Gastrointestinal Bleeding
Upper Non-Variceal
Variceal Hemorrhage
Lower GI
Upper GI Bleeding, non-variceal
Summary of Consensus Recommendations
Annals of Internal Medicine. 2003;193:843-857.
Recommendations:
Hospital Structure and Organization:
1. Develop protocols for multidisciplinary management, with the availability of an endoscopist with training in endoscopic hemostasis. Poor evidence, but experts agree.
2. Have support staff available for urgent endoscopy. Poor evidence, but experts agree.
Patient Evaluation and Care:
3. Immediate evaluation and resuscitation are critical to proper management. Poor evidence, but experts agree.
4. In selected patients, place a nasogastric tube, findings may have prognostic value. Fair evidence supports this, quality of evidence is uncontrolled.
A positive aspirate confirms an upper GI source. Presence of blood on an in-and-out NG placement portends a poor outcome and the need for emergent endoscopy. Bright red blood on NG aspirate is an indicator of likely rebleeding. When early endoscopy is planned, lavage may be helpful to clear the stomach of excess clots, but is not proven to stop bleeding.
Risk Stratification:
5.1 Clinical stratification (not based on endoscopy) for low vs. high risk for rebleeding and mortality are important for good management. Use available prognostic scales. Fair evidence with uncontrolled data.
80% of bleeding will stop spontaneously without recurrence. Morbidity and mortality occur in the 20 % with recurrence most often. The goal is to identify those at high risk on the basis of clinical, lab, and scope findings.
Clinical indicators of increased risk for rebleeding from multivariate analysis in trials include: age>65 years; shock; poor overall health status; comorbid illness; low initial hemoglobin; melena; transfusion requirement; fresh blood on rectal exam, in the emesis, or in the NG aspirate.
Clinical indicators for an increased risk of death include age over 60 years; poor overall health status; comorbid illnesses; continued bleeding or rebleeding; fresh red blood on rectal examination, in the emesis, or NG aspirate; onset of bleeding while hospitalized for another reason; sepsis; or elevated urea, creatinine, or serum aminotransferase levels. Care setting and physician specialty may be a factor also.
5.2 Early stratification into high and low risk categories for rebleeding and mortality, based on clinical and endoscopic criteria, is useful for management.
Risk of rebleeding or continued bleeding is strongly associated with the hemorrhagic stigmata seen on endoscopy:
Clean ulcer base <5%
Ulcer with flat spot 10%
Ulcer with adherent clot 22%
Nonbleeding visible vessel 43%
Active bleeding (ooze,spurt) 55%Rockall, et al: Risk Score for Rebleeding and Mortality
Component Score 0 1 2 3 Age in years <60 60-79 >=80 Shock
Pulse Rate (BPM)
Systolic BP (mm Hg)No shock
<100
>=100Tachycardia
>=100
>=100Hypotension
<100...... Comorbidity None ....... Ischemic Heart Disease, CHF, any other major comorbidity Renal Failure, liver failure, or disseminated malignant disease Diagnosis Mallory-Weiss lesions or no lesion observed and no stigmata of recent hemorrhage All other lesions Malignant lesions of the upper GI tract ...... Stigmata of recent
hemorrhageNo stigmata or dark spot in the ulcer base ..... Blood in upper GI tract, adherent clot, visible or spurting vessel ......
