In which subgroup of neurocritically ill patients is stress ulcer prophylaxis recommended?

Comment by InpharmD Researcher

There is currently no standard requirement for stress ulcer prophylaxis (SUP) in neurocritically ill patients. While available data generally suggests a benefit for SUP in traumatic brain injury (TBI), there is limited to no data on the use of SUP in those presenting with aneurysmal subarachnoid hemorrhage (aSAH) or intracranial hemorrhage (ICH). With respect to aSAH, one multicenter observational study found a potential benefit of SUP for the prevention of gastrointestinal bleeding (GIB). Additionally, a meta-analysis reviewing patients with general neurologic injuries, including TBI and SAH, reported a significant reduction in GIB with SUP. Of note, limited studies in patients with intracerebral hemorrhage are conflicting regarding whether there is a benefit associated with the use of SUP, and it is unknown whether this can be generalized to ICH as a whole.

Background

A 2022 systematic review and meta-analysis identified 11 randomized controlled trials (RCTs) of 930 adults admitted to the intensive care unit (ICU) with primary neurologic injury to evaluate the efficacy of stress ulcer prophylaxis (SUP) in reducing gastrointestinal bleeding (GIB). Primary neurological conditions included traumatic brain injury (TBI), subarachnoid hemorrhage (SAH), intracranial hemorrhage (ICH) as well as other conditions. The interventions compared the use of histamine-2-receptor antagonists (H2RAs) or proton pump inhibitors (PPIs) against placebo, no prophylaxis, or each other. Enteral nutrition varied with 6 studies implementing early enteral feeding practices, 4 studies implementing fasting, and the remaining studies lacking data. [1]

A pooled analysis revealed that both PPIs and H2RAs significantly reduced the occurrence of GIB compared to placebo or no prophylaxis, with risk ratios (RR) of 0.37 (95% confidence interval [CI] 0.23 to 0.59; 0<0.001) and 0.42 (95% CI 0.30 to 0.58; p<0.001), respectively. However, no significant difference was observed between PPIs and H2RAs (RR 0.53; 95% CI 0.26 to 1.06; p= 0.07). Despite the observed reduction in GIB, there were no significant differences in 30-day mortality or incidence of nosocomial pneumonia across treatment groups. H2RAs showed similar mortality outcomes as placebo or no treatment (RR 0.77; 95% CI 0.55 to 1.07; p= 0.12), and nosocomial pneumonia rates were not significantly increased with either treatment. Median duration of treatment across all studies was 6 days. The included trials exhibited variability in definitions of GIB, follow-up durations, and feeding practices, and all were classified as high risk of bias, limiting the strength of clinical recommendations. Additionally, small sample sizes and low event rates further constrained the interpretability of the findings. Although suggesting SUP may reduce the incidence of GIB, the data underscore the need for larger, high-quality RCTs focusing on neurocritical populations. [1]

A 2015 meta-analysis investigated the risks and benefits of SUP in adult neurocritical care patients. A total of 8 studies (N= 829) included individuals diagnosed with TBI (n= 288, 4 studies) and intracerebral bleeding (n= 440, 4 studies). TBI regimens included ranitidine 50 mg IV Q6-8H titrated to maintain gastric pH ≥4, 6.25 mg/h continuous IV up to 3-5 days, or cimetidine 300 mg IV Q4H up to 3 weeks. Among all patients, there was a significantly lower incidence of UGI bleeding in treatment versus placebo groups (RR 0.31; CI 0.20 to 0.47; p<0.00001; I2=45%). Subanalyss of TBI studies included the following: UGI bleeding (RR 0.22; 95% CI 0.11 to 0.43), mortality (RR 0.88; 95% CI 0.49 to 1.57), and nosocomial pneumonia (RR 0.75; 95% CI 0.38 to 1.51). [2]

