GI Bleeding Summary
- 1. ED Management of GI Bleeds Summary of Curbside Consult with Dr. Raddawi
- 2. Defining GI Bleed Identifying the bleed source is important and does change management Classic characteristics: PUD – Pain stops once bleeding starts (only 40% of all UGIBs) Perforation – Pain concomitant with bleeding Blood = cathartic, if no BMs, likely no active bleeding Bright red blood… Per mouth = UGI source (always) Same for coffee-ground emesis Per rectum = LGI source (most of the time) BRBPR in a stable pt is NOT a UGIB
- 3. Melena Black, foul-smelling, tarry stool caused by bacteria in the small intestine digesting blood, signifying bleeding proximal to the cecum Requires 60cc of bleeding Requires 8h of transit time
- 4. NG Tube Insertion This is debunked…Dr Raddawi insists there is still a role for them in the ED, but the ED literature says no! +hematemesis = NGT insertion, -hematemesis = none Witting, Annals of EM, 2004, 43(4):525 Diagnostic NG aspiration sens 42%/spec 91%, PPV/NPV were 92%/64%, PLR/NLR were 11/0.6 The final diagnosis was correctly predicted by the results of NG aspiration in 66% of the patients NG aspiration was positive in only 69% of the patients with a definitive upper GI source of bleeding (20/29) CONCLUSIONS: These results illustrate the limited value of NG aspiration for the evaluation of GI bleeding in patients without hematemesis. A positive test is a good predictor of an upper GI source of bleeding, but a negative test provides virtually no useful diagnostic information!
- 5. NGT for Cirrhotics or Variceal Bleeds? Go for it, perfectly safe! Digestive Dis 1973;18(12):1032 Anesth Analg 1988;67:283 Variceal bleeds are often brisk, an NG tube is essential to avoid aspiration
- 6. Guiac Positive Only takes 2.5cc blood in the GI tract to turn the card positive Technically, development requires 2-3min (not secs) Iron, pepto makes stools look black, but are –guiac False positives… Red meat Turnips Vit C Horseradish
- 7. History In the ED, probably our best strategy at defining the source is to ask the patient… UGI Source Likely If… Alcoholic/heavy drinker Frequent NSAID/ASA/alka- selzer user History of PUD/gastritis Prior UGIBs High stress lifestyle Recent ICU admission, intubation, prolonged hospitalization
- 8. But Does Source Matter? It all depends on the patient… If they are… Stable – Probably not Going home – Probably not (they need to f/u anyway) Admitted (asx’ic anemia, high risk pt, multiple co-morbidites, etc) – Probably not Unstable – Probably so But the unstable hemorrhaging patient will get scoped above and below anyway; so once they are stabilized, you are done It depends on your belief as to whether medications help or not The research does not always adequately identify source!
- 9. This Score Might Glasgow-Blatchford Bleeding Score (GBS) Risk stratifies those patients at risk for requiring medical intervention (who stays and who goes) BUN sBP ≥6.5 – 8 2 100-109 1 ≥8.1 – 10 3 90-99 2 ≥10.1 – 25 4 <90 3 ≥25.1 6 Other Markers Hemoglobin HR>100 1 ≥12 – 13 (men) 1 +melena 1 ≥10 – 12 (men) 3 +syncope 2 <10 (men) 6 +liver dz 2 ≥10-12 (women) 1 +CHF 2 <10 (women) 6 Score >6 associated with 50% risk of complications Masaoka, et al. J Gast Hep, 2006
- 10. Medical Management Let’s get to the heart of the matter…ED interventions As previously stated, most GI bleeders are stable and likely can be discharged home without much more than a H2- blocker or PPI and lifestyle modifications with outpatient follow up If they are unstable, resuscitate them, inform GI emergently and watch the patient closely For everyone else…the data is suspect to say the least
- 11. Stomach pH Green, et al (1978): Acid and pepcin’s effects on platelet aggregation and prolonged bleeding (in vitro) @pH 7.4 – 70-80% platelet aggregation @pH 6.8 – 20% @pH 5.9 – 0% Similar trend with PT and PTT Lowe, et (1980): Gastric juices promote fibrinolysis Bottom Line: Increasing gastric pH>7.4 will facilitate hemostasis by disinhibiting platelet aggregation and fibrinolysis
- 12. What is Important?!?! Before we continue, it is important to ask, what is important? Morbidity, mortality, patient outcomes? To the ED doctor, this is true In the many research studies herein outlined, these don’t matter as much as things like… Quality of ulcer base on endoscopy (based on Forrest Classification which has poor inter-relater reliability among GI specialists) Rate of re-bleeding at 5 days Units of blood transfused, etc… Just keep this in mind!
