Expert Consensus on Emergency Diagnosis and Treatment Process for Acute Upper Gastrointestinal Bleeding (2020 Edition)

Acute upper gastrointestinal bleeding is one of the most common acute and critical illnesses in the emergency department, with an annual incidence of 100/100,000 to 180,000 to 100,000 adults and a case fatality rate of 2% to 15%.

01 method

The 2015 edition of the Expert Consensus is based on the evaluation of acute upper gastrointestinal bleeding, circulatory stabilization, drug selection, and hemostasis

This update refers to the latest evidence-based guidelines and literature at home and abroad, and combines the actual clinical practice of emergency medicine in China to reach a consensus statement

02 Consensus content

2.

This version of the consensus still adheres to the concept of emergency "ladder downgrade" thinking, and constructs the emergency diagnosis and treatment process for acute upper gastrointestinal bleeding according to "3 assessments, 2 treatments", and strives to meet the clinical operability and practicality for the reference of emergency physicians, see Figure 1

2.

2.

Airway assessment: Assess the risk

Respiratory assessment: assesses respiratory rate, rhythm, exertion, and oxygen saturation

Circulatory assessment: Monitors heart rate, blood pressure, urine output, and peripheral perfusion

2.

Statement 1: Patient consciousness, airway, breathing, and circulation

2.

Statement 2: Patients with acute upper gastrointestinal bleeding are treated in stratified according to the degree of risk, and dangerous bleeding should be treated in the emergency department (evidence level: high, consistency rate: 100%)

2.

Routine measures are "OMI", i.

Statement 3: Patients with high-risk upper GI bleeding should be treated urgently (level of evidence: high, consistency rate: 100%)
.

2.
3.
1 Volume resuscitation Hemodynamically unstable acute upper gastrointestinal bleeding should be actively volume resuscitated, but the specific strategy of resuscitation currently lacks evidence-based evidence
.
Referring to the resuscitation concept of traumatic hemorrhage, restrictive fluid resuscitation and permissible hypotension resuscitation strategies are used when bleeding is not controlled, and it is recommended that systolic blood pressure be maintained at 80 to 90 mmHg
.
Controlled bleeding should be actively resuscitated
according to the patient's underlying blood pressure level.
For patients with acute major bleeding, invasive hemodynamic monitoring, comprehensive clinical manifestations, ultrasound, and laboratory tests should be performed to guide volume resuscitation, and attention should be paid to preventing hypothermia, acidosis, coagulopathy, and exacerbation
of underlying disease.

There is currently no consensus on the amount and type of intravenous fluids
.
In hemorrhagic shock, volume resuscitation should avoid large crystal fluid infusions and minimize crystal fluid infusions (less than 3 L in the first 6 h).

Isotonic crystal fluid has no benefit other than temporarily expanding the amount of vascular
contents.
With massive infusions of isotonic crystal fluid, the risk of complications such as respiratory failure, septal syndrome (abdominal and limb), and coagulopathy increases
.
Artificial colloids or hypertonic solutions also did not show significant benefit
as early in-hospital treatment of severe bleeding.
Blood pressure returned to pre-bleeding baseline level with pulse <100>0.
5 mL/(kg · h), clear consciousness, no significant dehydration, arterial blood lactate return to normal and other manifestations suggest adequate
volume recovery.
In addition, varicose vein rupture bleeding infusions should be done with caution, as excessive infusions may worsen bleeding
.
In patients with concomitant cardiopulmonary and renal disease, be wary of heart failure or pulmonary edema
caused by excessive infusion volumes.

Statement 4: Hemodynamically unstable acute upper GASTRO bleeding should be promptly volume resuscitated to restore and maintain vital organ perfusion (level of evidence: high, consistency rate: 100%)
.

2.
3.
2 Blood transfusion Patients with massive blood loss need to be properly transfused with blood products to ensure oxygen supply to the tissue and maintain normal coagulation function
.
Blood transfusions should be considered in the following cases: Systolic blood pressure <90 mmhg = "" > 110 times / min ; Hb<70 g/L ； Hematocrit (Hct) < 25% or hemorrhagic shock
.
For acute mass bleeding, a local mass transfusion regimen is initiated immediately
.
Although the proportion of red blood cells, plasma, and platelets is currently inconclusive, a pre-set ratio of blood products (e.
g.
, a ratio of 1 to 1 to 1 for red blood cells, plasma, and platelets) and the use of adjuvant drugs such as calcium may provide survival benefits
.
Platelet transfusion is not required for inactive bleeding and hemodynamic stability, and platelet
transfusion should be done for active bleeding with a platelet count <50×109/L.

