Jump to content
Main menu
Main menu
move to sidebar
hide
Navigation
Main page
Recent changes
Random page
Help about MediaWiki
Special pages
Surgopaedia
Search
Search
Appearance
Create account
Log in
Personal tools
Create account
Log in
Pages for logged out editors
learn more
Contributions
Talk
Editing
Upper GI bleed
Page
Discussion
English
Read
Edit
Edit source
View history
Tools
Tools
move to sidebar
hide
Actions
Read
Edit
Edit source
View history
General
What links here
Related changes
Page information
Appearance
move to sidebar
hide
Warning:
You are not logged in. Your IP address will be publicly visible if you make any edits. If you
log in
or
create an account
, your edits will be attributed to your username, along with other benefits.
Anti-spam check. Do
not
fill this in!
Loss of blood from the alimentary canal proximal to the ligament of Treitz * 'massive GIB' - haemodynamic instability or transfusion requirement * 'occult GIB' - anaemia that persists or recurs after negative workup == '''Epidemiology''' == * About 60% of all acute GIT bleeds == '''Aetiology''' == * '''Most commonly EROSIVE (oesophagus, stomach or duodenum) or VARICES''' ** So if varices are not suspected, it's most likely to be erosive ** In liver patients (cirrhosis or portal hypertension), 90% of UGIB are caused by varices * '''Erosive/ulcerative''' ** Oesophagitis (13%) *** Chronic exposure to gastric acid leads to an inflammatory response that can result in chronic blood loss, perhaps also with ulcerations *** Other underlying causes: **** Infection (candida, HSV, CMV) **** Crohn disease **** Radiation **** Medications (NSAIDs, oral bisphosphonates, tetracyclines) *** Haematemesis very common, melaena less common *** Treat with PPI. Endoscopic coagulation is generally successful. Surgery seldom necessary. ** PUD (see separate topic and section below) - about 60% of UGIB *** 10-15% of patients with PUD will develop bleeding at some point *** Bleeding ulcers are more frequently associated with NSAIDs than H. pylori, but note that H. pylori tests have reduced sensitivity in bleeding, so this may be an underestimate *** 75% will stop bleeding with IVF and IV PPI *** Duodenal ulcers are more likely to stop bleeding spontaneously, and have lower morbidity/mortality rates than gastric ulcers *** Significant bleeding occurs when the erosion reaches a submucosal artery, or an even larger artery in penetrating ulcers (GDA or LGA) ** Gastritis *** See separate topic - stress gastritis ** Duodenitis * '''Complications of portal hypertension''' ** Oesophagogastric varices *** See separate topic *** Look for signs of portal hypertension (splenomegaly, thrombocytopaenia, ascites) and stigmata of chronic liver failure (jaundice, palmar erythema, gynaecomastia, spider angiomata, testicular atrophy, Dupuytren's contracture). *** Often blood in GIT precipitates decompensation in cirrhotics, leading to encephalopathy. ** Ectopic varices ** Portal hypertensive gastropathy *** Often smaller volume bleeding. Still see signs of portal hypertension/cirrhosis on exam such as splenomegaly. * '''Vascular lesions''' ** Angiodysplasia ** Dieulafoy's lesion *** Sudden-onset large volume painless bleeding *** See separate topic ** Gastric antral vascular ectasia (GAVE syndrome) *** Collection of dilated venules that appear as a linear red streak converging on the antrum, giving the appearance of a watermelon *** Associated with cirrhosis/portal HTN, collagen vascular disease (scleroderma), IHD, CKD, pernicious anaemia, and bone marrow transplant recipients *** Severe haemorrhage is rare, with most patients presenting with persistent iron deficiency anaemia from continued occult blood loss *** Endoscopic therapy (APC) indicated for persistent anaemia or transfusion-dependent bleeding. Success rate >90%. Typically 2-3 treatments are required, separated every 1-2 months. *** Consider antrectomy for those that do not respond to endoscopic therapy. * '''Traumatic/iatrogenic''' ** Mallory-Weiss tears (5-10%) *** See