 20%) receive a combination patients receiving enteral nutrition and H2RA therapy
of different medications, including PPIs and H2RAs.6,38,79 (OR 1.89; 95% CI 1.04–3.44; P = 0.04), suggesting that
Additionally, it has been shown that, although most SUP might be unnecessary and possibly deleterious in
of the patients receiving SUP in ICU present with at this patient population.1 Interestingly, studies originating
least one recognized risk factor, continuation of acid- from the same team have suggested that enteral feeding
suppressiv­e therapy during hospitalization outside the can either have no effect on clinically important bleeding
ICU occurs in >86% of patients and that 24.2% are (multiple regression OR 1.0; P = 0.99),4 or may decrease
being discharged from the hospital still receiving acid- the risk by 70% (OR: 0.30; 95% CI 0.13–0.67; P = ns).20
suppressive therapy.80 The cost implications of this issue Unfortunately, shortcomings in the identification of who
are discussed later. had received enteral nutrition and the lack of per-patient
Among patients with no risk factors it is indeed of analyses limit any conclusions that can be made.
concern that up to 68.1% are being placed on prophylaxis
at ICU admission; 60.4% continue treatment upon trans- Management of UGIB in the ICU
fer from the ICU and about 30% are being discharged Critical care physician perspective
home on an agent without a new indication.79,81 Most of the guidelines that have been published on man-
In conclusion, it seems reasonable to recommend aging gastrointestinal bleeding in the ICU are based on
SUP (preferably with PPIs) in high-risk patients, that is, the non-ICU literature as there is not enough supporting
those with well-defined risk factors such as mechanical evidence to specifically develop treatment paths.85,86 An
ventilation for at least 48 h and coagulation disorders. urgent need exists to produce evidence as some of the

102 | FEBRUARY 2015 | VOLUME 12  www.nature.com/nrgastro

© 2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

guidelines published outside the ICU setting 87 might not Enteral nutrition must be stopped and gastric content
be adaptable to patients bleeding in ICU. We describe removed with soft aspiration to enable gastroscopy, or
the management of UGIB in the ICU on the basis of prokinetic agents considered, as discussed later. For the
expert opinion. same reason, sedation and analgesia should be re­inforced
The management of patients with SRMD is similar to in patients receiving mechanical ventilation. In patients
that of other patients in the ICU presenting with non- on anticoagulants, coagulopathy is usually reversed, using
variceal UGIB, keeping in mind the commonly more vitamin K, vitamin‑K-dependent factors or fibrinogen;
severe disease acuity and multiple comorbidities often however, this process should not delay early endoscopy,
present in this ICU patient population. Detailed reviews defined as within 24 h of acute UGIB, keeping in mind
and contemporary consensus recommendations have that endoscopic haemostasis can be safely performed in
been published on the management of patients with patients with an elevated INR as long as it is not supra­
nonvariceal UGIB.85,87–89 The approach might, of course, therapeutic (that is, up to around 2.5).87,88 A platelet trans-
be altered if the critically ill patient is suspected to have fusion threshold of 50 × 109/l has been proposed for most
variceal bleeding. 90 Although airway, breathing and patients, with a target of 100 × 109/l for patients in whom
circulation remain the most crucial steps in the initial platelet dysfunction is suspected.98
assessment of patients presenting with acute UGIB, these Well-validated risk stratification scoring systems in
factors have usually already been optimized in an ICU the setting of UGIB can help identify the ideal time to
setting. A minimum blood work-up in all patients should perform endoscopy.99 The Rockall score can be calcu-
include bloodtyping and crossmatching for an appropri- lated using data for seven pre-endoscopic factors (clinical
ate number of units of packed red blood cells along with Rockall) and 11 postendoscopic factors.100 The use of pro-
determinations of haemoglobin, haematocrit, platelets kinetics before endoscopy might be considered in selected
and electrolyte levels, and coagulation time. Depending patients. Indeed, meta-analyses show that erythromycin
on the severity of the acute bleed and if it is indicated is associated with a decreased need for repeat endoscopy
(mainly when prothrombin time and factor V are <50% in patients with evidence of ongoing active bleeding and
of the normal range), prompt and appropriate resuscita- blood in the stomach (haematemesis, coffee ground vom-
tion with either crystalloids or colloids, packed red blood iting, or bloody nasogastric aspirate).101 However, use of
cells and fresh frozen plasma, is required.87,88 erythromycin has failed to change outcomes in terms
Vitamin K, vitamin‑K-dependent factors or fibrino- of length of stay, transfusion requirements and need for
gen are usually administered in case of anticoagulant surgery in patients with nonvariceal UGIB.
supratherapeutic dosing. Saline or Ringer acetate are pre- Although many patients might already be on a PPI in the
ferred to hydroxyethyl starch, as hydroxyethyl starch has ICU, it is nonetheless pertinent that a Cochrane system-
been shown to increase the need for renal-replacement atic meta-analysis found no evidence that pre-endoscopi­c
therapy in patients in the ICU and might even increase administration of PPIs led to a reduction in the most
the risk of severe bleeding (RR 1.52; 95% CI 0.94–2.48; important clinical outcomes, but this practice does result in
P = 0.09).91,92 A systematic review of 10 RCTs comparing a downstaging of high-risk endoscopic ulcers into low-risk
restrictive versus liberal red blood cell transfusion strate- lesions.102 Therefore, this approach might be cost-effective
gies in 1,780 patients with suspected UGIB from a variety when early endoscopy is not feasible or local expertise is
of clinical settings, including the ICU and patients with limited. The use of pre-endoscopic PPIs, however, should
SRMD, concluded that a restrictive approach led to a not replace appropriate initial resuscitation or delay the
42% reduction in the probability of receiving transfu- performance of early endoscopy.103 The optimal dose
sions with no effect on mortality, rates of cardiac events, remains unknown but many clinicians choose to use the
morbidity, or length of hospital stay.93 These data support high-dose intravenous PPI infusion of an 80 mg bolus fol-
a restrictive strategy with a haemoglobin transfusional lowed by 8 mg/h infusion,104 as the highest quality evidence
threshold value of <70 g/l for transfusion, which is comes from trials using this regimen.
further strengthened by the results of a recent large RCT
in patients with variceal and nonvariceal UGIB pub- The gastroenterologist perspective
lished in 2013.94 Nevertheless, in critically ill patients On the basis of RCT data, current management rec-
a haemoglobin transfusional level ~90 g/l has been ommendations for UGIB suggest the performance of
proposed, mostly in case of coronary artery disease,95 early endoscopy within 24 h of presentation.87,88 In the
although red blood cell transfusion increases oxygen ICU a more urgent endoscopy might be considered as
delivery but not oxygen consumption.96 The objective it has been suggested that in patients with a Glasgow–
is to maintain mean arterial pressure >65 mmHg and Blatchford score >12, mortality is reduced when endo­
urinary excretion >0.5 ml/kg/h. scopy is performed within the first 13 h.105 In patients
Placement of a nasogastric tube might be appropri- with massive haematochezia and/or in case of haemo-
ate in the ICU, mostly to assess for rebleeding as altered dynamic instability guidelines published in 2006 recom-
gastrointestinal motility and the use of vasopressive mend performing upper endoscopy as soon as possible
drugs might delay exteriorization of bleeding or haemo- in the presence of this symptom as it will be a presenting
dynamic changes. However, as the negative predictive feature in ~10% of patients ultimately found to have an
value of a nasogastric tube aspirate is low, around 60%, upper gastrointestinal source of bleeding.106 The treat-
its usefulness is questioned.97 ment of ulcers (if noted at endoscopy) is well defined

