This chapter deals with the investigations and resuscitation of massive gastrointestinal bleeding, where by "gastrointestinal" one refers to "things that go wrong in the lumen". Bleeding that is intraperioneal or retroperitoneal is obviously going to be a very different thing and will require a separate chapter.
In terms of relevance for the CICM fellowship, massive GI bleeding is probably not a high-yield subject matter (even though it forms a major part of our daily workload). Historically, it has appeared in several SAQs:
The best references for this would probably have to be large society guidelines, rather than review articles.
Question 1 from the first paper of 2017 asked for four causes of massive GI haemorrhage. There is a long possible list of such causes. The list offered here was generated mainly using Oh's Manual, but contains conditions which are not listed in the canonic Chapter 42 (pp. 487, "Acute gastrointestinal bleeding" by Joseph JY Sung).
Bleeding of non-gastrointestinal origin
Endoscopy is best. Barium investigations are useless. Why?
Generally, postero-inferior duodenal wall ulcers and high lesser curve of stomach ulcers tend to rebleed most vigorously, owing to the presence of large arteries nearby.
In order to catch an upper GI bleeder on an angiogram, the rate of bleeding should be greater than 0.5ml/min.
Identifying the high risk patient, whom to cart away to the endoscopy suite:
If endoscopy fails, or there is evidence of perforation, the next move is surgery.
Nobody seems to agree precisely when to make this call. In general, if you are taking a patient to theatre for an emergency salvage surgery after an upper GI bleed, their mortality will be around 15-20%.
All these methods seem roughly equivalent in their effectiveness.
It is generally thought that the use of all three methods on the same lesion gives the greatest reduction in rebleeding risk.
Question 1 from the first paper of 2017 asked specifically for "clinical indicators for risk of re-bleeding", which is a fancy way of asking for a list of risk factors. A good paper by Augustine et al (2010) actually presents an entire table of studies which investigated the various prognostic indicators which predict rebleeding in acute variceal haemorrhage. Re-bleeding in the case of many of these was rolled together into "5-day failure", a sort of composite endpoint together with mortality, which it obviously has some sort of effect upon. This was remixed into the following list of predictive features:
Tranexamic acid plays no role - there is good evidence that it is ineffective.
In general, drugs can prevent rebleeding post endoscopy, but they usually play little role in massive haematemesis, and on their own are ineffective.
For the purposes of having a firm definition, anything below the ligament of Treitz is taken as the "lower" tract.
Unstable patients: BSG recommend you stratify according to "stable" or "unstable", on the basis of whether or not the patient is shocked, which means most ICU patients will be definition be in the "unstable" category. Weirdly, the BSG use the shock index to determine whether the patient is shock, which is a strange thing to do as nobody else really uses this measure for anything. In case the reader is wondering about this metric (having themselves never encountered or used it before), it is a value is calculated by dividing the heart rate by the systolic blood pressure, where an SI of >1.0 suggests haemodynamic instability. It is a metric which could theoretically give a reassuring number of 0.7 for a patient with a heart rate of 68 and an SBP of 48 mmHg. The ESGE guidelines statement instead suggests that "no single risk score should be used in isolation to predict adverse outcomes" and recommends a gestalt clinical assessment to determine risk.
CT angiography is the recommended investigation for the haemodynamically unstable GI bleeder, from both the BSG and the ESGE. It was not even mentioned in the stem of Question 16 from the second paper of 2023, presumably because there was no point in asking about this uncontroversial first step which everyone is in agreement on. The guidelines quote the sensitivity and specificity as 79 %–95 % and 95 %–100 %. The most important role for this is to identify the site of the bleeding so as to guide the decisions around endoscopy and angiography.
First line colonoscopy is useless. You will never find the bleeder in that mess of blood and stool. Both the BSG and the ESGE seem to believe that unprepped colonoscopy in acute gastrointestinal bleeding is a waste of time. Both sources of guidelines agree that there is a role for colonoscopy at some stage during that admission, but not in the initial acute stages, unless the bleeder is very clearly localised by CT and is well-placed in some easily accessible area of the distal bowel. Rapid bowel prep (4–6 L of PEG solution within 3–4 hours, usually via NG tube because who could possibly), is also an option, with a view to perform colonoscopy within six or so hours. This unpalatable option may be viable where angioembolisation is not possible. Interestingly, both societies recommend upper GI endoscopy as the next stage, if CT has failed to identify an embolisable lower GI source of bleeding.
Angiography is the gold standard, as it can identify a bleeder 85% of the time, and allows one to launch a bunch of Gelfoam into the offending artery, thus putting an end to the bleeding.
In order to see the characteristic "blush" of contrast, you need to be bleeding at a rate of at least 0.5ml/minute. Time is generally accepted to be the most important factor determining outcome here, and both societies suggest the angio happen as soon as possible following the CTA (ESGE actually specify 60 minutes).
Tc-99 Red Cell Scan is a way of radiolabelling RBCs and observing where they go. Apparently, it is identifies a GI bleeder 80% of the time. However it is unlikely that this will be available within the timeframe required to control acute bleeding.
Colonoscopy is the ideal treatment option for patients in whom the bleeding is minor, and who have been prepped sufficiently. It is probably the least invasive treatment option.
Angioembolisation is the best option for patients who are bleeding too rapidly to wait for bowel prep. In these people, the rate of bleeding - which makes endoscopy futile - also makes angiography more accurate. The societies are broad;y in agreement that haemodynamically unstable patients should undergo angioembilisation as soon as possible, with the UK guidelines showing a slightly higher preference for this modality.
Surgery is the option of choice for diverticular disease, unmanageable polyps and malignancies, and in situations where AVMs or inflammatory bowel disease lesions are not amenable to endoscopic management. It is the option of last resort for patients with bleeding that could not be controlled by other methods. On one hand, definitive control of major bleeding can be achieved by bowel resection, and good samples can be collected intraoperatively to aid diagnosis (eg. of malignancy), but on the other hand it is a laparotomy in an unprepped and unstable patient, with all the attendant risks and morbidity. Risk of perforation, slow recovery, wound infection, etc - but also one needs to consider the possibility that the bit of the bowel they resect was not the one that was bleeding. Determining the source of endoluminal bleeding is actually rather challenging, and the surgeons do not always got it right.
Balloon-occluded retrograde transverse obliteration (BRTO) is mentioned by the college in their answer to Question 1 from the first paper of 2017 as a "new technique and still undergoing evaluation", which is weird because we have been doing this since 1996 (Haruta et al). It is basically a technique whereby the interventional radiologist accesses a portosystemic shunt vein and inflates a balloon in it. Usually the gastric fundus varices drain into the left renal vein via a gastrorenal shunt, and that makes it all very convenient. With the balloon inflated, the radiologist can then inject sclerosant in a retrograde fashion into the the varix. Unfortunately, the usual sclerosant (ethanolamine oleate again) can cause anaphylaxis, renal failure and pulmonary oedema. Once the balloon is released, some sclerosant may make its way into the systemic circulation. This does not seem to impact much on outcomes, as Park et al (2015) reviewed the evidence and had found this technique "safe and efficacious".
Factor VIIa is also mentioned by the college in their answer to Question 1 from the first paper of 2017, though they themselves moderate their enthusiasm by pointing out that it has "questionable efficacy". If one reads the papers (eg. Bosch et al, 2008) this would certainly seem correct (there was no effect on any primary endpoints). The mention of this option in the college answer is itself questionable, as it is promoted as an option for controlling a re-bleed even when trial results "do not support the routine use of rFVIIa in this setting"