A new American Society for Gastrointestinal Endoscopy guideline published in Gastrointestinal Endoscopy suggests colonoscopy as the initial test for acute lower GI bleeding in patients who are hemodynamically stable at presentation or after resuscitation, and contrast-enhanced CT, with or without angiography, for those who remain unstable.
The guideline considers either approach acceptable as a first test. Prepared by ASGE’s Standards of Practice Committee, it also covers treatment of diverticular, vascular, and procedure-related bleeding.
The choice between colonoscopy and CT angiography “should largely be guided by a patient’s initial presentation with lower GI bleeding,” first author Nauzer Forbes, MD, MSc, FASGE, of the University of Calgary in Alberta, Canada, told GI & Hepatology News. “This is probably the single most important decision node when it comes to investigating and managing these patients.”
“Existing ASGE guidance was somewhat outdated regarding the management of lower GI bleeding,” Dr. Forbes said.
The guideline states that upper endoscopy should be performed in all patients presenting with hematochezia and initial hemodynamic instability to rule out an upper GI source, unless imaging directly shows a lower GI source. In a randomized trial of 85 patients with hematochezia and hemodynamic instability or a rapid hemoglobin drop, an upper GI source was identified in 15% of those who underwent upper endoscopy within six hours, most commonly peptic ulcers or varices.
On the choice of initial test, a meta-analysis of eight retrospective cohort studies found similar rates of bleeding identification and initial control with CT and colonoscopy. Patients who underwent CT first were older, had more comorbidities, presented with lower blood pressures, and received more transfusions. The panel cautioned that patients with the most severe presentations may have been preselected for CT, making the CT group’s higher mortality and longer hospital stays difficult to interpret. The recommendation is conditional, based on very low-quality evidence.
The guideline suggests performing colonoscopy once patients are stable enough for the procedure and able to tolerate a full bowel preparation, regardless of timing — a conditional recommendation based on moderate-quality evidence. In a pooled analysis of four randomized trials involving 463 patients, recurrent bleeding did not differ between urgent and nonurgent colonoscopy; three trials involving 385 patients showed no difference in 30-day mortality.
In a sensitivity analysis limited to the two trials that defined urgent colonoscopy as within 24 hours, urgent colonoscopy was associated with about three times the odds of recurrent bleeding. It was also associated with shorter hospital stays, which the panel characterized as roughly a one-day reduction. The panel found no differences in need for surgery, radiologic intervention, or transfusion, or in adverse events.
Full bowel preparation carries the guideline’s only strong recommendation, despite very low-quality evidence. The panel limited exceptions to a strongly suspected distal source or iatrogenic bleeding at a known, easily identifiable location, such as after polypectomy or endoscopic resection. Patients with established radiation proctopathy should still receive a full bowel preparation before argon plasma coagulation, the panel noted, given the risk of colonic explosion with higher levels of intestinal gas.
The guideline suggests choosing between band ligation and clipping for diverticular bleeding based on the lesion’s location, clinical expertise, and device availability. It suggests band ligation for lesions in the transverse or proximal colon when feasible and through-the-scope clipping for distal lesions. In seven observational studies involving 2,329 patients, banding was associated with about one-third the odds of recurrent bleeding compared with clipping, although the benefit appeared limited to right-sided lesions.
Banding was also associated with less radiologic intervention and a nonsignificant trend toward more adverse events. All comparative studies were conducted in Asia. Much of the research used band ligators mounted on gastroscopes, and colonoscope-specific banding devices have limited availability in North America. The panel advised attempting banding only when the endoscopist is experienced with the approach and a colonoscope-compatible device is available.
When clipping is used, ASGE suggests placing clips directly on the bleeding vessel rather than closing the diverticular opening, particularly in the proximal colon. Nine observational studies associated direct clipping with 42% lower odds of recurrent bleeding and 50% lower odds of radiologic intervention than indirect clipping. Initial hemostasis, transfusion, surgery, and adverse events were similar. The panel noted that direct clipping may not be feasible with small-mouthed diverticula or when no clear vessel is visible.