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | FEBRUARY 2015 | 103

© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

and outlined later. The treatment of other lesions, such the hospital setting.114 The second decision model con-
as Dieulafoy ulcers or mucosal vascular abnormalities cluded that PPI prophylaxis was the most cost-effective,
has been less well studied, but in the presence of one in fact economically dominant, prophylactic strategy.115
or multiple lesions that are thought to be the source of However, in both cost-effectiveness analyses, the prob-
bleeding, an assessment of the presence of high-risk stig- ability of ventilator-associated pneumonia occurrence
mata of the lesion(s) is performed. Whether gastroduo- affected the aforementioned conclusions even though
denal ulcers, erosions or oesophagitis are noted, if only it remains controversial as to whether or not acid sup-
low-risk stigmata are present, single intravenous PPI pression increases the risk of this iatrogenic complica-
therapy is indicated without endoscopic haemostasis. tion. The discrepancy in these conclusions highlights
Endoscopic therapy is recommended only for high-risk the poor quality, small sample sizes and limited gen-
endoscopic lesions and is the cornerstone of manage- eralizability of contemporary care trials, emphasizing
ment with modalities that include injection, mechani- the need for prospective controlled assessments of both
cal and thermal treatments.87,88 Haemostatic powders bleeding and possible infectious complications.
have now been adapted for use in the gastrointestinal A number of society guidelines have been published
tract,107 resulting in substantial reductions in rebleeding over the past 15 years in an attempt to optimally guide
in select patients, but reported benefits on mortality and acid-suppression use for SRMD prophylaxis, with most
rates of surgical intervention are variable, although the published before the flurry of meta-analytical studies
data are purely observational and patient numbers are discussed above. These society guidelines include an
limited.87,88 Following successful endoscopic haemosta- authoritative set of recommendations by the American
sis, PPI therapy has further improved outcomes. A high- Society for Health-System Pharmacists that were devel-
dose intravenous PPI approach has been favoured by oped for the most part before the PPI era.116 In 2008, the
guidelines (80 mg bolus if the patient has not been on a Eastern Association for the Surgery of Trauma guidelines
PPI, followed by 8 mg/h for 72 h); however, the optimal suggested no differences in efficacy between H2RAs and
dose and route of administration remain unclear.87,88 PPIs.117 In 2013, although no specific trials have inves-
The acute management of patients on antiplatelet tigated the benefit of SUP in patients with sepsis, the
agents is based on weighing risks and benefits, but RCT Surviving Sepsis Campaign guidelines favoured the use
data support the early reintroduction of acetylsalicylic of PPIs over H2RAs for SRMD prophylaxis.96
acid within 3–5 days of the acute haemorrhage in patients Although uncertainty exists about the optimal strat-
with bleeding ulcers who are taking acetylsalicylic acid egy for effective SRMD prophylaxis in high-risk patients,
for secondary cardiovascular disease prevention.108 The it is well established that antisecretory therapy is not
theoretical interaction between clopidogrel and PPI warranted in patients at low risk, as discussed earlier.
seems unlikely to be of clinical importance and should Although the definitions of inappropriate use have varied
not deter physicians from administering PPIs, if clini- among studies and with the publication of the latest
cally indicated.109 The long-term secondary prophylaxis meta-analytical data, inappropriate in-hospital use is
of patients with SRMD-related bleeding has not been common. Outside the ICU, in-hospital PPI use has been
studied, but it is reasonable to at least consider applying shown to vary between 50% and 60% of patients, with
the same recommendations as for bleeding ulcers occur- 18–34% of patients erroneously discharged on a PPI into
ring outside the ICU. However, discussion of this idea is the community with attendant inpatient and outpatient
beyond the scope of this Review.87,88 costs.77 A retrospective chart review of 1,769 patients
admitted to family medicine and general internal medi-
Cost considerations and guidelines cine teaching services demonstrated an inappropriate
In 1995, a Canadian matched cohort study identified in-hospital acid-suppressive agent use of 22.1% (pre-
clinically important bleeding to be associated with per sumably in large part because of unjustified concerns of
patient costs of ~CND$12,215.110 Schupp et al.111 subse- SRMD).118 In total, 54% of these patients were discharged
quently performed an economic analysis, but unfortu- on antisecretory medication that were deemed unneces-
nately limited the costs solely to drug acquisition. More sary upon audit review, with an estimated annual inpa-
important cost-drivers are known to include possible tient and outpatient cost of US$67,000. A 2010 analysis
complications of SRMD and its prophylaxis,112 daily costs of a US national administrative database demonstrated
and duration of hospital stay, especially in the ICU. a ~69% rate of inappropriate prescribing for SRMD
Two decision models have assessed the cost- prophylaxis in ICU or Coronary Care Unit.119 Resulting
effectivenes­s of SRMD prophylaxis using either H2RA overall hospital and discharge medication costs might
or PPI therapies during hospitalization. One study reach US$753,267 per year nationally over the 30 days
concluded that H2RA therapy seems to reduce costs after leaving the hospital.118
with comparable survival benefit compared with PPIs Although early interventional studies suggested
for SUP.113 However, this study adopted mortality as a that the implementation of guidelines might impact
primary effectiveness measure, an outcome that has favourably on practice and resource utilization in the
not been shown to be affected by SRMD prophylaxis, as ICU,120,121 a 2013 national cluster RCT demonstrated no
discussed earlier.10,12 Also, this analysis included the pos- differences in hospital use of PPIs for nonvariceal UGIB
sible complication of C. difficile infection, even though following a structured multidisciplinary multifaceted
this association with PPI use remains highly debated in educational intervention.122