separate topic ** FB ingestion ** Post-surgical anastomotic bleeding ** Post-instrumentation bleeding *** Endoscopic sphincterotomy - inject with adrenaline and gold probe ** Aortoenteric fistula *** Separate topic ** Cameron Lesion *** Erosions or ulcers occurring in the sac of a hiatal hernia *** Usually incidental, but can cause bleeding *** Surgery to repair hiatus hernia can be considered in patients with recurrent bleeding despite conservative management * '''Tumours''' ** Upper GI tumours such as GIST *** Usually associated with chronic bleeding rather than acute *** Often successful in initial control, but with a high rebleeding rate *** Therefore, resection is indicated when this is diagnosed **** Palliative: wedge resection **** Curative: standard oncologic operation * '''Miscellaneous''' ** Haemosuccus pancreaticus *** See separate topic ** Haemobilia *** See separate topic under biliary ** Inflammatory bowel disease *** Very rare to present as UGIB *** Can get duodenal ulcer - treat in the usual way ** Swallowed blood (nosebleed) == '''Aetiology of post-op bleeds:''' == * PUD/stress ulcer (more likely with sicker pts) * M-W tear * Varices * Anastomotic * All other causes still possible == '''Identifying anatomic location''' == * Factors predictive of UGI source ** Pt-reported melena (LR 5) ** Melena on exam (LR 25) ** BUN:serum creat >30 (LR 7.5) *** This references mg/dL units for both urea and creatinine *** If using urea in mmol/L and creatinine in micromol/L like in Australia, ratio >0.121 predicts UGI source ** Blood/coffee grounds on NGT (LR 9) ** No blood clots in stool (LR 0.05) * Imaging ** CTA *** Sensitivity 100%, specificity 90% *** Ideally three-phase scan with just IV contrast *** Can apparently detect bleeds as slow as 0.3mL/min ** Technetium-99-m-labelled RBC scintigraphy *** Can detect bleeds as slow as 0.04mL/min * Gastric lavage ** Inject 50mL of saline to the NGT and aspirate ** Bile-stained fluid is fairly sensitive ** If no bile, could be duodenal bleed with pyloric spasm, which happens commonly == '''Clinical manifestations''' == * Melaena ** Malodourous, black, tarry stool ** By-product of haemoglobin degradation by digestive enzymes and intestinal bacterial flora into haematin ** Can occur with as little as 50mL of bleeding ** Generally means proximal to ligament of Treitz (90%), but consider nasopharynx through to right colon as sources * Haematochezia (usually means lower GI) * Haematemesis ** Red blood means ongoing/severe ** Generally means proximal to ligament of Treitz * Don't usually have pain - blood is alkaline and serves as an antacid == '''Approach to significant UGIB''' == === Resuscitate === ## ABCD ## 2x IVC + IVF ## Bloods - VBG, FBE, UEC, LFT, coag, x match (aim Hb > 70, or >90 if CAD/active bleeding) ## Antiemetics, analgaesia ## May need to call consultant/OT/anaesthetist/ICU ## Transfuse (once 1L crystalloid has been given) and MTP if needed (EMST - 10 units in 24 hours or 4 units in 1 hour). Give at least 1 unit FFP per 2 units pRBC. In general, stick to standard transfusion targets, but can give more blood early due to unclear true degree of blood loss. ## Sistaken-Blakemore tube if truly unstable (can leave for 36 hours; generally very effective) === Initial management for admission === ## ?ICU ## NBM ## TEDs, no enoxaparin ## IV PPI 80mg BD (leads to fewer Forrest Ia to Iia ulcers on endoscopy) ## Antibiotic prophylaxis in patients with cirrhosis ## May need NGT ## ?IDC ## Stop antihypertensives, anticoagulants and antiplatelets ## Consider correcting coagulopathies ## ?thiamine ## If suspected variceal bleeding (any symptoms or findings suggestive of portal hypertension)Β - octreotide - not useful otherwise unless endoscopy unavailable - see below section on varices ## Probably no role for tranexamic acid (HALT-IT trial, 2020 - no benefit to mortality) ## AIMS65 score to predict in-hospital mortality from UGIB ### One point for each of: #### Albumin <30 #### INR >1.5 #### Altered mental