Skip to content
Upper Gastrointestinal Bleed episode artwork

Mar 17, 2017 32 min

Upper Gastrointestinal Bleed

from Take Aurally Hosted by Take Aurally

Dr Jamie Thomas (@mcdreaamie) and Dr Charley Peal (@charleypeal) discuss Upper GI Bleed. You can find Take Aurally on both Facebook and Twitter where we will put links to any guidelines mentioned.

Episode details come from the podcast's RSS feed.

Embed this episode

Ready to play

Upper Gastrointestinal Bleed

0:00 32:13
of matches

Transcript Auto-generated, may contain errors

Thank you for downloading and welcome to Take Orally, the emergency medicine podcast recorded at Dream, Queen's Medical Centre at Nottingham. In this episode we'll be discussing upper GI bleed. As ever, all information is correct at the time of recording. Any and all guidelines mentioned are correct for Nottingham University Hospital's NHS Trust. Other trust guidelines may vary. All views and opinions are the speaker's own. Hello once again and welcome to Take Orally. My name is Jamie. I'm one of the teaching fellows in emergency medicine. You can find me on Twitter at McDreamy. And I'm Charlie. I'm one of the teaching fellows in medicine and I don't know how to use Twitter.

Excellent. You are on there. I am on Twitter. So this podcast comes from a request from one of our students who requested a podcast on upper GI bleeding. So you asked, we delivered. Thank you very much for coming back to the booth to help me with this.

potentially a serious problem upper GI bleeding something that could be quite dramatic and quite difficult to manage yeah absolutely there's a huge spectrum from the sort of what appears to be an upper GI bleed that then isn't or people that don't even need the endoscopy because it seems unlikely although they presented with a degree of sort of what appeared to be haematemesis all the way through to sort of um sort of peri-arrest patients who have a systolic of 40 who are sort of hemorrhaging blood from both ends basically so potentially life-threatening not always um but something not to let catch you out basically yeah i completely agree with that i mean in a and e everything from the care home resident who it looked a bit brown their vomit query coffee ground the query coffee ground vomit all the language should never be used yes absolutely as a gastroenterologist said you never use that phrase okay um so i mean the reason we have charlie here she's a fantastic medrage upper gi bleed medical problem yep lower gi bleed we'll deal with separately surgical absolutely so we are focusing on upper gi bleed today fresh we are bleeding not for me thank you not for uh not for charlie um so

How is our, I suppose we've already alluded to this, there is a spectrum, how is our patient with a child bleed going to present? How are we going to recognise it? What are the key bits of history we want? Yeah, so I think history is one of the most important things here. And not just the sort of history of events, but past medical history, which you may or may not know. So it's important to pick out the important factors and really ask important questions as well. So anyone with a history of liver disease or even just a sort of chronic alcohol history, just always be wary of cirrhosis, portal hypertension and therefore varices. Variceal bleed is potentially catastrophic.

Make sure you're getting that history. You'll have to ask it potentially in a sensitive way. But just be prepared for that as a possible cause of your GI bleed. The other thing that's really important to be asking about is their drug history. And it may be that you've got to do a bit of Sherlock Holmes investigative work to try and find out what this is. Nothing but Sherlock Holmes. Absolutely. But anything over the counter. People can get ibuprofen over the counter. So ask about NSAIDs. Ask about brand names as well. Neurofen. Some people might not know it to be called ibuprofen. So it's making sure you're asking the relevant questions in a way that people can understand. SSRIs are...

also culpable in causing sort of a gastritis or a peptic ulcer GI bleed. So really important to know about that. Steroids are another sort of culprit. And certainly in the elderly patients, you want to be using GI protection when using any of these medicines. There's also the absence of a PGI. Absolutely, yeah. And obviously the other thing that's really important to be establishing is the presence of anticoagulation or coagulopathy. So again, think about your liver patients, synthetic derangement of your liver function, lack of clotting factors, or warfarin, apixaban, dabigatrin, rivaroxaban, the ones that people are less familiar with.