104 | FEBRUARY 2015 | VOLUME 12  www.nature.com/nrgastro

© 2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

New approaches still associated with substantial mortality and costs, and
New therapeutic approaches might be assessed based on prospective trials comparing modern-day prophylactic
an improved understanding of the underlying mecha- approaches, including enteral nutrition, are needed. The
nisms. For example, increasing gastrointestinal micro- almost universal (>90%) prescription of SUP in critically
vascular blood flow might be beneficial. As discussed ill patients must be questioned and the risk:benefit ratio
previously, improved management of cardiac output in reassessed because of the decrease in incidence of SMRD.
the ICU can explain the decrease in SRMD over time, In addition to randomized trials, there is a need for pro-
although the link between recognized risk factors, spective cohort studies to better quantify and understand
such as sepsis, and SRMD might not involve decreased outcomes in the real-word setting.
microcirculation.123 As reported in a 2014 review, several Treatment strategies need to be evaluated on a case-
studies have attempted to increase microcirculatory flow by-case basis, taking into account the expected benefits
parameters but with limited efficacy.124 Other pharmaco­ of therapy balanced against the patient’s risk factors
logical options to prevent SRMD can be hypothesized for bleeding and the risk of nosocomial pneumonia or
based on its pathophysiology. For example, gastric C. difficile infection in the local epidemiological context.
bicarbonate secretion is stimulated by nitric oxide in a Generally speaking, SUP is not warranted in non-ICU
process that involves cyclic GMP, under the control of patients, except where confirmed clinical indications
phosphodiesterase 1 and 5; phosphodiesterase 5 inhibi- exist. Educational programmes are needed to decrease
tors, such as sildenafil, might increase bicarbonate secre- unjustified overuse of SUP in patients in ICUs127 and
tion. Additionally, phosphodiesterase 5 inhibitors might promote appropriate treatment discontinuation after
increase mucosal blood flow, as it has been shown that ICU and upon hospital discharge.
vardenafil protects the digestive tract after ischaemia–
reperfusion injury.125 As HSPs (mostly HSP70) protect
the gastric mucosa through inhibition of apoptosis, pro- Review criteria
inflammatory cytokines and cell adhesion molecules, We conducted a literature search using the OVID,
regulators that enhance HSP70 expression could rep- MEDLINE, EMBASE, PubMed, and ISI Web of Knowledge
resent interesting agents for the prevention of SRMD. 4.0 databases to identify articles published in English or
Unfortunately, most of the HSP70-targeted compounds French from January 1994 to 15th October 2014. A highly
that are being developed are inhibitors.126 Moreover, it sensitive search strategy was used to identify randomized
is highly challenging to design and power clinical trials controlled trials, cohort and case–control studies
conducted in adults, using combinations of search terms,
aimed to compare new therapeutic options to PPIs or
including: “stress ulcer bleeding”, “stress related mucosal
H2RAs. Indeed, with an actual prevalence of SMRD lesion”, “stress related mucosal disease”, “stress
of ~1%, more than 10,000 patients would have to be ulcer”, “gastroduodenal damages”, “intensive care unit”,
included to prove a 50% relative risk reduction. “critically ill”, “epidemiology”, “prevention”, “prophylaxis”,
“guidelines”, “histamine–2 receptor antagonists”,
Conclusions “sucralfate” and “proton pump inhibitors”. In addition,
Stress-ulcer bleeding is an increasingly less frequent con- recursive searches and cross-referencing were performed
dition in the contemporary era of markedly improved and manual searches of the reference lists of articles
identified in the initial search were completed.
overall care for critically ill patients. However, SRMD is

