Vomiting of blood, its source usually proximal to the DJ flexure and
generally indicating a rapidly bleeding lesion.
Passage of tarry black stools, the colour and characteristic smell
being due to bacterial degradation of blood which usually originates
the stomach, duodenum or small bowel.
D I A B M I M
50% Peptic ulcer disease (mortality <5%)
10% Gastric erosion / stress ulcers
10% Mallory Weiss tear.
10% Oesophageal Varices (mortality 30%).
10% Oesophagitis / Gastritis
- 1ml/min adds to 1.5L / 24
1% Upper GI cancer.
5% Vascular malformations.
Rest: multiple pathologies / other causes.
Age and comorbidity are major determinants of prognosis.
75% of PUD bleeds stop spontaneously or with medical intervention.
arise on posterior duodenum; can penetrate gastroduodenal
D2 ulcers frequently bleed
1-4cm tear in mucsoa near GJ following vomiting.
- 75% confined to stomach, 20% straddle jx, 5% in distal esophagus.
- many have a hiatal hernia.
Acute Haemorrhagic Gastritis
Often in systemically / critically ill patients due to Curling ulcer type pathophysiology.
D I A B M I M
- coexistence suggests severe haemorrhage
Bright red blood PR is possible with profuse bleeding.
May be stable, transient responders, or unstable.
vs Slow Trickle
may be only sign
- how much?
- pud symptoms?
- previous episodes / ulcers?
- liver disease / alcohol?
- family history (common in pud)?
- alcohol and smoking (smoking>alcohol for PUD causation)
- drugs (eg NSAIDs, anticoagulants)?
- aortic surgery (fistula assumed)?
- liver disease / scars.
Typically forceful retching then blood.
90% stop with ice water gastric lavage.
- sometimes endoscopic intervention, occasionally surgery.
D I A B M I M
Hb: unreliable for
Urea: urea/creat ratio
highly predictive of upper GI cause.
D I A B M I M
More precise blood transfusion, undertransfusion.
Late decision to operate.
Normal BP & pulse, Hb>100.
Admit, hrly obs.
Drink from outset.
Normal diet within 24hrs.
- once clear evidence no more bleeding.
- repeat Hbs.
- remember to fast pending endoscopy
Think about H.pylori, NSAIDs, acid suppression.
Pulse >100, BP<100sys, Hb<100, Age>60
Most have significant comorbidity.
2 large leurs
Catheterise if reqd.
Large bore cannulae
- (preferably 2, but don't delay resuscitation to gain second)
- blue = 31ml/min; pink 55, green
gray 170, orange 275.
- X-match 4-6units
- LFGS, U&Es, PR.
IV fluids, catheter as appropriate
- cautious resuscitation, controlled hypotension.
In 70-80% bleeding will stop spontaneously
Want Hb 100-110.
If >3units blood reqd
--> surgical consideration needed.
- or if suspected perf / ongoing bleeding.
- FFP x2 units may be reqd.
- rarely plasma reqd.
What if they are on dabigatran?
Won't wear off for 24 hrs, at least.
If renal impairment, make than 72.
Ideally within 24 hrs, especially
- earlier if hypotension persisting / large volume.
Stigma of ongoing risk:
- active pulsatile bleeding (high risk)
- visible vessel / clot
- ongoing oozing (lower risk)
- varices or cancer
--> surgical consideration needed; high risk of rebleed (Greath)
Reasonable to defer biopsy in presence of active bleeding.
If endoscopy normal, consider haemobilia.
Control bleeding; residual oozing is a worry.
Adrenaline injection or thermocoagulation reduces rebleed risk by
- 20% will still rebleed within 72hrs.
Indications for PUD Repair
Exsanguinating haemorrhage that can't be controlled endoscopically
Initial sentinel bleed is controlled, but subsequently have a
Risk of not operating becomes greater than risk of operating in
<60 years, no comorbidities.
OR >6 units of blood in any 24 hour period.
OR >12 units of blood total.
>60 years, comorbidities.
1 re-bleed (may attempt endoscopic control first again, perhaps in
OR >4 units of blood in any 24 hour period.
OR >8 units of blood total.
Primary goal is to stop bleeding.
Secondary goal is to prevent rebleeding.
Gastric ulcer surgery
As for perforation - localized excisions
If bleeding site unknown (no prior endoscopy):
- anterior gastrotomy and find source
- in very rare instances, ligation of gastric blood supply (except
short gastrics) or even near-total gastrectomy warranted; high
May simply under-run the bleeding point and manage on medical
May need to excise ulcer and close defect.
Protects the clot rather than helping the ulcer.
Stability of clot is reduced in an acid environment.
- ph>6 needed for platelet aggregation
- clot lysis occurs in pH<5.
Daneshmend et al originally conducted a placebo-controlled RCT,
- however flawed by heterogenous inclusion of varices, tumours,
2 RCTs have since been done looking at ulcers with visible
- (Khuroo et al, Lau et al)
- reduces rate of rebleeding (6.7% vs 22.5%)
- less transfusions, hospital stay, endoscopies, operations.
Omeprazole 40mg q6hrly po.
- if allowed oral intake.
- may take a day or two to get ph up.
Omeprazole 80mg IV stat
40mg q6hrly IV
- given in 100ml N saline over 20-30mins.
(costs $20 a day).
Stable infusion form would be best, but not available in NZ.
Second best is IV dosing.
- whether or not varices suspected.
- esp if endoscopy going to be delayed.
50mcg IV stat, then 25 mcg/hr in N saline.
Consider prior to endoscopy
- encourages gastric emptying.
IV 250mg, 30-60m prior to endoscopy.
Vigorously correct coagulopathy.
Broad spectrum ABs
Consider early B-blockers.
Sengstaken/Blakemore tube +/- intubation may be necessary.
D I A B M I M
NEJM editorial 343:358-359.
Daneshand et al. BMJ 1992;304:143-7.
Khuroo et al. Comparison of Omeprazole and Placebo for
PU. NEJM 336 : 15 April 97.
Hill J. Surgical Emergencies.
Lau et al. NEJM 2000;343:310-16.