so yeah drug history really important now you may not get that history though um you may get your patient presenting an extremist hypotension tachycardic look for your shock features um look for the signs of anemia as well you're a very pale patient postural symptoms dizziness lightheadedness collapses certainly with any sort of this history of liver disease just be aware you might not see that blood yet be prepared to do a pr look for melina It may not present as a haematemesis. And I think the San Francisco syncope rules mentions low haematocrit, to reduce haematocrit as part of their criteria, I think because of an occult GI bleed, I think is probably what they're thinking about when they're going to that. Yeah. So, yeah, I mean, presentation, as we said, can be incredibly varied. You may get something as straightforward as a patient with liver disease who presents with sort of frank haematemesis, but it may not be as obvious as that. Your other clues are going to be in your vitals and your examination at that point. So, as I say, consider those shock features, consider abdominal pain, consider that PR. The other thing, I mean, I know we're going to...

primarily be talking about your acute upper GI bleeds and management of that but you may get people presenting having had occult bleeds so sort of slow chronic bleeds which they may not even have noticed so this is just really highlight that you might not always see that blood loss you may get sort of that sort of malignant related occult blood loss in your faeces which presents with an acute anaemia and somebody may equally be sort of shocked and symptomatic from their anaemia so you mentioned portal hypertension yeah so obviously you can't see into a patient to see if they have varices just from the end of the bed what other features might there be from the end of the bed and an examination of portal hypertension so I think I mean we're going to go into the realms of chronic liver disease a little bit but it's to be familiar with those stigmata so someone with portal hypertension will more or less have some ascites so examine for ascites they may have evidence of varices elsewhere so you may see rectal varices they may have the capet medusa out there sort of varices on the sort of around the umbilicus on the abdominal wall you may see spider nevi as well jupiter's contracture palmarine erythema they may be in cephalopathic and certainly a liver patient who's decompensated will have some of these sort of stigmata of chronic disease and that evidence of decompensation

So we've briefly mentioned about liver disease and alcohol and drugs as a cause. What other causes are there of our patient presenting with I'm vomiting blood doctor? Yeah, so going back to the history, the etiology is going to identify what type of bleed it is really and almost highlight the urgency in your management. So your liver patients may present with a variceal bleed. They've got sort of these torturous dilated vessels in their esophagus, very prone to tearing, to bleeding and quite catastrophic bleed, even though it is a venous bleed.

They equally can have gastritis from alcohol consumption. The other thing to think about are your malary vice tears. So again, going back to your history, history, history, history. Was the blood on the first vomit or has it happened after several episodes of violent vomiting? And then they produced a bit of blood. Even in your liver patients, you can get any number of these. Just to think about other causes, I've mentioned malignancy. More likely to present as a chronic anemia, but you may get catastrophic bleed, certainly if there's been an erosion into a vessel.

seen that as well sort of ulcers that have eroded into vessels and you can get your malignant ulcers obviously um don't forget ibd ibd is a medical condition so your inflammatory bowel diseases which may present with bloody diarrhea so i think it's easy enough to differentiate within the history if people have got either known history of ibd or sort of inflammation of symptoms they've got frequency of going to the toilet and they're now passing bloody diarrhea you're going to think more ibd than worrying about upper gi bleed but just to be aware of blood coming from anywhere in the GI tract really yeah absolutely it's a long tract yeah and that we're gonna get graphic here that color and consistency of the blood will sort of direct you towards the source as well so if you've got bright red blood PR you're thinking more hemorrhoids and something lower GI yeah but if you're passing that black sticky really quite sort of sweet smelling melina I think we're all familiar with that well you say this I've mentioned this to students have you ever smelled melina and it is actually I say it's probably A&E humour it is a key learning point as a student. I think once smelt, it's a bit like C. diff, once smelt, never forgotten. And pseudomonas. Yeah, those are those three. But I can walk into a ward and go, yeah, somebody's got melina. It is a characteristic sickly sweet sort of smell. And it is so characteristic as well. Melina, upper GIV. Don't call your surgeons about melina.