Risks and benefits of transfusion should be weighed individually, with a generally restrictive transfusion strategy, with a recommended target value of 70 to 90 g/L
for Hb.
Varicose vein bleeding, in addition to the hepatic function Child-Pugh C grade, requires strict restriction of transfusion indications Hb<70 g/L, otherwise it may increase the case fatality rate
.
However, in patients with advanced age, underlying cardiovascular and cerebrovascular disease (e.
g.
, acute coronary syndrome, stroke, or transient ischemic attack), hemodynamic instability, or persistent massive bleeding, the restrictive transfusion strategy is not appropriate, and the transfusion indications can be relaxed to Hb<90 g/L or more to avoid exacerbation
of the underlying disease due to massive blood loss.

In patients with coagulation dysfunction, changes in coagulation indicators or thromboelastograms are dynamically observed to assess coagulation status
in real time.
For active bleeding, fresh frozen plasma (FFP) should be transfused if the prothrombin time (or international normalized ratio) or activated partial thromboplastin time is greater than 1.
5 times normal, and if fifibrinogen (FIB) levels are still below 1.
5 g/L after FFP, FIB infusion or cryoprecipitation
is recommended.
Active variceal bleeding from cirrhosis should be given FFP
if FIB < 1 g/L.

Massive transfusions can lead to transfusion complications such as hypocalcemia and coagulopathy, and calcium should be given empirically (eg, 1 g of calcium chloride after transfusion of 4 blood products) and closely monitored for ionized calcium levels
.
The procedure of mass transfusion also requires attention to possible hypothermia, acidosis, and hyperkalemia
.

Statement 5: Weigh the risks and benefits of transfusion and adopt the optimal transfusion strategy (level of evidence: high, consistency rate: 97.
7%)
.

2.
3.
3 Vasoactive drug application Vasoactive drugs can be used
in severe persistent hypotension due to hemorrhagic shock.
However, there is a lack of high levels of evidence to support it
.

Statement 6: Persistent hypotension remains despite aggressive volume resuscitation, and vasoactive drugs are an option to ensure minimal effective perfusion of vital organs (level of evidence: medium, consistent rate: 100%)
.

2.
3.
4 Initial pharmacotherapy For acute upper gastrointestinal bleeding of unknown risk, although insufficient evidence is lacking to support it, "empiric combination therapy" may be used in cases where emergency gastroscopic intervention may be delayed to maximize the possibility of reducing bleeding, serious complications and death, and to create conditions
for endoscopic or other follow-up treatment.

Acute upper gastrointestinal bleeding is often caused by non-variceal bleeding, so proton pump inhibitor (PPI)
is recommended before endoscopy when the cause is unclear.
In addition, PPI
is also recommended in patients with a history of liver disease or cirrhosis who cannot exclude ulcer bleeding.

Patients with cirrhosis, a history of chronic liver disease, or signs of portal hypertension are more likely to have varicose bleeding, and such patients tend to have heavy bleeding, a high early case fatality rate, and require medications
including vasoconstrictor drugs before endoscopic diagnosis is confirmed.

Somatostatin is indicated for the treatment
of severe acute esophageal variceal bleeding, severe acute gastric or duodenal ulcer bleeding, and complicated by acute erosive gastritis or hemorrhagic gastritis.
Therefore, for dangerous acute upper gastrointestinal bleeding, the cause is unknown, and the combination of PPI and somatostatin can be applied, and the cause is clear and then adjusted
.

Because prophylactic antibiotics for varicose vein bleeding can significantly improve prognosis, antibiotics
should be used prophylactically when there is a high suspicion of varicose vein bleeding.

Statement 7: When the cause of dangerous acute upper gastrointestinal bleeding is unknown, it can be treated with a combination of intravenous PPI and somatostatin, and then adjusted after the cause is clear (level of evidence: low, consistency rate: 98.
9%)
.

Statement 8: Prophylactic antibiotics are recommended when there is a high suspicion of varicose vein bleeding (level of evidence: high, consistency rate: 83%)
.