- 13. H2-Blockers Levine JE, et al. Meta-analysis: the efficacy of intravenous H2-receptor antagonists in bleeding peptic ulcer. Aliment. Pharmacol. Ther. 2002;16(6):1137-1142. Placebo vs. H2 during UGIB, n=4000 No difference in mortality, 2.5-3% ARR in re-bleeding (p=0.053) for H2, need for surgery 2.5% ARR (p=0.057), 7% ARR in re- bleeding and need for surgery of specifically gastric ulcers, including trend toward decreased mortality for H2B (sub- group analysis) Leontiotis has Astra-Zeneca ties (makes protonix), works for Cochrane Underpowered review
- 14. H2-Blockers Levine recommend stopping H2B usage and use PPIs, citing rate of H2 tolerance as cause Netzer Am J Gastroent, 1999, compared stomach pH’s and showed no difference in pH at 24h b/w H2B and PPI, but lower pH in PPI group at 48-72h But does this matter in the ER? We need to reduce the acid to reduce bleeding rate, and H2Bs are effective in first 24h! Bottom Line: H2Bs do reduce stomach pH and do reduce re- bleeding rate, need for surgery, and possible reduce patient mortality compared to placebo
- 15. IV vs PO Pepcid? IV (20mg) better than PO for gastric acid secretion Ryan, JR et al. Comparison of effects of oral and intravenous famotidine on inhibition of nocturnal gastric acid secretion. Am J Med 1986;81(4), 60-64 After 1h, gastric pH ~7.2 for pepcid 20mg IV
- 16. PPIs (i.e. Protonix) This is touted as the standard of care, but with little supporting evidence Theory: PPIs more specifically blunt gastric acid secretion, greater onset of action, no tolerance effect by patients
- 17. PPIs Sreedharan A, et al. PPI treatment initiated prior to endoscopic diagnosis in UGIB. Cochrane Rev. 2010;(7):CD005415 Took undifferentiated ED GIB pts and compared PPIs vs placebo before endoscopy, n=2000 No difference in mortality, no benefit for rate of re-bleeding, no difference in need for surgery PPIs improved characteristics of ulcer base at time of endoscopy (9% over placebo) based on Forest Classification Note no value in giving PPI prior to EGD Bottom Line: Prior to EGD, PPIs have no benefit on clinically significant outcomes
- 18. Are PPIs Toxic?!?! Daneshmend TK, et al. Omeprazole versus placebo for acute UGIB: randomised double blind controlled trial. BMJ. 1992;304(6820):143-147 n=1147, pts get EGD 24h after enrollment from ED Rate of re-bleeding and need for transfusion no different between PPI and placebo Mortality (death over next 2-3wks)… 6.9% (PPIs), 5.3% (plac), p>0.05 (not sig) Author states, “high dose PPIs might be toxic!” Bottom Line: WTF?!?! PPIs have possibly greater (or equal) mortality over doing nothing!
- 19. More on PPIs Lau JY, et al. Omeprazole before endoscopy in patients with GIB. N. Engl. J. Med. 2007;356(16):1631-1640 Industry-supported study through author affiliations, n=638 “No sig diff b/w omep and placebo group in transfusion requirements, rebleeding, mortality, or pts needing emergent surgery, however, fewer clean-based ulcers and non-bleeding ulcers on EGD (stigmata).” This study is listed as “pro PPIs” in most reviews, but… The author’s initial protocol iteration defined “mortality” as primary outcome, not “need for EGD tx”…mid-study primary outcome change!!! (BAD!)
- 20. Does Anyone Benefit From PPIs? Leontiadis GI, et al. PPI treatment for acute peptic ulcer bleeding. Cochrane Rev. 2006;(1):CD002094 Study doesn’t stipulate, but this implies that you know the source of the bleed as peptic (so likely, post-endoscopy, in the ED, we don’t know!) n=4400, equal morality, 6.5% (p<0.05) benefit in re-bleeding vs “control groups” (not all placebo), 3% (p=0.05) fewer going to OR Asians have mortality benefit to receive PPIs (no non-Asian sub-group analysis performed)
- 21. PPIs for Non-Asians? Post-hoc analysis of the following studies show… Daneshmend study (1992), 5% increase in mortality in PPIs for bleeding PUD pts Hasselgren (1997)…stopped early for harm (p>0.05 in mortality for older pts, 6% vs 1%), met only 75% of enrollment goal before study stopped (using low-dose PPI, 20mg PO) Schaffalitzky (1997)…no mortality difference, ~8%, only enrolling younger patients, stopped early along with Hasselgren because of similar design features Bottom Line: For non-Asian UGIB pts with the source being a peptic ulcer, a trend toward increased mortality exists 1. Daneshmend TK, et al. Omeprazole versus placebo for acute UGIB: randomised double blind controlled trial. BMJ. 1992;304(6820):143-147 2. Hasselgren G, et al. Continuous intravenous infusion of omeprazole in elderly patients with peptic ulcer bleeding. Results of a placebo-controlled multicenter study. Scand. J. Gastroenterol. 1997;32(4):328-333 3. Schaffalitzky de Muckadell OB, et al. Effect of omeprazole on the outcome of endoscopically treated bleeding peptic ulcers. Randomized double- blind placebo-controlled multicentre study. Scand. J. Gastroenterol. 1997;32(4):320-327.
