Accuracy of Remote Interpretation of Pediatric Emergency Ultrasound Over Third Generation Networks Across Continents

Objective This study aimed to evaluate live-stream videos for evaluation of pediatric emergency medicine ultrasound using a third generation (3G) network. Methods The interrater reliability of a remote 3G network live-streaming assessment of ultrasound scans for abdominal free fluid, intussusception, and hip effusions was evaluated. Fifty-eight deidentified 6-second video clips in .mp4 format of focused assessment with sonography in trauma scans were downloaded to a laptop. A "bedside" ultrasound expert evaluated each scan and marked it positive or negative for the presence of free fluid. Using an iPhone 3G, Skype was used to connect to the second ultrasound expert located in Addis Ababa, Ethiopia, who reviewed the same scans. When the remote expert returned to the United States, evaluation of the scans was repeated as a second bedside reviewer. Interrater reliability was assessed using the kappa statistic and percent agreement. This process was repeated for 44 intussusception scans and 28 hip effusion scans. Results The kappa values (95% confidence interval) between the bedside and remote evaluators for focused assessment with sonography in trauma, intussusception, and hip effusion were 0.748 (0.576-0.92), 0.816 (0.648-0.984), and 0.764 (0.519-1.0), respectively. The percent agreements after adjusting for chance were 86%, 80%, and 88%, respectively. The kappa values between the bedside assessments by each evaluator were 0.851 (0.71-0.992), 0.8 (0.616-0.984), and 0.747 (0.479-0.884), respectively. The percent agreements after adjusting for chance were 92%, 81%, and 88%, respectively. Conclusions Live-stream video conference using cellphones over 3G network is a feasible and accurate method for consultation of ultrasound scans from a remote location.