2.
4 Comprehensive Assessment

2.
4.
1 Presumed Etiology of Bleeding After the life-threatening condition of active bleeding or major bleeding has been temporarily controlled, fluid resuscitation and drug therapy has begun, or when the condition is mild and vital signs are stable, a comprehensive evaluation and inference should be made to infer the cause and location
of bleeding.
Early identification should be needed for suspected variceal bleeding, which can be assessed
based on signs and risk factors for portal hypertension.

The causes of acute upper gastrointestinal bleeding are divided into two categories
: acute non-variceal bleeding and varicose bleeding.
Most are acute non-variceal bleeding, the most common causes include gastroduodenal peptic ulcer, upper gastrointestinal tumor, stress ulcer, acute and chronic inflammation of the upper gastrointestinal mucosa, other causes are cardia mucosal tear syndrome, upper gastrointestinal arteriovenous malformations, Dieulafoy lesions, etc.
, iatrogenic factors include: taking nonsteroidal anti-inflammatory drugs (NSAIDs), especially antiplatelet drugs (such as aspirin), endoscopic mucosectomy / dissection (EMR/ESD
。

Statement 9: A comprehensive assessment should be made after the initial disposition to determine the cause of bleeding (level of evidence: high, consistency rate: 100%)
.

2.
4.
2 Dynamic monitoring Indicators such as vital signs, blood routine, coagulation function and blood urea nitrogen should be continuously and dynamically
monitored.
In addition, blood lactate levels should be dynamically monitored to determine whether tissue ischemia improves and fluid resuscitation efficacy, and to optimize fluid resuscitation regimens
.
Active bleeding should be considered in the following cases :(1) Increased number of hematemesis and melena, vomiting from coffee to bright red or excreted feces from black dry stool to dark red dilute stool, or with active bowel sounds; (2) There is more fresh blood in the gastric tube drainage fluid; (3) After rapid infusion of blood, the performance of peripheral circulating perfusion did not improve significantly, or although it temporarily improved and worsened again, the central venous pressure still fluctuated, and then decreased again after a little stabilization; (4) The red blood cell count, Hb and Hct continued to decrease, and the reticulocyte count continued to increase; (5) In the case of adequate hydration fluid and urine output, blood urea nitrogen continues to be abnormal or increased
again.

Statement 10: Dynamic monitoring of changes in the condition and determination of active bleeding (level of evidence: high, consistency rate: 100%)
.

2.
4.
3 Severity, need for clinical intervention and prognosis Assessment Comprehensive assessment of disease severity, treatment intervention needs, and prognosis based on bleeding manifestations, vital signs, changes in Hb, and risk factors
.
Risk factors include age >60 years, advanced tumor, cirrhosis or other serious concomitant disease, previous history of severe upper gigadiological bleeding or instrument insertion, hematemesis, coagulation dysfunction (INR>1.
5), absence of liver and kidney disease but persistent elevation of
blood urea nitrogen.

The risk-scoring scale can be broadly divided into two categories
.
One is used pre-endoscopy to assess the risk of death based on the need for or without intervention of clinical intervention, and the other is primarily used to determine prognosis, some of which include endoscopic findings
.
Some of the scales can be used in general
.
Because the pre-endoscopic scoring scale can help with subsequent clinical decision-making, it is more commonly used
.

Commonly used pre-endoscopic scores are GBS, pre-endoscopic Rockall, and AIMS65 (albumin, international normalized ratio, mental status, systolic blood pressure, age >65 years).

However, an international multicenter prospective large sample study showed that the scoring scale for most acute upper gastrointestinal bleeding was not highly
accurate.
Although the study suggests that GBS is the best indicator of early prediction of the need for clinical intervention (transfusion, endoscopic therapy, or surgery) or death, and the GBS score of ≥7 is the best choice for predicting endoscopic therapy, its clinical value is still limited
.
This is because all risk-scoring scales, including GBS, do not accurately identify high-risk patients
.
The higher clinical value is in very low-risk patients
with a GBS score of ≤ 1 that accurately predict survival and do not require emergency clinical intervention.

Statement 11: Clinical assessment of severity, need for treatment intervention, and prognosis (refer to the GBS scale) (level of evidence: Medium, consistency rate: 98.
9%)
.

2.
5 Further diagnosis and treatment

After a comprehensive evaluation, the emergency physician should reasonably choose the next step of diagnosis
and treatment based on the results of the assessment.