but it is the vomiting of bloody substances which becomes a little hard to differentiate frank blood in your vomit yes you're going to think more variceal and certainly much higher but some blood will sit around in the stomach that's what gets past as melina and it is a gastric irritant as well so that's why you get multiple episodes of this sort of loose stool as well but it may get digested before you vomited again which is why we're kind of used to this dark coffee ground vomit in inverted commas but it's not a term really I would suggest using I really want to think, is this blood? Is this the chocolate cake they had for lunch? Yes, I was once called as an F1 urgently to assess a patient who vomited brown and it had been their, I just happened to look at their bowl that was in front of them and it was their pudding that was chocolate cake and I was like, I think they just vomited their chocolate cake. Again, Jamie Sherlock Holmes. Absolutely, I wish I was Sherlock Holmes, I love Benedict. Yes, and I also knew a girl at uni called Melina, which is an unfortunate name when you're at medical school learning these names and things like that. I think there's some drink in the international aisle at Sainsbury's called Melina as well. Oh dear, I would stay away from that.

okay um and uh yes i think history history history um from an a and e point of view as well i'd point out that um i've seen patients presenting with a malary vice tear um anyone with a prolonged period of vomiting also presented with chest pain any form of esophageal perforation I would be worried about yes it's very rare but I have seen a few cases and nearly been caught out oh it's just a bit of vomiting and actually it was something else always to keep that in the back of the mind and also yes as you said a blood is a potent gastric irritator I assessed a gentleman for haematemesis but it was actually the vomit he'd swallowed from epistaxis because he'd been swallowing swallowing swallowing it and his stomach was like I don't like this anymore out it came And you will get people who struggle to differentiate between whether or not they're coughing up bloody sputum or whether or not it's coughing that then causes them to vomit. So it does come down to history and really, really trying to nail down on those facts. So we're happy, well not happy, but we are content that our patient has a hematemesis. What's our immediate management plan going to be then for our patients who's an extremist? Yeah, so first of all, go down your parallel assessment route. You're going to ABCDE that patient.

initially obviously we would think of airway problems so is that airway patent but also think about their risk of aspiration someone who is shocked hypotensive drowsy vomiting their risk of aspiration so you want to be protecting that airway um in terms of respiratory again make sure that oxygen's on they're reducing their circulating volumes they probably do need some oxygen but i've seen people vomiting oxygen masks that's not pleasant either um shock is going to be the most important thing in your emergency resuscitation based on blood pressure and heart rate is important. Crystalloid is all well and good but you will dilute their circulating volume so whilst you might use it as a holding measure to bring the blood pressure up for a little bit get those blood products as soon as possible. As I always say to my students the oxygen carrying capacity of the crystalloid is still zero. Yes absolutely so you don't really want to make their Hb of 6 fall because the rest is saline. Along with that think about reversing what you can so even though obviously we can't do endoscopy and A&E yet but reverse their end calculation give them some VITK plasma, FFP, cryoprecipitate, octoplex, whatever it might be, have that conversation with haematology and decide what's best. Your liver patients may have platelets of 10, they may have a PT of 40, so reversing that coagulopathy is a really important part of that sort of management. There is some argument for tranexamic acid, I think jury's still out a little bit, but the idea of preventing the clot breakdown. So emergency resuscitation, really important. There is slightly different management based on whether or not you think it's variceal or non-variceal bleed, which we'll go on to.

But in terms of your initial diagnosis management of this patient, make sure you're getting your blood tests off as well. So you're going to do your full blood count, their HB. You may see that on a VBG as well. To your UNEs, your urea is really important here, and it's a good indicator of GI bleed. And as I say, coag screen is going to be a really important part as well. And that urea is all part of that protein meal, isn't it? Yes. You're digesting blood. It's quite a strong factor in predicting GI bleed, and it's part of our risk stratification as well, is if they've got disproportionately high urea compared to their creatinine, that idea they've got digestive blood. Absolutely.