1. Marik, P. E., Vasu, T., Hirani, A. & Pachinburavan, M. 7. Spirt, M. J. Stress-related mucosal disease: stress ulcer prophylaxis in critically ill patients:
Stress ulcer prophylaxis in the new millennium: risk factors and prophylactic therapy. Clin. Ther. a systematic review and meta-analysis. Crit. Care
a systematic review and meta-analysis. Crit. Care 26, 197–213 (2004). Med. 41, 693–705 (2013).
Med. 38, 2222–2228 (2010). 8. Stollman, N. & Metz, D. C. Pathophysiology 13. Lin, P. C., Chang, C. H., Hsu, P. I., Tseng, P. L.
2. Laine, L., Takeuchi, K. & Tarnawski, A. Gastric and prophylaxis of stress ulcer in intensive & Huang, Y. B. The efficacy and safety of proton
mucosal defense and cytoprotection: bench to care unit patients. J. Crit. Care 20, 35–45 pump inhibitors vs histamine-2 receptor
bedside. Gastroenterology 135, 41–60 (2008). (2005). antagonists for stress ulcer bleeding prophylaxis
3. Laine, L. et al. Lower gastrointestinal events in 9. Masson, S. C., Mabasa, V. H., Malyuk, D. L. among critical care patients: a meta-analysis.
a double-blind trial of the cyclo‑oxygenase‑2 & Perrott, J. L. Validity evidence for FASTHUG- Crit. Care Med. 38, 1197–1205 (2010).
selective inhibitor etoricoxib and the traditional MAIDENS, a mnemonic for identifying drug- 14. Peura, D. A. & Johnson, L. F. Cimetidine for
nonsteroidal anti-inflammatory drug diclofenac. related problems in the intensive care unit. prevention and treatment of gastroduodenal
Gastroenterology 135, 1517–1525 (2008). Can. J. Hosp. Pharm. 66, 157–162 (2013). mucosal lesions in patients in an intensive care
4. Cook, D. J. et al. Risk factors for gastrointestinal 10. Barkun, A. N., Bardou, M., Pham, C. Q. unit. Ann. Intern. Med. 103, 173–177 (1985).
bleeding in critically ill patients. Canadian & Martel, M. Proton pump inhibitors vs. 15. Skillman, J. J. & Silen, W. Acute gastroduodenal
Critical Care Trials Group. N. Engl. J. Med. 330, histamine 2 receptor antagonists for stress- “stress” ulceration: barrier disruption of varied
377–381 (1994). related mucosal bleeding prophylaxis in critically pathogenesis? Gastroenterology 59, 478–482
5. Reintam Blaser, A. et al. Gastrointestinal ill patients: a meta-analysis. Am. J. Gastroenterol. (1970).
symptoms during the first week of intensive care 107, 507–520 (2012). 16. Zandstra, D. F. & Stoutenbeek, C. P. The virtual
are associated with poor outcome: a prospective 11. Leontiadis, G. I. et al. Systematic reviews of absence of stress-ulceration related bleeding in
multicentre study. Intensive Care Med. 39, the clinical effectiveness and cost-effectiveness ICU patients receiving prolonged mechanical
899–909 (2013). of proton pump inhibitors in acute upper ventilation without any prophylaxis. A prospective
6. Buendgens, L. et al. Administration of proton gastrointestinal bleeding. Health Technol. Assess. cohort study. Intensive Care Med. 20, 335–340
pump inhibitors in critically ill medical patients 11, 1–164 (2007). (1994).
is associated with increased risk of developing 12. Alhazzani, W., Alenezi, F., Jaeschke, R. Z., 17. Duerksen, D. R. Stress-related mucosal
Clostridium difficile-associated diarrhea. J. Crit. Moayyedi, P. & Cook, D. J. Proton pump inhibitors disease in critically ill patients. Best Pract. Res.
Care 29, 696.e11–696.e15 (2014). versus histamine 2 receptor antagonists for Clin. Gastroenterol. 17, 327–344 (2003).