Intracerebral hemorrhage regimens included ranitidine 50 mg IV Q6-8H, omeprazole 40 mg Q12H with cimetidine 300 mg IV Q6H, esomeprazole 40 mg/day, lansoprazole 40 mg/day, ranitidine 150 mg/day, or famotidine 40 mg/day. Subanalysis of intracerebral hemorrhage studies included the following: UGI bleeding (RR 0.31; 95% CI 0.14 to 0.72), mortality (RR 0.63; 95% CI 0.41 to 0.95), and nosocomial pneumonia (RR 0.98; 95% CI 0.29 to 3.31). Overall, the findings suggest that SUP may be more effective than placebo or no prophylaxis for UGI bleeding and all-cause mortality in neurocritical care patients, but the benefits in TBI and intracerebral patients are limited to secondary analyses. [2]

References:

[1] Daou M, Dionne JC, Teng JFT, et al. Prophylactic acid suppressants in patients with primary neurologic injury: A systematic review and meta-analysis of randomized controlled trials. J Crit Care. 2022;71:154093. doi:10.1016/j.jcrc.2022.154093
[2] Liu B, Liu S, Yin A, Siddiqi J. Risks and benefits of stress ulcer prophylaxis in adult neurocritical care patients: a systematic review and meta-analysis of randomized controlled trials. Crit Care. 2015;19:409. Published 2015 Nov 17. doi:10.1186/s13054-015-1107-2

Literature Review

A search of the published medical literature revealed 4 studies investigating the researchable question:

In which subgroup of neurocritically ill patients is stress ulcer prophylaxis recommended?

Level of evidence

C - Multiple studies with limitations or conflicting results  Read more→



Please see Tables 1-4 for your response.


 

Outcomes of Patients with Traumatic Brain Injury After Stress Ulcer Prophylaxis: a Retrospective Multicenter Study

Design

Retrospective study

N= 407

Objective

To assess the association between universal stress ulcer prophylaxis (SUP) administration and complication rates in neurocritical care patients

Study Groups

SUP group (n= 336)

Non-SUP group (n= 71)

Inclusion Criteria

Neurocritical care patients aged ≥18 with moderate or severe traumatic brain injury (TBI) or spinal cord injury (SCI), admitted to intensive care unit (ICU), and a Glasgow Coma Scale (GCS) score <13

Exclusion Criteria

ICU stay <2 days, prior SUP medication use, pre-existing SUP diagnoses

Methods

Patients in the SUP group received prophylaxis, typically with proton pump inhibitors (PPIs) prescribed as once-daily doses. The length of PPI prophylaxis was not specified.

A clinically significant gastrointestinal bleed (CSGIB) required an overt gastrointestinal (GI bleed confirmed by upper GI endoscopy and a bleeding event within 24 hours.

Duration

Study period: October 1, 2020, to September 30, 2021

Median SUP treatment duration: 3.3 days

Outcome Measures

Primary: Clinically significant gastrointestinal bleeds (CSGIBs)

Secondary: Pneumonia, in-hospital mortality

Baseline Characteristics

 

SUP group (n= 336)

Non-SUP group (n= 71)  

Age, years

45 53  

Female

22% 31%  

Admission diagnosis

Traumatic brain injury

Spinal cord injury

Both

 

60%

4%

37%

 

80%

4%

15%

 

Injury severity score

1-9

10-15

16 or more

 

5%

8%

87%

 

11%

15%

73%

 

Antithrombotic use prior to arrival

Anticoagulants

Antiplatelets

 

8%

5%

 

3%

9%

 
Length of stay, days

14

4

 
Intensive care unit length of stay, days

8

3

 
Mechanical ventilation

89%

63%

 

Results

Endpoint

SUP group (n= 336) Non-SUP group (n= 71)

p-Value

CSGIB

0.9% 4%

0.06

All-cause pneumonia

Ventilator-associated pneumonia

21%

10%

1%

0

<0.001

0.006

Clostridium difficile

Helicobacter pylori

2%

0

0

0

0.61

In-hospital mortality

17% 23% 0.24

Adverse Events

Higher rates of pneumonia in SUP group; no significant differences in Clostridium difficile infection or in-hospital mortality.

Study Author Conclusions

SUP may benefit patients with severe TBI by reducing CSGIBs, but increases pneumonia risk, suggesting the need for personalized SUP practices.