status #### SBP <90 #### Age >65 ### Each point is equally weighted and gives increasing mortality: 1, 5, 10, 15, 25% ## Glasgow-Blatchford score to identify low-risk UGI bleeds ### ## If big bleed, consider gastric lavage prior, or at least having a tube ready === Identify cause === ## Aim endoscopy at 8-24 hours (unless meets indications for immediate endoscopy - below) ### Consider IV metoclopramide 10mg or erythromycin 3mg/kg, 1 hour prior to endoscopy as a prokinetic (erythromycin has more convincing data behind it to reduce need for second-look endoscopy, but metoclopramide is probably better than nothing) ### Consider putting a nasogastric lavage tube on standby if active bleeding and going to endoscopy ### If can't see anything because of blood - try repositioning the patient upright ## Indications for immediate endoscopy: ### Suspected variceal bleeding ### Non-responder or transient responder to resuscitation ## PUD - will need endoscopy. If suspected bleeding ulcer should be within 24 hours. If serious bleeding, 12 hours. Endoscopy within 8 hours can often mean a bad view and worse outcomes! So aim 8-12 hours. ## Oesophagitis/gastritis ## Varices - look for evidence of portal hypertension - see below for specific management # Post-procedure care ## A conservative approach would be to keep to clear fluids for 48 hours if bleeding was found ## Consider abdo USS to exclude portal vein thrombosis ## Classically an NGT on suction would be used to assess for re-bleeding, but this is unnecessary and causes disproportional discomfort === '''Normal gastroscopy - the next step''' === * CTA if it is likely to show anything * If not, prep the patient and do a colonoscopy (35% GIT bleeds from large bowel) ** Can do 'rapid prep' - 4L polyethylene glycol given within a period of 4 hours (NGT), followed by colonoscopy within 1-2 hours * If colonoscopy is normal, evaluate small bowel (5% of acute GIT bleeds) ** CTA/RBC scan/pillcam/push endoscopy * Obscure-overt bleeding (no cause after all these investigations, but ongoing) should have a repeat upper and lower endoscopy == '''Bleeding ulcers''' == === '''Classification:''' === ** {| class="wikitable" |Forrest classification |Stigmata |Risk of re-bleeding on medical management |- |I a |Brisk bleeding/spurting |90% |- |I b |Oozing |10-20% |- |II a |Non-bleeding visible vessel |50% |- |II b |Adherent clot |25-30% |- |II c |Flat pigmentation spot |5-10% |- |III |Clean-based ulcer |3-5% |} *** Iic ulcer ** Intervene on category I, II a and II b ulcers ** Other high-risk factors - ulcers >2cm, posterior duodenal wall ulcer, lesser gastric curve ulcer === '''First-line: endoscopic intervention''' === ** Wash off adherent clot if present ** inject with ~10mL total of 1:10,000 adrenaline spread between all four quadrants, AND another modality ** Either thermal (gold) probe, haemoclip (IIa ulcers), APC or sclerosant (alcohol) injection *** Push gold probe into base of ulcer and fire - bleeding is most often caused by an artery in the base, and you can collapse the walls then fire the probe. Fire it a few times. ** Important thing is to use dual modality === '''Second-line: embolization''' === ** If no resolution possible on endoscopy, proceed to transcatheter arterial embolization (TAE) - seems to have equal outcomes with surgery and lower complication rates ** Better for duodenal rather than gastric bleeds ** Especially useful with haemorrhage into biliary tree or pancreatic duct ** Most common bleeding vessel is GDA, left gastric, right gastric, splenic arteries ** Higher rates of re-bleeding from TAE than surgery, but good in high operative risk patients ** Efficacy for UGIB 44-100% ** Previously vasopressin infusion may have been used, but not deployed any more === '''Third-line: surgery (needed in 5-10%)''' === ** Indications: *** Failed endoscopy x2 *** Haemodynamically unstable despite vigorous resus (>3 units) **** Previously, >= 6 units in index 24 hrs was indication for surgery *** Recurrent haemorrhage