TOP 25 - Bladder Point-of-Care Ultrasound to Assess Patient Readiness for Pelvic Ultrasound with the use of Artificial Intelligence: A Randomized Controlled Trial

Publication Number: 4728.TOP 25

Callie Alt, Nicklaus Children’s Hospital, Miami, FL, United States; Paul Khalil, Nicklaus Children’s Hospital, Miami, FL, United States

Background: Ultrasound (US) is the preferred diagnostic method in the assessment of pelvic pain in women presenting to the Emergency Department (ED). The transabdominal pelvic US (TAPUS) requires a full bladder for sonographic visualization of the pelvic organs. Patients’ subjective reporting of bladder fullness can be inaccurate and lead to delays in diagnosis and management. Timely imaging is critical to preserve ovarian function. Prior studies have shown that the use of POCUS to evaluate bladder fullness can decrease time to TAPUS. POCUS is considered highly operator-dependent. Artificial Intelligence (AI) can reduce the variability between sonographers and assist in making more objective measurements in real-time.

Objective: To determine whether the use of a POCUS machine with AI capabilities (EXO Iris) to assess for bladder volume can expedite time to TAPUS, quicker than subjective sensation of bladder fullness.

Design/Methods: We are performing a Randomized Controlled Trial in a pediatric ED with Level I Trauma designation. We are enrolling patients aged 8 to 21 years with lower abdominal pain requiring a TAPUS. Patients are randomized, in a 1:1 ratio, to the standard care group or to the AI-POCUS group with bladder volume measurements every 30 minutes. Patients are subsequently sent for pelvic US once the bladder is considered subjectively full or until 350-mL has been reached (whichever comes first). The primary outcome is the time to successful pelvic US. The secondary outcomes are ED LOS and the rate of successfully completed pelvic US on first attempt. The study is approved by the IRB.

A sample of 264 patients per group is required to detect a 30-minute difference between the methods (α 0.05, independent two-tailed test, power 80%). The analysis of the primary outcome will be based on estimating the absolute difference in time to successful US between the two study groups, along with 95% CI and p-values. The same approach will be used to assess the ED LOS. The frequencies of successfully completed US on first attempt will be compared by obtaining risk differences, along with 95% CI. Analyses will be conducted using the Intention to Treat approach.