InpharmD Researcher Critique

Limitations include the study's retrospective nature, potential missed lower-grade GI bleeds, and lack of differentiation in SUP administration routes.



References:

McGraw C, Briscoe A, Reynolds C, et al. Outcomes of patients with traumatic brain injury after stress ulcer prophylaxis: a retrospective multicenter study. Trauma Surg Acute Care Open. 2024;9(1):e001285. Published 2024 Feb 23. doi:10.1136/tsaco-2023-001285

 

Stress-Related Gastrointestinal Bleeding in Patients with Aneurysmal Subarachnoid Hemorrhage: A Multicenter Retrospective Observational Study

Design

Retrospective observational study

N= 627

Objective

To determine the incidence of stress-related mucosal bleeding (SRMB) in aneurysmal subarachnoid hemorrhage (aSAH) patients, evaluate the effect of acid suppression on SRMB, and identify specific risk factors for SRMB

Study Groups

Gastrointestinal (GI) bleed (n= 31)

No GI bleed (n= 596)

Inclusion Criteria

Age 18 to 85 years and admitted for acute, nontraumatic aSAH

Exclusion Criteria

Active GI bleeding on admission, hospital length of stay <48 hours, or significant gaps in documentation related to data collection

Methods

All included patients were divided based on the incidence of a clinically relevant GI bleed, defined as any instance of the following <24 hours of reported GI bleeding: blood transfusion requirement, hemoglobin reduction of 2 gm/dL or greater, or abrupt systolic blood pressure reduction of ≥20 mmHg, where outcomes were retrospectively assessed.

Duration

January 1, 2013 to June 30, 2017

Outcome Measures

Clinically important GI bleeding, correlation between acid suppressant agent-use and GI bleed occurrence, other risk factors associated with GI bleeds

Baseline Characteristics

 

GI bleed (n= 31)

No GI bleed (n= 596)

p-value 

Age, years

54.6 ± 10.4 56.8 ± 13.1 0.363

Female

74.2% 63.6% 0.230 

History of GI bleed

3 (9.7%) 10 (1.7%) 0.002 

Hunt & Hess Score (IQR)

3 (1 to 5) 2 (1 to 5) 0.038

Fisher Score (IQR)

4 (1 to 4) 3 (1 to 4) 0.038

Aneurysm intervention

Coil

Clip

Other

None

 

18 (58.1%)

5 (16.1%)

1 (3.2%)

7 (22.6%)

 

253 (42.4%)

134 (22.5%)

43 (7.2%)

166 (27.9%)

0.367

-

-

-

-

Abbreviations: GI=gastrointestinal; H2RA=histamine-2 receptor antagonist; IQR= interquartile range

*some patients received a combination of PPI and H2RA during their admission

**number of patients reported with PPI, H2RA, or no prophylaxis prior to GI bleeding episode

Results

Endpoint

GI bleed (n= 31)

No GI bleed (n= 596)

p-value

SRMB prophylaxis

PPI*

H2RA*

No agent

Corticosteroid use

 

23 (74%)**

25 (81%)**

6 (19.4%)**

12 (38.7%)

 

215 (36%)

339 (57%)

118 (20%)

288 (48.3%)

 

<0.0001

0.009

0.952

0.296

Length of stay, days (IQR)

20 (5-81) 15 (2-175) 0.004

ICU length of stay, days

17 (5-44) 12 (0-82) 0.005

Clinically relevant risk factors

Mechanical ventilation

Coagulopathy (INR>1.5)

any corticosteroids

Any ICP>ICP

Any CrCl<60 mL/min

Cerebral  vasospasm

 

25 (81.6%)

5 (16.1%)

12 (38.7%)

17 (54.8%)

14 (45.2%)

21 (67.7%)

 

349 (58.6%)

23 (3.9%)

288 (48.3%)

170 (28.5%)

93 (15.6%)

267 (44.8)

 

0.015

0.001

0.296

0.001

0.0002

0.012

 