after haemostasis with endoscopy (reasonable to try twice with endoscopy) *** Ongoing transfusion requirement >3 units per day *** Unable to transfuse (rare blood type, refusal of transfusion) ** Technique *** Upper midline laparotomy (bleeding duodenal ulcers can technically be treated laparoscopically, but difficult) *** Examine for external signs of ulcer - will often be able to see serosal inflammatory changes at the site of a chronic ulcer *** If any doubt as to location, use intra-operative gastroscopy ** Duodenal ulcers: intervention will usually be to oversew or ligate bleeding vessel, which is the '''GDA''' at the base of a posterior ulcer *** May need a Kocher maneuvre to mobilise the duodenum and get manual control *** Longitudinal duodenotomy/pyloroduodenotomy from pylorus through to anterior wall of D1/D2 *** Remove blood, isolate bleeding *** Apply pressure and time for anaesthetics to catch up *** Underrun bleeding vessel with 3-point U-stitch technique of 2/0 Vicryl **** Careful of CBD - can insert a probe through the ampulla if necessary - although Mosche says he has never heard of it being sutured over **** Classically need to suture ligate GDA superiorly and inferiorly (12 and 6 o'clock), with another suture at 3 o'clock for transverse pancreatic branch *** Close the duodenotomy transversely (like a stricturoplasty) *** Consider vagotomy or other acid-reducing operation, but this is generally unnecessary nowadays ** Gastric ulcers *** Use the simplest possible operation that will stop the bleeding *** First gastrotomy and underrun the bleeding ulcer (30% rebleeding risk) *** Wedge resection may be a good option for greater curvature ulcers, but this gives a 10% rebleeding risk, so distal gastrectomy is favoured *** Bleeding proximal ulcers near GOJ are quite difficult to manage - proximal or near-total gastrectomy has very high mortality in setting of haemorrhage. Perhaps a distal gastrectomy with a tongue of proximal stomach, or a wedge resection/buttressed oversewing of the ulcer? === '''Rebleeding''' === ** Repeat endoscopy if technically possible. Surgery is better for haemostasis but more complications === '''Medical management''' === ** ICU? ** IIb and above ulcers should stay on IV PPI for 72 hours then can be discharged (can be BD high-dose - no benefit to infusion in most cases) *** There doesn't appear to be much difference between IV and PO PPI even while still in hospital, but guidelines still recommend staying on IV ** Pause aspirin, clopidogrel, NSAIDs and SSRIs for 24 hours with ulcer bleeding *** Aspirin can be resumed after 24 hours - rationalise ongoing need (stop it if it was primary prevention, restart for secondary prevention 1-7 days after bleeding cessation in most cases) *** Clopidogrel can be resumed after three days in patients with stents *** SSRIs five days *** COX-2 selective inhibitors combined with PPIs have a low rebleeding risk, so that is an acceptable combination if anti-inflammatories must continue. ** Don't give tranexamic acid ** Clear fluids first 24 hours then normal diet. IIc and below can go straight to normal diet. ** Discharge - IIc and below can go 24 hours after endoscopy. Anyone that needs treatment for a bleeding ulcer should stay for 72 hours. ** All patients must have H. pylori status established and eradicated if present. Once it has been confirmed eradicated, there is no need for long-term PPI as it doesn't decrease the re-bleed risk of 1.3%. [[Category:UGIS]] [[Category:Intern education]]
Summary:
Please note that all contributions to Surgopaedia may be edited, altered, or removed by other contributors. If you do not want your writing to be edited mercilessly, then do not submit it here.
You are also promising us that you wrote this yourself, or copied it from a public domain or similar free resource (see
Surgopaedia:Copyrights
for details).
Do not submit copyrighted work without permission!
Cancel
Editing help
(opens in new window)
Search
Search
Editing
Upper GI bleed
Add topic