We're going to get started. I think we're going to go through two sessions, GI bleeding and hepatic failure back-to-back. No financial conflicts. I'll acknowledge Dr. Nanchao, who is the primary orator for the sessions that you guys may have heard in the pre-study sessions. We'll start with the first case, 75-year-old patient presented to the ER with multiple episodes of hematemesis for a day, associated symptoms of lightheadedness, abdominal pain and dizziness, slightly hypotensive, tachycardic, respiratory rate of 15, afebrile, and has got other clinical signs of volume deficit, dry mucous membranes, hemoglobin's 10, rest of the labs are normal. What is your choice of access? Once this patient comes into your ICU, would you ask your nurses to try to get a 14, 16, or an 18-gauge catheter in, or would you jump and put in a central line right away? Peripheral IV access, 16-gauge or 18-gauge peripheral IVs are always the best resuscitation catheters compared to a quad lumen or a triple lumen. Introducer is a different discussion at that point. What is the bolus fluid of choice? I know we went through this discussion earlier today. A bolus fluid of choice when the patient just rolls into your ICU is always crystalloids. It is never PRBC. You don't ever want to wait for a PRBC to be delivered from the blood bank at that point. Just as a bolus fluid of choice, that's never the right approach. Coming to the transfusion strategy, we're going to go through some of these discussions. In a non-cardiac patient, studies done with the restrictive strategy or the liberal strategy with hemoglobin of 7 or more versus 10 or more, the restrictive strategy always has a better outcome. Hemoglobin goal of 10 or higher is associated with increased blood product use, obviously, but also poor patient outcome. Some of that is related to the decreased functional capacity of the blood that has been stored in the blood banks. It's come down to the GI bleed part. Most common cause of upper GI bleed is varices, peptic ulcer disease, malarie-wise steer, or esophagitis. Right. Peptic ulcer disease, that is the most common bread and butter etiology for GI bleed. Unless you are in a specialized center, then variceal bleed becomes more prominent if it's a liver center. Various epidemiological data for peptic ulcer for upper GI bleed is 35% to 50% versus diverticulosis being the most common for the lower GI bleed. Risk stratification scores. This patient comes in, and there are various ways of stratifying how sick these patients are, whether you intervene with admission to the ICU, admission to the step-down ICU, and there are some that are on the extremes, which are always easy to identify, but there are three different types of scores. There's the Glasgow Blatchford score. There's Rockwell score and the AIM-65 score that can be used to identify the right location for that patient. Again, clinical judgment supersedes all of these scores. The Blatchford score is predominantly used to identify the lowest risk patients that can ideally be discharged from the ER. Most of us would never see these patients. This is predominantly helpful in case you're going through a board question and they identify these scores. They are predominantly used to stratify where these patients should be located, not in terms of therapeutic management. So 80% of these GI bleeds, they do stop spontaneously without a re-bleed. Blatchford score predicts transfusion intervention, re-bleed, or death. So peptic ulcer disease in early EGD does reduce the length of stay. Endoscopic therapy does reduce the risk of re-bleeding, surgery, and mortality. So just a conservative treatment with a PPI versus an endoscopic intervention. Endoscopic intervention is beneficial. If a patient does re-bleed after a first EGD, the recommendation is to reconsider EGD at that point. This is the magic number of two. You always want to go through two EGDs before you get someone else involved, whether it be interventional radiology or your surgeons, depending on the etiology. If EGD fails times two, that is when we call IR for embolization of a gastroduodenal or a short gastric vessel, depending on the location of the ulcer. Pre-EGD proton pump infusions, it reduces the need for EGD interventions. It can help in reducing the risk stratification of the ulcer. Continuous versus intermittent PPI infusion, that always is a question. Sometimes it's based on the availability of the drug. Sometimes it's based on the availability of optimal IV access, because you're utilizing fluids, antibiotics, or something else. There is no evidence that continuous infusions are better compared to intermittent PPI. Risk of re-bleeding is based on the endoscopic evaluation of these ulcers. Type 1A, which is an actively bleeding ulcer, the re-bleed without any therapy is obviously 100%. 1B, oozing ulcer without a visible vessel, the risk of re-bleed is 10 to 27. A 2A is non-bleeding but a visible vessel, re-bleed risk is approximately 50%. 