Odds ratio (95% confidence interval)

p-Value

Received acid suppressant agent

0.39 (0.18 to 0.83) -

Multivariate analysis of risk factors

Any ICP >20

Any CrCl <60 mL/min

Cerebral vasospasm

 

2.27 (1.03 to 4.97)

5.05 (2.31 to 11.03)

2.51 (1.09 to 5.79)

 

0.82

1.62

0.92

Adverse Events

All subjects (N= 627)

p-Value

Pneumonia incidence

Overall

PPI use

No PPI

H2RA use

No H2RA

 

147 (23.4%)

69 (46.9%)

169 (35.2%)

102 (69.2%)

262 (54.6%)

 

-

-

0.01

-

0.001

Clostridium difficile diarrhea

Overall

PPI use

H2RA

 

15

2

13

 

-

-

-

Abbreviations: GI=gastrointestinal; H2RA=histamine-2 receptor antagonist; ICU=intensive care unit; PPI=proton pump inhibitor; SRMB=Stress related mucosal bleeding; IQR= interquartile range; ClCr=creatinine clearance; ICP=intracranial pressure

*some patients received a combination of PPI and H2RA during their admission

**number of patients reported with PPI, H2RA, or no prophylaxis prior to GI bleeding episode

Adverse Events

See Results

Study Author Conclusions

Clinically important GI bleeding occurred in 4.9% of patients with aSAH, similar to the general critical
care population. Risk factors associated with GI bleeding were prolonged mechanical ventilation (>48 hours), creatinine
clearance<60 ml/min, presence of coagulopathy, elevation of intracranial pressure, and cerebral vasospasm. Further
prospective research is needed to confirm this observation within this patient population.

InpharmD Researcher Critique

The study was limited by its retrospective design and lack of a complete review of adverse reactions with acid suppression therapy. The true incidence of SRMD remains unknown as outcomes included GI bleeding, but SRMD was unable to be confirmed as there were no routine esophagogastroduodenoscopies performed in all patients.



References:

Ali D, Barra ME, Blunck J, et al. Stress-Related Gastrointestinal Bleeding in Patients with Aneurysmal Subarachnoid Hemorrhage: A Multicenter Retrospective Observational Study. Neurocrit Care. 2021;35(1):39-45. doi:10.1007/s12028-020-01137-5

 

A Randomized Controlled Study Comparing Omeprazole and Cimetidine for the Prophylaxis of Stress-related Upper Gastrointestinal Bleeding in Patients with Intracerebral Hemorrhage

Design

Single-center, randomized, placebo-controlled study 

N= 165

Objective

To evaluate the effectiveness of omeprazole and cimetidine compared with a placebo in the prevention and management of stress-related upper gastrointestinal (UGI) bleeding in patients with intracerebral hemorrhage (ICH)

Study Groups

Omeprazole (n= 58)

Cimetidine (n= 54)

Placebo (n= 53)

Inclusion Criteria

Aged ≥18 years; admitted to the neurosurgical intensive care unit (ICU) with CT-proven ICH within 72 hours of ictus and negative results for gastric occult blood testing

Exclusion Criteria

Aarteriovenous malformation or aneurysmal hemorrhage; history of peptic ulcers; likely to swallow blood; antiplatelet and anticoagulation therapy; renal insufficiency requiring hemodialysis; thrombocytopenia <30,000/mL; death within 72 hours after ictus

Methods

Consciousness was assessed using the Glasgow Coma Scale, and stroke severity was assessed according to the Canadian Neurological Scale. The pH of nasogastric tube contents was recorded and aspirates underwent gastric occult blood testing. Patients with negative gastric occult blood testing were randomized to receive omeprazole 40 mg IV every 12 hours, cimetidine 300 mg IV every 6 hours, or placebo solution every 12 hours. All drugs were given for 7 days until UGI bleeding occurred. 

Patients with positive gastric occult blood tests were treated with high-dose omeprazole 80 mg bolus plus 8 mg/hr for 3 days, followed by 40 mg IV every 12 hours for 7 days.