2.
5.
1 Medication management (1) Acid-suppressing drugs
.
Acute non-varicose upper gastrointestinal bleeding often requires acid suppression therapy
.
Commonly used clinical acid-suppressing drugs include PPIs and H2 receptor antagonists
.
PPI is currently the preferred acid-suppressing drug
.
Although some studies have shown that the use of PPI before endoscopy does not affect the rate of rebleeding, surgery or mortality, it has also been found that the use of PPI before endoscopy can reduce the signs of high-risk endoscopic bleeding and the need for endoscopic intervention, and the author still recommends the use of PPI
before endoscopic intervention, combined with the possible delay or incompleteness of emergency endoscopy.

PPI
should be given as appropriate after endoscopic interventions.
Non-varicose upper gastrointestinal bleeding (eg, peptic ulcer, corrosive esophagitis, gastritis, duodenitis) or esophageal cardia mucosal tear syndrome associated with gastric acid should be treated with
PPI.
The duration of peptic ulcer PPIs is 4 to 8 weeks
.
Low-risk rebleed peptic ulcer (Forrest II.
c- III.
, basal flat and clean) is given an oral PPI
of 1 /day.
In patients at high risk of peptic ulcer (active bleeding, visible blood vessels, adherent clots), a meta-analysis confirmed that receiving a high dose of PPI of 72 h (first dose of 80 mg intravenously followed by continuous infusion of 8 mg/h for 72 h) after successful endoscopic therapy may reduce rebleeding and mortality.

One randomized controlled trial (RCT) showed that sequential oral PPIs 2 times/d compared with 1 dose/day to two weeks after high-risk patients received high-dose PPIs significantly reduced the risk
of rebleeding.
Current domestic guidelines recommend that for high-risk patients, high-dose PPIs should be changed to standard-dose PPIs for intravenous infusion, 2 times /day, and oral standard-dose PPIs after 3 to 5 days until ulcers heal
.

Statement 12: For acute non-varicose upper gastrointestinal bleeding, PPI should be considered before and after endoscopic interventions (level of evidence: medium, consistency rate: 97.
7%)
.

(2) Drugs
that reduce portal vein pressure.
Patients with EGVB have a higher early case fatality
rate.
In patients with varicose vein bleeding, pharmacotherapy is preferred to reduce portal pressure and reduce active bleeding [11].

Therapeutic agents include somatostatin and its analogues (octreotide) and vasopressin and its analogues (terlipressin).

Somatostatin is a cyclic active tetradecylide composed of multiple amino acids with an elimination half-life of about
3 minutes.
Octreotide is a synthetic octapeptide somatostatin analogue with an elimination half-life of approximately 100 min
.
Somatostatin and octreotide reduce portal vein pressure
primarily by reducing portal blood flow.
Vasopressin and terlipressin can cause visceral vasoconstriction, increase mesenteric vascular resistance, and reduce portal vein blood flow by activating vascular smooth muscle V1 receptors, thereby reducing portal vein pressure
.
Due to the strong constriction effect of vasopressin, adverse reactions of cardiac and peripheral vascular ischemia will occur, so its clinical application is limited
.
Terlipressin is a synthetic vasopressin analogue that can effectively reduce portal vein pressure for a long time, has little effect on systemic hemodynamics, and the most significant adverse reaction is peripheral acromelia
.

Somatostatin Usage: The first dose of 250 micrograms is given intravenously followed by a continuous intravenous infusion
of 250 micrograms/h.
Octreotide Usage: The first dose of 50 micrograms is followed by a continuous intravenous bolus
of 50 micrograms/hour.
Trispressin Usage : The starting dose is 1 mg/4 h slowly intravenously, the first dose can be doubled
.
Bleeding may be changed to 1 mg/12 h after cessation
.
The course of treatment of the above three drugs is generally 2 to 5 days
.
Somatostatin (octreotide) or vasopressin (terepressin) have been shown to improve endoscopic hemostasis and reduce the rate
of recent rebleeding after endoscopic therapy.
Octreotide-assisted endoscopic therapy (2 to 5 days) may prevent early rebleeding
of EGVB.
There was no significant difference
in the efficacy of somatostatin, octreotide, and terlipressin in reducing bleeding.
If somatostatin or octreotide fails to control bleeding, a combination of terlipressin may be considered, but the efficacy of the combination needs to be further verified
.

Statement 13: For acute varicose upper gastrointestinal bleeding, somatostatin (or its analogue octreotide) or vasopressin (or its analogue terlipressin) is recommended for up to 5 days (evidence level: high, consistency rate: 95.
5%)
.