2B, non-bleeding but an adherent clot is 8 to 35. And then everything else, which is a clean base or a black pigmented location, the risk of bleed is less than 10%. So early endoscopy, as we discussed, is the way to go. It helps with early discharges, obviously helps with throughput, decreases length of stay, and decreases the transfusion requirements. Low risk stigmata, clean base, or a pigmented spot does not benefit from an endoscopic therapy. So these patients still need the EGD, but they do not need endoscopic therapy of CLIPS, epi-infusion, or a fibrin spray. Endoscopic therapy versus medical management doesn't happen that often, but occasionally if it's a weekend or a Friday evening, and if there's a lack of GI person available locally, they would say, well, just do a PPI, patient's not actively bleeding, you don't have symptoms, you can probably just do medical management. But endoscopic treatment is recommended because it helps with all four of these, reduces the risk of re-bleeding, reduces the risk in surgical intervention, decreases mortality, and also benefits mostly with the high-risk lesions. Dual therapy, if this is an ulcer one, or level one or level two ulcers, then dual therapy with either epi-injection and CLIP, or a CLIP and a spray is better than just performing monotherapy. Let's complicate this case a little bit. I think standard patient coming in with simple GI bleed is rare. Most of these patients come with some form of an add-on. So this patient also has a history of AFib and is on a PIXABAN for 10 years. Patient has undergone therapy for his peptic ulcer disease for the last four days and is likely to be discharged tomorrow. Patient is questioning whether they should resume a PIXABAN at this point. What is your recommendation for resuming anticoagulation? Resume in two weeks from stabilization? Patient hasn't bled for four days, doing fine, you can ask them to resume a PIXABAN today. You could say, well, we saved you once, can't save you every time, so do not ever resume a PIXABAN. Or you send them over to cardiology, ask them to get a left atrial appendage CLIP done within the week, and then you don't need a PIXABAN. So we'll start with the bottom one, consult cardiology. This has happened, at least at our center, a lot of times these patients come in with at least two or more instances of GI bleed. The next step is you send them over for a Watchman's Procedure. Even when they get a Watchman's Procedure or a left atrial appendage CLIP, the recommendation is to continue anticoagulation for up to six weeks before you would ask them to stop anticoagulation. Never resume, we have it written that the CHAD score is two, that means they have to be on some form of anticoagulation, so that's definitely not the right answer. Resume today versus resume in two weeks from stabilization. In this situation, it would be resume in two weeks from stabilization. Resuming today is more defined on if there are other risk factors of a pending DVT or a patient with an existing acute PE. We'll go into the details, so I won't dig too much into the details here, we already had a lecture earlier today on the anticoagulation part. So let's come to this part. How would you approach this in terms of discussion with the patient? It's a busy slide, but I'll try to make it easier. So we've identified that your patient, does the patient have more than one risk factors or the clinical indications that apply here? If you have a patient that has AFib with a CHAD score of less than one, that's an easy one, you can just say, well, you don't meet the risk factors to be on Eloquus or Apixaban or any form of anticoagulation, you can go ahead and stop it. If the anticoagulation was based on a temporary indication or like orthopedic patients that go home on some form of a therapeutic DVT prophylaxis, a therapeutic dose just for a DVT prophylaxis, those can be discontinued at that point in time. If the patient does have high risk of re-bleeding or death from re-bleeding, our patient does not want to re-initiate anticoagulation therapy. That is easy because there are high risk for death. You can just say, why don't we hold it for seven to 14 days and then reassess at that point and reconvene, either be an outpatient discussion with primary care, outpatient discussion with cardiologists, or outpatient discussion with a post-ICU clinic. If patient does not have any of those two things, you could suggest restarting anticoagulation. If it's Coumadin or it's DOAC, resuming within four to seven days is recommended under both those situations. Let's come to the lower GI bleed therapy. Injection of epinephrine. So endoscopic examination is recommended as the first approach for lower GI bleeds. That is the recommendation. That is the approach that is considered the right answer in the boards. I can tell you that clinically does not happen for a lower GI bleed. It's difficult to have a gastroenterologist be the first person to go in with a colonoscopy. Rightfully so. The concern is it's an inadequately prepared bowel. The visibility is going to be minimal. But if they do intervene, injection with epinephrine plus mechanical or thermal cautery is recommended. There's no RCTs that support this. Evidence of efficacy is largely based on observational studies and case series for diverticular bleeds. And diverticulosis, angio-dysplasias, and post-polypectomy bleeds are most likely to benefit from a colonoscopy. We already discussed in detail earlier during the Q&A session for the angiography and intervention. Case series are present that demonstrate high rate of success. Bowel ischemia is definitely a complication that can happen depending on the level of arteries that you're embolizing. A bowel does have a lot of collateral circulation built in it. So large bowel ischemia is a rare finding. Surgery is recommended if you can't identify the location. If the CP has failed and patient is requiring more and more transfusions. A