1 00:00:10,110 --> 00:00:18,240 I and I've been honoured to be the Graham Plant professor, the Royal College of Radiologists and the BSI, 2 00:00:18,240 --> 00:00:25,770 ah, have award me this honour, but Graham Plant was one of the senior figures in the UK. 3 00:00:25,770 --> 00:00:34,050 He worked out of Basingstoke, so some of you may have know him have met him, but he was one of the father figures of interventional practise. 4 00:00:34,050 --> 00:00:38,100 You're probably more likely to know what he Fletcher, who was one of our own, 5 00:00:38,100 --> 00:00:46,890 who was an interventional radiologist out of Oxford, and he founded the BSR and International Practise in the UK. 6 00:00:46,890 --> 00:00:56,870 And just by way of background, I just want to share with you where we are because it was 55 years ago that interventional radiology really started, 7 00:00:56,870 --> 00:01:06,690 and this chap here challenged doctor who carried out the first intellectual procedures here is to carry out the angioplasty, 8 00:01:06,690 --> 00:01:15,180 which was the first in the world which resulted in the healing of this Lady al-Aziz ulcer and saved her leg. 9 00:01:15,180 --> 00:01:21,580 And it was a first. It really transformed the way we think about coming out into practise. 10 00:01:21,580 --> 00:01:32,160 This sadly didn't always get on with his baska surgical colleagues, and he was called the Mad Dog Charlie or Charlie Crazy Charlie. 11 00:01:32,160 --> 00:01:36,930 I think partly because some of his surgical colleagues didn't entirely agree. 12 00:01:36,930 --> 00:01:42,630 I don't understand why it's called mad, but he was then closely followed by a couple of other people, 13 00:01:42,630 --> 00:01:47,700 and he developed the first balloon technology that we use. 14 00:01:47,700 --> 00:01:55,620 And both he and Charles Dossa really talked about the evolution of air as a as a speciality, 15 00:01:55,620 --> 00:02:00,630 which can deliver a minimum of six treatments using small tubes and catheters, 16 00:02:00,630 --> 00:02:06,810 including what I'm going to talk about today dealing with patients who are bleeding, as well as opening up blood vessels, 17 00:02:06,810 --> 00:02:15,420 potentially blocking them off and produce these tubes on his kitchen table, firstly to get access to blood vessels. 18 00:02:15,420 --> 00:02:24,780 And he did for some of the first country work and cardiac angioplasties, and the first balloon, which was about eight or nine French to begin with. 19 00:02:24,780 --> 00:02:33,270 And then we had Holbert, who then developed the sort of more typical sausage shaped angioplasty balloons and the smaller catheters, 20 00:02:33,270 --> 00:02:39,690 and then eventually things developed from these big Teflon donators, stiff wires, 21 00:02:39,690 --> 00:02:43,770 wires with little balloons on to the sort of more modern technologies we have today, 22 00:02:43,770 --> 00:02:53,310 with the more refined as balloon stent stent grafts and getting ever, ever smaller so that we're now using small, full French systems. 23 00:02:53,310 --> 00:02:56,760 And that means we can get pretty much anywhere in the body, 24 00:02:56,760 --> 00:03:06,180 and it means we can carry out treatments in arterial and non arterial pathologies pretty much anywhere in the body. 25 00:03:06,180 --> 00:03:10,740 So GI bleeding. There are several guidelines out there. 26 00:03:10,740 --> 00:03:17,820 The most recent of the and support guidelines, and quite a lot of what I'm going to say is come from there. 27 00:03:17,820 --> 00:03:27,270 This is a difficult and complex area, and as you be aware, GI bleeding is often a serious and life threatening conditions. 28 00:03:27,270 --> 00:03:34,770 Clinically, we want to differentiate between upper and lower GI haemorrhage, although it's becoming apparent we need a common pathway really. 29 00:03:34,770 --> 00:03:39,930 For both metastases, Molina suggests bleeding from the upper GI tract. 30 00:03:39,930 --> 00:03:43,860 Lower GI bleeding presents with fresh rectal blood, as you know. 31 00:03:43,860 --> 00:03:50,850 But of course, profuse bleeding from anywhere in the tract can present with bright red blood and can be a confusion. 32 00:03:50,850 --> 00:03:54,210 If you're not thinking the protocols now, 33 00:03:54,210 --> 00:04:01,380 suggest that we should be thinking about marrying using some of these patients as outpatients so patients with upper GI bleeds. 34 00:04:01,380 --> 00:04:05,940 Think about early discharge if the young the stable at presentation. 