Upper GI Bleeding

April 2014 Regional Training Update – Author Greg Cranston

Disclaimer: Summary of key points only. This is intended to supplement your own work on these curriculum items and not provide a complete evidence base on which to change your practice. All drug choices and dosing calculations should be checked with external resources.

 

Upper GI Bleeding

(HAP16 – Haematemesis and malaena)

Intro

• Defined as bleeding proximal to Ligament of Treitz (Duodenojejunal flexure)
• Vomited blood obviously diagnostic, but MUST confirm malaena PR or seeing stool ourselves
• 70% can stop spontaneously, but 20% re-bleed 24-72hrs
• They die from comorbidities / cirrhosis. Exsanguination is rare.

 

• Guidelines:                                                                                                                                          

SIGN 105 Sept 2008

                    NICE CG141 June 2012

 

Overall:

Resuscitate first -> Then diagnose -> Then treat definitively

Initial management / resuscitate

o   If INR>1.5 or fibrinogen <1 – give FFP

o   If on Warfarin – FFP / Vitamin K

o   If on Warfarin and ongoing bleeding – Prothrombin complex concentrate

o   If Platelets <50 and ongoing bleeding – give platelets

o   Scoring systems: predict high risk re-bleed and mortality. For us, Rockall pre-endoscopy is out, Glasgow-Blatchford is in. A score of 0 probably means discharge is safe. (http://www.mdcalc.com/glasgow-blatchford-bleeding-score-gbs/ )

o   PPIs? Not pre-endoscopy. It won’t do any harm if you really want to, but it’s unlikely to change anything in a big bleed, so basically unnecessary. Post-endoscopy, if stigmata of recent haemorrhage, it will be started.

 

Endoscopy

o   To define specific cause for a) prognosis and b) haemostasis. But beware complications in those actively bleeding, and can be life-threatening in the unstable or with comorbidities. Immediately following resuscitation in those with severe bleeding.

o   Inject with adrenaline plus a) Thermal coag or b) Mechanical e.g. clips/staples

o   Repeat endoscopy if re-bleeds, at high risk for re-bleed, or doubt about haemostasis.

o   If all fails and patient unstable: interventional radiology, or surgery if not available.

 

Extra info

o   Stop aspirin / clopidogrel temporarily / (NSAIDS possibly permanently). Locally they restart aspirin within 1 week – it lowers mortality / fewer ischaemic events, despite more re-bleeding. Discuss clopidogrel though.

o   Ranitidine / PPI as primary prevention for critical care admissions was mentioned

o   H.pylori is becoming less prevalent. Stool antigen testing / Breath testing is best, serology exists but is variable. CLO test at endoscopy also used.

 

Variceal Bleeding – additional points

o   12% of those undergoing endoscopy. 60% of ascites will have varices. 40% mortality per bleed.

o   Consider airway protection

o   Terlipressin 2mg iv. Depending on brand, given every 4hr for up to 5 days or until bleeding stops (off-license after 2-3 days). Reduces mortality, better haemostasis, and less additional procedures needed.

o   Antibiotic prophylaxis e.g. Tazocin (but local guidelines obviously)

o   Endoscopy: Band ligation, or glue

o   TIPS (Transjugular Intrahepatic Portosystemic Shunts) considered if above fails

 

Sengstaken-Blakemore tube

• Last resort device to try and control variceal bleeding.
• Consider securing airway first.
• Goes through mouth or nose. Only inflate Gastric balloon initially (not oesophageal).
• Pull it back to the gastro-oesophageal junction, 1litre bag of fluid on the end for traction.

Further reading:

St Emlyn’s : Glasgow Scores – Not just for coma anymore  by Gareth Hardy

Thinking Critical Care (Philippe Rola) : “But Doctor, he’s vomiting blood” a summary of the NEJM paper on transfusion in UGIB