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | FEBRUARY 2015 | 105

© 2015 Macmillan Publishers Limited. All rights reserved
REVIEWS

18. Alhazzani, W., Alshahrani, M., Moayyedi, P. 37. Fennerty, M. B. Pathophysiology of the upper 57. Samloff, I. M. Peptic ulcer: the many proteinases
& Jaeschke, R. Stress ulcer prophylaxis in gastrointestinal tract in the critically ill patient: of aggression. Gastroenterology 96, 586–595
critically ill patients: review of the evidence. rationale for the therapeutic benefits of acid (1989).
Pol. Arch. Med. Wewn. 122, 107–114 (2012). suppression. Crit. Care Med. 30 (Suppl.), 58. Schiessel, R., Feil, W. & Wenzl, E. Mechanisms
19. Zimmerman, J. E., Kramer, A. A. & Knaus, W. A. 351–355 (2002). of stress ulceration and implications for
Changes in hospital mortality for United States 38. MacLaren, R., Reynolds, P. M. & Allen, R. R. treatment. Gastroenterol. Clin. North Am. 19,
intensive care unit admissions from 1988 to Histamine-2 receptor antagonists vs proton pump 101–120 (1990).
2012. Crit. Care 17, R81 (2013). inhibitors on gastrointestinal tract hemorrhage 59. Krag, M., Perner, A., Wetterslev, J., Wise, M. P.
20. Cook, D. et al. Risk factors for clinically and infectious complications in the intensive care & Hylander Moller, M. Stress ulcer prophylaxis
important upper gastrointestinal bleeding in unit. JAMA Intern. Med. 174, 564–574 (2014). versus placebo or no prophylaxis in critically ill
patients requiring mechanical ventilation. 39. Chierego, M., Verdant, C. & De Backer, D. patients. A systematic review of randomised
Canadian Critical Care Trials Group. Crit. Care Microcirculatory alterations in critically ill clinical trials with meta-analysis and trial
Med. 27, 2812–2817 (1999). patients. Minerva Anestesiol. 72, 199–205 sequential analysis. Intensive Care Med. 40,
21. Schuster, D. P., Rowley, H., Feinstein, S., (2006). 11–22 (2014).
McGue, M. K. & Zuckerman, G. R. Prospective 40. Bailey, R. W. et al. The fundamental 60. Kotlyanskaya, A., Luka, B. & Mukherji, R. A
evaluation of the risk of upper gastrointestinal hemodynamic mechanism underlying gastric comparison of lansoprazole disintegrating tablet,
bleeding after admission to a medical intensive “stress ulceration” in cardiogenic shock. lansoprazole suspension or ranitidine for
care unit. Am. J. Med. 76, 623–630 (1984). Ann. Surg. 205, 597–612 (1987). stress ulcer prophylaxis in critically ill patients
22. Beejay, U. & Wolfe, M. M. Acute gastrointestinal 41. Yasue, N. & Guth, P. H. Role of exogenous acid [abstract 194]. Crit. Care Med. 7 (2008).
bleeding in the intensive care unit. The and retransfusion in hemorrhagic shock-induced 61. Drasar, B. S., Shiner, M. & McLeod, G. M. Studies
gastroenterologist’s perspective. Gastroenterol. gastric lesions in the rat. Gastroenterology 94, on the intestinal flora. I. The bacterial flora of the
Clin. North Am. 29, 309–336 (2000). 1135–1143 (1988). gastrointestinal tract in healthy and achlorhydric
23. Steinberg, K. P. Stress-related mucosal disease in 42. Cook, D. J. et al. The attributable mortality and persons. Gastroenterology 56, 71–79 (1969).
the critically ill patient: risk factors and strategies length of intensive care unit stay of clinically 62. Thorens, J. et al. Bacterial overgrowth during
to prevent stress-related bleeding in the intensive important gastrointestinal bleeding in critically ill treatment with omeprazole compared with
care unit. Crit. Care Med. 30, 362–364 (2002). patients. Crit. Care 5, 368–375 (2001). cimetidine: a prospective randomised double
24. Daley, R. J., Rebuck, J. A., Welage, L. S. 43. Mutlu, G. M., Mutlu, E. A. & Factor, P. GI blind study. Gut 39, 54–59 (1996).
& Rogers, F. B. Prevention of stress ulceration: complications in patients receiving mechanical 63. Wang, K. et al. The effect of H2-receptor
current trends in critical care. Crit. Care Med. 32, ventilation. Chest 119, 1222–1241 (2001). antagonist and proton pump inhibitor on
2008–2013 (2004). 44. Martindale, R. G. Contemporary strategies for the microbial proliferation in the stomach.
25. Quenot, J. P. et al. Bedside adherence to clinical prevention of stress-related mucosal bleeding. Hepatogastroenterology 51, 1540–1543 (2004).
practice guidelines in the intensive care unit: the Am. J. Health Syst. Pharm. 62 (Suppl. 2), 11–17 64. Zedtwitz-Liebenstein, K. et al. Omeprazole
TECLA study. Intensive Care Med. 34, 1393–1400 (2005). treatment diminishes intra- and extracellular
(2008). 45. Perry, M. A., Wadhwa, S., Parks, D. A., Pickard, W. neutrophil reactive oxygen production
26. Quenot, J. P., Thiery, N. & Barbar, S. When should & Granger, D. N. Role of oxygen radicals in and bactericidal activity. Crit. Care Med. 30,
stress ulcer prophylaxis be used in the ICU? ischemia-induced lesions in the cat stomach. 1118–1122 (2002).
Curr. Opin. Crit. Care 15, 139–143 (2009). Gastroenterology 90, 362–367 (1986). 65. Giuliano, C., Wilhelm, S. M. & Kale-Pradhan, P. B.
27. Frandah, W., Colmer-Hamood, J., Nugent, K. 46. Stupak, D. P., Abdelsayed, G. G. & Soloway, G. N. Are proton pump inhibitors associated with the