35 00:04:05,940 --> 00:04:12,630 No obvious haemorrhage at the time for committees and then not a current inpatient and for large eye bleeds again. 36 00:04:12,630 --> 00:04:19,470 If there's no fresh rectal bleeding and there's an obvious anal rectal cause, then one can treat or manage. 37 00:04:19,470 --> 00:04:23,260 Then again, these could be considered for discharge. 38 00:04:23,260 --> 00:04:35,220 Now I'm going to mention the Oakland school because Oakland with various colleagues Richard Guy and Mike Murphy myself. 39 00:04:35,220 --> 00:04:40,710 We did a study with look, two and a half thousand patients, which we published in The Lancet recently, 40 00:04:40,710 --> 00:04:45,890 and we produced this score where you use clinical parameters such as the age, 41 00:04:45,890 --> 00:04:54,300 gender, systolic pressures and haemoglobin to risk stratify patients again thinking about early discharge for these patients. 42 00:04:54,300 --> 00:05:01,530 Using this score, you could strongly discriminate for those patients who was, it was safe or not safe to discharge. 43 00:05:01,530 --> 00:05:07,230 And this seems to be more sensitive and specific than the currently available risk stratification guidelines, 44 00:05:07,230 --> 00:05:13,250 many of which are modified from upper GI bleeding. Whereas this is much more specific to large eyes. 45 00:05:13,250 --> 00:05:21,090 So a score of less than eight predicts a 95 percent probability of safe discharge if the patients is 46 00:05:21,090 --> 00:05:26,600 older and they're unstable and there's active bleeding and a history of liver disease with viruses. 47 00:05:26,600 --> 00:05:31,610 And clearly these patients do need to be admitted and simply for low GI bleeds, 48 00:05:31,610 --> 00:05:35,900 they're more likely to have major bleeds with non-steroidal and aspirin. 49 00:05:35,900 --> 00:05:44,690 And if they have severe co-morbidities now, one of the guidelines at the RNC put is that these patients, regardless of upper lower GI bleed, 50 00:05:44,690 --> 00:05:53,240 should be admitted to a dedicated unit and there is a team leader with a multidisciplinary approach to managing these patients. 51 00:05:53,240 --> 00:05:59,060 It does take a big team of people to look after these patients, and most appropriately, 52 00:05:59,060 --> 00:06:04,970 the upper GI bleeding the rock hall store is pretty much the standard. 53 00:06:04,970 --> 00:06:09,560 Shocked patients require some degree of volume replacement, but that really should be. 54 00:06:09,560 --> 00:06:20,210 Chris Lloyd's red cells should be reserved until they've either got a 30 percent volume loss or drop the haemoglobin to less than seven platelets. 55 00:06:20,210 --> 00:06:26,780 Again, think about transfusing if it goes to below 50 millimetres per litre, 56 00:06:26,780 --> 00:06:30,860 proton pumps should not be started until they've had their upper endoscopy. 57 00:06:30,860 --> 00:06:35,330 This is, of course, for upper GI bleed. No anticoagulants can be really difficult. 58 00:06:35,330 --> 00:06:41,050 More and more of our patients are already on anticoagulation to various degrees. 59 00:06:41,050 --> 00:06:46,340 And again, the guidelines suggest we need to really think about why they're on those anticoagulants. 60 00:06:46,340 --> 00:06:53,270 And as appropriate, we may be worth reversing, particularly the iron on very high of two point five. 61 00:06:53,270 --> 00:07:01,040 But in some instances, we might be able to modify it, but not completely stop because of the potential consequences. 62 00:07:01,040 --> 00:07:09,470 We need to test for peptic ulcer disease and instigate treatments if that's positive for upper GI ulcers and bleeding. 63 00:07:09,470 --> 00:07:13,430 And then we need to remember that we used to test for H. 64 00:07:13,430 --> 00:07:19,460 Pylori while they're off PPIs, 65 00:07:19,460 --> 00:07:25,940 and then they should go into pieces after the initial treatment and up and GI 66 00:07:25,940 --> 00:07:32,780 endoscopy stop the non-steroidal and only reintroduce if it's absolutely necessary. 67 00:07:32,780 --> 00:07:40,430 And then with a, of course, start in the eye once you've started or completed treatment with four H. 68 00:07:40,430 --> 00:07:46,060 Pylori, if that's appropriate. If you think about the incidence of bleeds, 69 00:07:46,060 --> 00:07:49,960 it's much more common to have upper GI haemorrhage around one hundred and three 70 00:07:49,960 --> 00:07:55,420 per 100000 the UK now endoscopy is pretty much the mainstay of treatment. 