- 22. PPIs & Stomach pH Metz DC, et al. Lansoprazole regimens that sustain intragastric pH>6.0: an evaluation of intermittent oral and continuous intravenous infusion dosages. Aliment. Pharmacol. Ther. 2006;23(7):985-995 Their goal was to raise pH>6, achieved this ~33% of the time, mean pH=5.45 Wang C-H, et al. High-dose vs non-high-dose PPI after endoscopic treatment in patients with bleeding peptic ulcer: a systematic review and meta-analysis of randomized controlled trials. Arch. Intern. Med. 2010;170(9):751-758 Hi-dose = 80mg bolus w/ 8mg/h gtt (higher doses do achieve higher pH), but hi vs low has no bearing on OR, need for transfusion, mortality outcomes, etc No difference between PO vs IV Bottom Line: PPIs are equivalent IV or PO, higher doses achieve higher pH, but no studies show the pH exceeding 7.4!
- 23. Official GI Society Guidelines on GIBs Barkun AN, et al. International consensus recommendations on the management of patients with nonvariceal upper gastrointestinal bleeding. Ann. Intern. Med. 2010;152(2):101- 113 Pre-endoscopic PPI’s may be considered to (1) downstage the endoscopic lesion and (2) mitigate need for endoscopy, but should not delay EGD Bottom Line: PPIs have no clear benefit and might be harmful! All patients with a UGIB need endoscopy, early if sick, later if stable!!!
- 24. Helping the Endoscopist So you have a patient who had hematemesis at home (none currently), HR 120, BP 80/50, guiac+ tongue and stool, you have fluids running… No need to insert an NGT (not actively vomiting) This pt needs a scope, and the best outcomes are achieved by the cleanest views of the bleeding ulcer…so clear out the stomach so the endoscopist can see!
- 25. Erythromycin Motilin receptor agonist, promotes gastric emptying 1x dose at 3 mg/kg given 20 to 120 minutes prior to procedure more effective than NGT and placebo! Coffin B, Pocard M, Panis Y, at al.. Erythromycin improves the quality of EGD in patients with acute upper GI bleeding: a randomized controlled study. Gastrointest Endosc 2002;56:174-9. Frossard JL, Spahr L, Queneau PE, at al.. Erythromycin intravenous bolus infusion in acute upper gastrointestinal bleeding: a randomized, controlled, double-blind trial. Gastroenterology 2002;123:17-23. Bottom Line: Note the dates on these studies, this is NOT new research; erythromycin works, give it if your patient needs an emergent endoscopy!
- 26. Octreotide The last of the commonly prescribed drugs needs its time in the spotlight Theory: Synthetic somatostatin with greater potency and longer half-life than endogenous somatostatins which function to reduce splanchnic blood flow Cirrhotics have baseline splanchnic dilation secondary to excessive CO levels, making them pressor-resistant as well Note: Octreotide is NOT FDA approved for treatment of variceal bleeding Note: UCB (pharma company manufacturing octreotide) failed to release data from a clinical trial comparing octreotide vs placebo because (and this is a direct quote) “UCB is an ethical company and that all data are proprietary solely to the UCB”!!!
- 27. Class IA Recommendation ACGE 2007 recommended octreotide for all variceal bleeds (class IA recommendation) Only looked at a meta-analysis of 8 clinical trials, only concerned about 5d re-bleed rate on pts treated with octreotide (was some benefit, NNT=6), but no other factors had clinical significance Many of the included studies had major problems… Clinical heterogeneity (Corley, 2001) Inaccurate mortality data that was incalculable from the numbers presented (Besson, 1995) Noted failure of included studies to report calculated data (D’Amico, 2002) Failure to demonstrate a mortality benefit
- 28. Flawed Pathophysics? Møller S, et al. Effect of octreotide on systemic, central, and splanchnic haemodynamics in cirrhosis. J. Hepatol. 1997;26(5):1026-1033 Showed that central and mesenteric arterial blood volume decreases with octreotide and mesenteric vasoconstriction occurs, but hepatic venous pressure gradients and blood flow do NOT change Escorsell A, et al. Desensitization to the effects of intravenous octreotide in cirrhotic patients with portal hypertension. Gastroenterology. 2001;120(1):161-169 Portal HTN pts w/ varices are desensitized to octreotide’s effects. Pt’s had a decrease in portal pressures and increase in arterial pressures, but effect lasted only 5 minutes and then reversed (the effects weren’t even sustained with a drip)
- 29. Failure to Launch ACGE recommendations are based on morbidity data of a 5d re-bleeding benchmark, but octreotide failed to show in any of the included studies any mortality benefit, differences in blood transfusion reuirements, length of hospital stay, etc. Bottom Line: Octreotide has a Class IA recommendation for use in patients with a variceal bleed, but a close look at the literature shows a lack of support for this recommendation for any ED-specific, clinically relevant outcomes!
- 30. Summary Critical bleeding from a gastric source is exacerbated by low stomach pH NG tubes are not needed during the diagnostic process H2 Blockers do have utility in the acute stages of bleeding but tolerance develops, more studies are needed PPIs don’t have data that shows patients benefit from receiving it in the ER and might increase mortality, some benefit in Southeast Asian populations Octreotide shows decreased incidence in 5d rebleeding, not much other benefits It is important to constantly look at the things we do “all the time” in medicine to make sure we are always doing what’s best for the patients.