Introduction
0
Gastrointestinal hemorrhage is a common reason for ICU admission. The approach to severe upper GI bleeding is relatively straightforward (figure below). A predictable approach facilitates planning ahead, and anticipating who needs to be contacted for help when.
0
0
0
Unfortunately, the approach to severe hematochezia is often less clear. Below is a description of how these cases often unfold. The diagnostic evaluation is frequently inconclusive. Fortunately, most cases of lower GI bleeding are due to diverticulosis or angiodysplasia and these generally stop without specific intervention.
0
Unfortunately, the approach to severe hematochezia is often less clear. Below is a description of how these cases often unfold. The diagnostic evaluation is frequently inconclusive. Fortunately, most cases of lower GI bleeding are due to diverticulosis or angiodysplasia and these generally stop without specific intervention.
0
0
Building Blocks: Performance of various tests
0
Diagnostic Nasogastric Lavage
0
Historically, diagnostic NG lavage has often been over-utilized in a broad range of patients with GI bleeding. For example, a recent article described the low yield of NG lavage in patients presenting with melena (Kessel 2015). To confuse matters further, most studies of NG lavage have combined patients presenting with either melena or hematochezia. Patients with an upper GI bleed presenting with hematochezia have a much brisker bleed than patients presenting with melena, and thus NG lavage might be expected to have a higher sensitivity in hematochezia.
0
Byers 2007 performed a prospective observational study of patients presenting to the emergency department with hematochezia who underwent NG lavage. Among 114 patients, 10% had a positive lavage and this had a high specificity for correctly identifying an upper GI source as confirmed upon endoscopy. Although this study does not define the sensitivity of NG lavage, it supports that NG lavage may have reasonable yield and high specificity in this context.
0
The sensitivity of NG lavage among patients presenting with hematochezia has not been studied. Based upon pooled studies of NG lavage of diverse presentations of GI bleeding, an estimate might be 50% (Palamidessi 2010). Duodenal bleeding can be missed. The specificity depends on the quality of material removed by the NG tube; a lavage demonstrating blood or coffee-grounds has a positive likelihood ratio of ~10 for upper GI bleeding (Srygley 2012).
0
The primary drawback of NG lavage is that it is very uncomfortable, although this can be alleviated with topical anesthesia (e.g., see the ALIEM blog). However, it has the advantages of being fast and inexpensive, with a reasonable yield and specificity (Anderson 2010).
0
Esophagogastroduodenoscopy (EGD)
0
EGD is potentially one of the more important tests in evaluation and management of hematochezia. 10-15% of patients with severe hematochezia may have an upper GI source with rapid intestinal passage. EGD has high sensitivity for identifying these patients and also allows for immediate therapy.
0
EGD does not have perfect specificity due the rare occurrence of multiple sources of bleeding. For example, a patient may have a minor gastric ulcer combined with active diverticular hemorrhage. There may be a risk of finding the gastric ulceration and ceasing further diagnostic efforts (“satisfaction of search”).
0
The main drawback of EGD is that it is an invasive test requiring conscious sedation, a gastroenterologist, and an endoscopy nurse. Logistically this may take anywhere from 30 minutes to several hours to organize. Given that most patients with hematochezia will nothave an upper GI source, this can cause significant delays in arriving at the correct diagnosis.
0
Colonoscopy
0
Unlike the stomach and upper gastrointestinal tract, it is difficult to suction and clear the colon of blood and stool during active bleeding. Therefore, for a critically ill patient with active hemorrhage, colonoscopy will often be impossible or nondiagnostic. Some studies and guidelines recommend emergent colonoscopy for patients with lower GI bleeding, either without bowel preparation or following emergent preparation. However, in our experience, this doesn't seem to work well and is not utilized for severe hematochezia.
0
Tagged RBC scan
0
Tagged RBC scan is frequently unhelpful. Its use in an emergency is limited due to time required to set up the study and acquire images. Even when it is positive, the image produced by extravasated blood is often unclear and doesn't locate the bleed with certainty. Up to 25% of bleeding scans suggest an incorrect location of bleeding, due to rapid luminal migration of blood (Ghassemi 2013). Tagged RBC scans have already been replaced by CT angiography at several centers (ASGE Guideline 2014).