(3) Hemostatic drugs
.
One RCT study reported that tranexamic acid for acute upper GI bleeding helped reduce the need for emergency endoscopy, but did not improve
mortality and rebleed rates.
Because tranexamic acid is at risk of thromboembolism, it should be used with caution before its safety is confirmed by large-sample RCT studies
.
Systemic and topical use of hemocoaguloids, oral or gastric duct topical use of thrombin, Yunnan Baiyao, sucralfate, or ice norepinephrine saline, the efficacy is uncertain
.
There have been no reports
of vitamin K1 for acute upper gastrointestinal bleeding in patients with acute and chronic liver disease.

Statement 14: Acute upper gastrointestinal bleeding should be used with caution with hemostatic drugs
.
(Level of evidence: low, consistency rate: 92%)
.

(4) Antibacterial drugs
.
The risk of infection in patients with cirrhosis with acute varicose vein bleeding can be assessed
by Child-Pugh grading.
The higher the Child-Pugh rating, the higher the risk of
infection.
Patients with Grade A alcoholism or alcohol consumption are also at high risk for infection after varicose vein bleeding
.
In patients with cirrhosis with acute upper gastrointestinal bleeding, prophylactic antibiotics are beneficial to stop bleeding, reduce the incidence of rebleeding and infection, and have a lower
case fatality rate at 30 days.
Antibiotics
should be reasonably selected based on local bacterial resistance.
The results of an RCT study showed that in patients with bleeding from advanced cirrhosis, intravenous ceftriaxone was more effective than oral norfloxacin in preventing bacterial infection; Another RCT study found no significant difference in efficacy between ceftriaxone and 7 d courses
.

Statement 15: Prophylactic antimicrobial therapy should be given to patients with cirrhosis with acute upper gastrointestinal bleeding (level of evidence: high, consistency rate: 83%)
.

(5) Antithrombotic drugs
.
Antithrombotics include antiplatelet and anticoagulant therapy
.
Whether to discontinue antithrombotics after acute upper gastrointestinal bleeding is a challenging clinical decision
.
It is recommended to weigh the risk of bleeding and ischemia with a specialist to complete an individualized assessment
.
It is generally not advisable to routinely discontinue all drugs
.
One retrospective study showed that discontinuation of antithrombotics after bleeding was associated with
an increase in thrombotic events and a decrease in survival.
A small RCT study showed that the case fatality rate was significantly higher than that of maintenance therapy patients who discontinued upper gastrointestinal bleeding after 8 weeks of discontinuation of aspirin as a secondary prevention, the main cause of death was a thrombotic event, and there was no statistically significant
difference in rebleed rates between the two groups.

Antiplatelet therapy after acute upper gastrointestinal bleeding needs to be considered
from both the necessity of drug use and the risk of bleeding.
If the drug is not necessary, such as aspirin as a primary prevention of cardiovascular events, it should be discontinued and evaluated
when clinically necessary.
Secondary prevention with aspirin alone or dual antiplatelet therapy should be individualized, depending on the risk of endoscopic bleeding signs, and can be given first discontinuation and then recovery, non-discontinuation or other treatment
.
For patients with acute coronary syndrome who are treated with dual antiplatelets, Chinese experts recommend that mild bleeding does not need to be discontinued, aspirin should be discontinued first for obvious bleeding, and if there is life-threatening active bleeding, all antiplatelet drugs should be discontinued, and antiplatelet therapy
should be resumed as soon as possible after effective hemostasis and stabilization.
Clopidogrel is generally restored after effective hemostasis for 3 to 5 days and aspirin after 5 to 7 days
.
For acute non-varicose upper GASTRO bleeding that cannot be discontinued from antiplatelet therapy, continuous PPI therapy is indicated
.

Patients taking warfarin should be discontinued if there is active bleeding or hemodynamic instability, and anticoagulant effects
can be reversed with prothrombin complex and vitamin K.
The anticoagulant effect of new oral anticoagulants (dabigatran, rivaroxaban, apixaban) disappears within 1 to 2 days, so there is generally no need to supplement prothrombin complexes, and other treatments that reverse anticoagulation are controversial
.
If the risk of thrombosis is high after the hemostasis is confirmed, restart of anticoagulation should be evaluated as soon as possible
.
Patients with high-risk cardiovascular disease may consider the use of heparin or a low molecular weight heparin transition
during discontinuation of oral anticoagulants.