For bleeding after polypectomy or endoscopic resection, the panel suggested through-the-scope clips or thermal coagulation, with hemostatic powder as an additional or rescue treatment. Argon plasma coagulation was suggested for angioectasia and radiation proctopathy. For bleeding from malignancy, the panel suggested hemostatic powder or Nd:YAG laser therapy to achieve temporary hemostasis, either for palliation or before surgery.
Surgery “should be considered a last resort,” the guideline states, reserved for bleeding that is not amenable to endoscopic or radiologic treatment or that continues after both fail.
The panel used the Grading of Recommendations Assessment, Development, and Evaluation framework. It noted that most evidence came from retrospective studies prone to selection bias, that some data were older or from outside the U.S., and that definitions of urgent colonoscopy varied. All comparative studies of diverticular bleeding therapies were conducted in Asia, and the clipping studies included much higher proportions of right-sided lesions than are seen in Western settings.
ASGE funded the guideline. Dr. Forbes and several coauthors reported financial relationships with medical device and pharmaceutical companies.
Expert Insight
Dr. Forbes shared what physicians should take from the updated guidance.
How should physicians decide between colonoscopy and CT angiography as the initial test for acute lower GI bleeding?
Dr. Forbes: The ASGE guideline suggests that the decision to pursue colonoscopy versus CT angiography should largely be guided by a patient’s initial presentation with lower GI bleeding. This is probably the single most important decision node when it comes to investigating and managing these patients. If a patient is stable (or initially unstable and then responsive to initial resuscitation), then endoscopic evaluation is usually indicated to evaluate and ideally treat a bleeding lesion, if possible. If a patient is persistently unstable despite initial resuscitation, then usually, an urgent CT angiogram is advised to help guide therapy alongside ongoing resuscitation and supportive management.
Which patients, if any, are most likely to benefit from urgent colonoscopy?
Dr. Forbes: In our review that informed this guideline, a total of 18 studies addressed the question of colonoscopy timing in the setting of lower GI bleeding, with four of them being randomized trials. Most of the included studies characterized any colonoscopy performed within 24 hours of presentation as “urgent.” From the randomized trials, there were no differences in key outcomes including mortality, recurrent bleeding, need for surgery, need for radiologic intervention, or transfusion requirements. Therefore, our panel opted to issue a conditional recommendation suggesting that colonoscopy be performed when the patient is stable enough for the procedure, and when they can tolerate a full bowel preparation. The actual timing of the procedure seems to be much less important than these two issues, and there appears to be no basis in the evidence to push for procedures to be done more urgently in this clinical setting.
What is the preferred endoscopic treatment for diverticular bleeding, and how should local expertise influence that choice?
Dr. Forbes: Diverticular bleeding was a particular area of focus in our guideline. Banding versus clipping were compared as endoscopic management strategies. Both are mechanical hemostatic modalities, but pooled data from seven observational studies showed that recurrent bleeding was lower following banding compared to clipping. Most of these studies showed benefit primarily for patients with proximal diverticular disease, with less or even no benefit for left-sided lesions. Given this, and given that there are currently no widely available banding devices in North America that can be loaded over colonoscopes, either banding or clipping is likely acceptable for all bleeding diverticular lesions at this time. This area is an interesting area for future study. We also assessed direct clipping versus indirect clipping for diverticular bleeding. While indirect clipping involves closing the entire diverticulum shut, direct clipping requires visualization of a bleeding vessel within a diverticulum, with one or more clips applied to the vessel directly. Nine observational studies were identified that showed that direct clipping was associated with significantly lower odds of recurrent bleeding and lower odds of needing radiologic intervention. This led the panel to issue a conditional recommendation suggesting direct clipping, if feasible.