Effective Withdrawal Time and Adenoma Detection Rate

This study prospectively evaluated the role of EWT versus SWT on adenoma detection rate (ADR) and other key quality metrics. In this prospective single-center study, patients undergoing colonoscopy were enrolled. EWT was calculated in real-time using an AI system with endoscopists blinded to the results. We performed multivariable analyses to assess the association of EWT and SWT with binary (e.g., ADR) and count outcomes (e.g., adenoma per colonoscopy \[APC\]), after adjusting for patient and procedural characteristics.

This was a prospective, single-center observational study designed to determine if an AI-powered metric, Effective Withdrawal Time (EWT), is a superior predictor of colonoscopy quality compared to the traditional Standard Withdrawal Time (SWT). All colonoscopies were performed by qualified endoscopists using high-definition white light video scopes. During the procedure, the scope is first advanced to the start of the large intestine (the cecum). The critical examination phase-the withdrawal-begins as the endoscopist slowly pulls the scope back out, meticulously inspecting the colon lining for abnormalities like polyps. It is during this withdrawal that the key metrics were measured. While SWT is a simple duration timed manually, the AI-measured EWT specifically quantifies the time of high-quality mucosal inspection, automatically excluding periods when the camera view is blurry, obscured, or moving too quickly. A crucial aspect of the methodology was that the endoscopists were blinded to the live EWT measurements to prevent the Hawthorne effect, where individuals alter their behaviour because they are being monitored. The study enrolled adults aged 40 and over, excluding patients with conditions that could confound the findings. The primary goal was to assess the independent impact of EWT on the Adenoma Detection Rate (ADR), a key benchmark based on the detection and removal of precancerous polyps for analysis. To achieve this, researchers used multivariable regression models to isolate EWT's effect from other variables and employed correlation tests to statistically compare whether EWT had a stronger relationship with detection quality than SWT