Statement 16: Weighing bleeding against ischemic risk, individually administering antithrombotics (level of evidence: high, consistency rate: 97.
7%)
.

2.
5.
2 Triple-lumen dicyscapsular tubes For EGVBs, if the bleeding volume is large and endoscopically difficult to treat, a three-lumen dicapsular tube may be placed as a temporary measure
for short-term bleeding control and transition to definitive therapy.
The three-chamber bicyscapsular tube should not be placed for more than 3 days, deflated once every 8 to 24 h according to the condition, and the time of extubation should be 24 h
after successful hemostasis.
Generally, deflators are observed for 24 h, and if there is still no bleeding, the tube can be extected
.
Treatment of the three-chamber dicystic tube is prone to rebleeding and some serious complications, such as esophageal rupture and aspiration pneumonia, which require attention
.

Statement 17: The triple-lumen, dicapsular duct is only used as a temporary transitional measure for the treatment of endoscopically difficult-to-treat EGVB (level of evidence: high, consistency rate: 95.
5%)
.

2.
5.
3 Emergency endoscopy Endoscopy is the preferred key test to determine the cause of acute upper gastrointestinal bleeding, and plays an important role
in disease risk stratification and treatment.
Emergency physicians should actively stabilize the patient's circulatory condition, do a good job of airway protection, and create conditions
for the successful completion of endoscopy and treatment.
Bedside endoscopy
may be performed in the emergency room or under close ICU supervision when the patient is critically ill or unfit for transport.
If the first endoscopy does not completely stop bleeding, repeat endoscopic therapy may be considered if necessary [76].

(1) Timing
of endoscopic examination.
For acute non-varicose upper GI bleeding, current guidelines recommend endoscopy within 24 hours of bleeding if no
contraindications are contraindicated.
Delayed endoscopy of more than 24 hours in patients with acute upper gastrointestinal bleeding is associated
with increased mortality.
Patients with persistent hemodynamic instability after aggressive resuscitation should undergo emergency endoscopy
.
A recent RCT study showed that endoscopy in patients with a high risk of further bleeding or death from acute upper gastrointestinal bleeding but with a stable hemodynamic range did not increase
the case fatality rate within 6 to 24 hours compared with 6 to 24 hours after consultation.
Varicose vein bleeding is often a major hemorrhage, with transfusions and fluids at a rate much lower than the rate of bleeding, and endoscopy should be performed within 12 hours
.
Notably, some studies have shown that the vast majority of deaths after acute upper GASTRO bleeding are caused by underlying complications rather than blood loss, so early recovery and management of complications before endoscopy are also crucial
.

Statement 18 : Dangerous acute upper gastrointestinal bleeding should be endoscopic within 24 hours after bleeding ; Emergency endoscopy should be performed if hemodynamic instability persists after aggressive resuscitation; If hemodynamically stable, endoscopy can be performed within 24 hours
.
Suspected variceal bleeding should be endoscopic within 12 hours (level of evidence: medium, consistent rate: 98.
9%)
.

(2) Precautions for
endoscopy.
There is high-level evidence that for acute upper gastrointestinal bleeding, erythromycin infusion prior to endoscopy reduces gastric blood volume, improves endoscopic visibility, and significantly reduces secondary endoscopy rates and endoscopic time
.

In addition, the available evidence does not support that drainage of blood in the gastric tract prior to endoscopy improves endoscopic vision [82].

For those taking anticoagulant drugs, the INR should be corrected to less than 2.
5 before endoscopy
.
In addition, during endoscopy, airway protection should be done to prevent reflux aspiration and avoid aspiration pneumonia, especially in elderly patients
with dialysis, stroke history and long surgery.

Statement 19: Intravenous infusion of erythromycin 250 mg 30 to 120 min prior to endoscopy may be considered to improve endoscopic vision (level of evidence: high, consistency rate: 80.
7%)
.

2.
5.
4 Abdominal CTA and other tests Endoscopic contraindications or negative examinations may be treated empirically and other diagnostic methods may be selected
.
Depending on the condition, you can choose between abdominal enhancement CT, CT angiography (CTA), angiography, colonoscopy, radionuclide scan, or laparotomy to determine the cause
.

