Capsule endoscopy (CE) is a widely used modality for small-bowel evaluation, but its performance may be affected by technical failures, incomplete examinations, or clinical complications such as retention. The aim of this study was to assess the diagnostic yield, technical limitations, and safety profile of CE in routine clinical practice over an eight-year period in a tertiary referral center.

This retrospective study included all patients who underwent CE between 2017 and 2024 in the Digestive and Hepato-Gastroenterological Functional Explorations Department at Ibn Sina University Hospital, Rabat, Morocco. CE was performed using PillCam SB3 or CapsoCam devices following standardized bowel preparation. Indications included obscure gastrointestinal bleeding (overt or occult), suspected or known Crohn’s disease, celiac disease, malabsorption syndrome, Rendu–Osler disease, and small-bowel lymphoma. Technical limitations were defined as equipment-related disruptions, and clinical limitations as events related to patient factors or physiological conditions. Data were analyzed using SPSS.

A total of 125 patients were included (mean age 56.2 ± 19.6 years; 54.4% female). Obscure gastrointestinal bleeding was the most frequent indication (76.8%), followed by Crohn’s disease (11.2%) and celiac disease (7.2%). CE completion rate, defined by cecal visualization before battery depletion, was 88%. Median small-bowel transit time was 3 hours and 10 minutes. The overall diagnostic yield was 65%, with angiodysplasia identified in 32.8% of patients and small-bowel ulcerations in 17.6%.

Clinical limitations and complications were observed in a subset of examinations. Incomplete examinations occurred in 11.2% of patients, mainly due to delayed gastric or small-bowel transit. Capsule retention occurred in 2.4% of cases. Two patients had strictures related to Crohn’s disease requiring surgical retrieval, while one patient with celiac disease passed the capsule spontaneously after corticosteroid therapy. Four cases of gastric retention required endoscopic deployment; three progressed into the small bowel while one remained in the stomach despite multiple attempts.

Capsule ingestion difficulties occurred in four patients, three of whom had pre-existing swallowing disorders, including dysphagia or a history of cerebrovascular accident. In one elderly patient with severe dysphagia, CE could not be performed. Inadequate visualization occurred in 10.4% of examinations, predominantly due to residual food debris, particularly affecting duodenal and jejunal segments. Blood and clots impaired visualization in six cases, mainly among patients with overt gastrointestinal bleeding.

Technical failures were rare, with a single case (0.8%) of premature battery depletion in the proximal jejunum.

CE demonstrated high diagnostic performance with a low rate of serious complications. The main limitations were related to clinical factors such as incomplete transit, capsule retention, swallowing disorders, and inadequate mucosal visualization, rather than true technical failures. Capsule retention remained infrequent and predictable in high-risk groups. Improving patient selection, optimizing bowel preparation, and ensuring careful evaluation of dysphagia risk may enhance CE performance and safety. Technological improvements, including extended battery life, real-time localization, and better image cleansing algorithms, could further reduce failures and improve diagnostic accuracy.