Background:
The current GI guideline recommendations for early diagnosis and lesion localization of patients with severe LGIB are to perform computerized tomography angiography (CTA) or labelled red blood cell (RBC) scanning and not urgent colonoscopy. However, only active bleeding (e.g. contrast extravasation) is diagnostic whereas non-bleeding stigmata of recent hemorrhage (SRH) such as non-bleeding adherent clots, visible vessels (NBVV), or flat spots can not be detected by these imaging techniques nor is a lesion diagnosis usually possible. Based upon urgent colonoscopy findings in definitive diverticular hemorrhage, only 25-30% of SRH are active bleeding whereas 70-75% are non-bleeding SRH. Improvements in early lesion diagnosis and bleeding lesion location are needed to guide patient management of LGIB.

Purpose:
To compare the diagnostic yield of urgent colon capsule endoscopy (CCE) with either CTA and/or RBC scanning in patients with severe hematochezia (e.g. LGIB) suspected to be from the colon or distal small bowel.

Methods:
17 patients with 18 hospitalizations for severe LGIB received both urgent colon capsule endoscopy and CTA and/or RBC scan in a prospective 2 center Institutional Review Board (IRB) approved study. Both tests were performed soon after a GI team saw the patient in the ER or hospital for severe hematochezia. Results were blinded for study investigators. All patients also had colonoscopy which was considered to be the gold standard. For statistical comparisons p ≤ 0.05 was considered significant.

Results:
Refer to Table 1 for baseline demographics and final diagnoses based upon colonoscopy and all imaging studies. The majority of patients were old, males, taking medications associated with GI bleeding, had significant co-morbidities, and were severely anemic after an acute LGIB. Refer to Table 2 for comparisons of diagnostic yields for all LGI patient hospitalizations (N = 18) and those with documented diverticular hemorrhage (N = 10). The diagnostic yield (e.g. bleeding site localization and/or lesion diagnosis) was significantly higher with CCE than standard imaging for all LGIB patients (77.8 vs. 11.1%) and also those with documented diverticular hemorrhage (80% vs. 10%).

Conclusions:
For patients hospitalized with severe LGI hemorrhage: 1. Urgent colon capsule endoscopy had significantly higher diagnostic yields than imaging with CTA and/or RBC scanning. 2. For the largest group of patients with documented diverticular hemorrhage, the rate of diagnosis with CCE was also significantly higher than standard imaging. 3. There are major practical advantages to CCE compared to currently recommended imaging which include convenience of testing at the bedside, lack of sedation, lower cost, and capability of remote interpretation by experts. 4. Further studies are warranted to confirm and expand upon these results.

Christopher Bradley

Appearances