Prospective evaluation of ERCP performance in an Italian regional database study.

BACKGROUND: Prospective studies about endoscopic retrograde cholangio-pancreatography (ERCP) in a community setting are rare. AIM: To assess success and complication rates of routinely-performed ERCP in a regional setting, and the priority quality indicators for ERCP practice. METHODS: Prospective region wide observational study on consecutive patients undergoing ERCP during a 6-month period. A centralized online ERCP questionnaire was built and used for data storage. Primary quality indicators provided by the American Society of Gastrointestinal Endoscopy (ASGE) were considered. RESULTS: 38 endoscopists from 18 centers performed a total of 2388 ERCP. The most common indication for ERCP was choledocholitiasis (54.8%) followed by malignant jaundice (22.6%). Cannulation of the desired duct was obtained in 2293 cases (96%) and ERCP was successful in 2176 cases (91.1%). Success and ERCP difficulty were significantly related to the experience of the operator (p = 0.001 and p < 0.001, respectively). ERCP difficulty was also significantly related to volume centers (p < 0.01). The overall complication rate was 8.4%: post-ERCP pancreatitis (PEP) occurred in 4.1% of procedures, bleeding in 2.9%, infection in 0.8%, perforation in 0.4%. Mortality rate was 0.4%. All the ASGE priority quality indicators for ERCP were confirmed. CONCLUSIONS: The procedural questionnaire proved to be an important tool to assess and verify the quality of routinely-performed ERCP performance in a community setting.

Development of a novel index to discriminate bulge from mass on small-bowel capsule endoscopy.

BACKGROUND: Submucosal malignant masses (SMMs) and innocent bulges look similar on small-bowel capsule endoscopy (SBCE). In a previous observational study, 4 criteria associated with innocent bulges were recognized. OBJECTIVE: To devise and validate an index based on these criteria (smooth, protruding lesion index on capsule endoscopy [SPICE]) to discriminate SMMs from innocent bulges. DESIGN: Single-center, prospective study. SETTING: General hospital in Busto Arsizio, Italy. PATIENTS: This study involved 25 of 424 consecutive SBCEs performed on as many patients having SBCE findings of smooth, round, protruding lesions. INTERVENTION: Patients' evaluation up to the final diagnosis. At study entry, a short video clip of the lesion was obtained and deidentified for blind SPICE calculation. MAIN OUTCOME MEASUREMENTS: SPICE accuracy, using the final diagnosis of each patient as the criterion standard. RESULTS: Six patients had SMMs (4 GI stromal tumors, 2 neuroendocrine tumors), and 19 had innocent bulges. SPICE scores ranged from 0 to 4; they discriminated SMMs from innocent bulges (P = .002). A SPICE value >2 had 83.3% sensitivity and 89.4% specificity, and the area under the curve was 0.90 (95% confidence interval, 0.72-0.98; P < .001) for the detection of SMMs. LIMITATIONS: Single-center study; small sample size; no invasive ascertainment in 36% of patients. CONCLUSION: SPICE is easy to calculate and useful for distinguishing SMMs from innocent bulges. An index >2 is predictive of SMM.