71 00:07:55,420 --> 00:08:02,320 It'll identify and manage and treat these patients in over 90 percent of patients. 72 00:08:02,320 --> 00:08:10,300 There's around 10 percent where upper GI endoscopy fails it, despite having a second look and repeat endoscopy. 73 00:08:10,300 --> 00:08:16,660 And that's where radiology has got involved with more and more in the past. 74 00:08:16,660 --> 00:08:19,780 There's a short list of other causes for upper GI bleeding. 75 00:08:19,780 --> 00:08:27,520 The one thing I will highlight is that even in the presence of viruses, about a third of patients will still have an arterial cause. 76 00:08:27,520 --> 00:08:33,760 So we mustn't forget that to look for an actual issue in these patients as well. 77 00:08:33,760 --> 00:08:42,250 Low GI bleeding As I say, it's about a quarter of upper GI bleeds and increases significantly as you get older. 78 00:08:42,250 --> 00:08:48,160 Up to 200 fold by the time you're 80 diverticular, disease is the most common. 79 00:08:48,160 --> 00:08:56,170 Cause is closely followed in the literature by Andrew Dysplasia. Now here's again another short list of other causes. 80 00:08:56,170 --> 00:08:59,840 One of the areas that we often get caught out by is bleeding piles. 81 00:08:59,840 --> 00:09:07,270 You might think it's not very frequent, but it can be up to 15 percent of studies showing bleeding paths. 82 00:09:07,270 --> 00:09:13,420 And this is one of our patients. And a simple anal rectal examination would have resolved it. 83 00:09:13,420 --> 00:09:21,430 This patient went all the way to angiography and to determine that this was coming from bleeding piles. 84 00:09:21,430 --> 00:09:28,600 So this is where it's important just to have a quick look at the tail end before we start instigating a whole 85 00:09:28,600 --> 00:09:34,840 host of treatments is much easier to get to it from below than to try and get a case all the way round. 86 00:09:34,840 --> 00:09:42,220 The and tear voluntary and separated branches about five to 30 percent will have an 87 00:09:42,220 --> 00:09:48,460 underlying cancer and three to five percent of bleeds are actually from the small bowel. 88 00:09:48,460 --> 00:09:56,410 Most of these will be under dysplasia not easy to get to with endoscopy, and this is where other techniques emulsify. 89 00:09:56,410 --> 00:10:03,190 C.T. and more recently, capsule endoscopy have come into their for their really useful for picking up 90 00:10:03,190 --> 00:10:07,870 these slow intermittent bleed or where they're in difficult to get places. 91 00:10:07,870 --> 00:10:13,720 The good news for low GI bleeding is the vast majority will stop on their own without any specific treatment. 92 00:10:13,720 --> 00:10:18,970 A third will bleed once, and half of these will bleed for a third time. 93 00:10:18,970 --> 00:10:26,230 The published mortality is actually quite low for around three point six percent, but it's significantly higher, 94 00:10:26,230 --> 00:10:32,770 as I'll show you a bit later for patients who are already inpatients now in the past. 95 00:10:32,770 --> 00:10:39,600 The role of radiology was to help localise the bleeds before aggressive treatments such as surgical 96 00:10:39,600 --> 00:10:45,070 U.S. could be undertaken by nuclear imaging was really an excellent technique is still used lies, 97 00:10:45,070 --> 00:10:52,790 but not so frequently. It's very sensitive compared to conventional and angiography and C.T. and it pick up 98 00:10:52,790 --> 00:10:59,980 bleed rates as low as point zero for two point five miles per minute on animal studies. 99 00:10:59,980 --> 00:11:04,540 They're very sensitive and specific, but you know, 100 00:11:04,540 --> 00:11:13,150 we can get false areas of positivity because of leaching of the agent and moving of the agents through the bowel. 101 00:11:13,150 --> 00:11:21,940 During the course procedure, but can help you in terms of localising prior to angiography and or potentially surgery. 102 00:11:21,940 --> 00:11:26,320 Scans can be done very rapidly and quickly. Only lasts about four hours. 103 00:11:26,320 --> 00:11:34,420 Labelled red cell scans can be scanned for up to 24 hours until the red cells are eaten up by the spleen, 104 00:11:34,420 --> 00:11:38,770 and this can allow you to pick up smaller and intermittent bleed. 105 00:11:38,770 --> 00:11:44,140 The main issue with nuclear scanning is that it's really pretty much nine to five Monday to Friday, 106 00:11:44,140 --> 00:11:53,080 and the agents may not even be available during the daytime because they have to be in and ready to be utilised or produced. 