Duration

April 2006 and December 2008

Outcome Measures

Primary: occurrence of UGI bleeding (as manifested by hematemesis, aspiration of coffeeground material from the nasogastric tube, or melena) within 15 days of ictus or death within 30 days of ictus

Baseline Characteristics (of prophylaxis groups)

 

Omeprazole (n= 58)

Cimetidine (n= 54)

Placebo (n= 53)

Age, years

<40

40-60

>60

 

10

44

4

 

13

33

8

 

10

40

3

Female

27 (46.6%) 20 (37.0%) 18 (34.0%)

GCS score

≤6

>6

 

38

20

 

30

24

 

35

18

CNS score 

<3

≥3

 

46

12

 

34

20

 

34

19

Location of hematoma

Supratentorial

Infratentorial

 

54

4

 

46

8

 

50

3

Hematoma size

Small (< 20 mL)

Medium (20-40 mL)

Large  (> 40 mL)

 

10

18

30

 

5

25

24

 

13

15

25

Ventilator, hours

≤48‡

>48

 

45

13

 

46

8

 

38

15

ICU stay, days

≤7

>7

 

35

23

 

25

29

 

24

29

Sepsis 

22

17

18

Abbreviations: CNS= Canadian Neurological Scale; GCS= Glasgow Coma Scale

‡ Including those without mechanical ventilation

Results

 

Omeprazole (n= 58)

Cimetidine (n= 54)

Placebo (n= 53)

Clinical data of prophylaxis groups

UGI Bleeding*

Death

Nosocomial Pneumonia

 

9 (15.5%)

17 (29.3%)

14 (24.1%)

 

15 (27.8%)

14 (25.9%)

12 (22.2%)

 

24 (45.3%)

20 (37.7%)

8 (15.1%)

* p< 0.05

Stress-related UGI bleeding in 15.5% omeprazole vs. 27.8% cimetidine and 45.3% placebo was significant (p= 0.003).

A total of 51 patients died, 21 of whom had UGI bleeding (p= 0.022). There were 17, 14, and 20 deaths in the omeprazole, cimetidine, and placebo groups, respectively; deaths were non-significant between groups.

No adverse events or clinically significant drug interactions were found in this study related to omeprazole. There was no significant difference between groups for nosocomial pneumonia.

In 118 patients with UGI bleeding with high-dose omeprazole initiation, UGI bleeding stopped within the first 3 days in 103 patients (87.3%). There were a total of 70 patients with a positive gastric occult blood test at admission and 48 which developed UGI bleeding in the prophylaxis group. In patients with arrested bleeds, there were 49 deaths and 33 cases of nosocomial pneumonia. There was no significant difference between patients with UGI bleeding at admission and UGI bleeding during/after prophylaxis.

Adverse Events

See results

Study Author Conclusions

Omeprazole appears to be effective and safe in reducing the morbidity of stress-related UGI bleeding in patients with ICH without increasing the risk of nosocomial pneumonia, compared with cimetidine and no treatment; however, it could not reduce 1-month mortality or the length of ICU stay. Currently, high-dose omeprazole is the candidate drug of choice for patients presenting with UGI bleeding, although this needs to be further investigated in ongoing and future prospective randomized studies.

InpharmD Researcher Critique

The study was not blinded, introducing the possibility that the groups were not randomly treated. Additionally, endoscopy was not performed to document the cause of UGI bleeding. Therefore, there is no direct evidence to support that prophylaxis medication is better than no medication to decrease the rates of stress-related ulcers or erosions.

References:

Liu BL, Li B, Zhang X, et al. A randomized controlled study comparing omeprazole and cimetidine for the prophylaxis of stress-related upper gastrointestinal bleeding in patients with intracerebral hemorrhage. J Neurosurg. 2013;118(1):115-120. doi:10.3171/2012.9.JNS12170

A Randomized Placebo-Controlled Trial of Ranitidine Versus Sucralfate in Patients with Spontaneous Intracerebal Hemorrhage for Prevention of Gastric Hemorrhage 

Design

Randomized, placebo-controlled, trial 

N= 141

Objective

To evaluate the usefulness of ranitidine and sucralfate in preventing gastric hemorrhage (GH) in patients with intracerebral hemorrhage (ICH)