Gastroenterologists always want to know that. Are they in keeping or is there a much range? Shout out also to the research team at DREAM who's a part of the HALTIT trial. I was hoping you were going to chime in with that one. I couldn't remember. But yes, a lot of work done about tronicamic acid in the context of geobleeds as well as major trauma. So what's going to push us towards thinking about an urgent endoscopy versus one that could potentially wait a bit longer? Say we're out of hours basically.

your resuscitation initially is always going to be sort of what it is you're going to be trying to resuscitate this patient with ideally blood products stabilize them what you want to do is reduce the risk of an intervention with this patient as much as possible and if you can stabilize them to do an in-hours endoscopy in the endoscopy department that's ideal but for an unstable patient who has ongoing bleeding and is hemodynamically unstable you might be thinking more about your emergency out-of-hours endoscopy and generally speaking that will probably happen with anesthetic support in theatres now so again It's an extremely high risk procedure because you're now anaesthetising, you're giving a general anaesthetic to this patient who is already unstable with an airway risk, low blood pressure, comorbid, whatever. So it's not without consideration that you go for that, but there is that spectrum again of how you can address this.

so this is a question of a few my students have asked me over the time whenever we discuss Jadley when to break out the Blakemore tube when do we well have you ever seen it in Amy? I haven't in this Amy no I've only ever seen well I've only briefly seen one once when I was in F1 but I was quickly ushered out of the room anyway so I didn't get the learning at the moment well if you put James and my experience together which is quite a bit now, neither of us have actually seen a staying star can make more cheap use. So it is infrequent, it's in the extreme situations, and it is for a virusy or bleed. So for patients who have ongoing catastrophic bleeding unstable, it may be a holding measure. That's all it is. It's not a treatment, it's a holding measure, and you need to be prepared to go forth with intervention after you've done it.

Generally speaking, this patient will be in intensive care as well. It essentially tamponades the vessels, so it goes all the way down into the stomach. You blow up that balloon and pull back, and it's an incredibly high pressure pullback, and it just tamponades those dilated, tortuous vessels in a way to prevent that bleeding. it's the equivalent of asking the patient to pinch their nose if they've got epistaxis basically while you ask where the MCs quickly rush down a lot more dramatic yes I think that's why I was rushed out of the room so we've mentioned varices so there is a distinction between a variceal and a non-variceal bleed and we might know from our suspicion that some patient may be high risk of varices they may have had an endoscopy before we know they have varices so what is the distinction between the management of a variceal bleed versus a non-variceal bleed So, I mean, both pre and during endoscopy. Yeah, so varices are something which are kept under surveillance. So people, obviously, cirrhosis has many etiologies. So it's not just to alcoholics. It may be non-alcoholic fatty liver disease or autoimmune liver disease, whatever it be. Many other things. And they may have surveillance endoscopy. So we may well know that they've got these torturous varices. And they may have had prophylactic treatment in the past. Now, when it comes to an acute presentation of the GI bleed,

Variceal bleeds, it's important to get prophylactic antibiotics in. So here I think we use tazosin. As long as they're not penicillin allergic. And the other thing that's different is we give them telepresin. And this basically reduces that porcelain venous pressure. So again, trying to reduce that bleeding. And again, it has proven reduction in mortality. The other interesting thing is for a variceal bleed, we have a lower HB requirement. So we aim for an HB around 70, over 70. Whereas for non-variceal bleed, we aim for an HB around 80 or above. So there is that distinction. We have more of a sort of...

permissive hypotension permissive anemia in these patients and that's really just to try and minimize the bleeding and obviously they tolerate a slightly lower hb as well needless to say i mean gastro would certainly be involved with your variceal bleeds probably all the way through in the sense that they're more likely to be liver patients and post endoscopy there's a possibility of a procedure called tips so the idea of a portosystemic shunt basically to reduce that portal pressure It's almost a little bit like management and the surveillance of an aortic aneurysm, the idea of blood pressure control, keeping an eye on it and things like that. Sorry, I interrupted you. Do you want to talk about endoscopy? Yes, let's go on to endoscopy. So the management for variceal bleed and endoscopy. So there is a difference between your variceal and non-variceal bleed and endoscopy. Obviously, this is something you would be asking a senior gastroenterology colleague to be performing. This is never something that would be done by a junior staff member. But it's interesting to be aware of your possible therapies endoscopy because whilst it's a diagnostic strategy, and we use it very often for diagnosis.