107 00:11:53,080 --> 00:12:00,130 If you're looking at rates, those scans now angiography many years was said to be the gold standard for picking up bleeds, 108 00:12:00,130 --> 00:12:05,680 and you can pick up quite slow bleeds of about 0.5 to one minute per minute. 109 00:12:05,680 --> 00:12:11,620 The sensitivity and specificity for upper and lower GI bleeding is highly variable, 110 00:12:11,620 --> 00:12:17,750 and that's not surprising because the patient has to be bleeding when you actually do the examination. 111 00:12:17,750 --> 00:12:25,090 Unlike nuclear scans, we have a large window here. The patient, when you do, the contrast injection has to be bleeding at the time. 112 00:12:25,090 --> 00:12:30,370 So the fact you don't see it doesn't mean they're not bleeding just means you've not done the time. 113 00:12:30,370 --> 00:12:35,860 And that's why we always ask that the patient is unstable and actively bleeding when we see them. 114 00:12:35,860 --> 00:12:42,270 And then there's other issues. There's bound peristalsis and overlying loops, which can make it very. 115 00:12:42,270 --> 00:12:49,320 You could see small bleeds as he may need to repeat it several times to actually find the bleed site. 116 00:12:49,320 --> 00:12:52,810 You need people who are skilled and geographers and have a good knowledge of the anatomy. 117 00:12:52,810 --> 00:13:00,030 There's multiple variants so that you can not forget about potential collateral sites of bleeding, 118 00:13:00,030 --> 00:13:06,810 which can occur if you can't see it on the first runs. There are techniques to try and improve on geography. 119 00:13:06,810 --> 00:13:11,220 Stopping peristalsis with glucagon and bus CPAN can be quite helpful. 120 00:13:11,220 --> 00:13:21,480 Provocative techniques so giving heparin thrombolytic agents, or vaisseaux dilate, is to try and reduce spasm, 121 00:13:21,480 --> 00:13:27,240 which often occurs at the time of bleeding, which is how often the vessels stop the blood loss. 122 00:13:27,240 --> 00:13:33,810 And then you allow yourself the opportunity to find that bleed and treat it. 123 00:13:33,810 --> 00:13:44,040 Using longer injections and CO2 in the last six or seven years has been advocated, and we've used a few times where because the CO2 as it comes out, 124 00:13:44,040 --> 00:13:51,360 as a gas expands, it gives you a much bigger area and makes it much more obvious where the bleed is coming from. 125 00:13:51,360 --> 00:13:58,340 Of course, if I've had endoscopy, then placing clips of some sort of marker could be quite helpful for targeting the angiogram and of course, 126 00:13:58,340 --> 00:14:05,050 good standard techniques using multiple oblique views to try and throw off overlying vessels. 127 00:14:05,050 --> 00:14:11,190 Now I've put anaesthetic support at the bottom. This is something we hardly ever have and I think is really crucial. 128 00:14:11,190 --> 00:14:19,350 These are by the time they get to, it's very sick patients and one wouldn't really contemplate trying to make these patients 129 00:14:19,350 --> 00:14:25,290 when they are shocked and bleeding on the table and you're trying to carry and embolisation, 130 00:14:25,290 --> 00:14:30,780 it's really important to have some form of medical and ideally assessing support. 131 00:14:30,780 --> 00:14:32,040 They don't have to be anaesthetised, 132 00:14:32,040 --> 00:14:40,620 but somebody needs to be keeping them safe and managing their airway if necessary when we're trying to do these patients. 133 00:14:40,620 --> 00:14:46,680 This is not one of my patients, and I think this caused a lot of excitement when it did happen. 134 00:14:46,680 --> 00:14:52,230 So this was a fistula, as you might have seen from an aneurysm into the rectum, 135 00:14:52,230 --> 00:14:58,650 giving the patient lots of fluids and probably also brown fluid to the angiography as well. 136 00:14:58,650 --> 00:15:07,120 When he saw these pictures, and fortunately, he was able to analyse the entire country which saved this patient's life. 137 00:15:07,120 --> 00:15:18,800 Okay. Now, now, with the advent of Multi City, this is really transformed the way we image patients with GI bleeding city, 138 00:15:18,800 --> 00:15:24,920 unlike all the other techniques, is almost universally available in all hospitals now in the UK. 139 00:15:24,920 --> 00:15:32,330 It is very sensitive. Animal models will pick up bleed rates lower than that for standard angiography, around 0.3 mils per minute. 