& Raj, R. Patterns of use of prophylaxis for Motility disorders of the upper gastrointestinal development of community-acquired
stress-related mucosal disease in patients tract in the intensive care unit: pathophysiology pneumonia? A meta-analysis. Expert Rev. Clin.
admitted to the intensive care unit. J. Intensive and contemporary management. J. Clin. Pharmacol. 5, 337–344 (2012).
Care Med. 29, 96–103 (2014). Gastroenterol. 46, 449–456 (2012). 66. Filion, K. B. et al. Proton pump inhibitors and the
28. deFoneska, A. & Kaunitz, J. D. Gastroduodenal 47. Haglund, U. Stress ulcers. Scand. J. Gastroenterol. risk of hospitalisation for community-acquired
mucosal defense. Curr. Opin. Gastroenterol. 26, Suppl. 175, 27–33 (1990). pneumonia: replicated cohort studies with meta-
604–610 (2010). 48. Itoh, M. & Guth, P. H. Role of oxygen-derived analysis. Gut 63, 552–558 (2014).
29. Ferrer-Ferrer, M. et al. Polymorphisms in genes free radicals in hemorrhagic shock-induced 67. Eom, C. S. et al. Use of acid-suppressive drugs
coding for HSP-70 are associated with gastric gastric lesions in the rat. Gastroenterology 88, and risk of pneumonia: a systematic review and
cancer and duodenal ulcer in a population 1162–1167 (1985). meta-analysis. CMAJ 183, 310–319 (2011).
at high risk of gastric cancer in Costa Rica. 49. Smith, P., O’Brien, P., Fromm, D. & Silen, W. 68. Dial, S., Delaney, J. A., Barkun, A. N. & Suissa, S.
Arch. Med. Res. 44, 467–474 (2013). Secretory state of gastric mucosa and Use of gastric acid-suppressive agents and the
30. Tahara, T. et al. Role of heat-shock protein (HSP) resistance to injury by exogenous acid. Am. J. risk of community-acquired Clostridium difficile-
70–72 genotype in peptic ulcer in Japanese Surg. 133, 81–85 (1977). associated disease. JAMA 294, 2989–2995
population. Hepatogastroenterology 59, 426–429 50. Marrone, G. C. & Silen, W. Pathogenesis, (2005).
(2012). diagnosis and treatment of acute gastric 69. Tleyjeh, I. M. et al. Association between proton
31. Takahashi, T. et al. Correlation of heat shock mucosal lesions. Clin. Gastroenterol. 13, pump inhibitor therapy and Clostridium difficile
protein expression to gender difference in 635–650 (1984). infection: a contemporary systematic review and
development of stress-induced gastric mucosal 51. Dzienis, H., Gronbech, J. E., Varhaug, J. E., meta-analysis. PLoS ONE 7, e50836 (2012).
injury in rats. J. Clin. Biochem. Nutr. 47, 64–73 Lekven, J. & Svanes, K. Regional blood flow 70. Shivashankar, R. et al. Clinical factors
(2010). and acid secretion associated with damage and associated with development of severe-
32. Hawkey, C. J. & Rampton, D. S. Prostaglandins restitution of the gastric surface epithelium in complicated Clostridium difficile infection. Clin.
and the gastrointestinal mucosa: are they cats. Eur. Surg. Res. 19, 98–112 (1987). Gastroenterol. Hepatol. 11, 1466–1471 (2013).
important in its function, disease, or treatment? 52. Durham, R. M. & Shapiro, M. J. Stress gastritis 71. Janarthanan, S., Ditah, I., Adler, D. G. &
Gastroenterology 89, 1162–1188 (1985). revisited. Surg. Clin. North Am. 71, 791–810 Ehrinpreis, M. N. Clostridium difficile-associated
33. Flemstrom, G. & Turnberg, L. A. Gastroduodenal (1991). diarrhea and proton pump inhibitor therapy:
defence mechanisms. Clin. Gastroenterol. 13, 53. Synnerstad, I., Johansson, M., Nylander, O. a meta-analysis. Am. J. Gastroenterol. 107,
327–354 (1984). & Holm, L. Intraluminal acid and gastric mucosal 1001–1010 (2012).
34. Phillipson, M., Atuma, C., Henriksnas, J. integrity: the importance of blood-borne 72. Kwok, C. S. et al. Risk of Clostridium difficile
& Holm, L. The importance of mucus layers and bicarbonate. Am. J. Physiol. Gastrointest. Liver infection with acid suppressing drugs and
bicarbonate transport in preservation of gastric Physiol. 280, 121–129 (2001). antibiotics: meta-analysis. Am. J. Gastroenterol.
juxtamucosal pH. Am. J. Physiol. Gastrointest. 54. Szabo, S. Mechanisms of gastric mucosal 107, 1011–1019 (2012).
Liver Physiol. 282, 211–219 (2002). injury and protection. J. Clin. Gastroenterol. 13 73. Pant, C., Madonia, P. & Minocha, A. Does PPI
35. Flynn, R. et al. Stress ulceration and gastric (Suppl. 2), 21–34 (1991). therapy predispose to Clostridium difficile
mucosal cell kinetics: the influence of 55. Miller, T. A. Mechanisms of stress-related infection? Nat. Rev. Gastroenterol. Hepatol. 6,
prophylaxis against acute stress ulceration. mucosal damage. Am. J. Med. 83, 8–14 (1987). 555–557 (2009).
J. Surg. Res. 55, 188–192 (1993). 56. Goldin, G. F. & Peura, D. A. Stress-related 74. Khorvash, F., Abbasi, S., Meidani, M.,
36. Robert, A. Cytoprotection of the gastrointestinal mucosal damage. What to do or not to do. Dehdashti, F. & Ataei, B. The comparison
mucosa. Adv. Intern. Med. 28, 325–337 Gastrointest. Endosc. Clin. N. Am. 6, 505–526 between proton pump inhibitors and sucralfate
(1983). (1996). in incidence of ventilator associated pneumonia