crude cutoff for cardiology, for a surgical consultation in this situation is considered as six PRBCs transfused in less than 24 hours with an active sign of bleeding. Risk factors for mortality in the upper and lower GI bleeds is age more than 60, hemodynamic instability, severe comorbidities with end-stage renal disease, cirrhosis, or congestive heart failure, ongoing need for blood transfusion, emergency surgeries, onset of bleeding in the hospital, or an existing renal failure. Let's get to the last case for the GI bleed part. So 60-year-old patient presents to the ER with multiple episodes of bloody vomit since last night. Patient presented with lightheadedness, abdominal pain, and dizziness. Social history of daily 12-pack of beer, seldom hard liquor. On exam, patient's hypotensive, tachycardic, respiratory rate is 15, dry mucous membrane. Patient is very catchy-tick looking, hemoglobin is 10, rest of the labs are normal. So similar patient, as we had talked about in the first one, added the social history of alcohol and underlying cachexia. With those two histories, the common etiology at this point would still be a peptic ulcer disease because that is still the most common, but varicial bleed is a close second. So patient rightfully undergoes endoscopy and is noted to have esophageal varicial bleed. Scanning is performed and bleeding is controlled. Patient is now, well, it should be now, not not. So patient is now stable from a hemodynamic standpoint, awake and following commands. Which of the interventions will improve patient's long-term mortality? So what would we do? Antibiotics for SBP prevention, how many feel that that is what will improve mortality? So octetide infusion for 72 hours, that does not improve mortality, that does reduce the risk of re-bleeding. PPI infusion for three days is again something that reduces the risk of re-bleeding. And ventilator support with low PEEP to minimize venous pressures. That is a physiological approach. It does not assist with GI bleed in a practical way. So initial management, right, we would again start with a large bore IV access, intubation especially in patients who are developing hepatic encephalopathy. Ventilator should be placed on a low tidal volume and a low PEEP to minimize the post-sinusoidal hypertension. Cautious volume and restrictive transfusion. Coagulopathy, no current evidence for management with the help of Teg or, like, no longer treatment should be based on your PT, PTTs, you should go down the route of Teg or Rotin. If you're using massive transfusion, then follow the massive transfusion protocols. Aim for a platelet count to be closer to 50,000. And hepatic Doppler is recommended predominantly to rule out a portal vein thrombus. So what are the drugs that we can use? Octretide and terliprasin being the splanctin vasoconstrictors, efficacy has been shown in multiple trials. Survival benefit is shown with terliprasin. At least at our center, we don't have access to that at this point. Octretide, which is the drug that most of us use, does not have mortality benefit attached to it. Recommendation is duration of five days. And drug with endoscopic therapy is considered to be superior than either alone. Use of ceftriaxone, one gram daily for a total of five days, has been, sorry, two grams daily for a total of five days, has been associated with mortality and a re-bleeding risk. And then PPIs, there's question about the utility of PPIs in this situation. The benefit is post-banding ulcers, and then there's also benefit of blood present in the stomach causing irritation and gastritis. You can reduce that risk. Three days later, patient is readmitted to ICU from a medical floor with recurrent hematomasis. Patient is alert, INR is stable, blood pressure is in the 80s, pulse is steady, and patient is tachypneic. What is the next step? Refer for tips, call hepatology for transplant, consult palliative care and consider for hospice, or consult GI again for a re-EGD. I think this is something we all need to just take home, two EGDs before we call anyone else. So even in this situation, we would consult GI again for a re-EGD, and when that fails, you go for the tips. So consideration for tips in a variceal bleed, recurrence of upper GI bleed following EGD times two, or a known large variceal bleed which was noted in the first EGD and could not be banded in the first place. You directly approach for tips. Contraindication for tips is venous thrombosis in the feeding vessel, or presence of a downstream organ failure which would be pulmonary hypertension or RV dysfunction, or a high MELD score at the onset. Complications from tips would be bleeding, shunt malfunction with an increased preload can cause RV failure, and definitely can cause worsening encephalopathy. So early tips in GI bleed situations or in a patient with advanced end-stage liver disease is associated with improved outcome and reduced risk of free bleeding. Okay. Open for questions. You said high MELD score is a contraindication, right? Yes. Is there a threshold? So a lot of centers have a different threshold depending on the level of expertise that they have. I don't think there is a hard cutoff, and it also depends on whether they feel that patient could be a candidate down the road for possible transplant. So you have to look at it from a multi-desk approach.

gastrointestinal bleeding

hepatic failure

peptic ulcer disease

risk stratification

endoscopic therapy

variceal bleeds