140 00:15:32,330 --> 00:15:38,840 So that's actually very good. With a modern bullseye scanners, we've got scan times of one to two seconds. 141 00:15:38,840 --> 00:15:44,360 So it reduces a lot of the artefacts that we get from movement and peristalsis and respiration. 142 00:15:44,360 --> 00:15:53,120 So that's really good. And even if you don't see active extrapolation, you might see the underlying pathology causing that place. 143 00:15:53,120 --> 00:15:56,840 So it's really improved the way we look at things. 144 00:15:56,840 --> 00:16:00,440 So here's an example of just a patient who's got an upper GI bleed. 145 00:16:00,440 --> 00:16:07,460 You can see the anatomy reconstructed the GDA show the branch vessel causing the bleed and 146 00:16:07,460 --> 00:16:14,480 then confirmed at the time of angiography and then selectively analysed with micro coils. 147 00:16:14,480 --> 00:16:18,650 Another patient here you can see this is actually bleeding from the sigmoid branch. 148 00:16:18,650 --> 00:16:26,210 You can see contrast translating for this branch, the Coronil oblique reconstruction showing the precise vessel. 149 00:16:26,210 --> 00:16:33,530 And then we can actually target this selectively and analyse this with multiple coils. 150 00:16:33,530 --> 00:16:39,080 This is a different cause. There's a tumour involving the superior mesenteric artery. 151 00:16:39,080 --> 00:16:44,330 You could see false aneurysms as a consequence. This is more harder to treat because you really can't. 152 00:16:44,330 --> 00:16:49,280 It blows the whole of the SMA, but you might be able to target small branches. 153 00:16:49,280 --> 00:16:57,830 This is another patient who's got an ambulance into the SMI. And again, it's an alternative cause to bleeds. 154 00:16:57,830 --> 00:17:00,770 This was this patient had upper GI bleeding. 155 00:17:00,770 --> 00:17:10,640 It could see bleeding blood in the Judean on endoscopy, but they couldn't see where it was coming from within the city. 156 00:17:10,640 --> 00:17:17,030 And then you can see here this is blush of contrast, sitting in the gallbladder with which also has some gas in it. 157 00:17:17,030 --> 00:17:22,320 And then inside the genome, you've got this lovely gallstone, which is laminated. 158 00:17:22,320 --> 00:17:31,850 And this patient had a coloured fistula causing bleeding sitting in the genome of this point and then angiography, 159 00:17:31,850 --> 00:17:36,140 we could find the cystic artery in the devastation. Here it is. 160 00:17:36,140 --> 00:17:41,630 Get the catheter into it and then embolism with multiple coils into this patient. 161 00:17:41,630 --> 00:17:48,230 Another patient here with this is that part of the transit colon. You can see this abnormal enhancement of the city. 162 00:17:48,230 --> 00:17:56,180 This is what it looks like on colonoscopy. And geographically, this is an early draining vein. 163 00:17:56,180 --> 00:18:01,220 As a patient with Andrew Dysplasia, another patient with reconstruction with under dysplasia, 164 00:18:01,220 --> 00:18:06,950 again showing a nice draining vein early with multiple abnormal vessels. 165 00:18:06,950 --> 00:18:12,350 And again, a list of 100 dysplasia type vessels in the colon. 166 00:18:12,350 --> 00:18:19,950 As I alluded to earlier, whichever type of imaging you use, the patient needs to be actively bleeding at the time. 167 00:18:19,950 --> 00:18:24,350 So one of the markers is that they've had a significant volume of transfusion 168 00:18:24,350 --> 00:18:31,370 over the previous 24 hours and that they should be hemodynamically unstable. 169 00:18:31,370 --> 00:18:39,060 That means a shock index of greater than one, so a systolic of less than 100 hundred and tachycardia greater than 100. 170 00:18:39,060 --> 00:18:47,000 Why? Because if they if the patient is stable, that almost invariably means they've stopped bleeding. 171 00:18:47,000 --> 00:18:54,410 It's not a hundred percent, but usually no matter how quickly you put the blood in in the fluids, they will still remain hypertensive. 172 00:18:54,410 --> 00:19:03,050 If the active bleeding, once you stabilise them and we often get this, the patients are actually controlled. 173 00:19:03,050 --> 00:19:11,060 We've got fluid again, but they actually are stable and they're stabilised in either i2 or head up on the ward, 174 00:19:11,060 --> 00:19:18,200 which means by the time we get down, the bleeding has stopped. So I really would prefer them down control with somebody with them, 175 00:19:18,200 --> 00:19:23,600 but unstable so we can actually get on to see what the bleed is and then we can treat them. 176 00:19:23,600 --> 00:19:26,300 Based on our experience with C.T., 177 00:19:26,300 --> 00:19:36,260 we published this protocol in one international radiological journals where we put pretty much at the forefront of investigations. 