0
0
jjj00
CT Angiography (CTA)
0
0
0
Advances in multi-detector helical CT scanning have allowed for the development of an IV contrasted CT scan which is highly accurate for locating bleeding anywhere along the GI tract. CTA typically consists of a series of three scans: an unenhanced CT scan of the abdomen, an arterial-phase contrasted CT scan, and a delayed venous-phase CT scan. Together, these scans provide a wealth of information about the patient's anatomy and the location and character of any bleeding. Meta-analysis revealed a sensitivity of 85% and specificity of 92% for identifying the bleeding source (Garcia-Blazquez 2013). With severe active bleeding the performance is better (sensitivity >90%; Geffroy 2011) . CTA has five major advantages compared to more traditional approaches:
0
(1) Detection and characterization of obscure bleeding sites
0
CTA has the ability to identify common sources of bleeding (both upper and lower) as well as more obscure sources of bleeding (e.g., aortoenteric fistula, small bowel sources, hemobilia). It may also provide information characterizing an underlying lesion (e.g. identification of diverticula, tumors, etc.). For example, the following images are from a CTA obtained in a patient with hemobilia due to a gangrenous gallbladder. CTA localizes bleeding to the gallbladder and also characterizes underlying biliary and vascular pathology, expediting appropriate management (in this case, cholecystectomy).
0
0
(2) Diagnosis of other abdominal pathologies that present with hematochezia
0
Patients presenting with hematochezia and shock are generally assumed to have hemorrhagic shock. However, a variety of disorders can mimichemorrhagic shock, for example infectious colitis causing septic shock, ischemic colitis due to cardiogenic shock, or mesenteric ischemia causing systemic inflammatory response syndrome. CTA will rapidly reveal these intestinal pathologies, immediately re-directing the management of these patients.
0
00…
(3) Speed and availability
0
Aside from NG lavage, CTA is often the fastest and most available study. Only intravenous contrast is utilized, so this test may be performed in the emergency department in under 10 minutes (Copland 2010). For a critically ill patient, this may facilitate immediate triage to a curative procedure (e.g., angiography), rather than performing a series of time-consuming tests (e.g. EGD first, then tagged RBC scan second when EGD is negative, then angiography third).
0
“CTA should be the standard of care for assessment of patients presenting with acute lower GI bleed”
– Chan et al. 2014, John Radcliffe Hospital, Oxford UK.
0
0
(4) Ability to target invasive angiography or surgery
0
When positive, CTA reveals the location and often the precise vascular anatomy leading up to the lesion. This may facilitate the speed and success of a subsequent invasive angiography procedure to embolize the bleeding site. If surgical resection is required, it may provide an adequate level of certainty that the surgeon will resect the appropriate segment of bowel. Tagged RBC scans do not provide this level of precision.
0
0
(5) Immediate prognostication and triage
0
CTA cannot detect very slow bleeding (i.e., < 0.3-0.5 ml/min). Thus, although CTA may miss some cases of bleeding, it will miss the slowestsources of bleeding. Indeed, although a negative scan doesn't reveal the source of bleeding, it still provides useful prognostic information.
0
Lower GI bleeding has a mortality rate of 2-4%, significantly lower than upper GI hemorrhage. Nonetheless, hematochezia may be quite visually impressive and this can provoke anxiety leading to over-transfusion and unnecessary ICU admission. A negative CT angiogram may be a helpful clue that bleeding may have stopped spontaneously. Chan 2014found that among patients presenting with lower GI bleeding and negative CTA, 77% had no recurrence of bleeding. Thus, a patient with a negative CTA who is otherwise stable may be appropriate for admission to the ward rather than the ICU.
0
0
Drawbacks: Safety concerns
0
CTA does involve exposure to contrast dye, and if the patient requires invasive angiography this will involve two contrast exposures. However, the existence of contrast nephropathy with modern contrast dyes is questionable (discussed further here). CTA requires 100-125ml of IV contrast, which for comparison is less than half of what may be required for a complex cardiac catheterization procedure (Artigas 2013). Overall, if the patient does not have severe renal failure and a safer contrast dye is utilized, this is unlikely to cause a problem.
0
CTA does also involve radiation exposure, which is concerning primarily among younger patients. Younger patients overall are more likely to have an upper GI source of hemorrhage (most causes of lower GI bleeding such as diverticular bleeding and angiodysplasia become more common with age). Therefore, it may be reasonable to try to utilize EGD rather than CTA as the initial test for younger patients, on the basis of both yield and avoidance of radiation exposure.
0
Invasive Angiography
00
Angiography is one of the most useful procedures for lower GI bleeding. It has the capability to diagnose the source of bleeding, although this requires a faster bleeding rate compared to CTA (e.g., >0.5-1 ml/min) rendering it somewhat less sensitive. Most importantly, it can provide therapeutic embolization.