106 | FEBRUARY 2015 | VOLUME 12  www.nature.com/nrgastro

© 2015 Macmillan Publishers Limited. All rights reserved
 REVIEWS

in critically ill patients. Adv. Biomed. Res. 3, 52 93. Jairath, V. et al. Red cell transfusion for the 113. MacLaren, R. & Campbell, J. Cost-effectiveness
(2014). management of upper gastrointestinal of histamine receptor-2 antagonist versus proton
75. Vincent, J. L. et al. The prevalence of nosocomial haemorrhage. Cochrane Database of Systematic pump inhibitor for stress ulcer prophylaxis in
infection in intensive care units in Europe. Reviews, Issue 9 Art. No.: CD006613. http:// critically ill patients. Crit. Care Med. 42, 809–815
Results of the European Prevalence of Infection dx.doi.org/10.1002/14651858.CD006613. (2014).
in Intensive Care (EPIC) Study. EPIC International pub3. 114. Freedberg, D. E. & Abrams, J. A. Does
Advisory Committee. JAMA 274, 639–644 94. Villanueva, C. et al. Transfusion strategies for confounding explain the association between
(1995). acute upper gastrointestinal bleeding. N. Engl. J. PPIs and Clostridium difficile-related diarrhea?
76. Lewis, S. C., Li, L., Murphy, M. V., Klompas, M. Med. 368, 11–21 (2013). Am. J. Gastroenterol. 108, 278–279 (2013).
& Epicenters, C. D. C. P. Risk factors for 95. Hebert, P. C. & Carson, J. L. Transfusion 115. Barkun, A. N., Adam, V., Martel, M. & Bardou, M.
ventilator-associated events: a case-control threshold of 7 g per deciliter—the new normal. Cost-effectiveness analysis: stress ulcer
multivariable analysis. Crit. Care Med. 42, N. Engl. J. Med. 371, 1459–1461 (2014). bleeding prophylaxis with proton pump
1839–1848 (2014). 96. Dellinger, R. P. et al. Surviving sepsis campaign: inhibitors, H2 receptor antagonists. Value Health
77. Barletta, J. F., El-Ibiary, S. Y., Davis, L. E., international guidelines for management of 16, 14–22 (2013).
Nguyen, B. & Raney, C. R. Proton pump inhibitors severe sepsis and septic shock, 2012. Crit. Care 116. ASHP Therapeutic Guidelines on Stress Ulcer
and the risk for hospital-acquired Clostridium Med. 41, 580–637 (2013). Prophylaxis. ASHP Commission on Therapeutics
difficile infection. Mayo Clin. Proc. 88, 1085–1090 97. Palamidessi, N., Sinert, R., Falzon, L. and approved by the ASHP Board of Directors on
(2013). & Zehtabchi, S. Nasogastric aspiration and November 14 1998. Am. J. Health. Syst. Pharm.
78. Eastwood, G. M. et al. Opinions and practice of lavage in emergency department patients 56, 347–379 (1999).
stress ulcer prophylaxis in Australian and New with hematochezia or melena without 117. Guillamondegui, O. D. et al. Practice management
Zealand intensive care units. Crit. Care Resusc. hematemesis. Acad. Emerg. Med. 17, 126–132 guidelines for stress ulcer prophylaxis [online],
16, 170–174 (2014). (2010). https://www.east.org/resources/treatment-
79. Barletta, J. F. et al. Pharmacoepidemiology of 98. Razzaghi, A. & Barkun, A. N. Platelet transfusion guidelines/stress-ulcer-prophylaxis (2008).
stress ulcer prophylaxis in the United States and threshold in patients with upper gastrointestinal 118. Heidelbaugh, J. J. & Inadomi, J. M. Magnitude
Canada. J. Crit. Care (2014). bleeding: a systematic review. J. Clin. Gastroenterol. and economic impact of inappropriate use of
80. Murphy, C. E. et al. Frequency of inappropriate 46, 482–486 (2012). stress ulcer prophylaxis in non-ICU hospitalized
continuation of acid suppressive therapy after 99. Gralnek, I. M., Barkun, A. N. & Bardou, M. patients. Am. J. Gastroenterol. 101, 2200–2205
discharge in patients who began therapy in the Management of acute bleeding from a peptic (2006).
surgical intensive care unit. Pharmacotherapy ulcer. N. Engl. J. Med. 359, 928–937 (2008). 119. Thomas, L. et al. Longitudinal analysis of the
28, 968–976 (2008). 100. Rockall, T. A., Logan, R. F., Devlin, H. B. & costs associated with inpatient initiation
81. Farrell, C. P., Mercogliano, G. & Kuntz, C. L. Northfield, T. C. Risk assessment after acute and subsequent outpatient continuation of
Overuse of stress ulcer prophylaxis in the critical upper gastrointestinal haemorrhage. Gut 38, proton pump inhibitor therapy for stress
care setting and beyond. J. Crit. Care 25, 214–220 316–321 (1996). ulcer prophylaxis in a large managed care
(2010). 101. Barkun, A. N., Bardou, M., Martel, M., organization. J. Manag. Care Pharm. 16,
82. MacLaren, R., Jarvis, C. L. & Fish, D. N. Use of Gralnek, I. M. & Sung, J. J. Prokinetics in 122–129 (2010).
enteral nutrition for stress ulcer prophylaxis. acute upper GI bleeding: a meta-analysis. 120. Devlin, J. W., Claire, K. S., Dulchavsky, S. A.
Ann. Pharmacother. 35, 1614–1623 (2001). Gastrointest. Endosc. 72, 1138–1145 (2010). & Tyburski, J. G. Impact of trauma stress ulcer
83. Hurt, R. T. et al. Stress prophylaxis in intensive 102. Sreedharan, A. et al. Proton pump inhibitor prophylaxis guidelines on drug cost and