178 00:19:36,260 --> 00:19:40,820 It doesn't stop doing other imaging and other investigations, 179 00:19:40,820 --> 00:19:49,160 but it means that we can get definitive information much more quickly in a significant number of patients. 180 00:19:49,160 --> 00:19:53,690 We also used CT to risk stratify, and we published this paper, 181 00:19:53,690 --> 00:20:03,230 which showed that you could risk stratify patients and those with GI bleeds if they had a negative CT were unlikely to bleed again, 182 00:20:03,230 --> 00:20:08,730 as opposed to upper GI bleed invariably and needed something further. Doing so means we. 183 00:20:08,730 --> 00:20:13,470 To be much more aggressive with Upper GI Bleed, as I alluded to earlier, 184 00:20:13,470 --> 00:20:19,650 so we're much more involved now in the therapeutic side in treating patients and trans arterial embolisation 185 00:20:19,650 --> 00:20:25,800 is in many senses become the standard of care went up a gentle squeeze felt and almost first line. 186 00:20:25,800 --> 00:20:35,610 When we're talking about lower GI bleeds and for the reasons we talked about, we can see the bleeds, extrapolation, the anatomy and direct treatment. 187 00:20:35,610 --> 00:20:41,970 And the reason we are able to do this now is because we've got really excellent kids. 188 00:20:41,970 --> 00:20:48,900 We've got catheters which are three and four or five French micro catheters, which are two in three French, 189 00:20:48,900 --> 00:20:53,550 which means we can get out to very small branches and deal with those microwaves. 190 00:20:53,550 --> 00:21:02,130 And the further out you can get into those branch vessels, the less likely to cause ischaemia and give an effective embolisation. 191 00:21:02,130 --> 00:21:13,680 We've got a variety of embolic agents available. And what you use depends on many factors, including what the anatomy is like, what the findings are, 192 00:21:13,680 --> 00:21:18,840 where you can get your catheter to, and often, of course, what the operator is used to. 193 00:21:18,840 --> 00:21:20,400 Some people like using glue. 194 00:21:20,400 --> 00:21:27,870 Some find it very difficult and actually is dangerous if you don't know what you're doing, so choose what you're really familiar with. 195 00:21:27,870 --> 00:21:33,050 And, of course, the vessel diameters most commonly, people use coils and gel foam. 196 00:21:33,050 --> 00:21:41,040 Now, if you only use coils, there's a significant increased incidence of bleeding or bleeding, where if you combine the agents, 197 00:21:41,040 --> 00:21:50,820 it seems to be a lower risk of bleed glue, as I said, and onyx are increasing use, but you need to be familiar with them and stent grafts. 198 00:21:50,820 --> 00:21:58,260 So things like hepatic aneurysms and splenic aneurysms where we've had bleeding, we've often put in stent grafts, 199 00:21:58,260 --> 00:22:04,780 preserves the vessel, which makes it safer, but also, of course, is effective at stopping the bleeding. 200 00:22:04,780 --> 00:22:14,700 So these can be really, really useful in different positions. This is a paper we published a few years ago about personalisation for upper GI bleeds. 201 00:22:14,700 --> 00:22:19,770 This is just highlighting that even if you don't see active externalisation at the time, 202 00:22:19,770 --> 00:22:23,130 you can count on embolisation, which would control the bleeding now. 203 00:22:23,130 --> 00:22:30,310 The important thing is empiric is a bit of a misnomer because we often have an idea of the territory of the bleed. 204 00:22:30,310 --> 00:22:31,530 So, you know, 205 00:22:31,530 --> 00:22:39,240 I've had an endoscopy or they've had C.T. or something to tell you that actually there's blood in the third part of the genome of the second part. 206 00:22:39,240 --> 00:22:43,260 So you can have a pretty good idea. And other studies have. 207 00:22:43,260 --> 00:22:48,940 Similarly, I can confirm subsequently that actually this is this is very effective. 208 00:22:48,940 --> 00:22:54,030 So just trying to analyse what you think is likely to be works, 209 00:22:54,030 --> 00:23:05,220 this is less so for lower GI bleeding because it's much more critical to get the right vessel and and often it's impossible to predict. 