0
Angiography is usually not used as an initial test, except in cases of exsanguinating lower GI bleeding. Without knowledge of where the bleeding is coming from (e.g. based on CT angiography or endoscopy), blind angiography is harder to perform as this requires sequential injection of multiple arteries searching for the bleed. Angiography also requires mobilization of an interventional radiologist and the interventional radiology suite, which further limits its ability to be used as a first-line investigation.
0
Proposed approach
0
0
0
Above is flexible approach to severe hematochezia incorporating CT angiography and clinical judgment. This is not truly “new,” as various CTA-based approaches have been advocated for several years and are already utilized in many centers (e.g. Copland 2010). However, knowledge translation has often been sluggish.
Above is flexible approach to severe hematochezia incorporating CT angiography and clinical judgment. This is not truly “new,” as various CTA-based approaches have been advocated for several years and are already utilized in many centers (e.g. Copland 2010). However, knowledge translation has often been sluggish.
0
The first goal of the algorithm is evaluating for upper GI hemorrhage, since these patients have the highest mortality and benefit most from intervention. For patients at high likelihood of upper GI hemorrhage, it is sensible to proceed directly to EGD (as is currently recommended in many algorithms for all patients with hematochezia). However, older patients without risk factors for upper GI bleed probably have a rate of upper GI bleed <10%. If such a patient has a negative NG lavage, then their risk of having an upper GI bleed may be <5%. At that pre-test probability, it may make more sense to proceed to CTA rather than EGD. Mis-directing a patient with upper GI bleed to CTA should not cause the upper GI bleed to be missed for too long, since CTA is sensitive for upper GI bleeding as well as lower GI bleeding (1).
0
This algorithm eliminates both colonoscopy and tagged RBC scan from the initial approach to severe hematochezia (similar to the algorithm by Marion 2014). Both of these tests are time-consuming and often low-yield. Delaying other tests may allow intermittent bleeding sources to stop, reducing the diagnostic yield. In contrast, CTA provides immediate information about the rate and location of bleeding anywhere in the GI tract.
0
This algorithm does utilize NG lavage for some patients. Some authors have recommended skipping NG lavage and proceeding directly to CT angiogram (Sun 2012). However, NG lavage may occasionally be useful because if positive it will facilitate expedited management (allowing omission of CTA and proceeding directly to endoscopy). A reasonable approach might be to try passing an NG tube with topical analgesia, but if this is not tolerated or unsuccessful not to persist with excessive attempts at NG passage.
0
0
- Abdominal CT angiography is a fast test with high performance to reveal bleeding anywhere in the GI tract. CTA has already replaced tagged RBC scanning in many centers.
- An approach incorporating physician judgment, NG lavage, and CTA may allow for thorough evaluation of hematochezia without subjecting every patient to an upper endoscopy (EGD).
- In situations where endoscopy is not immediately available, CTA may allow for rapid and accurate evaluation of hematochezia. This may help identify which patients require immediate intervention and which patients can be safely observed.
0
This post was co-authored with Dr. Paul Farkas, my father and senior consultant in Gastroenterology.
Additional Reading
…
…
- Copland A et al. Integrating urgent multidetector CT scanning in the diagnostic algorithm of active lower GI bleeding. Gastrointestinal Endoscopy, 2010; 72(2) 402-405.
- Artigas JM et al. Multidetector CT angiography for acute gastrointestinal bleeding: Technique and findings. Radiographics 2013; 33: 1453-1470.
Notes
(1) Additionally, an upper GI bleed with a negative NG lavage presenting with hematochezia is likely to represent a penetrating duodenal ulcer (often involving the gastroduodenal artery). It is not uncommon for this type of ulcer to fail to respond to therapy by EGD and require angiography. Therefore, obtaining a CTA in this situation is not necessarily the “wrong” approach but instead it may prove useful in guiding angiography if EGD fails to achieve hemostasis.
Josh is the creator of PulmCrit.org. He is an associate professor of Pulmonary and Critical Care Medicine at the University of Vermont.
Latest posts by Josh Farkas (see all)
- PulmCrit Wee: Michelin Chest Syndrome - March 15, 2025
- PulmCrit: ADAPT and SCREEN trials are full of sound and fury, signifying little - December 13, 2024
- PulmCrit: How to quickly create a useful professional account in BlueSky - November 28, 2024
Great post. Thank you for sharing.