For major or active bleeding, if endoscopy is not possible or endoscopy does not confirm the cause, an abdominal CTA may be selected to help determine the source and cause
of the bleeding.
Abdominal CTAs typically detect bleeding at a rate of 0.
3 to 0.
5 mL/min, which makes it sensitive
to both arterial and venous bleeding.
It can also be used to observe diseases of the intestinal wall, such as vascular malformations and masses
.
However, it should be noted that even if there is heavy bleeding, the bleeding can be stopped quickly, resulting in a negative test result
.
Therefore, to improve the rate of positive detection of abdominal CTAs, the delay in examination should be minimized
.
In addition, abdominal CTA is not a treatment and requires a trade-off
between the benefits of adjuvant diagnosis and the risk of delay in treatment.
Interventional therapy may be directly selected in cases where the risk of treatment delay is high
.
It is also important to note that contrast allergy and contrast-induced nephropathy
may occur on CTA examination.

Statement 20: Abdominal CTAs may be used to look for potential bleeding causes if endoscopic contraindications or examination negatives are still active (level of evidence: medium, consistency rate: 98.
9%)
.

2.
5.
5 Interventional therapy For patients with acute non-varicose upper gastrointestinal bleeding, selective angiography can be performed to determine the source of
the bleeding site.
Angiography routinely selects the left gastric artery, the gastroduodenal artery, the splenic artery, and the pancreatic duodenal artery
.
Treatment includes injection of vasoconstriction drugs in bleeding vessels or transcatheter arterial embolization (TAE).

Transjugular intrahepatic portosystemic shunt (TIPS)
may be considered after drug and endoscopic hemostasis failures.
Severe recurrent varicose bleeding, Child-Pugh grade C (<14 points), or B grade with active bleeding may be considered for early TIPS to reduce bleeding recurrence
.

Statement 21: Patients who are contraindicated or negative for endoscopy still have active bleeding, or who have failed drug and endoscopic treatment, or whose abdominal CTA suggests bleeding, can be treated with emergency interventional examination (level of evidence: medium, consistency rate: 98.
9%)
.

2.
5.
6 Multidisciplinary diagnosis and surgical intervention Acute upper gastrointestinal bleeding is mostly diagnosed in the emergency department
.
The diversity of etiology and the urgency of the condition often require the collaboration of physicians of different specialties, but it is often difficult to achieve effective collaboration and successful treatment using traditional monodisciplinary treatment and consultation, especially for refractory haemorrhage
.
A retrospective study showed that the implementation of multidisciplinary diagnostic strategies can improve the efficiency of diagnosis and treatment and reduce the case fatality rate
.
Surgical exploration may be considered in patients who are unable to stop bleeding with pharmacological, endoscopic, and interventional therapy, conditions permitting
.

Statement 22: For persistent bleeding that is difficult to control with medications, endoscopy, and interventional therapy, multidisciplinary care can be initiated, with surgical intervention if necessary (level of evidence: medium, consistency rate: 97.
7%)
.

2.
6 Prognostic assessment

Prognosis is assessed
after stabilization of acute upper gigaditic bleeding.
Assessments include vital organ function and the risk
of rebleeding and death.
Vital organ function can be assessed
according to clinical data.
Patients with acute non-variceal upper GI bleeding are at increased
risk of rebleeding if they are older than 65 years of age, severe comorbidities, shock, low haemoglobin concentrations, blood transfusions, and endoscopic ulcer basal blood clots and vascular exposure.
Acute varicose upper GASTRO bleeding is prone to rebleeding, with rebleeds occurring in 60% to 70% and case fatality rates as high as 33%
within 1 to 2 years after the first bleeding.
The risk of death is primarily empirically assessed based on patient risk factors, and the presence of risk factors described in the comprehensive assessment often suggests a poor
prognosis.
The risk-scoring scale was used
to determine the risk of rebleeding, length of hospital stay, or death.
Studies have shown that although the PNED (Progetto Nazionale Emorragia Digestiva) score of ≥ 4 and the AIMS65 score of ≥ 2 are the best indicators of predicting death, their clinical value is limited
due to their low accuracy.
After the prognosis assessment is complete, depending on the cause and the results of the assessment, the patient is recommended to be transferred to a specialty for further diagnosis and treatment or follow-up
after discharge.

Statement 23: Prognosis is assessed after acute upper gastrointestinal bleeding is stable, and the clinical value of the risk-scoring scale is limited (level of evidence: medium, consistency rate: 94.
3%)
.