210 00:23:05,220 --> 00:23:11,240 So there'll be several papers who've looked at the outcomes of embolisation relatively small numbers. 211 00:23:11,240 --> 00:23:20,790 The clinical success is perhaps being slightly more disappointing for upper GI bleeding compared to low GI bleeds, but it is very effective. 212 00:23:20,790 --> 00:23:25,710 They seem to be better for low GI bleeds for diabetics. These compared to anjou dysplasia. 213 00:23:25,710 --> 00:23:35,940 The main reasons are you can't get selectively enough into that vessel, and inexperienced and torturous is also another major factor. 214 00:23:35,940 --> 00:23:42,330 Overall, reblogged rates vary widely depending on whether you talk about upper and lower GI bleeds and the 215 00:23:42,330 --> 00:23:49,800 cause of bleeding between nine and 47 percent for low GI bleeds is less generally 10 to 15 percent. 216 00:23:49,800 --> 00:23:55,050 30 day mortality again varies hugely depending on the site and cause of bleeding. 217 00:23:55,050 --> 00:24:02,910 Predictors of three bleeds are if the coagulopathy they take a long time to get to angiography, 218 00:24:02,910 --> 00:24:09,570 previous surgery and trauma, cancers and of course, using cause as the only agent, a multi-organ failure. 219 00:24:09,570 --> 00:24:15,900 Some of the studies have used just vasopressin to try and less invasive or treat these patients, 220 00:24:15,900 --> 00:24:21,870 and it can be very effective in initially controlling bleeding because you're causing invasive spasm into the vessel. 221 00:24:21,870 --> 00:24:25,890 But unfortunately, as soon as you turn it off, there's a high rate bleed rate in these patients. 222 00:24:25,890 --> 00:24:30,540 So it can be if you've got no other options, an effective treatment, 223 00:24:30,540 --> 00:24:38,280 but you need to be prepared to go back into it again or do some different treatment. 224 00:24:38,280 --> 00:24:44,040 In terms of complications, a proper GI bleeds embolisation is considered very safe. 225 00:24:44,040 --> 00:24:49,020 There are some risk factors for a significant CV if they've had previous surgery, 226 00:24:49,020 --> 00:24:56,160 radiotherapy or atherosclerosis is marked or using liquid agents against glue seems to be a culprit here, 227 00:24:56,160 --> 00:25:00,870 and small particles causing much more distal embolisation. 228 00:25:00,870 --> 00:25:08,530 One small study reported stricture rates of around 16 percent for lower GI bleeds. 229 00:25:08,530 --> 00:25:18,010 You also get a ischaemia in around a quarter of patients with franking function varying widely between zero and 20 percent if they do. 230 00:25:18,010 --> 00:25:25,540 In fact, there isn't really much else to do than to get Chris and colleagues to come in and recite the colon. 231 00:25:25,540 --> 00:25:30,640 Here's a patient who's got active extermination from the sigmoid. 232 00:25:30,640 --> 00:25:40,330 Again, this is a dangerous place. It was obliged, and unfortunately, we've got an infarction of the bowel with ischaemic change. 233 00:25:40,330 --> 00:25:46,390 This is what it looked like resection compared to surgery. 234 00:25:46,390 --> 00:25:53,800 Some studies suggest similar efficacy rate, bleed mortality and morbidity rates. 235 00:25:53,800 --> 00:25:59,680 Other studies have suggested that actually the rates are significantly higher compared to surgery. 236 00:25:59,680 --> 00:26:06,760 Again, the similar risk factors that we've mentioned about Coyle's or gel foam as the only agents. 237 00:26:06,760 --> 00:26:13,180 This was the study I alluded to earlier with colleagues in Oxford looking at low GI bleeding. 238 00:26:13,180 --> 00:26:20,950 And it confirms many of the things we've already talked about that diabetic disease is the commonest cause, 239 00:26:20,950 --> 00:26:23,920 but benign conditions are also quite frequent. 240 00:26:23,920 --> 00:26:34,000 Almost 17 percent shock was pretty uncommon, but a lot of patients still got transfused, which was perhaps an appropriate flexible. 241 00:26:34,000 --> 00:26:42,370 Sigmoidoscopy was the most common investigation, but only around two percent actually went on to have a therapeutic. 242 00:26:42,370 --> 00:26:52,960 And he was Stacey's, which was a shame. Having found the bleed embolisation which surprised me, was also it was very rare. 243 00:26:52,960 --> 00:26:56,500 Surgery didn't surprise me so much, but embolisation did. 244 00:26:56,500 --> 00:27:08,740 Around just under one percent actually had embolisation, which I thought, I suppose you always kind of think about your own local practise, 245 00:27:08,740 --> 00:27:19,420 but clearly nationally, embolisation is not very widely available, and that's borne out by recent surveys. 