Study Groups

Ranitidine (n= 45)

Sucralfate (n= 49)

Placebo (n= 47)

Inclusion Criteria

Patients with computed tomography (CT)-proven ICH within 7 days of ictus

Exclusion Criteria

Patients with arteriovenous malformation; aneurysmal bleed; bleeding and coagulation disorders; hepatic and renal failure; history of peptic ulcer; on antiplatelet and anticoagulation therapy

Methods

Patients were randomized to receive either IV ranitidine 50 mg (Q8H),  sucralfate 1 g (Q6H), or a placebo (saline injection Q8H or starch solution Q6H). Assessment of consciousness utilized the Glasgow Coma Scale (GCS), while stroke severity was gauged using the Canadian Neurological Stroke (CNS) scale. Hematomas were categorized based on size: small (< 20 ml), medium (20-40 ml), or large (> 40 ml). Blood pressure was managed below 180/110 mmHg with hypertensive medications. Temperature was controlled under 38ºC by administering paracetamol (acetaminophen) injections and cold sponging. Soluble insulin was administered to maintain blood sugar levels below 160 mg/dl in cases of hyperglycemia. Seizures were managed with either phenytoin or carbamazepine.

Duration

2001 to 2003

Outcome Measures

Primary: Occurrence of gastric hemorrhages (gross blood, coffee ground aspirate from nasogastric tube, hematemesis or melena) during 15 days of ictus

Secondary: 1-month mortality

Baseline Characteristics

 

Placebo (n= 47)

Ranitidine (n= 45)

Sucralfate (n= 49)   

Age, years

< 40

40-60

> 60

 

4

25

18

 

6

25

14

 

3

25

21

 

Female

 13 16  11   

Duration

Up to 48h

> 48h-5 days

> 5 days

 

28

11

8

 

25

8

12

 

32

13

4

 

GCS

≤ 6

> 6

 

7

40

 

7

38

 

11

38

 

CNS scale

< 3

≥ 3

 

36

11

 

28

17

 

32

17

 

Location of ICH

Supratentorial

Infratentorial

 

39

8

 

40

5

 

47

2

 

Size

Small

Medium

Large

 

20

16

11

 

21

17

7

 

22

13

14

 

Midline shift

19 20 22  

Intraventricular extension 

22 24 25  

Septicemia

17 12 17  

Pneumonia 

5 2 5  

Abbreviations: CNS= Canadian Neurological Stroke; GCS= Glasgow Coma Scale; ICH= intracerebral hemorrhage 

Results

 

 

 

Endpoint 

Placebo (n= 47) Ranitidine (n= 45) Sucralfate (n= 49)  p-Value

GH occurrence* 

11 (23.4%) 5 (11.1%) 7 (14.3%) NS

1-month mortality**

13 (27.7%) 5 (11.1%) 12 (24.5%) -

Abbreviations: GH= Gastric hemorrhages; NS= not significant 

*The incidence of GH increased with delays in treatment initiation: 3.3% within 48 hours, 9.4% within 3-5 days, and 20.8% for over 5 days. There was no significant difference in GH frequency among these groups

**11 out of 30 deaths had GH. GF was noted to be significantly related to death (p= 0.001)

Adverse Events

N/A 

Study Author Conclusions

Ranitidine and sucralfate do not seem to significantly prevent GH or reduce 1-month mortality.

InpharmD Researcher Critique

While results indicate that ranitidine and sucralfate do not significantly prevent GH or reduce mortality, it is suggested that the timing of admission and initiation of therapy might better influence the incidence of GH. Due to the small sample size and single-center nature of the study, the findings may not be widely generalizable.

References:

Misra UK, Kalita J, Pandey S, Mandal SK, Srivastava M. A randomized placebo controlled trial of ranitidine versus sucralfate in patients with spontaneous intracerebral hemorrhage for prevention of gastric hemorrhage. J Neurol Sci. 2005;239(1):5-10. doi:10.1016/j.jns.2005.07.011