care unit patients and the role of enteral treatment initiated prior to endoscopic diagnosis frequency of major gastrointestinal bleeding.
nutrition. J. Parenter. Enteral. Nutr. 36, 721–731 in upper gastrointestinal bleeding. Cochrane Pharmacotherapy 19, 452–460 (1999).
(2012). Database Systematic Reviews, Issue 7 Art. No.: 121. Pitimana-aree, S. et al. Implementation of a
84. Ruiz-Santana, S. et al. Stress-induced CD005415. http://dx.doi.org/10.1002/ clinical practice guideline for stress ulcer
gastroduodenal lesions and total parenteral 14651858.CD005415.pub3. prophylaxis increases appropriateness and
nutrition in critically ill patients: frequency, 103. Barkun, A. N. Should every patient with decreases cost of care. Intensive Care Med. 24,
complications, and the value of prophylactic suspected upper GI bleeding receive a proton 217–223 (1998).
treatment. A prospective, randomized study. pump inhibitor while awaiting endoscopy? 122. Barkun, A. N. et al. Effectiveness of
Crit. Care Med. 19, 887–891 (1991). Gastrointest. Endosc. 67, 1064–1066 (2008). disseminating consensus management
85. Osman, D., Djibre, M., Da Silva, D. & 104. Lau, J. Y. et al. Omeprazole before endoscopy in recommendations for ulcer bleeding:
Goulenok, C. Management by the intensivist of patients with gastrointestinal bleeding. N. Engl. J. a cluster randomized trial. CMAJ 185, 156–166
gastrointestinal bleeding in adults and children. Med. 356, 1631–1640 (2007). (2013).
Ann. Intensive Care 2, 46 (2012). 105. Lim, L. G. et al. Urgent endoscopy is associated 123. Filbin, M. R. et al. The microcirculation is
86. Madsen, K. R. et al. Guideline for stress ulcer with lower mortality in high-risk but not low-risk preserved in emergency department low-acuity
prophylaxis in the intensive care unit. Dan. Med. J. nonvariceal upper gastrointestinal bleeding. sepsis patients without hypotension.
61, C4811 (2014). Endoscopy 43, 300–306 (2011). Acad. Emerg. Med. 21, 154–162 (2014).
87. Barkun, A. N. et al. International consensus 106. Rockey, D. C. Lower gastrointestinal bleeding. 124. Shapiro, N. I. & Angus, D. C. A review of
recommendations on the management of Gastroenterology 130, 165–171 (2006). therapeutic attempts to recruit the
patients with nonvariceal upper gastrointestinal 107. Barkun, A. New topical hemostatic powders microcirculation in patients with sepsis.
bleeding. Ann. Intern. Med. 152, 101–113 in endoscopy. Gastroenterol. Hepatol. (N.Y.) 9, Minerva Anestesiol. 80, 225–235 (2014).
(2010). 744–746 (2013). 125. Karakaya, K. et al. Mitigation of indomethacin-
88. Laine, L. & Jensen, D. M. Management of 108. Sung, J. J. et al. Continuation of low-dose aspirin induced gastric mucosal lesions by a potent
patients with ulcer bleeding. Am. J. Gastroenterol. therapy in peptic ulcer bleeding: a randomized specific type V phosphodiesterase inhibitor.
107, 345–360 (2012). trial. Ann. Intern. Med. 152, 1–9 (2010). World J. Gastroenterol. 15, 5091–5096 (2009).
89. Lu, Y., Loffroy, R., Lau, J. Y. & Barkun, A. 109. de Aquino Lima, J. P. & Brophy, J. M. Conflicting 126. Taldone, T., Ochiana, S. O., Patel, P. D.
Multidisciplinary management strategies for evidence: what’s a clinician to do? Ann. Intern. & Chiosis, G. Selective targeting of the stress
acute non-variceal upper gastrointestinal Med. 153, 413–415 (2010). chaperome as a therapeutic strategy. Trends
bleeding. Br. J. Surg. 101, 34–50 (2014). 110. Heyland, D. et al. Enteral nutrition in the critically Pharmacol. Sci. (2014).
90. Opio, C. K. & Garcia-Tsao, G. Managing ill patient: a prospective survey. Crit. Care Med. 127. Bulger, J. et al. Choosing wisely in adult hospital
varices: drugs, bands, and shunts. 23, 1055–1060 (1995). medicine: five opportunities for improved
Gastroenterol. Clin. North Am. 40, 561–579 111. Schupp, K. N., Schrand, L. M. & Mutnick, A. H. healthcare value. J. Hosp. Med. 8, 486–492
(2011). A cost-effectiveness analysis of stress ulcer (2013).
91. Myburgh, J. A. et al. Hydroxyethyl starch or saline prophylaxis. Ann. Pharmacother. 37, 631–635
for fluid resuscitation in intensive care. N. Engl. J. (2003). Author contributions
Med. 367, 1901–1911 (2012). 112. Barletta, J. F. & Sclar, D. A. Use of proton pump M.B. researched data for the article. M.B., J.-P.Q. and
92. Perner, A. et al. Hydroxyethyl starch 130/0.42 inhibitors for the provision of stress ulcer A.B. contributed equally to the discussion of content,
versus Ringer’s acetate in severe sepsis. prophylaxis: clinical and economic consequences. writing and, reviewing and editing of the manuscript
N. Engl. J. Med. 367, 124–134 (2012). Pharmacoeconomics 32, 5–13 (2014). before submission.

NATURE REVIEWS | GASTROENTEROLOGY & HEPATOLOGY VOLUME 12 | FEBRUARY 2015 | 107

© 2015 Macmillan Publishers Limited. All rights reserved