you actually can provide intervention and treatment during endoscopy for conditions such as this. So the things to be aware of is that you can band varices, so essentially putting an elastic band around the sort of tortuous part of the varics and again tamponading, blocking that sort of blood flow through that varics. as opposed to your non-variceal where you may have a bleeding vessel and an ulcer where we tend to use clipping so sort of clipping onto those bleeding vessels instead again there is a slight difference between using sclerotherapy which you'd use for varics so the idea of sclerosing that vessel rather than adrenaline injection around an ulcer base or a bleeding ulcer there is a difference there and the other thing which is a bit more available for ulcers than varices is going to be a diathermy so the use of hot stick basically to burn nose vessels And as I say, there is a spectrum of when people will go to endoscopy and there are guidelines for timing as well. And that is that for unstable patients after appropriate resuscitation, they should essentially be going immediately for an endoscopy. All other patients presenting with likely GI bleed is going to be within 24 hours, which basically means you can hold them until you can do a daylight procedure. And the role of PPIs then? Yeah, so we use PPI therapy purely in non-variceal bleeds. It doesn't have any role in variceal bleeds. And I mean, again, we tend to use a 72-hour infusion for...

non-variceal bleed here that's post endoscopy so you would load them with ameprazole IV beforehand and then they end up on a sort of three day treatment course of IV ameprazole i.e. they're in hospital for a further three days post endoscopy but that has no sort of role in a variceal bleed So it's all about the etiology, isn't it? Absolutely. And I mean, if you got in there and saw some gastritis, you may not need a full sort of 72 hours gastro, maybe a bit more sort of directed about what treatment they're suggesting. The other thing obviously to consider is that H. pylori can cause ulcers as well. It's not just sort of NSAIDs and SSRIs and aspirin and things. Do think about H. pylori as a cause and they will do tests and essentially that patient may need eradication therapy if those tests from the endoscopy prove positive.

It's a triple therapy. It's a high-dose PPI and two antibiotics. Shout out to the research team who gave themselves ulcers with H. pylori to prove that point. That's commitment to research, that is. Okay. So, I mean, here in our Amy, we...

We do have a system for low risk, upper GI bleeds, patients with upper GI bleeds, to go home to come back for an endoscopy. We book the endoscopy for the next day, but they go home to come back rather than staying in if they are at low risk. And this is based on the Blatchford score of risk stratification, which is something we'll talk about how we risk stratify our patients before and after endoscopy. Yeah, so the primary one you'll come across is going to be the Blatchford score, which should be used from first assessment, basically, of someone with... enough GI bleed and that takes into account both your hemoglobin and that is different for men and women but the other thing we talked about is the urea so the idea of disproportionately high urea and you get sort of scoring you score points based on the value of these the other sort of variable within that scoring system is your systolic blood pressure so again the idea that someone is shocked with an anemia and a high urea again you can see this building a picture of GI bleed There are some other non-specific markers. It does have a tachycardia in there, other sort of melina, syncope. But hepatic disease actually only carries two points, whereas you can get sort of six for a hemoglobin less than 100 in my hand. So you can see the weighting is a little bit sort of maybe not what you'd expect, but that's your first assessment. So that's the likelihood of them having an upper GI bleed and needing an endoscopy. So it scores more than six associated with a greater than 50% risk of needing an intervention. So less than six is possibly unlikely to make the other causes. But if you've got more than six, you're more than likely going to need an intervention. It's basically what it's saying.

again you need some clinical intuition and the recommendation certainly if you're thinking about going down with the endoscopy route you will be having that conversation with a gastroenterologist but keep those communications open it's not just a case of okay they're probably having GI bleed let's send them to D57 and leave them to it it's probably okay well their GBS score is 12 they've probably got GI bleed they've got a systolic blood pressure of 101 are we going to keep them here Do they need to go to HDU? Do they need to go straight to endoscopy? We're always going to have those conversations. And again, it goes back to that idea of knowing what you're asking when you're speaking to the relevant specialty, when you're trying to escalate patients. I should point out to anyone who's ever been to QMC that D57 is one of our acute medical wards. But yeah, I completely agree with that. I think I would say to my students, speak the language of the specialist you're speaking to. So if you can say to a gastroenterologist, their Glasgow Bachelor's score is whatever, you're speaking their language. And certainly our low-risk pathway here only works if you have a Glasgow Bachelor's score of zero the second you get anything.

