Predictive Factors of Significant Findings on Capsule Endoscopy in Patients with Suspected Small Bowel Bleeding

Background: Small bowel capsule endoscopy (SBCE) is an established non-invasive diagnostic modality for a variety of small bowel pathologies and has a significant role in altering the treatment course. The diagnostic yield of SBCE in the published literature varies widely between 45 and 75%. Furthermore, it is unclear if any patient-related factors predict higher diagnostic yield. The aim of this study is to report the diagnostic yield of SBCE for suspected small bowel disease and identify any predictive factors for identifying significant pathology on SBCE. Method: A retrospective study was conducted at Mubarak Al-Kabeer Hospital in Kuwait for patients who underwent SBCE between October 2013 and February 2022. All patients underwent upper and lower endoscopy prior to referral for SBCE. Patients' medical records were reviewed to determine SBCE indications, results, and complications. The significance of the SBCE finding was classified according to the Saurin system. A logistic regression was performed to characterize baseline predictors for identifying significant pathology on SBCE. Results: Overall, 210 patients underwent SBCE and were included in the analysis. The mean age was 57.9 years (SD 18.5), and 129 (61.4%) were males. The most common indication for SBCE was obscure occult gastrointestinal bleed (75.7%), obscure overt gastrointestinal bleed (28.6%), and investigating gastrointestinal symptoms (7.6%). Adequate bowel preparation was achieved in most patients (88.1%), imaging of the entire small bowel was achieved in 194 patients (92.4%), and no adverse events were recorded. The overall diagnostic yield of SBCE for small bowel disease was 68.1%. The most common findings were vascular lesions in the small bowel (40.0%), small bowel ulcers (22.9%), and erosions (22.9%). On multivariate regression analysis, melena at baseline was significantly associated with increased odds of identifying high-risk lesions (Saurin class P2) (OR 2.1, 95%CI 1.03-4.30, p = 0.04). Conclusions: SBCE is an effective and safe tool for investigating small bowel pathology with a diagnostic yield of 68.1% in carefully selected patients undergoing such a test. Melena at baseline is the strongest predictor of identifying high-risk lesions, and patients with which should be prioritized for SBCE.