Observed Rebleeding and Mortality by Rockall Risk Score
Score Total number of Patients (% of Total) Number of Cases
Rebleeding Deaths w/o rebleeding Deaths with rebleeding Total Deaths 0 143 (5.6) 7 (4.9) 0 0 0 1 278 (11.0) 9 (3.2) 0 0 0 2 323 (12.8) 16 (5.0) 1 (0.3) 0 1 (0.3) 3 402 (15.9) 49 (12.2) 4 (1.1) 4 (10.0) 8 (2.0) 4 450 (17.8) 62 (13.8) 10 (2.6) 9 (14.5) 19 (4.2) 5 367 (14.5) 62 (16.9) 16 (5.2) 13 (21.0) 29 (7.9) 6 238 (9.4) 70 (29.4) 16 (9.5) 20 (28.6) 36 (15.1) 7 202 (8.0) 80 (39.6) 12 (9.8) 28 (35.0) 40 (19.8) >=8 128 (5.1) 61 (47.7) 18 (26.9) 32 (52.5) 50 (39.1) Total 2531 416 (16.4) 77 (3.6) 106 (25.5) 183 (7.2) Endoscopic Therapy
6.0 Early endoscopy (defined as within the first 24 hours of admission) and endoscopic therapy, with risk stratification based on clinical and endoscopic findings, can distinguish low and high risk patients (see 5.1 and 5.2). Observational studies show that low risk patients may be discharged to outpatient settings with minimal risks of complications. Case control and cohort studies show improved resource utilization with appropriate risk stratification. Expert opinion supports improved outcome in high risk patients with appropriate risk stratification.
7.0 With low risk stigmata on endoscopy, no endoscopic therapy is needed. When a clot is found in the ulcer bed, you should attempt dislodgement by irrigation and treat based on the findings. A finding of high risk stigmata implies a need for immediate endoscopic hemostatic therapy. These recommendations are all supported by at least 1 randomized prospective controlled trial (RPCT).
8.0 No single injection solution for endoscopic hemostatic therapy is known to be superior to another. Based on RPCT. Tested injection solutions include: epinephrine; cyanoacrylate; epinephrine with ethanolamine or polidocanol; thrombin; sodium tetradecyl sulfate; ethanol.
9.0 No single method of endoscopic coaptive (literally "to put side by side") therapy is known to be superior to another. Based on RPCT. Available methods include: heater probe thermocoagulation, multipolar electrocoagulation, or neodymium yttrium-aluminum-garnet laser.
10.0 Either injection or thermal coagulation is an effective therapy for high risk stigmata, but both together are superior to either treatment alone. Fair evidence from RPCT.
11.0 The placement of clips is a promising endoscopic technique for high risk stigmata. Fair evidence from RPCT.
12.0 Routine second look endoscopy is not indicated. Good evidence by RPCT not to use this procedure.
13.0 In cases of rebleeding however, a second attempt at endoscopic therapy for control is generally indicated. Good evidence from RPCT.
14.0 Surgical consult should be obtained when patients fail endoscopic therapy. Evidence from cohort and case control studies, and common sense.
Pharmacotherapy
15.0 Proton pump blocking medications are the recommended medicines to treat acute upper gastrointestinal bleeding episodes. H2 blocker therapy is not preferred or recommended.
16.0 Somatostatin and octreotide are not recommended in the routine management of acute nonvariceal upper GI bleeding.
17.0 An intravenous bolus followed by continuous-infusion proton-pump inhibitor is effective in decreasing rebleeding in patients who have undergone successful endoscopic therapy. Good evidence by RPCT.
Works better than H2-blocker therapy.
Decreases rebleeding rates, mortality, and surgery.
Pantoprazole or omeprazole: 80 mg bolus IV, followed by 8 mg/hr IV x72 hours after successful endoscopic therapy.18.0 For patients awaiting endoscopy, expert opinion favors consideration for empirical therapy with a high dose proton pump inhibitor.
Data here are less clear. Studies may have been hampered by inadequate dosing so far. Some studies suggest an improvement in stigmata with proton pump therapy prior to endoscopy. Await further studies.
19.0 Patients considered low risk at endoscopy can be fed at 24 hours. Good evidence by RPCT.
20.0 Patients with UGI bleed should be tested for Helicobacter pylori and treated if positive. Good evidence by RPCT.
21.0 All of the above principles apply to patients with non-steroidal anti-inflammatory induced ulcers.