Melina, just one, for example, you're staying in. And then only then if you've got somebody at home to look after you and you're not elderly. So it's a very safety netted pathway that we agree with gastro. OK, so that's the first assessment of risk stratification. What scoring systems do we use after endoscopy? So the other scoring system you may come across, and probably used more by the gastroenterologist, purely because we use it after the endoscopy procedure, is the Rockel score. And this is purely for an upper GI bleed. And again, it has variables that you can get scored on. This actually includes things like age. That's a non-modifiable variable.

It's very briefly talked about shock, so the idea of no shock, either tachycardia, or the highest scoring is hypotension. Then talk about comorbidity. Diagnosis, so actually what they've seen at endoscopy. So obviously if there's any diagnosis that's caused to GIP, you get one, unless it's of malignant origin, in which case you get two. Mallory Weister actually scores zero. And then they look for other stigmata of bleeding. And your post-OGD score basically just indicates your prognosis. So score less than...

Three, it gives a good prognosis and an early discharge, greater than eight, really has a high risk of death. So if you're old with sort of hypotension and comorbidities and actually you've had a proven sort of GI bleed from an ulcer, your prognosis is actually pretty bleak. It also looks at your re-bleeding risk as well as your mortality doesn't seem to, that's useful as well. So, you know, with your medical hat on, where do patients tend to go then after endoscopy? A lot of them, in fact, will go back to the wards. It does depend, obviously, on their prognosis, risk, that sort of thing.

There is an element of clinical subjectivity, if that doesn't sound too fluffy. But we do get a lot of them on HDU. So we may get either people who need to come to HDU pre-endoscopy for that sort of, let's keep them stable, close management, close observation, they go to endoscopy and they either then go to the ward or they may come back to us. So there is a degree of variability again in where they go and sort of what's appropriate. And again, you're always going to weigh up their comorbid status as to whether or not that endoscopy is even appropriate.

The only other thing I was going to mention briefly was in terms of blood products. Sorry, we're going back a little bit. In terms of your emergency resuscitation, again, trying to target transfusion levels in these patients. So that target transfusion of either HB 70 or 80 is how you're going to get those blood products or your sort of clotting factors, other blood products if you need them. I think we all know what a group in SAVE is, where you can send off a sample of blood and they will cross...

get the blood type and the antibodies but you don't get anything from it cross match still takes sort of half an hour to an hour I think really to get any blood to you possibly even a bit more yeah but there are other ways of getting blood quicker if you're needing to emergency resuscitate these patients and this isn't just available for A&E you can get these on the wards as well so either type specific blood if they've already had a group and say well you get that sort of blood type or even requesting a negative blood and again that can come straight to the ward if this is happening on the ward not just A&E There will be major haemorrhage protocols in all institutions. I think a lot of them need consultant authorisation or at least haematology discussion. So they are not to be taken lightly. But in the context of a sort of haemodynamically unstable cirrhotic with a variceal bleed who's requiring a same start in the late more tube, you may need those blood products and plasma and clotting factors urgently. So it is an option that's there should you need it.

And I think just to go back right to the beginning is not to underestimate and it's always better to cannulate the well patient with their nice veins before they deteriorate and you're struggling. So I think to students, you know, get that coag screen, get the group and save and make sure they've got good access. You will be thanked further down the line if the patient is going to extreme. Absolutely. Thank you very much, Charlie. Thank you, Jamie.

That was the Take Orally Upper GI Bleed podcast. You can find us on Facebook and Twitter where we'll put up links to any guidelines mentioned and you can contact us to suggest topics you'd like to see covered in any future episodes. For more information on education and research opportunities with emergency medicine and acute medicine and major trauma, you can find any way stream on both Facebook and Twitter.

Frequently asked questions

How long is this episode of Take Aurally?

This episode is 32 minutes long.

When was this Take Aurally episode published?

This episode was published on March 17, 2017.

Is there a transcript available for this episode?

Yes, a full transcript is available for this episode. You can read the complete transcript on the episode page.

Can I download this Take Aurally episode?

Yes. Use the download control on the episode player to save the publisher-provided media file.