246 00:27:19,420 --> 00:27:26,740 And certainly, perhaps not people are not so aware of it, either. 247 00:27:26,740 --> 00:27:35,170 About just under 40 percent had re bleeds during the admission, and just under five were readmitted with bleeding in 28 days. 248 00:27:35,170 --> 00:27:42,520 The in-hospital mortality that we talked about was again very similar to previous literature, around 3.4 percent. 249 00:27:42,520 --> 00:27:53,350 And as I mentioned earlier, much higher in patients who are already inpatients just under 18 percent rates of around ten point one percent. 250 00:27:53,350 --> 00:27:55,930 Don't worry about this complex flowchart. 251 00:27:55,930 --> 00:28:06,490 He's just putting together all the things that we've talked about, but it just highlights that intervention and surgery still come very much later on. 252 00:28:06,490 --> 00:28:09,730 So just to finish off and conclude, 253 00:28:09,730 --> 00:28:20,320 I was just going to say that the most recent guidelines suggest and recommend that regardless of whether the upper and lower GI bleeds, 254 00:28:20,320 --> 00:28:26,680 they should be admitted to a dedicated unit which is used to dealing with these second shot patients. 255 00:28:26,680 --> 00:28:32,500 There should be a named lead clinician who is responsible for that treatment of that patient, 256 00:28:32,500 --> 00:28:36,550 whether it's over that 24 hour period or throughout the entire episode that is specified. 257 00:28:36,550 --> 00:28:41,680 But it needs to be somebody who knows that patient is going to manage them. 258 00:28:41,680 --> 00:28:45,100 They should undergo some formal risk stratification for upper GI bleeds. 259 00:28:45,100 --> 00:28:51,100 If this standard is Rockwall, there are those, of course, blanch for the straight, but Rocco is most commonly utilised. 260 00:28:51,100 --> 00:28:59,890 And of course, for larger bleeds, we would advocate using the Oakland school to try and risk stratify patients who 261 00:28:59,890 --> 00:29:06,050 could be discharged early and those that really need to be managed as inpatients. 262 00:29:06,050 --> 00:29:10,060 Clearly, patients will need a degree of resuscitation, 263 00:29:10,060 --> 00:29:16,030 but that needs to be balanced up by not over transfusing and initially they should be with Chris. 264 00:29:16,030 --> 00:29:22,810 Lloyd's endoscopy is very much the mainstay Robert GI bleeding and some 265 00:29:22,810 --> 00:29:29,140 assessment of the rectal tract from below colonoscopy is an excellent technique, 266 00:29:29,140 --> 00:29:33,910 but can be very difficult in the acute setting with his active bleeding. 267 00:29:33,910 --> 00:29:42,910 So often it may be delayed till the next day when there's a bit more visibility. 268 00:29:42,910 --> 00:29:49,930 Merseyside city now has become very much on the forefront of investigation of these patients because this can be done 24 hours a day, 269 00:29:49,930 --> 00:29:54,370 seven days a week and can be extremely useful. 270 00:29:54,370 --> 00:30:00,220 But regardless of whatever imaging we use, we need the patient to be bleeding at the time. 271 00:30:00,220 --> 00:30:06,400 So waiting to they stabilised to get them down to the department isn't always the ideal. 272 00:30:06,400 --> 00:30:11,860 Nuclear studies are pretty much set. And division in terms of trying to find slow, 273 00:30:11,860 --> 00:30:19,030 difficult intermittent bleeds rather than the massive active housing that we're trying to see capsule 274 00:30:19,030 --> 00:30:26,290 endoscopy again is a useful tool for appropriate patients more the sort of chronic slow bleeds, 275 00:30:26,290 --> 00:30:29,890 particularly in the mid gut and geography. 276 00:30:29,890 --> 00:30:34,780 As I say, it is even the second or third line in terms of trying to pick up. 277 00:30:34,780 --> 00:30:44,920 Please, we do do them if we have a negative message, but usually we need to have a good reason because we're very unlikely to see anything. 278 00:30:44,920 --> 00:30:50,740 And then, of course, some patients will need intellectual treatment and much less frequently. 279 00:30:50,740 --> 00:31:02,800 Now they're going off to surgery. I'm going to leave you with a picture of my Charlie, who started interventional radiology practise 55 years ago. 280 00:31:02,800 --> 00:31:12,085 Thank you very much for your attention.